# Script and Society

# Script and Society

The Social Context of Writing Practices in Late Bronze Age Ugarit

Philip J. Boyes

Published in the United Kingdom in 2021 by OXBOW BOOKS The Old Music Hall, 106–108 Cowley Road, Oxford OX4 1JE

and in the United States by OXBOW BOOKS 1950 Lawrence Road, Havertown, PA 19083

© Oxbow Books and the author 2021

Hardback Edition: ISBN 978-1-78925-583-6 Digital Edition: ISBN 978-1-78925-584-3 (ePub)

A CIP record for this book is available from the British Library

Library of Congress Control Number: 2020952173

An open-access on-line version of this book is available at: http://books.casematepublishing.com/ Script\_and\_Society. The online work is licensed under the Creative Commons Attribution 3.0 Unported Licence. To view a copy of this license, visit http://creativecommons.org/licenses/ by/3.0/ or send a letter to Creative Commons, 444 Castro Street, Suite 900, Mountain View, California, 94041, USA. This licence allows for copying any part of the online work for personal and commercial use, providing author attribution is clearly stated.

Some rights reserved. No part of the print edition of the book may be reproduced or transmitted in any form or by any means, electronic or mechanical including photocopying, recording or by any information storage and retrieval system, without permission from the publisher in writing.

Materials provided by third parties remain the copyright of their owners.

Printed in the United Kingdom by Short Run Press

Typeset by Versatile PreMedia Services (P) Ltd

For a complete list of Oxbow titles, please contact:

UNITED KINGDOM UNITED STATES OF AMERICA Oxbow Books Oxbow Books

Telephone (01865) 241249 Telephone (610) 853-9131, Fax (610) 853-9146 Email: oxbow@oxbowbooks.com Email: queries@casemateacademic.com www.oxbowbooks.com www.casemateacademic.com/oxbow

Oxbow Books is part of the Casemate Group

Cover illustration by Philip Boyes. Stele from Ugarit showing treaty-signing with possible writing-boards.

### Contents


*For Jennie*

### Acknowledgements

No research project is the work of a single person in isolation, and this book in particular has benefited from the support and assistance of a great many people. It was written between 2016 and 2020 under the aegis of the CREWS Project – Contexts of and Relations between Early Writing Systems, based at the Faculty of Classics, University of Cambridge and funded by the European Research Council under the European Union's Horizon 2020 research and innovation programme (grant agreement No 677758). This has, of course, been a time of great trauma and upheaval for the United Kingdom's relationship with Europe, and I am profoundly grateful to the European Union for all it has done, and continues to do, to support research such as this. I hope that Britain's self-imposed exile from this hugely beneficial combined effort will be a short one.

The CREWS Project has consisted of five members: Philippa Steele, the principal investigator, Robert Crellin, Natalia Elvira Astoreca, our administrator Sarah Lewis, and myself. The research in this volume has been massively enriched by the ideas and practical support these colleagues have offered, but I must single out Philippa for particular thanks. The best project leader anyone could hope for, she has offered constant advice, feedback and support over the last four years, including reading and commenting on multiple drafts of the manuscript.

Beyond CREWS, I must thank a number of other colleagues at Cambridge, in particular Martin Worthington, Selena Wisnom and Marie-Françoise Besnier, who all went above and beyond the call of duty in allowing me to join their Akkadian classes. Without them, my understanding of Semitic scripts and languages and of Mesopotamian culture would be very much poorer. Augusta McMahon served as my mentor in postdoctoral research and has also given helpful thoughts and advice. I must also thank Nicholas Postgate for many helpful comments on my research at seminars and conference presentations, and for very generously giving me his volumes of *Le Palais royal d'Ugarit*. My friends and colleagues at the Faculty of Classics, Annie Burman, Anna Judson, Matthew Scarborough and Daniel Unruh, have all been great sources of advice and support on research and navigating a postdoctoral career. I would also like to thank the University's 'support' staff – especially the librarians, computer officers and open access advisers – who have done a great deal to facilitate my work and its dissemination. The non-academic staff at universities rarely get the credit and appreciation they deserve, and I am very grateful to them.

Several fellow scholars have allowed me to reuse theirimages or offered thoughts, questions or suggestions when I have presented my research which have shaped my subsequent thinking. Where appropriate, these are acknowledged in the text. Others have sent me articles, pointed me in the direction of references or answered questions I had. Sophie Hardach helped me with some German.

Two reviewers read and commented on a draft manuscript of this work. Their suggestions were extremely constructive and helpful, and I am extremely grateful for their time and advice.

A draft version of this volume was largely completed by spring 2020, but the process of review and revision coincided with the coronavirus pandemic that took hold early in that year. I was fortunate to have completed the bulk of the research before the effects of this crisis became severe, but even so, there has been an impact. It hasn't always been possible to chase up references or contact people or organisations for image permissions. This is not to blame the virus for any mistakes or deficiencies in what is presented here, but to admit openly that there are things I would have liked to have included or elaborated in the later stages of revision which I have not been able to.

Finally, my deepest thanks and love to my wife Jennie, for everything she does.

*Fig. 0.1. Map of the city of Ugarit. Redrawn by the author after Saadé (2011, fig. 46).*

*Fig. 0.2. Map of the Kingdom of Ugarit. Drawn by the author after Calvet (2012, fig. 1).*

### Part I

Background, theory and methods

### Chapter 1

### Introduction: Ugarit and its scripts

Writing is a social practice. It fitsinto the same category as cooking a meal, performing one's daily routines, worshipping a deity. It is a thing that people do, and that they do according to patterns within which they have been socialised or which they have cultivated. It is not usually a solitary practice but one of communication and interaction, even if that occurs at some remove in time and space and even if the interlocutor in some cases is only imagined.

This, I would hope, is self-evident, since it forms the foundational premise for this book. It is also somewhat at odds with how writing in the ancient world can often be approached by scholars. Much work on writing has focused on writing *systems*, on the abstracted and self-contained workings of the scripts themselves, their development, spread and the techniques of their use. Research of this kind, while useful, can be strikingly unpeopled: it is filled with systems, graphemes, phonemes and styles. It is immaterial, austere and often mechanistic, and can seem divorced from the other practices of human life, from beliefs and agendas, from choices and agency. The goal of this work is to redress that balance, to reintegrate writing practices with other aspects of human practice and human social life, to situate them within their specific historical, cultural and material contexts – in this example using the case study of the Late Bronze Age kingdom of Ugarit, a small but prosperous trading city on the coast of what's now Syria. It is, in short, to produce an archaeology of writing practices at Ugarit, fully integrated into the rest of the polity's archaeology.

As such, this book isn't a comprehensive guide to the languages and scripts of Ugarit. Such works already exist by scholars eminently more qualified than me to describe the linguistic and palaeographic details.1 Nevertheless, since I hope this volume will be of interest to archaeologists and non-specialists who may not be as well acquainted with the principles of the main scripts we'll be discussing, this introductory chapter will lay them out in brief summary, after providing a grounding in the site of Ugarit and the history of research there.

<sup>1</sup> See, in particular, Schniedewind and Hunt 2007; Huehnergard 1989; 2012; Bordreuil and Pardee 2009.

#### **Introducing Ugarit**

A 5 heures de l'après-midi, lorsque le soleil couchant transformait les montagnes alaouites à l'est du tell en une frange dorée, j'observais l'un de mes ouvriers qui arrêta son travail pour examiner ce qui à distance avait l'aspect d'une petite brique. Mohamed Moursal, un Turcoman de Bordj Islam, bon ouvrier, mais préférant l'effort plutôt que le travail délicat de dégager des objets fragiles crachait sur sa trouvaille et avec la paume de sa main droite frottait dessus pour enlever la pellicule de terre qui masqua la surface.2

Thanks in part to colourful retellings such as that quoted above, written by Claude Schaeffer almost thirty years after the event, the tale of the discovery of the site of Ugarit and its archives of tablets has acquired something of the quality of legend, a true-life tale of chance discovery and buried treasure. In 1928, we are told, a local farmworker was ploughing fields near Minet el-Beida, around 12 km north of modern Latakia. His plough struck a stone and revealed the opening to a vaulted chamber tomb. His find swiftly attracted the attention of the interest of Service des Antiquités of the French mandate in Syria and Lebanon, under the direction of Charles Virolleaud.3 Excavations began the following year in 1929, led by Schaeffer, both at the necropolis at Minet el-Beida and the tell of Ras Shamra, around a kilometre inland. It was here that Mohamed Moursal made his important discovery, though of course it's Schaeffer and Virolleaud's names that would be remembered by history.

The importance of Ras Shamra was well established by the early campaigns, and its copious documents allowed it to be swiftly and securely identified as ancient Ugarit, whose magnificence was alluded to in the Amarna Letters:

[The king of Tyre's] property is as great as the sea. I know it! Look, there is no mayor's palace like that of the palace in Tyre. It is like the palace in Ugarit. Exceedingly great is the wealth in it.4

Excavations continued until the outbreak of the Second World War. This work focused particularly on the area of the acropolis, including the important archives of the House of the High Priest. Full-scale excavations resumed in 1950, in the area of the Royal Palace. Schaeffer remained director until 1970, when he was replaced by Henri de Contenson. After him, work at the site continued under Jean Margueron, Marguerite Yon, Yves Calvert and Bassam Jamous. In 2005 the archaeological investigations were

<sup>2</sup> 'At 5 o'clock in the afternoon, when the setting sun began to transform the Alawite Mountains to the east of the tell into a fringe of gold, I observed one of my workers who had stopped his work to examine what at a distance had the appearance of a small brick. Mohamed Moursal, a Turk from Bordj Islam – a good worker, but preferring effort more often than the delicate task of extracting fragile objects – spat on his find and, with the palm of his right hand, rubbed on it to remove the layer of earth which masked the surface.' (Schaeffer 1956, 1, my translation).

<sup>3</sup> Albanèse 1929; Day 2002, 37.

<sup>4</sup> EA 89, 47ff. Translation adapted from Moran 1992.

formally shifted from a French operation to a joint Franco-Syrian undertaking. They are currently co-directed by Valérie Matoïan and Khozama Al-Bahloul. At the time of writing, the civil war in Syria has interrupted archaeological investigation at the site, but to date it appears that Ugarit has fortunately largely escaped the large-scale damage and looting that has devastated many of the country's other historic sites.5

#### *The site*

The tell of Ugarit covers about 28 ha and rises around 17–20 m above the surrounding terrain at its highest point, the acropolis where the city's principal temples stood. The elevation of the site is, however, extremely uneven, with a marked depression towards the south. Deep sondages at the site have indicated that it was likely occupied since the Neolithic, in the eighth millennium BC, but for the most part archaeological work has proceeded outwards rather than down, uncovering an extensive area of the Late Bronze Age city but providing us with relatively little diachronic information. There have been some features excavated that have been dated earlier, such as various tombs, the so-called Hurrian Temple or the North Palace, all of which have at various times been assigned to the Middle Bronze Age; however, these are either poorly published, as with the funerary evidence, or have been shown by more recent work to belong to the Late Bronze Age, as is the case with the 'North Palace'. Sondages and Middle Bronze Age finds point to the important temples of Baʿlu and Dagan having existed at this period, but the surviving remains are fragmentary and provide little to go on.6 It does appear that there were a number of major construction horizons within the Late Bronze Age; the Royal Palace, for instance, evidences several destructions and rebuildings, including one that excavators have been keen to link with the partial destruction by fire alluded to in Amarna Letter EA 151 (hence, mid-fourteenth century), and a second around a century later that is paralleled across much of the rest of the site and is generally seen as due to an earthquake. The latter phase of rebuilding and restructuring is particularly important for our purposes as it coincides with the adoption of the alphabetic cuneiform script.

Excavators at Ugarit have delineated an assortment of broad districts. It's not necessary to explore each of these in detail here,7 but a general sketch gives a helpful overview of the character of the site. The two main focuses of elite activity are the Royal Palace and the Acropolis. The former is a massive complex in the north-west of the city, covering around 10,000 square metres of palace and associated structures. In keeping with its political status, it appears to have been somewhat segregated from

<sup>5</sup> A report by the American Association for the Advancement of Science in December 2014, based on satellite photography, noted that the coastal area where Ugarit is situated has not been a site of major fighting and did not find visual evidence of major damage (https://www.aaas.org/page/ancienthistory-modern-destruction-assessing-status-syria-s-tentative-world-heritage-sites-7, accessed 16 March 2017).

<sup>6</sup> Callot 2011.

<sup>7</sup> A useful and accessible summary is available in English in Yon 2006.

the rest of the site, with relatively few, closely controlled, connections between them and its own monumental fortified gatehouse in the western rampart of the tell. The Acropolis, in the north-eastern corner of the site, is most famous for being the home of Ugarit'stwo most prominenttemples,to Baʿlu and Dagan, and forthe so-called residence of the High Priest between them, from which have been recovered a number of literary and religious texts, including the celebrated Baʿlu epic.

Ugarit was not, however, characterised by a high level of urban planning – it was densely occupied, with labyrinthine and narrow streets (sometimes as little as around 1 m wide). Beyond the royal district, there was not rigid zoning by function or status. Certainly, there seem to be more high-status residences close to the palace, but these jostle with smaller houses; there's general residential occupation on the Acropolis right up to the temples. Large residences belonging to senior officials pop up amid the smaller homes of ordinary Ugaritians. Shops, workshops and smaller temples are interspersed in and among the warren of domestic habitation. Buildings of different function and status are jumbled together in a chaotic hodgepodge of human life. It can be helpful for modern scholars to talk about the 'South Acropolis' or the 'City Centre', but these should not be taken to imply the existence of well-defined correlating districts in the ancient city. This appearance of disorganisation extends to the deposits of written material. Collections of inscribed materials have been found throughout the city, and these include a wide range of scripts, languages and genres in various relationships with the ruling authorities. The so-called House of ʾUrtenu, for example, is in the South-central area, some distance from the Palace and not far from the main north–south thoroughfare that ran through the heart of the general residential area. Nevertheless, ʾUrtenu seems to have been an extremely high-ranking official and his archive includes a wide array of diplomatic and other official texts, including royal correspondence.

As is well known, Ugarit was destroyed at the end of the Late Bronze Age, an act usually attributed to the so-called Sea Peoples. There are limited signs of subsequent occupation, including small-scale use of the site by non-sedentary populations during the Iron Age, and a certain amount of inhabitation in the Persian and Roman eras, but unlike many similar Levantine sites Ugarit was not rebuilt or reoccupied on a large scale (see Chapter 12).

#### *The Kingdom of Ugarit*

The territory ofUgaritisrelatively well-defined thanksto the surviving textual material (see Fig. 0.2 at the start of this volume).8 In the north it was bounded by the mountains thatstretched inland from Mt Ṣapanu (modern Jebel al-Aqra), where Baʿlu was believed to have his palace. The western boundary was, of course, the sea. In the east, the Jebel al-Ansariyeh mountains provide an obvious natural boundary for most of Ugarit's

<sup>8</sup> The literature on the geography and topography of the kingdom of Ugarit is extensive, but good starting points are the contributions to Yon *et al.* 1995; van Soldt 2005; or, more recently, Calvet 2012.

territory, although the question of the Nahr al-Kabir valley has been debated. This route into the Orontes valley was the only connection between Ugarit and the Syrian interior, and scholars have taken differing stances on the extent of Ugarit's control along it. The maximalist position was proposed by Michael Astour, who attempted to identify locations well east of the Orontes with the toponyms listed as being assigned to Ugarit from its neighbour Mukiš in the diplomatic texts RS 17.340 and RS 17.237.9 Astour's suggestion has not been generally accepted, and most scholars see Ugaritian control as ending on the banks of the Orontes at the furthest; probably further west, between the mountains.

The southern borders are also fuzzy. It's clear from diplomatic correspondence that a separate kingdom existed to the south, centred on the cities of Siyannu and Ušnatu. For a while this was a vassal of Ugarit, but at its own request it was separated off by the Hittite authorities during the reign of Niqmepaʿ (*i.e.*, the late fourteenth or early thirteenth century) and placed directly under the overlordship of Karkemiš. Exactly where the line was drawn is unclear, but administrative records seem to point to the port town of Gibala (Tell Tweini) being within the Kingdom of Ugarit, so it is likely this was one of its southernmost holdings.

The Kingdom has not been extensively investigated compared to the capital: there has been no systematic archaeological survey. Many modern place-names are evidently descended from towns and villages recorded in Ugarit's administrative texts, which points to a general continuity of occupation, but there has been little or no archaeological work in these smaller settlements. Aside from textual data, our knowledge of the Kingdom beyond the capital comes mainly from four larger centres that have received archaeologicalstudy. The port of Minet el-Beida (ancient Maʾḫadu) is of course one. Another is Ras Ibn Hani on the promontory a little way down the coast. The palatial complex that was found there is thought to have belonged to the royal family of Ugarit and produced a collection of tablets (see below), and publication is still ongoing.10 The third site that has been excavated is Ras al-Bassit, on the coast north of the capital.11 Finally, and most recently, the southern port of Gibala/Tell Tweini has been excavated. This project has been particularly interesting not just as another point of comparison with metropolitan Ugarit, but also because of the contrasting methodology of the archaeologists. Unlike the principally textual and architectural focuses of the Ugarit campaigns, the Tweini excavators have undertaken a number of scientific analyses that provide valuable data on features such as ancient climatic changes.12

<sup>9</sup> Astour 1969; 1981a; 1981b; 1995; Singer 1999, 635.

<sup>10</sup> Arnaud and Kennedy 1979; Bordreuil and Caquot 1979; 1980; Bounni *et al.* 1979; 1981; 1998; Lagarce *et al.* 1983; Bordreuil *et al.* 2019.

<sup>11</sup> Published in preliminary form only by Courbin (1986), whose mention of the material culture focuses almost exclusively on Mediterranean imports.

<sup>12</sup> Bretschneider *et al.* 2004; 2008; Bretschneider and Van Lerberghe 2008; Kaniewski *et al.* 2015; Bretschneider and Jans 2019a.

Beyond the kingdom of Ugarit proper, it's worth mentioning the site of Tell Sukas, just the other side of the southern border. Since it probably fell within the territory of Siyannu-Ušnatu, the site provides useful comparative data both during and after that kingdom's time as an Ugaritian vassal. Tell Siyannu itself has also been investigated, but has not yet produced Late Bronze Age levels, possibly because of levelling in the Iron Age.13

One important observation from this relatively limited archaeological investigation of Ugarit's territory is that these second-order centres do not seem to have been abandoned at the end of the Late Bronze Age like Ugarit itself was. Ras Ibn Hani, Ras el-Bassit and Tell Tweini all show continuity of occupation into the Iron Age, with rebuilding after the destructions of the end of the Bronze Age.14 This makes the abandonment of Ugarit all the more curious.

#### *Outlining the history of Ugarit*

One of the consequences of the excavation strategy adopted by successive teams at Ugarit – uncovering the Late Bronze Age phases extensively while only probing deeper in limited sondages – is that much discussion of the city and its culture tends to adopt a highly synchronic perspective.15 The numerous discussions of social organisation and political economy16 draw from fourteenth and thirteenth century material to form a rather static composite picture of how Ugaritian society functioned. Attempts to introduce questions of social change into this discussion have centred mainly on exploring the processes leading to the city's destruction at the end of the Bronze Age,17 rather than on longer-term and more incremental social transformations of the kind expected under the model of structuration outlined in the next chapter.

That isn't to say that discussion of Ugarit has been entirely lacking a chronological dimension. Far from it: there is a flourishing sub-discipline concerned with Ugarit's political history,18 drawn overwhelmingly from the textual records of the site and of others with which it corresponded. There is not the space to review this in detail here, but a brief outline will be helpful in establishing the political background for the discussions of social context to come.

Because of the lack of excavation, Ugarit's history before the Late Bronze Age is little known. Much of the discussion of its early history has been rather ethnonationalist in focus, concerned with whether its origins are 'Canaanite' or 'Amorite'.

<sup>13</sup> Riis *et al.* 1996; Bretschneider *et al.* 2004, 220; 2019.

<sup>14</sup> du Piêd 2006–2007.

<sup>15</sup> Some earlier material is summarised in, for example, chapter 1 of Schaeffer (1939b), but just as there has been little impetus to uncover material from earlier than the Late Bronze Age, the material that has already been excavated has been at the back of the queue for re-evaluation and full publication.

<sup>16</sup> *e.g.* Heltzer 1999; Schloen 2001; Monroe 2009.

<sup>17</sup> Liverani 1987.

<sup>18</sup> Singer (1999) is the most convenient starting-point, but see also Freu (2006) and countless shorter articles on particular aspects of political history.

This is ultimately futile given that we know virtually nothing about how the city's inhabitants thought about identity, their own or others', at this time. More recent work, such as Buck (2018, esp. 21) has sought to downplay the ethnic dimension of this debate and reframe it in terms of linguistic and material culture relationships.

As far as we can tell, Ugarit's political situation paralleled that of other important port cities on the Levantine coast. Lacking military strength or extensive natural resources beyond the timber forests of its mountainous hinterland, it relied on its commercial networks and deft political manoeuvring to negotiate its position within a Late Bronze Age international landscape dominated by the superpowers of Egypt, Ḫatti, Babylon and, for a while at least, Mitanni. Another important, but unresolved, debate is whether Ugarit fell within the Mitanni Empire in the first half of the Late Bronze Age. Most scholars don't think it was under direct Mitanni control; however, Ugarit's northern neighbours in Mukiš (Alalaḫ) were Mitanni vassals, and it has occasionally been suggested that the empire exerted more influence over Ugarit than has generally been assumed.19 Certainly there are clear cultural similarities between Ugarit and Mitanni vassals such as Alalaḫ, just as there were between Ugarit and its southern Levantine neighbours. We can reasonably assume close links, as is demonstrated by the fact that some of the earliest surviving Akkadian material relating to Ugarit is diplomatic correspondence with Alalaḫ (RS 4.449 and AT 4 from Alalaḫ itself). This needn't necessarily indicate that Ugarit was politically subject to Mitanni authority, however. This is an important issue, as it would likely have profound effects for how people in Ugarit might have viewed cuneiform writing culture, and for their relationship with Mitanni's enemies the Hittites; but at the moment, the evidence is simply not there to make a judgement one way or the other. For a more detailed discussion of the role of Mitanni and 'Hurrian' culture and identity at Ugarit, see Chapter 9.

Although it is occasionally mentioned in earlier texts from cities such as Ebla and Mari (the latter's king, Zimri-Lim, visited the city around 1765 BC),20 our first written material originating in Ugarit itself dates from the fourteenth century: first, at least two and probably more of the Amarna Letters, and then – slightly later – the earliest Akkadian texts found at Ras Shamra itself. These documents describe an important political transition for Ugarit around the mid-fourteenth century. The Amarna letters indicate a close relationship with Egypt – probably more in the vein of partnership and elite emulation than direct political control – but shortly afterwards this seems to have been curtailed by the expansion of the Hittite Empire into Syria under Šuppiluliuma I. A number of Akkadian documents at Ugarit record its incorporation

<sup>19</sup> Singer 1999, 619–620 and n. 51. Astour (1981a) judges that Ugarit was independent of Mitanni but had previously been under the control of the kingdom of Yamhad. Liverani (1979, 1297) considers Yamhad control possible but not demonstrable from the available evidence. Recently, Cancik-Kirschbaum *et al.* (2014) and, in the same volume, Otto (2014) have counted Ugarit as a Mitanni vassal, but this is merely taken for granted, with no arguments made or evidence cited to support it. Buck (2018, 127), citing Singer as her source, concludes that Ugarit was not under Mitanni's political control.

<sup>20</sup> The first possible reference to Ugarit is in a gazetteer from Ebla – ca. 2400 BC (Tell Mardikh 75.G.2231 col 1, l.5).

into the Hittite sphere, although not always entirely clearly. It seems to be the case that while some of its neighbours resisted, Ugarit's king, Niqmaddu, saw which way the wind was blowing and 'voluntarily' invited Hittite overlordship, for which he was rewarded with territory from those kingdoms that had had to be forcefully integrated. From around the mid-fourteenth century to its destruction in the early twelfth century, Ugarit wasformally a vassal of the Great King in Ḫattuša. Often, Hittite control seems to have been relatively hands-off, probably due to Ugarit's status as an important source of wealth from its mercantile enterprises. Mostly, political oversight was managed by the Hittite appanage kingdom of Karkemiš rather than the authorities in Ḫattuša directly, and Ugarit was at times permitted exemptions from obligations to military service.

So long as the annual tribute was paid and Ugarit's kings came to Ḫattuša periodically to reaffirm theirloyalty,theirrule was permitted to continue with relatively little interference.21 Although Ugarit was obliged to fight alongside the Hittites at Qadeš, there are signs that it continued to feel more cultural connection with Egypt than with Anatolia. Aegyptiaca continue to comprise an important element of Ugaritian elite display, in contrast with the relative scarcity of imported or emulated Anatolian material culture. Unlike many of its neighbours, there is no sign of Hittite or Luwian being used at Ugarit. The only texts in these languages and writing-systems originated elsewhere. In the later thirteenth century, during the détente following the peace of Qadeš, and as the Hittite Empire weakened, diplomatic correspondence between Ugarit and Egypt resumed – if it ever truly ceased.

Towards the end of Ugarit's existence, there's a sense in the documentation that it was increasingly testing the limits of its vassal status. Numerous letters from the Hittite court attest dissatisfaction with levels of tribute, laxness in royal visits to the Great King or other failures to comply with their obligations. This is a topic we will return to, since it is precisely within this climate of increasing Ugaritian assertiveness and self-possession that alphabetic cuneiform first appears. This brings us to the matter of the scripts in use at Ugarit.

#### **The principal scripts of Ugarit**

#### *Logosyllabic cuneiform*

We'll start with logosyllabic (or Akkadian) cuneiform, since that's the more widely known of Ugarit's two main scripts, as well as the older and the one attested first at the site. Cuneiform was, of course, not created originally for Akkadian, but for the unrelated Sumerian language. It began as a pictographic script during the late fourth millennium BC, and over time grew increasingly schematised to facilitate quick and efficient writing by pressing wedges into soft clay with a stylus. By the second

<sup>21</sup> Although Niqmaddu's successor, ʾArḫalba, has often been thought to have been forcibly replaced by his overlords for some sort of insubordination, due to the brevity of his reign.

millennium BC, some signs retained traces of their pictographic origins, such as (*qat* – hand) but most had become thoroughly abstracted.

The relationship between Sumerian and Akkadian is key to understanding how logosyllabic cuneiform works. Sumerian writing, even after the pictographic stage, was primarily logographic – each sign represented a single word. However, signs could have multiple meanings – such as where the same sign was used for homophones. When the system was adapted for the Semitic language of Akkadian, many of the original Sumerian logographic readings were retained and the original pronunciation was used as a syllabic value. So, for example, meant 'hand' in Sumerian so could be read as the Akkadian word *qātu* – hand. However, the Sumerian word for 'hand' was šu, so it could also be used syllabically with this value in Akkadian. Signs also continued to be able to stand for things that sounded the same as, or similar to, their primary value. Thus, could be read as a logogram for 'wood', which in Akkadian was *iṣu*, but it could also be read syllabically as the similar sounding *is*, *iṣ*, *iz*, *es*, *eṣ* and *ez*. In addition, signs could function as determinatives – unpronounced indicators of what class of object the following word belonged to; thus could indicate that the word that followed was something made of wood. It can often be unclear which way a sign is to be taken. To give a very simple example, the same signs could indicate *alu ú-ga-ri-tu* – 'the city of Ugarit' or alu*ú-ga-ri-tu* – 'Ugarit (which is a specified to be a city)'. The difference in meaning is subtle here, but in other situations much more significant alternatives are possible. To add additional complication, the same syllable might be rendered by a number of different signs, represented in transcription with diacritics or subscript numbers.

It goes without saying that this was an extremely complex system, with each sign having many possible readings, and each word or syllable able to be rendered by multiple possible signs. This resulted in a very large repertoire of signs, each with a wide range of potential meanings. Not all the hundreds of attested cuneiform signs were in use at the same time, and some were certainly more common than others, but it was nevertheless a complicated and difficult system that required a great deal of time to learn,22 as well as familiarity with the increasingly obscure dead language of Sumerian. To further complicate matters, the insular groups of highly-educated elite writers indulged extensively in complex wordplay, multilingual puns and even codes23 – not just within the content of a composition, but as a fundamental step to the correct decipherment of the signs. When working with Akkadian cuneiform it's hard to escape the sense that accessibility and readability were alien concepts within the writing culture that created it, or were even actively avoided in the interests of elitist obscurantism. On the other hand, texts such as the Old Assyrian letters between merchants and their families, found at Karum Kaneš, suggest that at least at certain times, cuneiform literacy wasn't entirely the province of the dedicated professional literati.24

<sup>22</sup> On literate education in Mesopotamia, see below, Chapter 5.

<sup>23</sup> Finkel 2010.

<sup>24</sup> Larsen 2015.

Akkadian cuneiform is inextricably associated with the clay tablet, on which the overwhelming majority of surviving texts are written. These varied greatly in size, though most fit within the palm of the hand. They are generally lentoid or pillowy in cross section. It's not uncommon for text to continue on to the edges. Unlike the modern practice of turning a page horizontally, cuneiform tablets were flipped vertically when the writer wished to continue on the reverse – the text often continues uninterrupted around the bottom edge and on to the other side. The clay is thought to have often been leather-hard when used, which helped avoid signs becoming distorted by shrinkage as the clay dried. Wedges were pressed in using a stylus: originally a reed but later also wood, bone, ivory or metal. There remains debate about the shape of the stylus-head, with some scholars favouring a triangular cross-section, while others believe they were square.25 This may in fact have varied from place to place and over time.

Despite the preponderance of the tablet, we shouldn't overlook the importance of other surfaces for cuneiform writing – something that will be discussed in more detail in Chapter 5. For example, there's a good deal of evidence, both textual and archaeological, for the use of wooden, wax-covered writing boards, including a well-preserved example from the Ulu Burun shipwreck (although the origin of this particular one is unknown, it does at least demonstrate that they were in use in the eastern Mediterranean/Near East in the fourteenth or thirteenth centuries BC).26 Although we rightly think of cuneiform as a script designed for pressing into soft surfaces, there are also some indications that it may also sometimes have been written in ink on perishable materials. For example, a Neo-Assyrian tablet fragment in the British Museum (Museum number K.11055) includes a colophon added in ink after the clay had dried, which may point to the existence of a cuneiform tradition using ink and perishable materials, but its nature and scope are unknown.

Logosyllabic cuneiform was used across a wide area over around two millennia, and while traditions were conservative, that still allows huge scope for geographical and chronological variation. And that's just in the script itself: when we add the use of language into the mix, there's even more diversity. Akkadian encompasses two main dialects – Assyrian and Babylonian, of which the latter was the main basis for the language as used at Ugarit. However, influences passed backwards and forwards between them, as well as elements from other languages such as Canaanite or Hurrian, especially in so-called 'peripheral' contexts.27 The clearest example of this is the Akkadian of Phoenicia and the southern Levant, which, as evidenced by the Amarna letters, is quite unlike that of Mesopotamia proper; indeed, while mostly Akkadian in vocabulary, its syntax and grammar is much closer to Canaanite, such that many

<sup>25</sup> Taylor 2011; Cammarosano 2014; Ernst-Pradal 2019, 23.

<sup>26</sup> Payton 1991; Symington 1991.

<sup>27</sup> Andrason and Vita 2016. The term 'Peripheral Akkadian' is often used as a catch-all for the various dialects spoken or written outside of Mesopotamia proper, but I avoid it here both for its Mesopotamia-centricness and because it risks obscuring rather than highlighting the linguistic diversity of the region we are interested in. See Chapter 5 for more discussion.

scholars doubt whether it should be considered Akkadian at all, rather than a new and distinct mixed language, or even pure Canaanite encoded Akkadographically.28

So when we talk of Akkadian or cuneiform culture being adopted by Ugarit, we should be clear that we're thinking in terms of the emergence of a hybrid set of practices, which, while on the face of it founded in extremely orthodox Mesopotamian traditions, are nevertheless distinct from them and specifically Ugaritian, even before alphabetic cuneiform arrived on the scene. In terms of dialect, John Huehnergard has described the idiosyncrasies of Akkadian at Ugarit.29 While it has much in common with that of nearby cities such as Alalaḫ or Qaṭna, it remains distinct from them, notably in the much lower degree of Hurrian influence. It stands distinct too from the 'Canaano-Akkadian' of Phoenicia and the south. Carole Roche (2010) has discussed the possibility that at least some of the Akkadian from Ugarit may actually have been read as Ugaritic, where it consists primarily of logograms and personal names, or else a combination of these with syllabically spelled words, which are commonly used as Akkadographic spellings of native words elsewhere in the Near East, such as in Ḫatti or at Elam (*e.g*., *ša*, *ina*, *ana*). We know too little about when Akkadian came to Ugarit, and consequently the social situation of the city at that time,30 to be able to explore in any detail the factors that contributed to the growth of Ugarit's specific local variety of the language; but it's clear that there is considerable scope for nuance in how we imagine the arrival of cuneiform culture in Ugarit, and that it is not simply a matter of 'Akkadianising' or 'Mesopotamianising'.

#### *Alphabetic cuneiform*

The second principal script used at Ugarit – and the one for which the site is best known – is alphabetic cuneiform. It's also often referred to as 'Ugaritic', but this risks conflating the script with the language. In this volume I use 'alphabetic cuneiform' for the script, 'Ugaritic' for the language, and 'Ugaritian' as the general adjective for people or things from, or relating to, the city or kingdom.

As the name suggests, alphabetic cuneiform combines inspiration from both the older logosyllabic cuneiform discussed above and the alphabetic writing systems gaining ground elsewhere in the Levant (see Chapters 3–4). As in logosyllabic cuneiform, signs are composed of wedges usually pressed into clay with a stylus and the writing direction is generally left-to-right, a point in common with logosyllabic cuneiform but in contrast to most other Levantine alphabetic practices (although the right-to-left direction we tend to associate with Semitic linear alphabets was not fully standardised until around the same time as alphabetic cuneiform, or perhaps slightly later). Many of the writing practices established for writing logosyllabic were retained for alphabetic cuneiform: clay tablets predominate and are essentially the same as their logosyllabic equivalents; the script seems to have been overwhelmingly used

<sup>28</sup> von Dassow 2004; Izre'el 2012.

<sup>29</sup> Huehnergard 1989. See also van Soldt 1991.

<sup>30</sup> For a fuller discussion of this, see Boyes 2019a.

by formally-educated literates, and the writing system seems to have been taught alongside Akkadian using methods borrowed from Mesopotamia.

There are, however, profound differences between alphabetic and logosyllabic cuneiform. The most obvious and important is the relationship between signs and sounds. The alphabetic script has a repertoire of thirty signs (plus a word-divider), most of which each correspond to a single consonant. The exceptions are three quasi-vocalic signs, a,, i and u. Properly speaking, these represent glottal stops, but the choice of sign is dependent on the vowel following (or less commonly preceding) that consonant, with the result that they can be seen almost as syllabic signs including a vowel: *ʾa*, *ʾi*, and *ʾu*. Rarely, they might even stand for a vowel where there is no glottal stop. This is highly unsual within Levantine alphabetic systems, which otherwise do not begin to note vowels in any form until considerably later. From the arrangement of the alphabetic cuneiform signs in abecedaries, it seems clear that a was originally a simple glottal stop or aleph, and that the other two were appended on to the end of the alphabet somewhat later – it's generally assumed as part of an expansion to better accommodate writing Hurrian words.

These pseudo-vowels aside, the alphabetic cuneiform consonantal system is extremely similar to that seen in linear alphabetic writing elsewhere in the Levant. In both phonemic repertoire and conventional letter order, it's almost identical, so there can be little doubt that the Ugaritian system was derived from, or modelled after,the earlierlinearscript. The primary difference between the linear and cuneiform alphabets is that the latter attests more signs than do examples of the former from the first millennium BC. The 27 original signs of alphabetic cuneiform represent almost the full complement of the reconstructed proto-Semitic phonemic system. There is some debate as to whether this meant Ugarit had retained an archaic phonemic repertoire and alphabet, or whether the sound mergers that resulted in the shorter Phoenician and Hebrew alphabets had also occurred in Ugaritic and these letters had been artificially 'restored' (see Chapter 4).

It should be noted that while the vast majority of alphabetic cuneiform inscriptions use a more or less standardised version of the script with a repertoire of 30 signs, there is a small but interesting sub-set written in various variant forms of the script. These might include different forms for some signs, right-to-left writing direction and a tendency to be written on objects other than clay tablets. Several of these inscriptions seem to utilise a shorter alphabetic repertoire broadly comparable with that of Phoenician or Hebrew. However, not all these non-standard inscriptions attest all features, and while they are sometimes lumped together as using 'the short alphabet', it's doubtful whether this was really a single, coherent thing. Strikingly, the majority of inscribed objects using non-standard varieties of alphabetic cuneiform have been found outside Ugarit. They come from Phoenicia31 and Israel, Cyprus and even one example from as far afield as Tiryns on the Greek mainland. One inscription

<sup>31</sup> 'Phoenicia' is used in this book purely as a geographical label referring to the coastal polities between Arwad and Tell Dor – see Boyes 2013. It should be recognised that it is a modern exonym and has no ancient ethnic or political significance for the inhabitants of the region. For a recent discussion of this, see Quinn 2018.

has been shown to have been written in Phoenician rather than Ugaritic. We will discuss the probable significance of these in more detail later, but for now it's sufficient to say that these seem very likely to have been created outside the formal, state-aligned literate bureaucracy of Ugarit.32

*Table 1.1. The repertoires of standard official alphabetic cuneiform, non-standard variants and the linear alphabetic compared.33*


<sup>32</sup> See Boyes 2019b.

<sup>33</sup> The non-standard alphabetic cuneiform here is based on the presentation in Dietrich and Loretz 1988. It should be noted that this only includes 21 signs.

There has been some debate about exactly when and where the alphabetic cuneiform script emerged. These issues are discussed in Chapter 4. For now it's enough to say that it is best attested in the second half of the thirteenth century at Ugarit, but that we can't conclusively rule out the possibility that it may have been developed somewhere else, probably slightly earlier. Alphabetic cuneiform is inextricably associated with the Ugaritic language, the local west Semitic vernacular that was closely related to the Canaanite dialects to the south, though nevertheless distinct from them. It was, however, also used for other languages too, most notably Hurrian but also occasionally Akkadian. Biscriptal and bilingual texts also exist – usually Ugaritic and alphabetic main texts with Akkadian summaries.

Alphabetic cuneiform was used at Ugarit alongside logosyllabic Akkadian, and there was broad (but by no means entirely rigid) separation between what they covered. Whereas Akkadian was largely used for diplomacy, international correspondence and much legal documentation, alphabetic cuneiform and Ugaritic were used for internal letters, literary and mythical texts, religion and administration.

#### *Other scripts at Ugarit*

As well as the two main scripts we've discussed, there also exist – in much smaller numbers – examples of several other kinds of writing at Ugarit: Egyptian hieroglyphs; Luwian hieroglyphs and Cypro-Minoan. The Hittite implementation of cuneiform also occurs at Ugarit. Of these, only the Egyptian hieroglyphic and Cypro-Minoan inscriptions are at all likely to have been produced locally. The Anatolian inscriptions are all on letters sent from elsewhere in the Hittite sphere of influence or imported material culture such as seals. For the Egyptian material, much can be considered to be imported, but not everything. With several pieces, and to differing degrees, local production is possible either by resident Egyptian-speakers skilled in hierolyphic writing or by craftsmen specially sent from Egypt for the task. Cypro-Minoan, since it is largely undeciphered, is considerably more enigmatic and the nature of the tablets found at Ugarit extremely uncertain. Nevertheless, it's widely believed that at least some of these are likely to be locally made, since they show rather more similarity to Near Eastern scribal practices than do examples from Cyprus itself. These are matters we'll return to in Chapter 9, where we'll discuss the questions of who wrote these 'minority scripts', where, what social significance they held in Ugarit, and what they tell us about Ugaritian interactions with, and attitudes towards, users of these writing systems and their associated languages.

#### **Research and publication**

Publication of research at Ugarit has been ongoing since the 1920s. Preliminary reports on both the archaeology and the texts appeared primarily in the journal *Syria*, with more substantial publications occurring in the book series *Ugaritica, Palais Royal d'Ugarit* (*PRU*) and *Ras Shamra-Ougarit* (*RSO*). The journal *Ugarit-Forschungen* has been published annually since 1969. Syntheses of these decades' worth of scattered publications have been produced by, among others, Wilfred Watson and Nicolas Wyatt (1999), Marguerite Yon (2006) and Gabriel Saadé (2011), although all of these are now out of date in places, to greater or lesser degrees.

As regards the physical remains, the principal focus of much of this publication has been architectural. The early publications of Schaeffer's campaigns confine themselves to often rather vague descriptions of structures like the palace, and when they do discuss the objects, they focus almost exclusively on prestigious elite art and foreign imports, rather than anything more quotidian and representative. Even recent publications have been more concerned with delineating the built spaces of Ugarit than detailing the objects they contained or attempting to answer social questions about the lives of the people who inhabited them. Much of this work has been very good for what it is – such as the important work carried out in the residential areas of the city34 – but it remains less useful than it could be since the vast majority of material culture from the site is still unpublished. There have been countless preliminary publications, and certain classes of object have seen fuller treatments,35 but by and large no systematic or comprehensive data has been published on non-epigraphic artefacts, even from recently excavated areas.

From its outset, the archaeological investigation of Ugarit has been overshadowed by the extensive and flourishing epigraphic enterprise. From the limited repertoire of signs attested in the unknown cuneiform on the tablets found in 1929, Virolleaud correctly concluded that they must represent an alphabetic script, in contrast to the syllabic and logographic system of Mesopotamia. Decipherment was accomplished swiftly, with credit generally shared between Hans Bauer, Paul Dhorme and Virolleaud himself, although it's been argued that Virolleaud's contribution has been exaggerated, not least by Virolleaud himself.36 By the beginning of 1932, decipherment of the script was essentially complete and it was clear that the majority of the alphabetic cuneiform texts were written in a local north-west Semitic language with considerable affinities to both Phoenician and Hebrew.

As it happened, the House of the High Priest, excavated early on, contained a number of literary and religious texts which cemented Ugarit's status as one of the most important sites in the Bronze Age Levant. These included legends of kings and heroes such as Kirta, Aqhat and Danel, which offered our first real glimpse of the mythology and poetry of the Levantine Bronze Age. Most fêted, though, were the texts relating to Baʿlu, the storm-god and evidently Ugarit's patron deity. These not only provided insights into Ugaritian religion, belief and culture but also displayed strong parallels with sections of the Old Testament that could not help but resonate at a time when Levantine archaeology and epigraphy was still overwhelmingly conducted from a religiously-motivated perspective.

<sup>34</sup> Yon 1987; Callot 1994.

<sup>35</sup> Gachet-Bizollon 2007.

<sup>36</sup> Day 2002.

Over the following years a great many more assemblages of tablets have come to light (see Chapter 6), producing texts not only in alphabetic but also logosyllabic cuneiform, principally in Ugaritic and Akkadian language (thousands of tablets, divided approximately evenly between them), but also in Hurrian, Sumerian and Hittite. As well as literary and religious texts, a plethora of other genres are covered, including letters, administrative texts, legal documents and, significantly, several related to scholarship and scribal education. These have afforded us a view of Ugarit's culture, economy and social structure unparalleled in the Bronze Age Levant.

The standard corpus for alphabetic cuneiform inscriptions is Dietrich, Loretz and Sanmartín (2013) *The Cuneiform Alphabetic Texts from Ugarit, Ras Ibn Hani and Other Places*, usually referred to as *KTU* after its original 1976 German edition *Die keilalphabetischen Texte aus Ugarit*, although the abbreviation CAT is also less commonly seen. KTU is not without its shortcomings, especially as regards the artefactual and contextual information relating to Ugaritic inscriptions; but nevertheless, it remains a convenient and comprehensive collection of the alphabetic cuneiform inscriptions.

Unfortunately, no such resource exists for inscriptions from Ugarit in Mesopotamian logosyllabic cuneiform or other writing systems. Publication of the texts is extremely dispersed across a number of journal articles, several volumes of *Palais Royal d'Ugarit*, <sup>37</sup> *Ugaritica*38 and, more recently, *Ras Shamra-Ougarit*. 39 Various works have published catalogues or lists of the Akkadian texts,40 but these do not seek to function as corpora proper and lack detailed information. They are also extremely out of date, lacking the extensive and important material published from the House of ʾUrtenu over the last couple of decades. Taken on its own terms, the publication of the Akkadian from Ugarit is generally regarded as very good for its time, especially Jean Nougayrol's work. Once again, however, it is almost exclusively focused on the texts themselves and contextual and material information is extremely lacking. The early publications do record *points topographiques* for much of the corpus, but there has never been a definitive description or map of the precise locations these relate to; much is now probably lost to the poor record-keeping of Schaeffer's campaigns. Nevertheless, scholars such as Wilfred van Soldt (1991) have made valuable efforts to consider the topographic distribution and archaeological contexts of the inscribed material. However, in the absence of any comprehensive publication of non-epigraphic material from these contexts, we're still a long way short of being able to undertake the kinds of detailed contextual analyses I advocate in the next chapter.

What's most frustrating is that the shortcomings in the availability of such contextual information continue to be a feature of even modern publications. The House of ʾUrtenu, mainly excavated in the late 1980s and early 1990s, ought to have been a perfect opportunity to provide the kind of rigorous, comprehensive publication

<sup>37</sup> Nougayrol 1955; 1956; 1970.

<sup>38</sup> Nougayrol *et al.* 1968.

<sup>39</sup> *e.g.* Lackenbacher 2002; Lackenbacher and Malbran-Labat 2016.

<sup>40</sup> Bordreuil and Pardee 1989; Huehnergard 1989; van Soldt 1991.

for an archival context which we missed out on for other important areas containing written materials. Certainly, detailed artefactual and contextual information seems to have been recorded – a preliminary publication by Pierre Lombard (1995) professes some very encouraging ambitions in this area, as well as providing detailed plans of the distribution of inscribed objects recovered during the early excavation seasons in the area. However, the major publication of the tablets' archaeological contexts promised by Lombard41 has never emerged and the recent publications of the texts themselves include only rudimentary contextual and material information. Non-epigraphic material has likewise never seen the light of day, and we are left with a situation very similar to that for the other parts of the site: an architectural description and plan, a thorough epigraphic publication, but only the most general information on how these two classes of evidence go together.

Due to the lack of a definitive and holistic edition of the inscribed material,scholarly convention is rather fragmented in how texts are referred to. For the alphabetic cuneiform inscriptions, it is most common to cite texts by their KTU number or by their excavation number (*e.g.* RS 15.134), or both. For Akkadian or material in other languages, usually the RS number is used, though it is also not unusual to cite material according to its original place of publication, as, for example, Huehnergard does in his 1989 book on the language and grammar of Ugarit's Akkadian. In the majority of cases, these original publications were made by Nougayrol in various editions of PRU, though earlier material tends to occur in journal articles.42 For the present volume, I'll refer to alphabetic cuneiform inscriptions with KTU numbers and everything else with RS numbers.

#### **A note on scribes**

Finally for this introduction, I want to discuss my reasons for departing from usual practice on an important point of terminology. The scribe is ubiquitous in discussions of ancient Near Eastern writing practices, but I avoid the term in this book. There are good reasons for this. 'Scribe' can be used in several ways in Near Eastern studies, which only imperfectly overlap – as a translation for any of a number of local words such as Ugaritic *spr*, Akkadian *ṭupšarru*, or the logograms DUB.SAR and A.BA;43 to refer to any writer; to refer to someone who has undergone formal literate training; a bureaucratic

<sup>41</sup> Lombard 1995, 230, n. 17.

<sup>42</sup> Nougayrol 1955; 1956; 1970; Nougayrol *et al.* 1968; Huehnergard 1989.

<sup>43</sup> Ugaritic *spr* appears mostly in colophons, where there is little sense that it is a job title of identity marker, as opposed to merely an indication that 'PN was the writer [of this]'. An exception is KTU 4.836, which is a biscriptal list of professions in alphabetic cuneiform and includes an entry for LÚ. MEŠ.DUB.SAR 4/*sprm[.]ʾarb[ʿ* – four 'scribes'. There is no indication of what their specialisation actually was, however. There are far too many examples of Akkadian *ṭupšarru*/DUB.SAR to list, but they occur mainly in colophons, where they could again indicate merely that the person concerned was the writer of that document. There are a few instances of specialised terms: in RS 16.142 there is *ṭupšarru emqu* – 'expert scribe/writer' – and in RS 16.185 *akil šangi* – 'overseer of the administators' (Nougayrol 1955, 236).

functionary or servant whose principal job is to write; a political or religious official with other duties who uses writing in carrying these out. Likewise, it's common to talk about 'scribal training' when what we really mean is literate education. As David Carr has pointed out, writing was a part of this education, but it was a means to an end, not the end in itself.44 To call this education 'scribal' is to misrepresent its breadth and purpose: literacy was no more the end goal of Bronze Age 'scribal training' than the ability to use a word processor or bibliographic software is the main purpose of a modern PhD. To call those who had completed such education 'scribes' is to blur together a wide range of professions and ranks, from the genuine professional writers whose purpose was to take notes or draft legal tablets, to diplomatic messengers-cum-ambassadors, priests and high priests, exorcists and diviners, senior politicians and administrators, professional philologists and other career scholars.45

It's important to ask whether all these Ugaritian writers believed they were all the same kind of thing. Was there a sense of common identity centred around literacy and the practising of writing for professional purposes? A great deal of research into the relationship between writing and identity in the ancient Near East has assumed this to be the case; in particular that there existed an elitist, restrictive *esprit de corps* among literates. As van der Toorn puts it in his discussion of Assyrian scribes, '[i]n today's world we would call them the "in crowd." Scribes in the first millennium were conscious of their membership in a social elite. They saw themselves as initiates, in that the lore of the texts was theirs alone.'46 This was bolstered, he argues, by elements such as an oath of secrecy.

We'll be considering questions of identity in more detail in later chapters, but it makes sense to address this one now. At first glance, it seems a bit of a stretch to suppose that high-ranking literate politicians or religious officials saw themselves as part of a single group with ordinary secretaries who took dictation, or lowly notaries-for-hire. Did their shared ability to write trump the otherwise quite varied range of roles, backgrounds and statuses they embodied? What evidence there is points to something more complex than a singular scribal identity. Perhaps, if identities were centred around writing, we should be thinking more in terms of multiple literate identities. Hawley, Roche-Hawley and Pardee have advanced the view that we might be able to discern a difference at Ugarit between 'traditional' users of Akkadian and the logosyllabic script, and more innovative users of alphabetic cuneiform.47 I will discuss this more in Chapter 8, but with a certain amount of caution: while the evidence they cite does point in the direction they suggest, that evidence is at present very limited. But if more comes to light and continues to support their hypothesis,

<sup>44</sup> Carr 2005.

<sup>45</sup> Arguments in this vein have been made for at least half a century. As well as Carr 2005, see Landsberger 1960 and Michalowski 1987, 51. Covering similar ground for Egypt is Pinarello 2018. On the various professions of Ugarit's 'scribes', see Mouton and Roche-Hawley 2015.

<sup>46</sup> van der Toorn 2007, 65.

<sup>47</sup> Hawley *et al.* 2015.

then this would suggest not a single 'scribal identity' at Ugarit but at least two. If that were the case, who's to say there were not more?

On the other hand, there's little sign of any 'scribal' identity at all being articulated through material culture in charged contexts such as burial. Admittedly, we have very little funerary evidence from Ugarit because of the disturbance of many of the tombs and the poor publication of their excavations. However, other Near Eastern sites have not produced examples of 'scribal' identities being expressed through mortuary assemblages and practices. The nearest we come is in Egypt. There, we do find people buried with writing kit, as well as autobiographies or memorials in which they are described using the words generally translated as 'scribe'. This also occurs on memorial stelae from non-funerary contexts such as that of Mamy from the Temple of Baʿlu (see Chapter 7). However, Pinarello has disputed whetherthis material culture really reflects a single scribal identity. It's only the presence of writing equipment, he argues, which unites otherwise extremely disparate collections of tombs and assemblages belonging to people with quite different roles. In his view, 'scribal identity' is not archaeologically detectable in Egypt; what we have instead are a wide range of classes of people who expressed themselves in various ways using funerary objects, all of whom used writing, but in different ways.<sup>48</sup>

If we cast our nets much further afield, the situation in ancient China can shed interesting light on both the Mesopotamian and Egyptian examples, and further highlights the lack of a coherent scribal identity. During the third century BC, writing practices were fairly widespread in China, and highly institutionalised. 'Scribal schools' trained apprentices for lives in the imperial bureaucracy in a similar way to what is generally assumed in Mesopotamia. As in the Near East, the title of 'scribe' masks an extremely diverse assortment of roles and specialisations, from simple notaries and secretaries to diviners, occult specialists, legal experts, medical practitioners, detectives and more. In fact, in this period, the simple term 'scribe' (史 – *shi*) was not used on its own; only in compounds specifying the specialisation.49 As in Egypt, writing equipment is found in burials, alongside collections of documents and autobiographies detailing the deceased's career. However, in China, literacy seems to be a much more salient dimension of identity as expressed in the funerary context:

Manuscripts, brushes, and ink stones were visible parts of the mortuary rites. Indeed, they were the integral aspect of self-representation. In the tombs of scribes (nos. 7, 9, 10, 13, 15, 22, 26, and 33 in the appendix), the rest of the tomb assemblages were quite generic: mostly lacquer and pottery containers, some furniture, and occasional weapons, in addition to zoomorphic and anthropomorphic figurines as well as some miniature models. Following the argument that actors choose from among a pool of various role identities the ones most beneficial to them, I suggest that the tomb occupants were deliberately presented to the funerary audience as literate beings. The hope was that everyone should see and most probably admire the fact the departed served as scribes.50

<sup>48</sup> Pinarello 2018.

<sup>49</sup> Selbitschka 2018, and esp. 465 for the parallels between Chinese and Near Eastern scribes.

<sup>50</sup> Selbitschka 2018, 464.

The fact that non-writing-related material is quite generic is the key point here, and suggests that literate identity as expressed through material culture in Chinese funerary contexts was rather more coherent and unified than in the Near East, despite otherwise apparently quite comparable writing cultures. This is not to say that it's wrong to think knowledge of writing and the lore it transmitted was jealously guarded in Mesopotamia and other parts of the Near East; but I think we must be cautious about assuming this translated into a single 'scribal identity'. It seems more likely that any *esprit-de-corps* among literates was expressed in smaller groups and cliques of people operating in similar specialisations and at similar levels of status. Whether literacy was then the important aspect of these identities, as opposed to, say, being a diviner, or a merchant, or whatever, is an open question. As we'll discuss in Chapter 8, there is also limited but noteworthy evidence of literacy outside the formal administration and even at sub-elite levels in the Near East at various times and in various places.

For these reasons, I prefer other terms over 'scribe'. The most general is 'writer', which implies nothing about a person save that they wrote, either in general or a specific text in question. In place of 'scribal training' I prefer 'literate education' and for those who have completed it, 'literate intellectuals' or else specific job titles. It will also follow from this that the existence of a 'scribal culture' or 'scribal tradition' is something of an oversimplification.

Having set the scene for our discussion, we must now consider its theoretical and methodological underpinnings. This is the focus of the next chapter.

### Chapter 2

### The social archaeology of writing

Recent decades, it sometimes seems, have offered up archaeologies of pretty much everything; a veritable *smörgåsbord* of theory and method for approaching any topic we might care to imagine. From childhood to death, encompassing gender, politics, status, religion, food and so much more in between. But, surprisingly, there doesn't appear to be a well-defined Archaeology of Writing. There have of course been a number of specific studies applying archaeological data or perspectives to writing, and countless more bringing insights from inscriptions or documents to bear on the material record in a given context, but there are, to the best of my knowledge, no dedicated, full-length general treatments of the theoretical and methodological issues inherent in approaching writing through the material record or integrating epigraphic evidence into wider sociocultural contexts derived primarily through the archaeological recovery and analysis of material culture.1 With a few notable exceptions, such as Joshua Engledhardt's edited volume on agency in ancient writing,2 this area of life and practice, often seen as so fundamental to the human condition, has been largely overlooked by theoretically-inclined archaeologists and left to the epigraphers, philologists and literary historians to define and consider. This omission is especially ironic given the important role of metaphors of language and literacy in shaping archaeological theory. The pervasive influence of Saussurean structuralism in processualist archaeology during the mid-twentieth century has subsequently been rejected, but how often do we continue to discuss 'reading the past', 'reading archaeological landscapes', 'writing the body' and suchlike?3

<sup>1</sup> A number of publications that profess to deal with the archaeology of writing in fact tend to focus more on one than the other (*e.g.* Houston 2004; Nylan 2005; Wells 2015), deal with the topics separately rather than as a single integrated study (*e.g.* Mizzi *et al.* 2017) or else focus on case studies while lacking more general theoretical or methodological discussions (Rutz and Kersel 2014). There is also often a measure of conceptual slippage in discussions between the archaeology of writing practices and the use of written sources in archaeology. While these are related, they are not the same thing.

<sup>2</sup> Contributions to Englehardt 2013, especially Englehardt and Nakassis 2013; Nakassis 2013.

<sup>3</sup> This mode of metaphor was especially common in the first couple of decades of postprocessual archaeology, *e.g.* Hodder and Hutson 2003; Yamin and Metheny 1996; Meskell 2000; Hutson 2003; and see

#### **Materiality**

This is not to say that there has been no consideration of the physical and material aspects of written evidence from the ancient world. On the contrary, 'materiality' has for some years now been an important consideration for many epigraphers and manuscript scholars.4 It is widely accepted that any good analysis of an ancient inscription or class of documents will consider not just the writing in isolation, but how it exists in relation to the physical characteristics of the object upon which it is written and the materials from which that is made. In contrast to the text-focused presentations of times gone by, it is now expected that epigraphic illustrations will attempt to render the whole object as well as merely the written signs. Together with the explosion in the availability of decent colour photography provided by digital cameras and the internet, it is now considerably less easy to fall into the practice of studying an inscription without at least a vague visual and material understanding of what it is inscribed on, even if one does not have the opportunity to handle the original first-hand. Further innovations, such as new digital imaging techniques, 3D-scanning and photogrammetry allowing for highly detailed and accurate interactive 3D models,5 and probably soon the widespread availability of affordable, high-resolution and straightforward 3D-printing, continue to expand the possibilities.

This work on materiality is invaluable and has provided many insights into the production and use of written materials in the ancient world. At a basic level, it makes a great deal of difference to our understanding of an inscription whether it is written on, say, a transport amphora or an object purpose-made for being inscribed, such as a clay tablet. This might seem obvious, but in the past there has often been a tendency to see writing as a thing in itself, a *text*, almost free-floating and whose physical

also Preucel 2006, 138–142. Hodder (2004, 31) offers an explicit discussion of why he believed reading should remain an important metaphor in archaeological analysis after the end of the 'linguistic turn':

In my view, taking these various criticisms into account, it remains important to retain 'reading' and interpretation as components of archaeological procedure. This is because we do not only read texts. As social actors we are involved in daily acts of making sense of, 'reading' what is going on around us. This wider sense of reading refers to the larger process of interpretation – including making sense of textures, sounds, smells, power dynamics, and so on. Reading is a wider process than interpreting words on a page. It involves being thoroughly engaged in a social context and interpreting that context through a variety of senses.

<sup>5</sup> Such as, for example, the Digital Nestor project led by Dimitri Nakassis and Kevin Pluta, aimed at digitising Linear B tablets from the Palace of Nestor at Pylos, Greece. The advantages of detailed imaging and 3D recreation of images has also been demonstrated by the work of Kathryn Piquette and Martina Polig, who joined the CREWS Project as visiting fellows while this research was ongoing.

<sup>4</sup> The literature on this is vast and growing, but see, for example, Eidem 2002; Pearce 2010; Taylor 2011; Ferrara 2012; contributions to Piquette and Whitehouse 2013; Ellison 2015; Balke and Tsouparopoulou 2016, all with extensive further bibliography.

manifestation is at best secondary.6 The benefits of materiality research frequently go far beyond this, however. To give a more technical example, scientific analysis of the materials from which an object is made, such as clay, can tell us much, such as where it was produced – particularly important in the case of items that travelled, such as letters or inscribed storage vessels – or the expertise required and strategies undertaken by producers to produce objects which suited their purposes.7

And yet the majority of these analyses can remain rather limited in scope. The amount of material information given in a primary publication is still quite variable, dependent on the interests of the individual editor or editors and the disciplinary traditions within the region being studied.8 Near Eastern epigraphy and philology – much like its archaeology – have tended to be highly conservative, in part because of the strong and enduring influence of biblically-motivated approaches. As a result, alphabetic cuneiform has fared rather poorly in the availability of material data: the standard corpus, KTU, is commendable in lots of ways but lacks illustrations of any kind. While it includes dimensions for published items, it does not systematically record other material information. If the kind of object that bears the inscription is recorded at all, it is usually only mentioned in passing under the heading for 'genre'; more often, we are simply left to assume that it is a tablet, as if these are all of a kind. A considerably better example is set by Bordreuil and Pardee's 2009 *Manual of Ugaritic*, which includes well-done drawings and photographs, albeit for a much smaller selection of tablets; information on fabric, context and so on remains absent, however.

But even when epigraphers are consciously 'doing materiality', the information and analysis offered can frequently be dedicated to a rather narrow technical description of the features and manufacture of the object bearing the text, or even just the individual signs. It is still the text that makes an object worthy of study and which is the scholar's primary interest. Even in literature explicitly on materiality we see inscribed objects referred to as 'material supports' or 'text vehicles',9 as if it is no more than a secondary prop added by necessity to substantiate the otherwise inconveniently intangible text. But in an archaeological approach, such material description is only the first step; of more interest is using this as a basis for answering wider contextual and social questions. Archaeology views the usefulness of objects in isolation as limited, however well understood their physical characteristics or methods of production are. Archaeological interpretation is instead fundamentally relational; it consists in the contexts of artefacts and the links between them and

<sup>6</sup> Tsouparopoulou 2016.

<sup>7</sup> Taylor 2011.

<sup>8</sup> Epigraphy, palaeography and philology are no more regionally homogeneous than is archaeology, and distinct differences in approach exist between different regional traditions. For example, the conservatism of the Near Eastern approach can be contrasted with the rather more receptive attitude towards theory and the integration of archaeological data and methodologies evident in Aegean-based research.

<sup>9</sup> *e.g.* Taylor 2011, 23.

the people who used them. It is not enough simply to record the fabric, dimensions, manufacturing processes or find-spots of inscribed objects; we must also consider their interrelationships with other objects found alongside them, with the stratigraphy and environment of their location, their symbolic as well as purely functional uses. In short, we must consider inscriptions and the objects they are found upon together, as *inscribed objects*, as part of the wider material and social culture of their context, not as a separate class of artefact subject to different methods of study, undertaken by independent specialists. For an example of the importance of this we can consider contemporary ethnographic work: among the modern Sora people of India, a cult exists in which their alphabet is worshipped as a scriptal embodiment of the god Jagannath. Objects bearing this script include stones inscribed with graphemes, which are believed to grow from the ground in shrines where milk dripped from a sacred tree. Any study that explored these items merely in terms of the size and material of the stones and the techniques used to inscribe them would wholly miss the significance of these items within the Sora culture.<sup>10</sup>

It's not my intention here to denigrate epigraphers or their work, or to exalt archaeologists as flawless exemplars of the way forward. In their willingness to wash their hands of objects and leave the business of studying them to others, archaeologists are certainly not off the hook. Rather, my aim is to add my voice to those highlighting the problem of this disciplinary divide and to help clarify the steps we must take to bridge it. What is needed is for archaeologists, epigraphers and philologists to work closely, hand-in-hand, with a much better understanding of each other's methodologies. Such an approach is fundamental to the CREWS Project and formed the basis for an ERC-funded interdisciplinary conference I organised under the Project's aegis in 2019.11 It's at the heart of what I hope to achieve in this book. It will, I hope, allow me to place writing at Ugarit within its full sociocultural context and explore how it intersects with human practice, beliefs and ideologies well beyond the world of the texts themselves.

The first step in this endeavour has to be to pull together, as best we can, a proper theoretical and methodological framework for a Social Archaeology of Writing.

#### **Key concepts: practice, society, agency and networks**

If writing is social practice, then any archaeology of writing must begin with an understanding of the relationship between practice and society. Within the social sciences, a great deal of research over the last few decades has taken its starting point on this subject from the theories of human practice and its relationship with society developed by scholars such as Pierre Bourdieu and Anthony Giddens.12 Both emphasised that human society is not a fixed system of mechanistically interacting social subsystems,

<sup>10</sup> Guillaume-Pey 2018; forthcoming.

<sup>11</sup> Boyes *et al.* forthcoming.

<sup>12</sup> Bourdieu 1977; 1990; Giddens 1979; 1984.

as earlier, functionalist, approaches had often portrayed it, but a dynamic structure that both influences and is affected by the choices made by people and groups acting within it. Humans don't have absolute freedom of action but are constrained and enabled by the learned ideas, norms, politics and so on of the society within which they are socialised – what Bourdieu calls the *habitus* and John Robb has glossed as 'an ingrained system of dispositions which provides the basis for regulated improvisation'.13 This habitus is in turn not an independently existing thing in itself, but is constantly reproduced and altered according to the actions and choices of the people. This dialectic – called by Giddens structuration – is a powerful idea which is linked to a number of other concepts and ideas that will prove vital to the discussions to come.

The first is the idea of agency. This rose to prominence within archaeological writing around the turn of the millennium and since has thoroughly permeated the literature, in every aspect of our understanding of practice and the construction of meaning.14 At its heart, agency is simply the recognition that humans make decisions that shape history, society and culture: not just the well-documented decisions of the 'great men of history', but the ubiquitous and quotidian choices, ideas and practice of every member of a society. This doesn't necessarily imply that all such decisions are assumed to be made consciously. On the contrary, practice may be subconscious and informed by norms and ideas that people have internalised, rather than conscious reflection. Practice and agency are also understood not just to be abstracted products of the human mind, but are embodied: mediated through and shaped by the senses, physicality and autonomous responses of the human body. We have our materiality just as everything else does. Agency is thus fundamental to structuration models of society and practice – social structures are formed from the accumulated results of the exercising of agency, and it is in relation to those structures that agency is exercised.

The ubiquity of agency within theoretically informed archaeological research has not been matched in the study of ancient writing. Joshua Englehardt's 2013 volume on the subject is an important contribution, particularly for its introduction – the only such work I know of to discuss the theory and method of applying this work to ancient writing. Within epigraphy, even basic, ready-salted-flavour agency with its message to remember the role of people, is unusual and necessary. One illustration of this is just how readily and near-universally epigraphic research succumbs to a habit Marcia-Anne Dobres warned about two decades ago: that of conceptualising technical practices as the work of disembodied and socially-isolated *hands*, robotically devoted to following techniques; rather than complete, socially-embedded, thinking *people*. Even in modern, materially-aware work, we're accustomed to reading of *scribal hands*, or

<sup>13</sup> Robb 2010.

<sup>14</sup> Literature on agency in archaeology is extensive, but see Dobres 2000; Robb 2004; 2010; 2015; Dobres and Robb 2005a; 2005b; Knappett and Malafouris 2008; Englehardt 2013, with copious further references. Within the anthropology of literacy, the idea of reading and writing as practices people engage in rather than merely skills people have or don't have has been prevalent since the 1980s – see Scribner and Cole 1981.

*Fig. 2.1. Disembodied hands writing alphabetic cuneiform. From Ellison (2015, 162). Reproduced with permission.*

just *hands*. 15 Statements such as 'One tablet from outside the Megaron, La 1393, is attributed to Hand 13'16 (picked essentially at random from the sphere of Mycenaean studies) are typical. Even when the producers of these artefacts are mentioned as entire people, they generally continue to be defined according to their relationship to the text – as *scribes* (see the discussion last chapter).

This might seem like nitpicking – after all, close-up images of the hands such as Figure 2.1 illustrate much more clearly the techniques

involved in making cuneiform signs than would a full-body image of a person at ease in a palace courtyard, chatting with their pals while carelessly pressing the latest economic records into the clay – but the cumulative effect of such imagery and language is to alienate writing and inscribed objects from other aspects of human life and from the archaeological and historical disciplines that tell us about them. The reconstruction I imagine above, or Figure 2.2, taken at a workshop I ran on Ugaritic at the Faculty of Classics, Cambridge, in 2017, tell us something about the social context of writing that an exclusive focus on hands and technique cannot. They raise questions of things like age, gender, physical abilities, social status, use of space and interactions between people and objects. I'm not suggesting that we replace a detailed and rigorous understanding of technique with airy imaginings or modern candid snaps, but that we must supplement and strengthen the former by remembering the importance of *people* and their contexts.

Conceptually, this is all very nice, but how should we go about applying agency theory to ancient writing? It's not without some justification that 'agency' has acquired a reputation for being a bit of a handy buzzword that can be slapped on to anything, without necessarily much substance behind it. The challenge is how we can hope to recover such a pervasive and yet nebulous and intangible concept from the material record. As Dobres and Robb put it,

if we consider agency to be a fundamental quality of human existence, then 'The Archaeology of Agency' begins to look somewhat like 'The Archaeology of Breathing' – a dynamic so universal and inescapable that, without further linkage to something more specific, it is difficult to see how it can illuminate particular aspects of the past.<sup>17</sup>

<sup>15</sup> *e.g.* Driessen 2000; Palaima 2011, or recently, and concerning Ugarit directly, Ernst-Pradal 2019, esp. chapter 2.

<sup>16</sup> Skelton 2010, 108.

<sup>17</sup> Dobres and Robb 2005b, 160.

*Fig. 2.2. A postdoctoral linguist and epigrapher practises an unfamiliar script in a seminar setting. Photo by the author.*

The answer lies in another point that follows from Bourdieu and Giddens' ideas: the close relationship between agency and socialstructure. Thisis notreified and isolated – a thing which can be found – but exists in the relationships between people, objects and parts of society. This means that it is fundamentally both network-based and contextual, historically and culturally situated within a particular social environment.18 And so in every stratigraphic feature, artefact or inscription there is the potential to illuminate these connections, to ask how the decisions and practice of actual people in relation to their habitus shaped the artefacts we see or the practices we reconstruct. We might not be able to see it directly, but it can be inferred – especially in creative fields such as writing.

Dobres proposes a detailed and specific methodology for going about this, the *chaîne opératoire*. 19 Traditionally, this long-established analytical approach has usually

<sup>18</sup> Robb 2010, 499.

<sup>19</sup> Dobres 2000. Dobres is writing about the archaeology of technology, but much of what she says translates very effectively to an archaeology of writing. Indeed, writing has frequently been treated as a technology by historians of the subject. For a sense of the poles between which research on

focused on the narrowly technical, charting in detail the specific steps and gestures involved in producing an object, and the 'underlying logic and syntax' that inform them.20 Done correctly, Dobres argues, this is not merely descriptivism, but provides detailed data from which to infer social structures and explore variation. Did variability occur more with certain types of material or classes of object? At certain places or stages of the production process? Was it open, tacit or covert? How much variation was tolerated or encouraged? How widely were techniques shared or restricted? At what scales were decisions about variation made – individuals, a few people or at groups-level? By asking these questions, and others, Dobres suggests that we can begin to delineate the contours of agency within the production process.21

The difficulty, from the perspective of the current study, is the nature of the evidence. Dobres' approach was designed for her research into bone and antler products in the palaeolithic Pyrénées, for which she was able to control what information was recorded about finds as they were recovered, ensuring that suitable and detailed data existed across her sample. In all, she recorded ninety specific attributes per specimen.22 It's less clear how one might apply it to a very large existing corpus such as written material from a site such as Ugarit (around five thousand inscribed objects, all told), which has been subject to extremely poor geospatial recording of find locations (on which see Chapter 6 below), inconsistent publication and is not all accessible for first-hand study.23 Such detailed, specific data is simply not available.

#### *Networks and connectivity*

Other recent work on agency and archaeology is more amenable to the situation at Ugarit, and brings us to our final key concept – the importance of a network-based model of society, culture and agency. In the last decade or so, research into the idea of relationality and context has given rise to increasingly network- or mesh-based models of the relationships between people, material culture, social structure and

<sup>22</sup> Dobres 2000, 197.

this subject tends to be situated, compare the descriptive and matter-of-fact entry in Singer *et al.*'s multi-volume *History of Technology* (Hooke 1954) with Houston's considerably more modern and theoretically-informed discussion half a century later (Houston 2004). For Ong (1986), writing was a 'technology that restructures thought'.

<sup>20</sup> Lemonnier 1976; Schlanger 1991; 1994; Leroi-Gourhan 2013 [1965]. A more sceptical reading of the *chaîne opératoire* is offered by Djindjian (2013), who laments the obscurantism of the term and the lack of adequate quantification and diagramming. Dobres, itshould be noted, while in favour of large-scale data underlying the analysis, is critical of overly schematic and functionalist diagrams (Dobres 2000, 174ff.).

<sup>21</sup> Dobres 2000, 179.

<sup>23</sup> At the time of writing, the political situation in Syria precludes access to the very many tablets and other material culture stored there.

agency.24 This doubles down on the idea that agency is context-dependent and conceptualises it in terms of 'heterogeneous relationships which span humans, collectivities, bodies of knowledge and material things.'25 One of the results of this is, as I hinted earlier, that theorists have become more and more reluctant to see agency as the exclusive preserve of sentient human actors, and instead now frequently talk about the agency of things:26

[W]hile agency and intentionality may not be properties of things, they are not properties of humans either: they are the properties of material engagement, that is, of the grey zone where brain, body and culture conflate.<sup>27</sup>

Lambros Malafouris discusses this in terms of the production of wheelmade pottery: the pot isn't wholly the result of the potter's action alone, but by the potter's interaction – collaboration – with the wheel and the clay. Depending on the quirks of the wheel and clay themselves, how they work together and how the potter uses them, the potter might have more or less control over the shape of the pot – either through choice or by accident.28 Noting Dobres' caveats about focusing unduly on the hands, we can broaden this out and think about the wider social and cultural relationships that shape the agency and affordances of the wheel and clay. Such entanglements can be with norms, laws and ideology as well as with people, collectivities and objects. Robb cites the example of someone killing another with a gunshot, where the very nature of the action itself can vary a great deal according to its context: a heroic act of war, a murder, an act of self-defence, an execution (legally-sanctioned or otherwise).29 Even the definition of what an act or practice *is* is thus socially and relationally constituted.

These kinds of network approaches are also extremely useful for thinking about the interconnections of polities, and even 'writing systems' themselves.30 We can see these things not as self-contained entities that are whole and identical across themselves, but as networks of practices, ideas and so on, which are themselves interlinked with other polities. Our view of Near Eastern regional society, then, becomes effectively one great mesh of interconnected parts, where everything is linked with everything else, however distantly, and thus everything is in some sense hybrid, partial or a local manifestation of a greater, diffusely-imagined concept.

This emphasis on networks and the fundamental interconnectedness of every aspect of human existence and practice has much in common with concepts of globalisation, and thus insights and approaches derived from globalisation studies can

<sup>24</sup> Latour 2005; Knappett and Malafouris 2008; Robb 2010; Hodder 2012.

<sup>25</sup> Robb 2010.

<sup>26</sup> Gell 1998; Latour 2005; Knappett and Malafouris 2008; Hodder 2012.

<sup>27</sup> Malafouris 2008, 22.

<sup>28</sup> Malafouris 2008.

<sup>29</sup> Robb 2010, 503.

<sup>30</sup> For a network approach to variation within alphabetic cuneiform and its connections with logosyllabic cuneiform, see Boyes 2019b.

prove useful here. The majority of this research, coming mainly from the disciplines of economics and international relations, has tended to approach the subject purely as a phenomenon of modern capitalism whose roots can hardly go much deeper than the emergence of nation-states in the Early Modern period.31 This is arguably due more to ignorance of and lack of interest in deeper historical applications for these methods than to their explicit rejection. Other examples acknowledge the potential utility of their perspectives for historical investigations without delving into the matter themselves.32 In fact, a growing number of studies by archaeologists, historians and classicists have productively applied many of the approaches of globalisation studies to the pre-modern and even ancient worlds. It has been argued that globalisation is not a singular condition of modernity, but that globalisations – in the plural – can be observed at several points throughout history, from Uruk-era Mesopotamia to the Hellenistic and Roman worlds or the Huari cultural sphere of south America.33 This plurality is important if we're to avoid telling teleological histories in which advancement to our current degree of global integration is just another version of the old Whiggish idea of unilinear Progress.34

No one would argue that these historical instances of globalisation – if that's what we choose to call them – are identical in kind to the one we're experiencing today. In innumerable ways they are not – the most obvious being that none of them are *literally* global, in the sense that the political, economic and cultural networks they involve enmeshed the entire world. What many of these scholars argue is that modern, capitalist globalisation is not necessarily the only possible type; its key features are not intrinsically confined to the modern, literally-global context. This obviously raises the question of what precisely those features are, which is something of a problem since there is no universally accepted definition. Nevertheless, many discussions do converge on a number of features which most agree are characteristic of globalisation or global culture. These are usefully listed by Justin Jennings (2011, chap. 7) as follows:


<sup>31</sup> *e.g.* Giddens 1990, *passim,* but esp. 63–78; Castells 2000; and see Jennings 2011, esp. chapter 1.

<sup>32</sup> *e.g.* Grewal 2008.

<sup>33</sup> These examples are mainly taken from Jennings 2011, as well as Pitts and Versluys 2014. On pre-modern globalisations, see further Hopkins 2002 and LaBianca and Scham 2006.

<sup>34</sup> Some contributions, particularly those by scholars working from a paradigm of social (neo-)evolutionism, do fall into this trap, such as, for example, Levy 2006.


Pitts and Versluys (2014, 17) cite and endorse Jennings' hallmarks and add a number of their own, such as increased connectivity, common markets, the idea of belonging to a single world, impacts from this integration on local markets, and cosmopolitanism.

It will, I hope, be evident (or else will become so in the course of the discussions in this book) that a number of these traits apply to the Late Bronze Age east Mediterranean and Near East. The establishment of roads, diplomatic protocols, treaties permitting safe passage and the widespread use of Akkadian and logosyllabic cuneiform compressed geography and allowed more rapid and regular communication. The Amarna letters and other Late Bronze Age diplomatic documents attest to a homogenised elite culture governed by standards of language and etiquette; further standardisation is apparent in weights and measures. Convergence is apparent in elite art-styles and taste, while pottery classes such as LH III C defy straightforward association with a single place.35 As we'll see later in this book (and as I have argued previously for Phoenicia),36 in several parts of the region – including Ugarit – there is a growth in the articulation of local identities and the stressing of local cultural distinctiveness towards the end of the Late Bronze Age. Finally, although there remains widespread disagreement over exactly what led to the major social upheavals that ended the Late Bronze Age, most explanations now involve a concatenation of multiple interlinked crises which attest to the vulnerability and mutual interdependence of the system.37

This interconnected Late Bronze Age world has been variously characterised as a 'world system',38 an 'International Style'39 or as simply 'connectivity' – the latter

<sup>35</sup> Feldman 2002b; 2006; 2015; Sherratt 2003.

<sup>36</sup> Boyes 2013.

<sup>37</sup> See, for example, Cline 2014 for a recent overview, and Chapters 10 and 12 for further discussion of the crisis as it pertains to Ugarit.

<sup>38</sup> Sherratt 1993; Parkinson and Galaty 2009. World systems approaches, originally formulated by Immanuel Wallerstein (1974; 1980; 1989) as a means of explaining the development of modern capitalism in the Early Modern and modern periods were subsequently taken up by others and extended, very much against Wallerstein's wishes (Wallerstein 1993), to historical contexts. With its focus on connectivity and economic integration, world systems approaches have been seen by many as precursors to current globalisation research (*e.g.* Giddens 1990, 67–69; Jennings 2011, 10–13; Pitts and Versluys 2014, 8–10), although they differ from it through their primary focus on economic exploitation of a periphery by a dominant core, in opposition to globalisation's more decentred perspective, and less strictly economic area of interest.

<sup>39</sup> Feldman 2002b; 2006.

in the case of Miguel Versluys, when arguing against seeing it as globalisation. His reasoning for this seems to be that it only involved the movement of goods, not people.40 I don't think this is true – it is very clear from administrative records at Ugarit and elsewhere that all kinds of people moved around (see Chapter 8 below), from merchants to mercenaries, settlers with families, as well as specialists sent between courts as part of international elite exchange. But this is, ultimately, beside the point. I'm ultimately not especially interested in taxonomy and whether we apply the label 'globalisation' to connectivity in the Late Bronze Age east Mediterranean. What I think is undeniable is that there is enough in common between what was going on in the region in our period of interest and the 'hallmarks of globalisation' listed above that aspects of globalisation theory can be aptly and productively applied.

There are two main strands of this research that will prove particularly useful in the discussions to come: the first is the concept of 'glocalisation'; the second is the discussion of cultural convergence and the social, economic and political dynamics underlying cultural homogenisation and the emergence of standards within globalised social, cultural and political networks.

The awkward term 'glocalisation' covers two pairs of seemingly contradictory features of globalisation. On the one hand, the combination of 'deterritorialisation' with the 're-embedding of the local'; on the other, the homogenisation *and* heterogenisation of culture. While at first glance paradoxical, the concept is less hard to grasp than these contradictions initially suggest: it simply highlights the strong interrelationship between the global and the local within globalised networks. Nothing is purely a feature of the network as a whole, but is realised and manifested locally in different ways and with different responses. A simple example offered by David Grewal (2008, 267) and others is that of perhaps one of the most emblematic icons of modern globalisation – the spread of McDonald's. As an overarching phenomenon, the chain is an obvious example of homogenisation – you can walk into a McDonald's anywhere in the world and see similar branding, architecture, décor, corporate culture and products. This is a key element of its appeal for those who frequent it. On the other hand, this homogeneity is still carefully tailored to local markets, tastes, regulations and behaviours. Different products may be available, or different ingredients used in those that are shared. Furthermore, regardless of what the company itself does or intends, its outlets might be used and appropriated by local populations in very different ways – Caldwell gives the example of Moscow, where McDonald's has becomes associated with local produce and is seen as a comfortable meeting place for friends and families to take their time over meals.41

As well as the appropriation and transformation of imported material culture and practices to suit local needs and agendas, processes long familiar to archaeologists and anthropologists, glocalisation also covers reactions in local politics,

<sup>40</sup> Versluys 2014, 144, 162.

<sup>41</sup> Caldwell 2008.

discourse or identities to global events or culture. Frequently, the growth of global interconnectedness and the real or perceived homogenisation of culture can result in a counter-impulse in which local distinctiveness is emphasised and politics of isolation, nationalism and discrimination can flourish. This is evident in the modern world in such innocuous everyday forms as supermarket food labels (or indeed, McDonald's marketing materials) proudly proclaiming products or their ingredients to be produced in the country where they're sold ('buy British!'). In its darker incarnation, it is also at work in the alarming rise of far-right populist nationalism in Europe, America and elsewhere, manifest in radical isolationist and xenophobic projects such as Brexit.

Neville Morley, despite being in general somewhat sceptical of the utility of globalisation as an approach to antiquity, has applied these ideas to the ancient context in discussing interactions within the Roman sphere of influence.42 He stresses the uneven distribution of power within political and cultural networks of this kind, as well as the potential absence of freedom to decide even when people ostensibly choose to adopt a given cultural practice or item of material culture.

In the case of most networks, including that of the Roman elite, the 'standards' for membership are never stable or clear-cut but constantly renegotiated, and acceptance into the network (at any rate for most people) is not necessarily a single one-off moment but a matter of having, time and again, to win recognition from fellow members as being 'one of us' by performing in a manner appropriate to that status. Further, power is rarely evenly distributed across the network: some individuals, especially those firmly entrenched at the centre of power through their birth or achievements, held far greater influence in determining the acceptance of others, and equally could afford to be significantly 'unRoman' in some of their practices, effectively rejecting some of network's standards, without losing their membership. Greek elites, because of the importance of the Hellenic tradition for the 'hybrid' Roman elite culture, might need to make fewer adjustments to their behaviour in order to win acceptance than would be expected of 'barbarians'; arriviste Gallic notables might need to be far more Roman than the Romans, whether they liked this or not, in order to gain admission. The idea of networks and standardsthus offers a way ofre-describing the development of a Mediterranean-wide elite culture, and its role in regulating social and political relationships, in a way that engages with its complexity and diversity.43

By understanding the power relationships within networks, which inform processes such as cultural homogenisation, hybridisation, resistance and negotiation, we can thus illuminate and better understand the interplay of impulses towards the global and the local, towards cultural homogeneity and distinctiveness, and changing forms of social identity. As we will see, this is extremely helpful in unpicking the social and cultural changes in a society like Ugarit whose political and cultural interactions were in many ways characterised by an ostensibly less powerful position, whether that be

<sup>42</sup> Morley 2014.

<sup>43</sup> Morley 2014, 62–63.

its politicalsubordination to the authority of Ḫattuša and Karkemiš orthe influence of the prestige cultures originating in Mesopotamia or Egypt (and, as later chapters will show, matters were rather more complex than that). Understanding these dynamics, especially as they involved Ugarit's elites, is critical to understanding changes in writing practices and writing culture, and especially the relationship between globalised practices such as 'cuneiform culture' and the emerging vernacular literacy in Ugaritic language and alphabetic cuneiform script (see Chapters 5 and 11).

#### **Outlining an archaeology of writing**

Over the past several pages we have surveyed, in necessarily compact form, several decades of thinking about the nature of human practice, agency, materiality and connectivity. My aim now is to lay out in specific terms what I believe an archaeology of writing should look like, and my methodology for the study that follows.

We can identify three stages in the archaeological interpretation of an inscribed object.44 These stages are not intended to be wholly discrete and sequential – no analysis proceeds in such linear terms, and even if one did, the order of these steps would depend on the nature of the specific evidence being analysed; for example, whether it's an item newly discovered in situ as part of an ongoing excavation, or a long-known part of an existing corpus. Nevertheless, it is still useful to enumerate them separately.

#### *Defining the object*

Our firststep isthe essentially object-focused and technical work that formsthe mainstay of much traditional epigraphic practice. The item must be identified and described; its materiality considered, analysed and recorded. This could be as simple as recording its dimensions and preparing good quality illustrations, photographs, 3D scans and so forth; but it would ideally include more detailed analyses of features such as fabric, surface treatment or, in the case of inscribed vessels, residue analyses of contents. Within this first step we can also place the initial reading, interpretation, transcription and, if possible, translation of the inscription. Ideally, these 'archaeological' and 'epigraphic' aspects should not be separate; even if they must necessarily be the work of different specialists, they should work together and with due consideration of each other's methodologies and findings, as much as possible. This is not a call for anything especially novel: while disciplinary integration may not be as close as we might wish,

<sup>44</sup> These are partially inspired by Robb's three categories of meaning possible for material culture (Robb 2010, 506): structural meanings, which derive from the habitus and people's understandings of how the world goes together; generic meanings, which derive from the specific field of action an object is part of; and contextual meanings, which emerge from the specific context within which an object is used in a given instance. The difficulty for applying Robb's schema to Ugarit is that we rarely have the detailed contextual information necessary. My own stages of analysis, it will be noted, do not map exactly to Robb's in either scope or theoretical grounding, though they nevertheless owe a debt to them.

one would be hard-pressed to find an inscribed object from a well-conducted modern excavation that was not dealt with more or less in line with these recommendations. The key point is that this can't be the entirety, or even the majority, of the analysis; it is only the beginning.

#### *Exploring meaning in context*

Through the steps detailed above, we will have determined much about the nature of an inscribed object, including, with luck, what the inscription itself actually says; but as will be clear from the discussion throughout this chapter, we are still some way short of determining its meanings, since these are not inherent in the text itself, or even the whole inscribed object; but instead are socially constructed and exist in its relationships to other objects, people, groups, ideas and social structures at multiple scales and in ways that are specific to the particular society and historical juncture at hand. It is at this point that it becomes impossible to proceed with a separation between the 'archaeological' or the 'epigraphic'; to understand these linkages we must explore them holistically, analysing the inscribed object as a 'total social fact'.45

There isn't a single correct methodology for doing this; it will depend on the specific research questions being asked, the nature of the evidence and the point in its modern lifespan that we encounter it. In some circumstances it may be that the *chaînes opératoires* championed by Dobres and many others before and since will offer a useful tool; but they are not applicable in all cases, and even where they are, they will be more useful for delineating some of an inscribed object's relationships than others – namely those relating to the production process. Fortunately, as I mentioned at the beginning of this chapter, there is by now an archaeology of almost everything, and we can – and must – integrate these existing methodological frameworks into our work. Since the vision of context I offer here encompasses the whole socio-cultural web within which an object is enmeshed, it would be impossible to list, much less to do justice to in describing, these many ideas and approaches here. The point is that the Archaeology of Writing I propose here must also include archaeologies of power, status, ideology, religion, gender, trade, economy, the body, landscape, urbanism, and so on *ad infinitum*. The key question is not what we should include in exploring the contexts of our inscribed objects, but where we should draw the line. The answer, of course, is that this judgement must be made carefully on the merits of each example; it will depend on the limitations and affordances of our data available. But even when the evidence doesn't permit us to say anything useful about how a particular aspect of society and culture might have structured the meanings of our inscribed object, we cannot assume that there was no relationship there. We must be explicit and open about the gaps of our knowledge and recognise that absence of evidence is not evidence of absence. To give a concrete example, one of the problems we will grapple with over the course of this book is what the general Ugaritian population – the majority of whom were likely non-literate – made of the introduction of the alphabetic

<sup>45</sup> Mauss 2005 [1954].

cuneiform writing system. Ultimately, the evidence only permits us to answer this in the broadest and most speculative terms; but this doesn't mean this was not an important aspect of analysing in the social and cultural meanings of writing in Ugarit. On the contrary, arguably in terms of the proportion of the population involved, this was more significant than the meanings of writing for the small literate elite.

If this totalising, expansive approach to reconstructing the contextual meanings of ancient writing seems daunting, then it should do. To think of and investigate every possible social and cultural relationship that might be important for an object is a hell of an ask. But it also offers us a much richer and methodologically grounded basis for understanding the place of our material within a society. Frameworks such as Robb's three categories (2010) – meanings derived from specific context, meanings stemming from the generic field of action, and meanings dependent on relationships with overarching structures of internalised habitus – can help us here, but ultimately the onus is on us to be comprehensive and open-minded, to think about what questions archaeologists ask of other kinds of material culture and to ask them of our inscribed objects too. And to reiterate once again, in this analysis the inscription *cannot* be disentangled from the object it is written upon: the relationship between writing, text and object is at the heart of the mesh of connections and meanings that ramifies out from them.

#### *Finding agency and implications for understanding wider society*

So we've thought about these questions and traced how things like physical location, relationship to other objects, ideologies of power, prestige, age and gender, notions of self and other, and so on contribute to form meanings for our inscribed object. Perhaps we have a horrifically messy sketched-out diagram of the network or mesh of relationship and meaning and are lying awake at night wondering how we will ever reduce this overwhelming complexity to something intelligible enough to be published without sacrificing the holistic nature that is its whole point. We're still not done. Because so far we've thought about objects and structures and ideologies, but in doing all this we might have lost sight of the people. A network diagram – even the miraculous example that is both decipherable and comprehensive – is only an abstraction and can risk reducing all this work back down to a functionalist machine blueprint of interacting systems and feedback loops. We need to remember that the structure we have delineated is not absolute and unchanging, but must be constantly reproduced by the actions and choices of human beings operating within it and in response to it.

As with Dobres' suggestion that the descriptive work of *chaînes opératoires* can be used to identify where and on what bases variation takes place, or Robb's emphasis on the importance of variation within similar genres or fields of action,46 our analysis must be the basisfor exploring the differences between variousrealisations ofsimilar acts within similar social contexts. How does this particular writing exercise tablet

<sup>46</sup> Robb 2010.

reproduce orreact against widerstructures of practice? How different isit from other comparable examples in similar contexts? In this way we can ensure that we don't lose sight of agency – both of humans and of things, and avoid our work becoming the kind of sterile, mechanistic abstraction that social theorists and archaeologists have for a long time now argued against.

As we move up into larger scales of analysis, first exploring how variations between different realisations of a type of inscribed object or writing-related practice help us delineate the agency of those involved, and then comparing different examples of agency in different contexts across a society, we reach the point where we can meaningfully talk about the social context of writing and of writing practices at a social or cultural level. This isn't to propose that we should aim for banal and reductively overarching conclusions that smooth out the important contours of difference according to time and context; but that alongside the specific and contextual it's possible to suggest big-picture conclusionsforthe roles ofspecific kind of material or culture or practice, such as writing, within a given society; or to draw conclusions about patterns in where and how agency tends to operate.

This chapter has been rather theoretical in nature and has presented something of an ideal for how the archaeology of writing practices should be approached. As we will see, however, applying this theory to existing, imperfect datasets can be far from straightforward. As we explore writing practices at Ugarit in the rest of this volume, we will encounter many instances when we are forced to compromise on the kinds of analysis I have laid out as desirable here. It's also true to say that aspects of this methodology have been applied to Ugarit by others. Nevertheless, I believe it has been valuable to lay out the theoretical and methodological ideas that have driven my research up-front and in explicit terms. In part, because it lays the groundwork for what I have aimed to achieve, even if that has not always been possible; and in part because I hope it will have wider relevance, especially for future research less constrained by the shortcomings of past exacavation and publication practices. However, it's not my goal for this work to be principally a theoretical disquisition. The next three chapters will explore how alphabetic writing developed and spread through the regional networks of the Bronze Age Near East, before we turn more closely to Ugarit itself and examine writing practices within its society and culture for the remainder of this book.

### Part II

Late Bronze Age writing practices in regional context

## Chapter 3

### Writing in the Bronze Age Levant

To place Ugarit and its writing practices in their proper context, and in particular to situate alphabetic cuneiform with respect to wider developments in the burgeoning of alphabetic writing in the Levant, we need to look beyond Syria and the Late Bronze Age and explore changes in writing and its social context across the coastal Levant in the centuries leading up to our period of interest. This has been a field rich in discussion and debate, especially in recent years, not least because there are so few firmly established facts. Material is scarce, lacunose and often without clear (or at least published) information on its archaeological context. Much is therefore down to the subjective palaeographic analyses of each researcher. Consequently, dates are frequently subject to very large margins of error, and at times there is even significant disagreement as to which items belong within the corpora. Added to this, there has been a slow but steady trickle of new discoveries, which prompt continual upheaval and re-evaluation. Gordon Hamilton wasn't exaggerating when he said in 2014, 'the scholarly consensus about the periods of early alphabetic scripts has virtually collapsed during the last decade'.1 He focuses on the alphabet here, which has always dominated discussions of writing in the Middle and Late Bronze Age Levant and will account for most of the discussion in this chapter too, but we can also add that the extent and nature of cuneiform usage in the region remains extremely ambiguous and subject to rather greatly differing estimations. For these reasons, it would be foolish to expect to offer much in the way of firm conclusions in this chapter. Rather, my aim is to lay out the state of the discussion and to establish the proposals I find most convincing and will be using as a basis for my discussion of Ugarit.

#### **The invention of alphabetic writing**

The last two decades have seen considerable expansion of the discussion surrounding the earliest phases of alphabetic writing. For most of the twentieth century, essentially

<sup>1</sup> Hamilton 2014, 30.

our only window on this apparent first phase of alphabetic writing was the so-called Proto-Sinaitic inscriptions, discovered by William Flinders Petrie in and around Egyptian mine-workings at the remote site of Serabiṭ el-Khadem in the Sinai Peninsula and subsequently identified by Alan Gardiner as alphabetic and West Semitic in nature. This changed around the turn of the current century with the discovery of two more similar inscriptions on a stone wall in the Wadi el-Ḥôl in Egypt's Western Desert.<sup>2</sup> Around the same time, a heddle jack – an implement used in weaving – which had been found at Lahun in the Faiyum by Petrie was rediscovered in the British Museum and its alphabetic inscription assigned to a similar date.3

With this expansion of the corpus and demonstration that alphabetic writing was spread well beyond the Sinai, Hamilton has recently argued that we should discard the traditional geography-based labels 'Proto-Sinaitic' and its Levantine successor 'Proto-Canaanite' in favour of a three-phase scheme for a single alphabetic tradition, with his 'Early Alphabetic A' corresponding to the former and 'Early Alphabetic B' to the latter.4 'Early Alphabetic C' relates to the Iron Age and so doesn't concern us here. Notwithstanding my suspicion that the traditional labels might by this stage be too ingrained to shift, this seems like a good suggestion and helps avoid some of the rather artificial dividing and taxonomising that can tend to bedevil the subject.

Early Alphabetic A, or Proto-Sinaitic if you prefer, is evidently a consonantal alphabet of much the same kind we see later in the Levant. However, its signs remain highly pictographic and betray their heavy indebtedness to Egyptian hieroglyphic and hieratic prototypes. Crucially, however, the Egyptian meaning of the signs was not borrowed together with the forms; rather, the inventors of the alphabetic script assigned new, consonantal values based on the acrophonic principle. For example, the water-sign, M, has the value /m/, because that is the first sound of the Semitic word for water (Hebrew, Phoenician *mym*, Ugaritic *mh/y*).

There are two main areas of debate surrounding Early Alphabetic A: the first is its date, and the second concerns the social context of its creation and early use. The majority of scholars place these inscriptions in the Middle Bronze Age, although exact positioning tends to vary.5 The mines at Serabiṭ el-Khadem are known to have been active in this period and Canaanites were employed by the Egyptian expeditions there, as well as by military caravans passing through the desert in the Wadi el-Ḥôl area; the sign-forms seem to be most consistent with Middle Kingdom hieroglyphic and hieratic prototypes, and radiocarbon dating of the Lahun heddle jack has given a date of around 2140–1940 (though based on other considerations such as the typology of the object and the floruit of the site, Hamilton advocates a slightly lower date for this object of ca. 1850–1700).6 The domination of Egypt by the Hyksos of Levantine origin during Egypt's Second Intermediate Period has

<sup>2</sup> Darnell *et al.* 2005.

<sup>3</sup> Cartwright *et al.* 1998; Hamilton 2006, 330–331; 2014, 33.

<sup>4</sup> Hamilton 2014.

<sup>5</sup> Goldwasser 2006; 2012; 2013; Hamilton 2006; Rainey 2009; Rollston 2010a; Haring 2019.

<sup>6</sup> Hamilton 2014.

also occasionally been linked with this horizon of writing, although most scholars would see the earliest inscriptions as pre-dating it. Even outside the period of Hyksos rule, however, there was a significant Levantine presence in Egyptian cities such as Tell el-Dabʿa (Avaris). With the early Middle Bronze Age as a *terminus post quem*, Hamilton has argued that the Sinai inscriptions belong to various phases over almost a thousand years, with the latest occurring around the thirteenth century.7 If the earliest of these extant inscriptions date around the beginning of the second millennium BC, then it's possible the actual invention of the alphabet might have occurred somewhat earlier, as Ben Haring speculates,8 to account for its dispersal over quite a wide geographical area.

While this all seems perfectly plausible and consistent, it does result in a very long 'start-up period' for alphabetic writing in which it seems to develop very little and gains little ground, before a sudden and rapid explosion in development and use around the end of the Late Bronze Age. Benjamin Sass has seen this 'boom' as deeply problematic and despite advocating the Middle Bronze Age date in his seminal 1988 work on the early alphabet,9 has in recent years argued for a dramatic lowering of the date for the creation of the alphabet to around 1300 BC.10 There's nothing, he argues, to pin down Early Alphabetic A/Proto-Sinaitic any more precisely than Middle-to-Late Bronze Age, and the later date would allow for a scenario where the 'genius' of the alphabet was swiftly recognised and it developed reasonably rapidly. While Sass is right to flag up the dubious and unknown provenances of many of these inscribed objects, which does allow for his dating, his argument nevertheless fails to convince since it must argue away a rather substantial body of Middle Bronze Age alphabetic inscriptionsfrom the Levant (see below), including the well-stratified Lachish dagger, which even he accepts has a Middle Bronze date 'beyond reproach' (his response is to cast doubt on whether its inscription is really part of the same alphabetic tradition, as opposed to some other, otherwise unrecognised writing). Even more puzzlingly, Sass's aversion to a rapid boom in the alphabet around the end of the Late Bronze Age is not matched by a similar attitude to the Iron Age. Instead, in a series of recent articles, he has sought to lower dates for the standardisation of the alphabet, its spread beyond the Shephelah and Philistia and its adoption by states such as Byblos until around the ninth century or even later,11 at which point we must believe it spread rapidly not just through the Levant, but well beyond, into Anatolia and the Mediterranean, evolving into daughter scripts such as Greek and Phrygian in a remarkably (implausibly) brief span of time. Even under the traditionally accepted dates, some scholars have already been voicing disquiet about the speed at which the Phoenician alphabet spread and spawned these significantly different daughter scripts.<sup>12</sup>

<sup>7</sup> Hamilton 2006, 289.

<sup>8</sup> Haring 2019.

<sup>9</sup> Sass 1988.

<sup>10</sup> Sass 2004–5.

<sup>11</sup> Finkelstein and Sass 2013; Sass and Finkelstein 2016.

<sup>12</sup> Waal 2019.

The second main area of dispute is perhaps even more pertinent for our discussion since it concerns the social context that gave rise to the alphabet and its relationship with existing elite writing cultures. There are a number of good reasons to associate early alphabetic writing not with officialdom and traditional literacy practices, but with the marginal,sub-elite and non-prestigious. The Sinai and Wadi el-Ḥôl inscriptions are, after all, found in remote locations associated with relatively ordinary labourers: miners on the one hand, military personnel on the other. They are relatively crudely inscribed and lack standardisation to the extent that most are still largely or wholly indecipherable. They are the work of an immigrant community of 'Asiatics', hardly a favoured or prestigious group in Egypt even when they were high-ranking. After this, as we've just mentioned and will go on to discuss in more detail below, we have no evidence for alphabetic writing being used at an official level by a state bureaucracy before the emergence of alphabetic cuneiform at Ugarit in the thirteenth century, and no evidence for the linear alphabet fulfilling such a role before the early Phoenician royal inscriptions at Byblos, the date of which is disputed but which are very unlikely to be much older than the tenth century. The scattered attestations of alphabetic writing from the Bronze Age Levant are mostly personal, experimental and unstandardised, consistent with the idea that alphabetic writing is marginal. In contrast, what clear examples we have of official use of writing in the Levant beyond Ugarit at this time are in Egyptian hieroglyphic and hieratic scripts, and in Mesopotamian logosyllabic cuneiform.

The bone of contention is not that Early Alphabetic A inscriptions were made by non-elite, non-professional writers, but whether or not such people could have been responsible for the script's creation. The two positions were defined in an online debate between Orly Goldwasser and Anson Rainey in 2010,13 in response to an article by Goldwasser. In a series of papers, Goldwasser has argued that not only was the alphabet created by people lacking formal training in Egyptian writing practices, their very illiteracy was crucial to its invention. In her view, their ignorance of how Egyptian hieroglyphs actually worked allowed them to 'think outside the box' and create a new writing system suited to their own language. Early Alphabetic A's departure from Egyptian writing practice – not just in the structure and sign-values of the script but also in stylistic conventions such as direction of writing, stance of signs and so on – would therefore not so much be a conscious and informed rejection so much as the fact that its users neither knew nor cared how the Egyptians would do it. We might compare it to the more recent example of the Cherokee syllabary (Fig. 3.1), developed in the early nineteenth century AD by Sequoyah, a silversmith whose work brought him into contact with literate Europeans but who did not have formal education in writing. The script he developed was inspired by observation of the alphabet but, as a syllabary, works along quite different lines. Like Early Alphabet A, it includes numerous signs borrowed or derived from the writing system that inspired its development, but they are assigned entirely different values.

<sup>13</sup> Rainey and Goldwasser 2010.


*Fig. 3.1. The Cherokee syllabary, incorporating numerous signs borrowed from the Roman alphabet, but assigned wholly different values. Image from Wikimedia Commons, Public Domain.*

There are some apparent higher-status uses of Early Alphabetic A: an inscribed sphinx (Sinai 345) is a dedication from the Egyptian temple of Hathor, as was a small statue base (Sinai 346) inscribed by the *rb nqbn* – chief of the miners. Being allowed to put a dedication in the temple is an honour that was unlikely to have been granted to just anyone, especially among Levantine miners. Even more suggestive is Stele 92 from outside the temple. This is inscribed on four sides in hieroglyphs. Two are competently done in typical Egyptian fashion; the others are rather poor and Goldwasser interprets them as the work of a Canaanite scribe. The content of the inscription mentions a Khebeded, 'brother of the ruler of Retenu' (*i.e.*, Canaan, or a part of it). This man is also depicted riding a donkey on another hieroglyphic stele, and was evidently a dignitary among the local Levantine population. Despite being in Egyptian hieroglyphs, the two lower-quality faces of Stele 92 have some peculiar sign-variants, which Goldwasser links to the alphabetic inscriptions. She sees this as evidence that the local elite within the Levantine community had become aware of the alphabetic script and that it was interfering with their writing of hieroglyphs. In other words, what higher-status use of, or influence from, Early Alphabetic A there was, was secondary and resulted from the elite becoming aware of a scriptal development that had already taken place among the rank-and-file miners.<sup>14</sup>

Rainey took a diametrically opposed view, arguing that the alphabet must have been the work of a single genius familiar with Egyptian writing, who adapted it for the Semitic language from an informed, literate perspective. Echoing an earlier

<sup>14</sup> Goldwasser 2006; 2011; 2012; 2013; 2015; 2016; Rainey and Goldwasser 2010.

suggestion by Alan Millard, he proposed that the most likely context for this was the commercial sphere at Tell el-Dabʿa/Avaris.15 In support of this 'elite' interpretation of the early alphabet, Rainey argued that the objects depicted in the pictographic signs of Early Alphabetic A are particularly centred around the military and hunting lifestyle of a Middle Bronze Age warrior aristocracy. Goldwasser objects to this, and points out that no Early Alphabetic material has been found at Tell el-Dabʿa. Perhaps her biggest objection to Rainey's position is his idea that, having been created within the scribal bureaucracy of the Levantine community in Egypt, the alphabet would have been used in a formal and relatively standardised form on perishable materials before it came to be crudely adopted by the workers of the Serabiṭ el-Khadem mines. This 'lost papyri' theory, she contends, is at present pure speculation without any actual evidence to support it.

Others have joined Rainey in arguing against Goldwasser's position, including prominently Christopher Rollston, who focused in particular on Stele 92 and the more prestigious instances of Early Alphabetic A. By showing that the alphabetic script was known by a Canaanite elite who were also literate in hieroglyphs, he argued, Goldwasser had undermined her own argument and proven just the opposite: that it cannot conclusively be attributed to a sub-elite social context.16 Brian Colless has also published a detailed refutation of many of Goldwasser's claims,17 and most recently Aaron Koller has argued that the alphabet was disseminated by professional writers – probably Egyptian ones – and 'based on current knowledge, the alphabet seems to have been kept in the hands of the scribes and benefited no one else'. He suggests this scribal dissemination may also imply it was invented by those 'scribes' too.18

There are strengths and weaknesses on both sides of the debate. If Rainey's statement that 'The alphabet was not invented to be scratched on the walls of a cave,'19 seems overly sweeping and elitist, Goldwasser's position almost certainly oversteps the available evidence for the reasons Rollston and Colless spell out. At any rate, all this discussion has been rather too narrowly focused, both geographically and methodologically. Commentators on both sides (but especially Goldwasser) have preoccupied themselves with the Serabiṭ inscriptions, and largely overlooked the discoveries from the Wadi el-Ḥôl and Lahun. Discussion has been almost entirely palaeographic and epigraphic, with little attention to what actually constitutes 'elite', 'educated' or ' prestigious' practice and material culture at this time, especially in marginal locations such as these. We know relatively little about how these expeditions were organised or how the people involved in them lived. Research at Serabiṭ has understandably focused on the alphabetic inscriptions and on the temple of Hathor; work on other Egyptian mining operations in the Sinai tends to concentrate on questions of production and

<sup>15</sup> Millard 1986, 394; Rainey 2009; Rainey and Goldwasser 2010.

<sup>16</sup> Rollston 2010b, with a response in Goldwasser 2012.

<sup>17</sup> Colless 2014.

<sup>18</sup> Koller 2018.

<sup>19</sup> Rainey and Goldwasser 2010.

industry.20 Although every indication isthat the operation at Serabiṭ waslong-running and large-scale, we don't have any remains from the miners' dwellings; apart from the temple, the only structures that have been identified are some low, dry-stone enclosures, which Petrie interpreted as shelters or windbreaks for ritual outdoor sleeping. This brings us to the question of religion, the importance of which is well documented for the mines in general and their inscriptions in particular. At Serabiṭ, Hathor in her guise as 'Mistress of Turquoise' was the pre-eminent deity, and it was equating her with the Semitic Baʿalat in the alphabetic inscriptions that allowed Gardiner to make the breakthrough and identify them as Semitic and consonantal in the first place. For Goldwasser, religious experience was a fundamental motivator in the impulse to write, and to write alphabetically,21 but the actual patterns of religious practice in these mining communities, its relationship to material culture and especially to writing surely demand considerably more research.

The argument remains open, but the balance of evidence and scholarly opinion suggests that Goldwasser's conclusions are at best premature. There seems little guarantee that Sinai was ground zero for the alphabet, rather than just a place where a significant number of inscriptions happen to have survived and been discovered. Even within Sinai, there doesn't seem to be any conclusive way of telling whether the higher-ranking Canaanites and their scribes picked up the script from the miners or vice versa. The exact sociology of prestige and writing within these communities remains opaque, but it does seem likely that alphabetic writing was in use across a broad spectrum of levels of prestige, including miners and possibly metalworkers. Even if we can't say that these people invented the alphabet, the importance of this broad usage shouldn't be underestimated in a Middle Bronze Age world where literacy was extremely limited and largely the preserve of the elites and their bureaucracies. The association of alphabetic writing with sub-elites in peripheral locations is remarkable in itself.

#### **Middle Bronze Age alphabetic writing beyond Egypt**

We've already touched on the sparse and difficult nature of the evidence for alphabetic writing in the Levant. This material has traditionally been called 'Proto-Canaanite', which implies a distinct separation from the 'Proto-Sinaitic' material we've been discussing so far, which isn't really warranted. It also falls afoul of the problems and ambiguities surrounding the term 'Canaan',22 so we will continue to follow Hamilton's

<sup>20</sup> Giveon 1974; Rothenberg 1987; Bloxam 2006.

<sup>21</sup> 'Was it the forlorn remote place,' she wonders, 'the pressure, the sudden acknowledgment of an option of "eternalizing the name," of "contacting the gods" that led the Canaanites to this great invention?' (Goldwasser 2006, 152). 'We must therefore surmise that the impetus for the invention of the alphabet was spiritual. The Canaanites wished to communicate with their gods, to talk to their gods in their own language and their own way.' (Rainey and Goldwasser 2010, 'Goldwasser's first rebuttal'.

<sup>22</sup> Discussed in Boyes 2013, 18–19, including n. 22.

new schema, which categorises the material typologically rather than geographically. Some of the material from the Levant is thus the relatively pictorial, incised Early Alphabetic A, some the more simplified and linear Early Alphabetic B, which tends to be painted on ceramic.23

When considering writing in the Middle Bronze Age Levant, the general tendency has been to focus on the alphabet inscriptions, again from a largely palaeographical perspective since, as we've already hinted, most are not from known archaeological contexts. For a fuller picture, however, we must also consider other writing practices in the region – primarily Egyptian hieroglyphic and hieratic and Mesopotamian cuneiform – and how these various traditions interacted both with each other and with changing social and political structures.

Let's take this last point first. The Middle Bronze Age was a period of quite rapid and dramatic social change in the Levant, characterised by significant urbanisation and centralisation within settlement patterns and increasing standardisation of material culture. Politically, the first half of the second millennium BC saw the crystallisation of city-state polities and the emergence of large-scale palatial architecture, which in many cases reached great heights of sophistication and wealth.24 There has been a temptation to think about this in terms of a somewhat homogeneous 'palace culture' similar to that in other parts of the Near East and in the Aegean – something exacerbated by the appearance of Aegean-style frescoes at sites like Qaṭna and Alalaḫ in Syria, Tel Kabri in northern Israel and Tell el-Dabʿa in the Nile delta. However, recent research has suggested such shared elements of elite display belie quite significant differences in how palaces and the elites that inhabited them operated from region to region. The excavators at Kabri have argued that palaces in the southern Levant were primarily focused on production and storage to meet their residents' needs rather than acting as manufacturing, storage and redistribution hubs for the polity as a whole.25 They portray them as 'oikos'-economies, exceptionally large households but essentially households nonetheless, and lacking large storage magazines or evidence of extensive administrative practices such as sealing or writing.

Writing would certainly have been known in the Levant for a long time through its close links with Egypt and Mesopotamia. Itzhaq Shai and Joe Uziel (2010, 70–71) argue that we might expect Egyptian writing to have been adopted at major Levantine administrative centres with close Egyptian trade connections, such as Megiddo or Tel Yarmuth, as early as the Early Bronze Age. The fact that we don't, they suggest, can only represent a deliberate rejection of the idea of literacy, perhaps because it was seen as an unwelcome element of Egyptian cultural influence. In general, they note, despite trade links with Egypt, Levantine cities of the Early Bronze Age appear to have looked more to Mesopotamia for inspiration. Differences in social change at the end of the Early Bronze may begin to explain the north–south divide of the

<sup>23</sup> Hamilton 2014.

<sup>24</sup> Ilan 1995; Cohen 2014; Yasur-Landau 2019.

<sup>25</sup> Yasur-Landau *et al.* 2015.

Middle Bronze Age. Many southern cities show a break in habitation at the end of the Early and beginning of the Middle Bronze Age which isn't apparent in their more northerly counterparts, leading to more of a discontinuity between Middle Bronze culture and what came before.26

#### *Logosyllabic cuneiform in the Levant during the Middle Bronze Age*

Only thirteen examples of cuneiform are known from the southern Levant, of which three are cylinder seals (Table 3.1). Eight of the remaining ten tablets are from Hazor, on the edge of what would later be considered Phoenicia and by far the largest centre in Israel/Palestine during this period. Academic texts are the largest genre, but not by a significant margin and numbers in any category are very small.27 Hazor is the only site to have produced more than a single example of cuneiform and even there the numbers are too low to indicate the large-scale use of the script in an administrative capacity.

The key point of contact for the Mesopotamian scribal world and the coastal Levant seems to have been Mari: Horowitz, Oshima and Sanders (2006, 12) detail a number of scribal peculiarities shared by the Hazor and Mari corpora and a king of Hazor appears in the Mari archives as a contemporary of Zimri-Lim (ca. 1775–1761). Since there are no earlier traces of cuneiform in the southern Levant, these authors suggest that cuneiform was introduced to the region at this time via Mari to facilitate trade.28 The same may be true of Ugarit: Zimri-Lim visited the city and the first allusion to Ugaritian writing is a letter from Aleppo to Mari in which the king of the former mentions that his Ugaritian counterpart has written to him expressing a desire to travel to Mari.29 The dynastic seal of the royal house of Ugarit, inscribed in Akkadian with the name of King Yaqaru, may also date to around this period.30 Another coastal city that features prominently in the Mari archives is Byblos, but here there is a hint at an earlier familiarity with cuneiform writing and scribal practices. A fragmentary Sumerian syllabary dated to the Ur III period (late third millennium BC) was reportedly found in material taken from near the Obelisk Temple, though unfortunately its original context has been lost.31 In general, however, cuneiform is rare in the West Semitic region before the Late Bronze Age, and while Levantine rulers and scholars may occasionally have used it, there is little evidence of systematic and large-scale bureaucratic use.

An intriguing complication to this picture has come in recent years, however. In 2009 Stephanie Dalley published a private collection of cuneiform tablets from the

<sup>26</sup> Cohen 2002, 5.

<sup>27</sup> Horowitz *et al.* 2006.

<sup>28</sup> Horowitz *et al.* 2006, 13.

<sup>29</sup> Schaeffer 1939a, 16.

<sup>30</sup> For a summary of scholarly opinions on the subject, see Di Paolo 2013, 77–78.

<sup>31</sup> Dossin 1969. For more on the potential relationship between Ugarit, Byblos and Mari in the Middle Bronze Age, see Loretz 1994.


*Table 3.1. Cuneiform from the southern Levant in the Middle Bronze Age. After Horowitz* et al*. (2006, 10, table 2).*

Sealand in south-east Babylonia, which, although of unknown provenance, are dated from around the end of the first dynasty – that is, around the sixteenth or fifteenth centuries BC, based on palaeography and content. Four of the tablets bear additional inscriptions on their edges in linear, alphabetic West Semitic.32 The inscriptions have been discussed briefly by Laurent Colonna d'Istria (2012) and in more detail by David Hamidović (2014). Both have pointed out that the alphabetic script most resembles examples from much later than the date assigned to the cuneiform texts. Colonna d'Istria draws particular parallels with the Tell Zayit abecedary, which Finkelstein and Sass recently dated as late as the ninth century.33 Hamidović cites the Phoenician arrowheads of the eleventh century as parallels, as well as first-millennium examples, but settles on a palaeographic date range of ca. 1200–1000.34 The mismatch between the most plausible dates of the cuneiform and linearinscriptions presents a significant chronological dilemma. The tablets were unfired in antiquity so it's notimpossible that someone added the linear texts later.35 However, Hamidović considers this the least likely possibility. It should be noted, for instance, that the complete linear inscription on tablet 149 contains a name, *ʾldnʾl*, which also appears in the cuneiform text (as Ali-dīn-ili). In Hamidović's view, these inscriptions are evidence of bilingualism and bigraphia among scribes in Middle Bronze Age south-east Mesopotamia.36 If correct,

<sup>32</sup> Dalley 2009.

<sup>33</sup> Colonna d'Istria 2012; Finkelstein and Sass 2013.

<sup>34</sup> Hamidović 2014.

<sup>35</sup> It is possible to soften portions of dried tablets using wet cloths, for the purpose of additions or corrections – see Pape *et al.* 2014.

<sup>36</sup> Hamidović cautiously speculatesthatthese inscriptions might be connected to the Amorites – nomadic peoples present in both the Levant and Mesopotamia – but there is little evidence that either these tablets or alphabetic writing in general were connected with Amorites. In general, the theory that the many innovations and changes in Levantine society during the Middle Bronze Age should be attributed to Amorites has declined substantially in popularity over recent decades (Cohen 2014). On the idea that Ugarit's royal dynasty may have been Amorite in origin, see Buck 2018.

this would be the first example of interaction between alphabetic and cuneiform traditions and the first instance of the alphabet being used on clay tablets, and therefore a clear precursor to the Late Bronze Age situation in Ugarit. More than that, it would mean that our entire understanding of the early history and development of alphabetic writing needs radical revision.

At the moment, caution is clearly warranted. Any alphabetic inscriptions in Mesopotamia during the Middle Bronze Age would be surprising even if the script looked appropriate for what we would expect for alphabetic writing at that time. In my view, the alternatives – that the dating of the cuneiform is wrong or that the inscriptions are much later additions – have not been sufficiently disproven that we should want to embark on a rapid tearing-down of what few points of knowledge we think we have about the history of the alphabet during the Middle Bronze Age. The readings of the inscriptions aren't always terribly clear and even when they are, the meanings are open to debate. In the case of tablet 149, the presence of the same name in the alphabetic and cuneiform inscriptions needn't necessarily indicate that they're contemporary. As Hamidović reads it, the linear text consists only of personal and place names. Perhaps this was a later label indicating the tablet's contents: 'Re: Ali-dῑn-ili, at/from TOPONYM, collated/filed by PN?' This is merely one of several imaginable explanations which should be discounted before we begin rewriting alphabetic history. As Hamidović concludes, further research is clearly needed.

#### *Hieratic and hieroglyphic in the Middle Bronze Age Levant*

Assessing the extent and nature of Egyptian writing in the Levant during this period is complicated by the fact that there isn't a comprehensive corpus or even much in the way of dedicated studies. This era of Egyptian-Levantine relations tends to receive less attention than the 'imperial' period of the Late Bronze Age, and what work there has been has not been focused on writing practices. Nevertheless, there is little doubt that both hieratic and hieroglyphs were known in the region, although once again a north–south divide is apparent. Small portable objects such as scarabs were, as ever, ubiquitous and several northern sites have produced quite significant collections of Egyptian and Egyptianising material, which often include inscriptions. Ugarit is among these, and for a fuller discussion of its Aegyptiaca see Chapter 9. Further south, Byblos has produced an impressive array of high-status Egyptian material. Several extremely fine items were recovered from the royal chamber tombs of the Middle Bronze Age, which also included hieroglyphic inscriptions in honour of the local kings. Notably, these don't just adopt the Egyptian language and writing system, but also Egyptian conventions in such matters as titulary: rather than adopting the pharaonic title of 'king', they style themselves using the Egyptian terms for Asiatic rulers: *ḥꜢtyw-ʿ*, 'governor' or *ḥkꜢ ḫꜢswt*, 'chief of foreign lands'.37 Monumental hieroglyphic inscriptions

<sup>37</sup> Incidentally, the latter title is generally agreed to be the origin of the term 'Hyksos' used for the rulers of Levantine origin who established themselves at Tell el-Dabʿa at the end of the Middle Kingdom.

were also found at the site outside the funerary sphere, particularly in and around its various temples. Some of these may have been dedicated by Egyptians (the tutelary deity Baʿalat Gebal was worshipped as a variant of Hathor) but others seem to have been created by the local elite.

Traditionally, Egyptian influence atsiteslike Byblos(and to a lesser extent Ugarit) in this period has been seen in terms of empire, with an assumption that Egyptian functionaries were present to oversee local rulers who were considered essentially part of an Egyptian provincial hierarchy. There is always a difficulty disentangling ideology and representation from political reality, however, and we might question whether Egyptian 'control' at these sites was quite so literal as that, or whether it was – as it seems to have been in the Late Bronze Age – more a matter of alliance and trade relationships backed up by an elite culture strongly orientated towards Egypt for inspiration, and a language of vassalage which could be strategically deployed by both sides when it suited them.38 In the southern Levant, Egyptian material culture and evidence for Egyptian presence in the Middle Bronze Age was much less than expected, and much less than in Lebanon and Syria.39 The fictional tale of Sinuhe implies Egyptian familiarity with the southern Levant and speaks of Sinuhe receiving Egyptian correspondence while in the region, but Egypt seems to have been less of a cultural and ideological force there than it was in the north, with consequences for the prevalence of Egyptian-influenced writing practices. Koller (2018) has even suggested that the alphabet may have been brought to the Levant by Egyptians as an 'unofficial' counterpart to hieroglyphic/hieratic, although this remains speculation.

Unlike with cuneiform, then, for which we have only fairly restricted evidence for state usage, Egyptian writing seems to have been enthusiastically adopted by certain Levantine elites as a medium not of administration, so far as we can tell, but of prestige display; a marker of their status, culture and close relationship with their powerful ally to the south. While Byblos and Ugarit seem to have particularly strong connections, we shouldn't make the mistake of assuming they were unique in this regard. Recent work at Sidon has brought to light large amounts of Egyptian material there, and other polities along the Levantine coast also show signs of engaging in the same international elite culture. It wouldn't be surprising to find the elite use of Egyptian scripts quite widespread and institutionalised throughout many of the wealthy palatial elites of the Levantine coast.

One area where Byblos is exceptional, however, is in the use of another script, known as Byblian pseudo-hieroglyphic or the Byblos syllabary. This is known from thirteen inscriptions, mostly incised into metal or stone. Since the corpus is so small, the script remains undeciphered and highly enigmatic. Because of the high number of signs (Maurice Dunand put the figure at 114, while Giovanni Garbini estimated 90),<sup>40</sup>

<sup>38</sup> For a discussion of these issues in the Late Bronze Age, see Boyes 2013.

<sup>39</sup> Cohen 2002.

<sup>40</sup> Dunand 1945; Garbini 1988.

most commentators have considered the writing system a syllabary, although George Mendenhall (1985) and following him Colless (1990; 1992) saw it as proto-alphabetic, a sort of syllabic implementation of the acrophonic principle whereby each sign had the value of the first syllable of the thing it represented. They thus argued that it stood as a 'missing link' between Egyptian hieroglyphic and Levantine alphabetic writing. Mendenhall and Colless's interpretation has failed to gain widespread support, like all other attempts at decipherment to date.

Visually, the script appears to be broadly pictorial, taking a good deal of inspiration from Egyptian hieroglyphs, although people have also seen connections between some sign forms and those of both Early Alphabetic A and Phoenician. It remains extremely unclear where the script sits within the framework of Levantine writing, or even merely within the social and scribal history of Byblos. Indeed, we don't even know when it dates from. It's often assumed to be Middle Bronze Age because of the strong Egyptian influence at Byblos at this time, but it could feasibly fit anywhere up to the emergence of the first Phoenician inscriptions around the beginning of the first millennium BC. A bronze spatula bearing a palimpsest early Phoenician inscription over vestiges of a pseudo-hieroglyphic one may indicate that the syllabic script was still in use around the time that the linear script was officially adopted in the city, but the object is hard to interpret and overall the status of Byblian pseudo-hieroglyphic is among the less certain and more puzzling aspects of the already highly contested early history of writing at the site. What we can perhaps say with some certainty is that the existence of the script is further evidence of the climate of scriptal experimentation in the Levant during the latter two-thirds of the second millennium and a willingness to blend influences from diverse sources in the creation of distinctive local scripts.

#### *Alphabetic writing in the Middle Bronze Age Levant*

There are four examples of alphabetic writing from this period in the southern Levant, although only one has a satisfactory archaeological provenance. Two are unstratified sherds from Gezer and Nagila, with the signs incised before firing. Although the script looks similar to the Early Alphabetic A inscriptions from Egypt,41 hence their usual placement in the Middle Bronze Age, they cannot be dated on anything other than palaeographical grounds, so more recently Sass has argued that their date range should be extended to cover the whole Middle and Late Bronze Age, in accordance with his low dating for the invention of the alphabet.42 The so-called Shechem Plaque, a fragment of stone relief with an inscription added at a later date to its border, is likewise from an unknown context and so hard to date. Sass suggests that the relief itself may stylistically be attributed to the end of the Middle or beginning of the Late Bronze Age, but this does nothing to pin down the inscription. In none of these cases can we say

<sup>41</sup> Sass 1988, 54–56.

<sup>42</sup> Sass 2004–5.

anything aboutthe social context of writing, beyond the factthatitis confined to short inscriptions on small objects, a contrast to the rock-cut alphabetic writing of Egypt.

Fortunately, we can do a little more with the final item, a bronze dagger found in a small, single-period tomb at Lachish. While small, this burial was evidently not low status. As well as this object, its grave goods included a second, similar, dagger and prestige objects such as scarabs and ostrich eggs. A toggle-pin and the pottery found in the tomb concur with typological dating of the dagger to place this in the Middle Bronze Age, and it is typically dated ca. 1750–1650.43 The script retains a high degree of iconicity, which is consistent with a dating close to the early alphabetic inscriptions from Egypt, although its signs don't exactly correspond, and its reading remains partially uncertain.

Overall, the picture of writing in the Middle Bronze Age Levant is patchy and hard to read. There are perhaps signs of a north–south divide, although it would be valid to wonder how much of this might be due to differences in excavation practices and research interests in different countries. In the south, the reticence towards writing apparent in the Early Bronze Age can justifiably be argued to continue. Cuneiform remains scarce and alphabetic inscriptions are very small in number. It's extremely likely that Egyptian writing systems were known in the region, but there doesn't seem to be large-scale elite adoption of them of the kind we see further north. Shai and Uziel have suggested that the lack of Middle Bronze Age writing in the southern Levant may stem from an ideological opposition to inequality and socialstratification. Citing the growing standardisation and uniformity of material culture in the period, they suggest that writing may have been seen as a 'tool of the elite' and so rejected by societies that wanted to encourage equality and integration.44 Unfortunately, this isn't very convincing. The idea of uniformity in material culture is not the same as a lack of social stratification, and a supposed ideological opposition to eliteness is hard to sustain in the face of growing palatial centres and spiralling displays of wealth. Instead, perhaps it has more to do with differing economic and social roles for the palaces and their elites, which placed less emphasis on fine-grained regional administration. Certainly this is an area that will require further research, both from an epigraphic perspective and from an archaeological one, since clarifying the structure and culture of southern Levantine society in the Middle Bronze Age is likely to be key to better understanding the patterns of writing we see.

A transition in attitudes towards writing seems to begin with Hazor, with areas to the north showing considerably more willingness to engage with and utilise writing. At Hazor and other large polities that traded with Mari, people began to engage with cuneiform culture in this period, although somewhat tentatively. We can reasonably assume it was used for international correspondence, and perhaps for a limited amount of scholarly work, as well as for elite purposes such as the Yaqaru seal from

<sup>43</sup> Sass 1988, 53–54; Haring 2019.

<sup>44</sup> Shai and Uziel 2010, esp. 73.

Ugarit. But we are evidently not yet dealing with a large-scale international standard for diplomacy and administration. Alongside this emergent cuneiform writing culture, we find localised but intense evidence for enthusiastic adoption of Egyptian language, scripts and writing practices by local Levantine elites, principally in the field of prestigious display and legitimation.

In part, the piebald nature of this picture reflects the inadequacies of the evidence available, but it also underlines that the Levant was never a single place. We must consider local trajectories and local ideologies alongside the overarching regional pattern. It's easy to be drawn into thinking of 'writing' or 'the alphabet' as coherent, single entities or traditions, when in fact it makes much more sense to think in terms of local or even individualrealisations of loosely-defined ideas. Within thisframework the lack of standardisation in alphabetic writing, varying degrees of enthusiasm towards Egyptian scribal practices, or local idiosyncrasies such as Byblian pseudohieroglyphic all make sense: they reflect local experimentsin and reactionsto the idea of writing within the specific social, political and ideological contexts of each place.

#### **The Late Bronze Age**

In a number of ways, Levantine writing in the Late Bronze Age continues the patterns we've been discussing for the preceding period: the expansion of logosyllabic cuneiform for administrative and diplomatic purposes alongside the continued use of Egyptian writing systems at certain centres, especially in elite display, and at the margins the rather engimatic and hard to quantify expansion of the linear alphabet. There are, of course, numerous differences and additional wrinkles to consider, however: the non-standard use of cuneiform in the Amarna letters from the southern Levant; the impact of Egyptian imperialism on the use and social connotations of hieroglyphic and hieratic; the arrival on the Levantine scene of the Hittite Empire and the distinctive Anatolian writing system it brought with it. For the sake of the present discussion, this complex blend of continuity and change centres around one crucial question: how widely used was the linear alphabet in the period leading up to its adoption by the Ugaritian state? Answering this is a vital precursor to being able to assess how and why alphabetic cuneiform developed, and indeed where: was it solely an Ugaritian phenomenon, or is its concentration there merely a result of its having been embraced by that city's elite? It will also help provide context for understanding the relationship between alphabetic cuneiform and the logosyllabic cuneiform tradition in Ugarit.

#### *Logosyllabic cuneiform in the Late Bronze Age Levant*

It is often said that by the Late Bronze Age, the Akkadian language and the logosyllabic cuneiform script that went with it were a lingua franca that united the Near East. This is certainly the first impression from the Amarna Letters, which are our most famous source on diplomacy in this period, and is supported by other archives from sites in Syria, Anatolia and Mesopotamia. As we see at Ugarit, many Syrian cities have produced large corpora of material written in Akkadian language and logosyllabic cuneiform script. It's clearthat cuneiform writing had been officially adopted by state authorities for administration, epistolography and justice, and this was backed up by local traditions of formal literate education adapted from the Babylonian model.45 There was, of course, adaptation of both language and script to the various exigencies of the local contexts, as well as interference from local vernacular languages, which means that we are not dealing here with 'good' core Akkadian, or with a homogeneous picture across the region. We would expect no different. But there is no doubt that cuneiform and many of its associated norms and practices were thoroughly institutionalised as part of the fabric of Syrian and Anatolian urban life.46

The situation is rather different further south, however. In Lebanon and Israel/ Palestine, cuneiform material is much rarer: Horowitz, Oshima and Sanders (2006, 16) list only 42 examples found in Israel/Palestine (plus six more uncertain cases), to which we can add a handful more from the Amarna letters sent to Egypt from the region. This corpus doesn't include sources from Lebanon, but these are not numerous (in general, the Late Bronze Age in Lebanon is relatively poorly known, especially in the major urban centres such as Byblos, Tyre and Beirut).47 As well as examples being considerably fewer in number, the use of cuneiform in the central and southern Levant also differs substantially in type from the north. Horowitz, Oshima and Sanders (2006, 16) suggest that the appearance of the signs is much closer to cuneiform from Egypt than to that of Mesopotamia, or even Syria. This seems to have been an important entry-point for cuneiform writing culture into the southern part of the region. Linguistically, the 'Akkadian' of the central and southern Levant is extremely non-standard, and debate exists as to whether it can really be considered Akkadian at all. Grammatically and syntactically, Akkadian features are heavily intermixed with elements of the local West Semitic dialect. There is now consensus that this should be seen as a mixed hybridising linguistic form in its own right, rather than merely a defective form of Akkadian. Within this broad agreement, numerous different terms or categorisations have been advanced, including pidgin, contact language, mixed language and interlanguage.48 Eva von Dassow has gone further and suggested that actually, the language used is merely Canaanite, but that it is encoded 'Akkadographically', that is, that Akkadian words and the logosyllabic writing system were used as a means of transcribing the local language, and that it is in that language that the cuneiform would have been read.49

With the small numbers of texts and high degree of linguistic and scriptal adapatation, it would be tempting to see cuneiform culture in the sourthern and central Levant as essentially skin-deep; a superstrate used for international diplomacy

<sup>45</sup> van Soldt 1995; 2011. See also Chapter 5 below.

<sup>46</sup> On the nature and history of 'peripheral' Akkadian and associated cuneiform writing practices in the west, see Foster 2015.

<sup>47</sup> But see Finkel 2006.

<sup>48</sup> Rainey 1996; Gianto 1999; Izre'el 2012.

<sup>49</sup> von Dassow 2004.

but otherwise not widely practised. Nevertheless, there are signs that cuneiform was used more broadly: letters were found at Taanach, Gezer, and Tell el-Ḥesi, there was a legal deed from Shechem that lists twelve witnesses, and administrative tablets from Taanach.50 Millard classes the majority of this material as 'palace texts', but not everything fits unambiguously into this category. Engagement with the Mesopotamian literary and academic traditions also existed, as shown by a fragment of the epic of Gilgamesh from Megiddo or a West Semitic translation of the *urra = ḫubullu* lexical texts from Aškelon.51 Horowitz, Oshima and Sanders (2006, 18) suggest that Syrian cities such as Ugarit most likely served as intermediaries for the spread of such material into the south.

#### *Egyptian writing in the Late Bronze Age*

The idea of an Egyptian empire in the Levant has persisted far more stolidly for the Late Bronze Age than the Middle. It has long been a matter of consensus that Egypt launched a military expansion into the region around the beginning of the fifteenth century BC under Thutmosis I, largely in response to period of Hyksos domination during the Second Intermediate Period. The main issue of debate has not been the reality of Egyptian imperialism or even its nature, but rather how far north it extended and the exact details of provincial administration. Some years ago, in my doctoral dissertation,52 I argued a relatively minimalist position here, that actual Egyptian presence in the Levant was quite minor even in the south, largely consisting of a relatively small number of officials and a limited number of garrison towns. For Lebanon, which was my main focus, my view was, and is, that Egyptian imperialism is essentially a figment:there may have been campaign forces passing through the region on their way north from time to time, but there's little sign of any permanent Egyptian presence or administrative control. The claims of loyal vassalage from Lebanese and more northerly rulers that appear in the Amarna letters and other diplomatic correspondence seem to be strategically deployed platitudes and flattery aimed at eliciting desired Egyptian responses, not statements of day-to-day political reality. One need only look at the vain attempts of the Byblian king Rib-haddu to secure a favourable Egyptian military intervention on his behalf to get a clear sense of Egypt's absence and of the pharaoh's substantial indifference. What was true in Byblos was almost certainly true in Ugarit too. Claims of a boots-on-the-ground Egyptian imperialism there are much fewer, but do still occur. The idea of 'special relationships' between Egypt and certain favoured partners with whom it had close historical, cultural and mercantile tiesseemsto fitthe evidence much better, with these being, at various times, doubtless as meaningful or meaningless as the 'special relationship' between the modern United Kingdom and USA.

<sup>50</sup> Millard 1999.

<sup>51</sup> Horowitz *et al.* 2006. On *urra = ḫubullu*, see Chapter 5 below.

<sup>52</sup> Boyes 2013.

These differences are reflected in the use of Egyptian writing systems. The Egyptian material from Ugarit is discussed in more detail in Chapter 9, but suffice to say for now that it continues to be the kind of elite diplomacy- and prestige display-related material that we saw in the preceding period. Likewise at Byblos, although we lack Late Bronze Age strata for the city itself due to the inadequacies of the excavation, hieroglyphic inscriptions without context or from the Late Bronze – Early Iron Age royal Tomb V are temple dedications and diplomatic gifts, again pointing to prestigious elite activity. Other examples of hieroglyphs from Phoenicia at this time are either military campaign stelae or small portable objects such as scarabs and statuettes. In the southern Levant, there are far fewer diplomatic gifts, a pattern perhaps best exemplified by the distribution of stone vessels bearing pharaonic cartouches (Fig. 3.2). On the other hand, monumental inscriptions associated with actual Egyptian presence are more common, such as a twelfth-century door lintel from Beth-Shean, inscribed in hieroglyphics and with Egyptian-style relief illustration that depicts a local Egyptian governor, Ramesses-Weser-Khepesh, as well as the names and titles of Ramesses III. There is evidence for the use of Egyptian scripts beyond the sphere of elite display. Several bowls have been found bearing hieratic inscriptions mostly concerning tax collection. Carolyn Higginbotham (1996, 165) sees these as combining an administrative and cult function, suggesting that the fact that these taxation records were inscribed on complete bowls indicates that they may also have been used in dedications. Another sherd, from Tel Seraʿ, apparently contains part of a legal document, indicating that Egyptian script was also employed in this area. There is also a notable sherd from Lachish with a hieratic inscription of the Egyptian word for a writer, *sš*, followed by a name which is possibly non-Egyptian.53 Alongside this tradition of hieratic written in ink on pottery, it's not unreasonable to assume it was also in common use on papyrus, although this is not at present directly attested.

The emerging dichotomy in how the northern and southern Levant responded to their encounters with foreign literacy and adapted it for official use, which we observed in the Middle Bronze Age, is thus intensified in the succeeding period, with Egyptian presence in the south considerably complicating matters. While in Syria, the Middle Bronze Age flirtation with the Mesopotamian cuneiform tradition was cemented into a deeply ingrained institutional and bureaucratic adoption of it (often adapted for local purposes, as we'll see in Chapter 5), the situation in Lebanon and further south is characterised by plurality, hybridisation and relatively limited spheres of usage. Far from logosyllabic cuneiform and its associated Mesopotamian-derived writing traditions being the standard and pervasive form of official literacy, that position was both shared – by Egyptian hieroglyphic and hieratic – and complicated by the fact that the language written in the logosyllabic script was so heavily influenced by the local vernacular that it's doubtful whether it can be considered Akkadian at all. Lebanon seems to have been something of a middle-ground, without quite the

<sup>53</sup> Goldwasser 1991.

*Fig. 3.2. Distribution of Egyptian stone vessels with royal inscriptions of the 18th and 19th Dynasties. Data from Sparks (2003), Ahrens (2006).*

wholesale appropriation of Mesopotamian traditions seen further north, but also without an Egyptian administrative presence driving the expansion of hieroglyphic and hieratic. It is perhaps not surprising, then, that in this environment the local Byblian syllabic script seems to have persisted.

It's fitting to pause at this point and consider the social and ideological connotations of these various writing practices in different parts of the Levant and among different social groups. With the possible exception of Byblian pseudo-hieroglyphic, all the scripts known to have been used at a state or official level in the region were foreign-derived, and it'slikely thatthis affected how they were viewed and the cultural meanings people attached to the writing practices associated with them. These issues were likely particularly pronounced in the south, where Egypt was, to some extent, an occupying power. Although we have no direct testimony as to how Egyptians or their writing were viewed by local populations, it wouldn't be surprising if there was a measure of hostility. Attitudes were, however, probably heterogeneous. How did they vary in cities with strong Egyptian presences compared to those without? Did the elites have different perspectives on Egyptian culture than the general population?

#### *State and status in Late Bronze Age alphabetic writing*

This brings us to the final topic I want to explore in this chapter: the relationship between writing and elites, and the question of whether alphabetic writing might have been more widely used at a state level than the scarce surviving evidence may initially suggest.

There are only a handful of linear alphabetic inscriptions from secure Late Bronze Age contexts in the Levant. Israel Finkelstein and Benjamin Sass place the number at only four: a ewer, bowl and bowl fragment all from Lachish (respectively from Fosse Temple III, a tomb and a twelfth-century pit east of a large public building), and another bowl found in a pit containing LB III pottery at Qubur el-Walaida. Gordon Hamilton points out additional examples from Tell el-ʿAjjul and Megiddo.54 A few more examples have palaeographic features which allows for Middle-to-Late Bronze Age dates, but lack archaeological contexts that would allow for greater precision. Even including these objects, Finkelstein and Sass argue that there are only around seventeen possible linear alphabetic inscriptions from this period.

This Late Bronze Age linear alphabetic corpus displays both continuity and differences with relation to its Middle Bronze predecessor. Inscribed objects continue to be eclectic both in type and in the palaeographic characteristics of the script they bear. They're still written overwhelmingly on non-purpose-made artefacts rather than dedicated writing materials such as tablets or papyrus. Unlike earlier examples, however, pottery is now the most commonly inscribed material, and we also see changes in how the writing is applied: signs are often written with ink or paint rather than incised, as can be seen, for example, on a cup found at Tell el-ʿAjjul (Sparks 2013, 79).

<sup>54</sup> Ussishkin 1983, 155; Finkelstein and Sass 2013, 153; Hamilton 2014, 50.

The status and other social connotations of these inscribed objects and the texts they bear is hard to assess. In themselves, they don't appear to be particularly high-prestige items; however, find contexts and the nature of the inscriptions may suggest a degree of prestige or ideological significance. For example, the Lachish bowl was found in the potentially charged context of a grave (which was said to be 'wrecked'; no other details were given) and the ewer came from a temple. The latter's inscription was interpreted by Frank Moore Cross as marking it as an ex-voto. The Lachish bowl fragment and the Qubur el-Walaida bowl both came from later pits, so were clearly not in their original contexts of use. The former was associated with an important public building, however, and in the latter case, Cross again interprets the inscription as votive (though it only consists of two personal names, so we should be careful here).55

Should we, then, see writing as 'elevating' otherwise mundane items in culturally important settings? Or was the writing merely marking an object that was already significant for some other reason – such as its object history or role in ritual?

Alongside the uncertainties, there's a paradox here. The surviving examples are few, eclectic and hard to interpret. While there are sketchy signs that linear alphabetic writing may have played a limited ritual role, nothing in the available evidence directly supports the idea that it was used at an official level for administrative, epistolary and literary purposes comparable to those seen at Ugarit. For this reason, it's often taken for granted that it was a rather marginal, low-status phenomenon. Perhaps developments in alphabetic writing were the result of experimentation by literate intellectuals as a sideline to their main work in cuneiform, hieroglyphic and hieratic (which might explain why it seems to develop so slowly); perhaps it was wholly outside the sphere of official writing culture. Finkelstein and Sass (2013, 183) note that alphabetic inscriptions from the southern Levant are often found in similar locations to hieratic ones and suggest this implies that the spread of the alphabet was in some way linked to Egyptian activity. It's also possible that Egyptian activity may have given rise to writing practices in a less direct way, as Angelika Berlejung has suggested regarding the Qubur el-Walaida bowl. In her view, Egyptian presence did not always bring with it people (or enough people) competent in Egyptian writing practices, especially in 'peripheral' areas; in such cases sheer practicality may have prompted local writers to employ the linear script they were familiar with for the administrative activities required by Egyptian officials.<sup>56</sup>

And yet, by the end of the second millennium, linear alphabetic writing must have been well on its way to developing into the standardised regional scripts we know from the first millennium, such as Phoenician, and was on the cusp of official

<sup>55</sup> Starkey 1935, 202; Cross 1980, 1–4; Berlejung 2010. The status of Qubur el-Walaida in the Late Bronze Age is somewhat unclear. Cohen (1978, 195) plays it down, calling the site a 'small, unfortified village', but Berlejung has suggested that there was a large Egyptian administrative building on the site, pointing to a degree of regional importance.

<sup>56</sup> Berlejung 2010, 275–276 and personal communication (29 November 2019).

adoption by elites, if not already being used in this way. Reinhard Lehmann has noted that by the time of the first extant official texts in the linear alphabet – the royal inscriptionsfrom Byblos around the turn of the first millennium – Phoenician already appears to have been subject to at least several generations' worth of standardisation and calligraphic experience.57

Clearly, there's more going on here than is evident from the inscriptions that have been preserved. We return, then, to the themes we saw at the beginning of this chapter: is the alphabet to be claimed for the elite or the proletarian masses, or a mixture of both? As with the Early Alphabetic inscriptions of the Sinai, the crux of this question is what role we imagine for writing on perishable materials, the most likely locus for the unseen development and expansion that must be reconstructed to take us from the Middle Bronze Age situation to the early Phoenician one.

Let's examine the idea of officially adopted alphabetic writing first. If we hypothesise that Ugarit wasn't the only Levantine state where alphabetic writing was adopted by the official bureaucracy, then there are two main possibilities: first, that other polities were using linear scripts on perishable materials, and second, that they were using the cuneiform alphabet just as Ugarit was. The first of these is the more commonly suggested hypothesis, and has been advanced in particular by Alan Millard and José-Ángel Zamora López.58 Despite the lack of direct proof, there's a good amount of circumstantial evidence that supports this. In two articles,59 Zamora López has argued this on the basis of a comparison between the surviving linear inscriptions of the Late Bronze Age and objects inscribed in 'short alphabet' variants of alphabetic cuneiform from Ugarit and the surrounding region (as well as a couple of further-flung examples on Cyprus and at Tiryns in the Aegean).60 Zamora López distinguishes two classes of inscribed object within the alphabetic cuneiform corpus: those objects purpose-made for writing, such as clay tablets, which he calls 'supports spécifiques' and which constitute the overwhelming majority, and those objects that would exist otherwise and merely have writing added to them, which he calls 'supports marginaux'. These would include items like inscribed storage vessels, dedicated prestige objects, ostraca and suchlike.

Zamora López points out that the corpus of 'supports marginaux' in alphabetic cuneiform – a motley assortment of inscribed tools, weaponry and vessels, mostly using the unstandardised 'short alphabet' – is very similar to the surviving Early Alphabetic B corpus. The find contexts and contents of the texts also appear similar: inscriptions involving personal names are common, and several items seem likely to be dedicatory, such as a foundation-deposit of inscribed bronze adze- or hoe-heads

<sup>57</sup> Lehmann 2008; 2019. The development of alphabetic writing during the Late Bronze/Early Iron Age transition is discussed in more detail in Chapter 12 below.

<sup>58</sup> Millard 1991; 2007; Zamora López 2007.

<sup>59</sup> Zamora López 2006; 2007.

<sup>60</sup> On the features of these variants and how they differ from standard alphabetic cuneiform, see Chapters 1 and 5.

found alongside numerous other bronze items at the foot of a staircase in the House of the High Priest on Ugarit's Acropolis. The inscriptions read simply *rb khnm* – 'the High Priest'.61 If the official texts of Ugarit had been on a perishable material rather than clay tablets, and so had not survived, alphabetic cuneiform would display almost exactly the same pattern of surviving material as we see with the linear alphabet. The obvious purpose-made writing material for linear alphabetic that would correspond to clay tablets for alphabetic cuneiform is papyrus or some other perishable material. The use of paint and ink for writing linear inscriptions on ceramic and the likely use – at least at certain centres – of hieratic for administrative purposes would also be consistent with the idea of southern Levantine states using linear alphabetic writing on perishable materials in a similar way to Ugaritian use of alphabetic cuneiform on clay.62 Some additional support can be gleaned from the existence later of historical traditions and king-lists stretching back beyond the tenth century, or the reference to written scrolls detailing previous diplomatic agreements between Byblos and Egypt in the Egyptian account of Wenamun (see Chapter 12).

If Zamora López is right, then Ugarit's uniqueness isn't in what it did – the adoption and standardisation of alphabetic writing for official purposes – but how it did it. Given the apparent association between linear alphabetic inscriptions in the south and Egyptian hieratic (see above), that script might fulfil a similar role in southern sites to logosyllabic cuneiform at Ugarit: an older, more prestigious script associated with dominant external powers and used for certain kinds of administration, with vernacular writing practices adopting similar materials and methods of inscription, albeit in a different script and in different contexts of use. This is an intriguing speculation with much to recommend it, and potentially narrows the gap considerably between Ugaritian writing practices and those of the rest of the Levant, bolstering the idea advanced in Chapter 1 that we should see these as a single integrated mesh of writing practices despite the at-first-glance significant differences in visual appearance and usage between them. Ultimately, however, Zamora López's idea has the drawback that there is an almost complete lack of hard evidence to support it. We still have, and are likely to continue to have, no direct evidence of large-scale use of perishable materials for writing in the Bronze Age Levant. It's one thing to reconstruct the use of hieratic for official purposes when surviving inscriptions on non-perishable objects, such as the tax documents, point in that direction; it's quite another to do so for linear alphabetic writing, for which we have not even that limited evidence. As Sass points out, even bullae with impressions of papyrus are absent before the ninth century.63

<sup>61</sup> Schaeffer and Dussaud 1929.

<sup>62</sup> Following Zamora López's argument, the link could be that sites accustomed to using linearised hieratic on papyrus would be most receptive to also writing alphabetic scripts in this way. Hieratic might thus be seen as broadly analogous to logosyllabic cuneiform at Ugarit.

<sup>63</sup> Finkelstein and Sass 2013.

The second possibility – that Ugarit wasn't alone in its use of alphabetic cuneiform – has been argued by Benjamin Sass. He suggests that bureaucracies in alphabetic cuneiform or other scripts distinct from the linear alphabet existed across the Levant and have simply not been found. In his view, the alphabetic cuneiform 'short alphabet' inscriptions, which often have poor archaeological contexts, may actually date later than Ugarit's destruction and thus be evidence of an otherwise unattested alphabetic cuneiform phase of literacy between the end of widespread use of Akkadian and the assumption of the linear script.64 At Byblos, he thinks the use of the local syllabary may have persisted up to the ninth century. In both cases, the evidence is, I think, very poor. As Sass himself mentions, traces of signs in the Byblos syllabary underneath alphabetic inscriptions from the site were supposedly observed by Martin (1961), but have never been seen by anyone else.65 More problematically, if official use of alphabetic cuneiform was widespread across the Levant at the end of the Bronze Age or beginning of the Iron Age, surely we would have found some evidence of it. It's true that excavations at sites like Tyre or Byblos have been too small or too poorlyhandled to say for certain there was nothing of the kind there, but there are plenty of other sites that have been extensively excavated under modern conditions where we would expect Ugarit-style alphabetic cuneiform material to have been observed, had it existed. One thinks particularly of Sidon, Tell Kazel, Tell Dor and Aškelon.

This suggestion by Sass is really only necessary to fill the gaps introduced by his own idiosyncratic views about the geography and chronology of linear alphabetic writing.

While I don't think that Ugarit had a monopoly on alphabetic cuneiform, and wouldn't object to dating some of the inscriptions from outside the city slightly later than Ugarit's destruction, the reconstruction of large-scale administrative use of alphabetic cuneiform for other parts of the Levant is not tenable.

In the next chapter we will take up several of the issues that have arisen in this one. How did alphabetic cuneiform emerge and how does standardisation operate in script and writing practices? How do local scripts and languages interact with those promulgated by the 'globalised' culture of the region? What were the social issues underlying and resulting from these interactions?

<sup>64</sup> Sass 2005, 53.

<sup>65</sup> Sass 2005; Finkelstein and Sass 2013.

### Chapter 4

### Standardisation, vernacularisation and the emergence of alphabetic cuneiform

In the last chapter we described the development and spread of writing practices in the Bronze Age Levant and the emergence of multiple traditions that interacted in complex ways. A broad north/south dichotomy was apparent in the use of cuneiform and clay versus ink and perishable materials, connected with the differing degrees of influence from Mesopotamia (and possibly Mitanni) and Egypt, respectively. However, we also observed the possibility of similarities across the region, particularly in how these foreign influences coexisted with and helped shape local manifestations of pan-Levantine alphabetic vernacular writing practices.

In this chapter we turn our attention to Ugarit and alphabetic cuneiform more closely, and consider how, where and why script came into existence, and the processes by which it was adopted for official use within Ugarit's bureaucratic, religious and literary establishment. In the previous chapter I suggested that linear alphabetic scripts may well have been being used in similar ways in other Levantine polities, and evidence does not survive simply because of their use of perishable materials. If this is correct then the Ugarit case study might also provide an insight into developments in writing, culture and politics that were taking place around this time in many parts of the region. As we saw, assessing the likelihood of this is at present difficult because of the scanty evidence, but I will work here from the assumption that what occurred at Ugarit could *at least possibly* have had parallels in other coastal polities, and that this may even be considered likely.

Connected to this question of parallels is the issue of the permeability of the boundary between the northern 'cuneiform' world and the southern 'linear' one. There were strong cultural, political and economic links between cities in Syria and those in Lebanon and Israel, and records such as administrative tablets, trade agreements and extradition requests amply attest to the movement of people back and forth between then, either on a temporary basis or for long-term settlement. It seems inconceivable that these connections didn't bring with them awareness of writing practices. We know that alphabetic cuneiform variants were used at least to a limited extent in the south from the finds of 'short alphabet' inscriptions, and letters from Tyre and Byblos to Ugarit in alphabetic cuneiform may point to a more official engagement with the script.1 By definition, evidence for the use of linear writing on perishable materials in the north will be harder to identify, but it is clear that the influence of linear script on alphabetic cuneiform was not simply a one-off matter of providing the initial inspiration. We discussed in the Introduction how three additional signs were added to the alphabetic cuneiform repertoire some time after the creation of the original twenty-seven. One of these, 1 *s*́, is unmistakably derived in form from the linear s *s*. Together with the initial creation of alphabetic cuneiform, this gives us at least two instances of alphabetic cuneiform borrowing inspiration from its linear counterparts. I think it's very likely that linear forms of the alphabet were well known in Ugarit, not just as an initial catalyst for the creation of the cuneiform script, but on an ongoing basis, being used by visitors from the southern and central Levant if not by Ugaritians themselves. It may be, as we'll see, that there was somewhat more than this.

#### **The development of alphabetic cuneiform**

The first thing we need to consider is the question of where alphabetic cuneiform was created. Is the script so often known simply as 'Ugaritic' actually native to Ugarit at all? Arguments in favour of Ugarit as the place of invention essentially boil down to the sheer number of tablets found there, and the complete absence of any evidence of alphabetic cuneiform being used for official purposes anywhere else. But simple numbers of surviving inscriptions show only that Ugarit was a site at which this form of writing received a large-scale official implementation, not that it was an Ugaritian invention. Although Dennis Pardee supports Ugarit as the place of invention, he has also pointed out some ways in which the fit between the graphemic and phonemic repertoires is somewhat imperfect. For instance, the signs for *d*̱and *ẓ* are rare and often replaced by more common alternatives, which would be consistent with these not being fully distinguished phonemes in Ugaritic at the end of the thirteenth century. Either the alphabet preserves an archaic phonemic repertoire or the script was borrowed from speakers of a dialect with a slightly different range of sounds.<sup>2</sup> Where this might have been remains entirely mysterious; no plausible alternative locations for the creation of the script present themselves.

Dietrich and Loretz3 believe that alphabetic cuneiform was introduced to Ugarit by immigrants from southern Arabia. This is an attempt to explain the (at least) three different alphabetic repertoiresthat are attested forthe script. As well asthe 'standard' 30-sign version, a 27-sign version exists, which lacks the three supplemental signs

<sup>1</sup> KTU 2.38 is from Tyre; 2.44 from Byblos. KTU and others speculate that these might also be translations of Akkadian/logosyllabic originals. Although that is the usual form for international correspondence, it is not universal, and the use of alphabetic cuneiform in Phoenicia is not intrinsically implausible.

<sup>2</sup> Pardee 2007, 183–184.

<sup>3</sup> Dietrich and Loretz 1988; 1989.

*ʾi*, *ʾu* and *ś*, and there is also the 22-sign 'short alphabet'. It is extremely clear, and universally agreed, that the 30-sign alphabet is a later expansion of the 27-sign one. However, there has been more debate about whether the 27- or 22-sign alphabet is older: did Ugaritic have a large, inherited Semitic repertoire that was shortened in certain contexts, especially when alphabetic cuneiform was used by speakers of other north-west Semitic languages? Or did Ugarit share the same reduced repertoire as these other languages in the region, with the script being artificially expanded for other purposes? Dietrich and Loretz favour the latter, suggesting an intrusive Arabian elite as the source of this expansion, since later south Arabian writing features a similar large repertoire. The evidence for this is very poor, however. Alphabetic writing is not attested in Arabia until much later, there is no textual or archaeological evidence whatsoever of Arabian immigrants at Ugarit or of the establishment of a new, ethnically distinct elite dynasty at the time of the advent of alphabetic cuneiform,4 and the 'extra' signs are distributed randomly throughout the alphabet rather than tacked on to the end as would be expected of late additions and is the case with the signs already mentioned that were added to aid with writing Hurrian. It thus makes far more sense – as most other scholars agree – to see the 27-sign alphabet as original and the 22-sign repertoire of the first–millennium linear scripts as resulting from a shortening of this due to sound-mergers and orthographic simplification.

The place of alphabetic cuneiform's invention remains up for grabs, then. Ugarit is certainly a possibility, but unlike most scholars, I don't think it's the only one. The next question we need to address is *when* the script emerged. This too is a more complex question than it first seems. The dating of the earliest alphabetic cuneiform tablets from Ugarit has been subject to much debate, but a consensus has now largely emerged. Originally, they were believed to be contemporaneous with the earliest Akkadian texts to survive from the site – namely, mid-fourteenth century BC – with the script remaining in use until the city's destruction in the early twelfth century. The basis for this dating was the colophon of the famous Baʿlu Cycle poems, whose scribe, ʾIlimilku, says that he is working during the reign of King Niqmaddu. This was originally assumed to be the fourteenth-century Niqmaddu II. More recently, however, an additional text by the same ʾIlimilku has come to light among the assemblage of the House of ʾUrtenu, which is securely dated to the final years of Ugarit's existence. The majority of scholars have consequently lowered the dating of the Baʿlu texts accordingly, to the reign of Niqmaddu III. The earliest firmly dated alphabetic cuneiform texts now belong to the time of ʿAmmiṯtamru II, that is, the mid-thirteenth century BC. There is a fragmentary tablet, KTU 3.11, which includes a reference to

<sup>4</sup> Such a phenomenon might be marked by the appearance of new, Arabian-related onomastics, food practices, burial practices, religion or elite material culture. While there is not, of course, a simple one-to-one relationship between material culture and the ethnic identity or geographic origin of the user, a significant demographic, political and ideological rupture such as the intrusion of a new, foreign-born elite might be expected to result in some changes to elite culture, practice and material culture preferences, which ought to be archaeologically detectable.

the preceding king Niqmepaʿ. In its extant form, this is the only occurrence of this king's name that isn't part of a later ruler's patronymics. However, Pardee (2010) restores the name of Niqmepaʿ's son, ʿAmmiṯtamru, in the broken first line and sees this tablet as dating from his time as crown prince, probably towards the end of his father's reign. This may push the origins of alphabetic cuneiform back a little further than ʿAmmiṯtamru's own reign, but not by much.

This revised dating means that the alphabetic cuneiform script was actually in official use at Ugarit only for a rather short span of time – less than a century, at the very end of the city's existence. Official usage need not correspond exactly with the actual invention of a script, of course. In fact, it's more likely that the script existed for a while before being adopted for official purposes, rather than that it was 'commissioned'specifically forsuch uses and implemented immediately. There'slittle sign of diachronic change among the surviving inscriptions, and no evidence of an early or less standardised period, which we might expect if a new and unfamiliar writing system was introduced for large-scale use. Alphabetic cuneiform thus seems to have been a 'mature' writing system before it was pressed into the service of the Ugaritian state, even if not necessarily a very old one. If we (somewhat arbitrarily) estimate this 'prehistory' for the script at perhaps a generation or so, this would place its creation some time around the end of the first quarter of the thirteenth century, with an official adoption around 1250.

This pattern would fit fairly well with other examples of vernacularlanguages and associated scripts rising to take their places alongside – or supplanting – previously established 'high' languages and scripts that were not those of the majority of the population. Sheldon Pollock distinguishes between 'literisation' – the first written use of a vernacular language – and its 'literarisation' – its first use for high-prestige literary forms. He is at pains to point out that it often takes centuries between these two instances.5 Christian Novetzke concurs, saying of his Marathi case study, '[t]his is a revolution measured in centuries, not days or years, and it moves in line with the pace of everyday life: consistent, constant, but cautious of change too rapidly enacted'. A gap between the first use of alphabetic cuneiform and its adoption for elite purposes would make sense, then, but we should note that the fairly brief separation I am suggesting is well short of the several centuries Pollock and Novetzke reconstruct fortheir first-millennium AD Indian examples. Thisistrue even if we take the 'literarisation' of Ugaritic and alphabetic cuneiform not as the first appearance of official texts but the apparently slightly later use of them for 'literary' material such as ʾIlimilku's mythological poems.

To understand this apparent discrepancy, it is necessary to explore the phenomenon of vernacularisation in more detail. Vernacularisation can apply both to language and script, and, while they often go together and ideas are transferable from one

<sup>5</sup> Pollock 2006, 318. In the case of Kannada, for example, the separation is around 500 years (336). In Iceland, which Pollock considers a 'telescoped' example, it is still around two centuries (440).

to the other, as with other areas it's important to maintain a conceptual separation between script and language. Most discussions of vernacularisation have concerned languages, but I'm not the first to talk about the rise of alphabetic cuneiform in terms of vernacularisation. Seth Sanders argued that 'Ugarit seems to be the first known society to have produced a written vernacular literature, and to have created a writing system especially for it. […] The cosmopolitan scribes of Ugarit deliberately, and uniquely, made their writing system local',6 while Robert Hawley, Dennis Pardee and Carole Roche-Hawley stated that 'in fostering and implementing the development of a local alphabetic written tradition for the vernacular language, they were also insisting on their apartness, in affirming their specific regional and cultural identity with respect to their neighbors'.7 We'll return to this important matter of vernacularisation's connection to identity later on, but for now, let's pursue vernacularisation as a cross-cultural phenomenon and consider the different processes by which it can come about.

There are several well-documented examples of vernacularisation processes in history that can shed comparative light on what may have been happening in Ugarit, including early medieval Europe,8 China in the nineteenth and twentieth centuries AD, and, as we've already mentioned, the first-millennium AD examples from southeast Asia discussed by Pollock, Novetzke and others. In all these cases, as with Ugarit, a hegemonic or high language and its associated writing culture dominated writing practices and especially the production of prestigious literature: Latin, Classical Chinese and Sanskrit respectively. Not all of these high languages were promulgated through imperial expansion or colonial encounters, though Latin stands as a clear example of the potential of such interactions to bring about diglossic situations such as these.

Looking at these other examples of vernacularisation, we can discern two broad categories – top-down and bottom-up. In the latter, the importance of the vernacular grows gradually as a result of various social factors, often over a long time, before specific socio-political circumstances arise, which prompt usersto create a vernacular literature. This seems to be what we have, for example, in the case of Marathi, where Novetzke argues that the vernacular arose as part of a more general religious and philosophical valorisation of the everyday and non-elite among particular sects, in part despite, and in part because of,the indifference of the ruling elites.<sup>9</sup> By 'top-down' vernacularisation, I am thinking more of politically-motivated programmes promoted by the authorities that seek to encourage popular use of a vernacular where that has been previously marginal or declining. State initiatives in post-colonial countries like Rwanda or South Africa are good examples. These types of vernacularisation would seem to have the potential to be much quicker than bottom-up processes, but are

<sup>6</sup> Sanders 2004, 46.

<sup>7</sup> Hawley *et al.* 2015, 236.

<sup>8</sup> Smith 2000.

<sup>9</sup> Novetzke 2016.

not necessarily successful. As Nkonko Kamwangamalu argues (2013), for populations to respond positively to such governmental efforts and for the vernacularisation to succeed, non-elite populations need to perceive some advantage to be gained, often in terms of economic prospects, from switching to the vernacular that is being promoted. In modern contexts, this can place officially-promoted vernaculars at something of a disadvantage in relation to global languages such as English or Arabic.

I don't want to pretend these two categories are entirely mutually exclusive. To be sure, any successful process of vernacularisation ultimately involves the vernacular becoming an accepted standard to some extent, which implies ultimately a degree of official acceptance or endorsement even in 'bottom-up' cases. We can see an example of this in modern Greece, where the prestige form of the language, katharevousa, was artificially imposed in the nineteenth century and maintained a dominance in literary production – especially varieties perceived as more prestigious – over many decades. The vernacular, dimotiki, nevertheless remained the spoken language and its written use continued to expand for much of the later nineteenth and twentieth centuries, 'from poetry to nonfictional prose and thence to creative literature more broadly, from the latter to serious critical writing, and thence to scholarly, technical, scientific, and official writing'.10 Nevertheless, despite this fairly clear example of 'bottom-up' vernacularisation, itstillrequired an officialrecognition – in this case the 1976 linguistic reforms following the overthrow of the Colonels – for the vernacular to be seen as fully established as the expected written form.

Another case that exists on the border between 'top-down' and 'bottom-up' is that of early medieval Europe. In reality of course, this should be seen as countless smaller cases, but for the sake of brevity I will generalise. Vernacular writing had for a long time existed in Europe, particularly around the fringes of the Roman Empire, as seen, for example, in the Ogham script or Scandinavian runes, which adapted the idea of the Roman alphabet for distinctly local texts and object types.11 As the Romance languages increasingly diverged from written Latin, this acted as a stimulus for the emergence of new vernacular forms, and for the reconceptualisation of the Romance languages as new vernaculars rather than Latin. To this extent, vernacularisation in Europe was 'bottom-up', but again an official imprimatur conferred a degree of legitimacy – such asthe Carolingian reformsto the use of Latin, which codified the divergence ofspoken Romance from classical Latin.12 The role of the Church adds further ambiguity. As the dominant force in medieval literary production, it exerted considerable power over what was written and how it was written, and thus over the written use of vernacular languages. In some regions, such as Anglo-Saxon England, the Church seems to have been much more open to the idea of the use of the vernacular than in others. These are clearly complex matters which have been extensively discussed within medieval studies, and this is obviously not the place to delve into them in detail. But

<sup>10</sup> Hudson 2002, 31.

<sup>11</sup> Forsyth 2019; Heier forthcoming.

<sup>12</sup> Pollock 2006, 392.

this summary will, I hope, be sufficient to illustrate that the boundaries between 'top-down' and 'bottom-up' vernacularisation can be less than clear when power is not concentrated in the hands of government. In the Middle Ages, the Church was arguably at least as powerful as the state, arguably more so in the world of writing. So vernacularisation both pursued and hindered by various branches of that Church defies easy placement within a scheme such as this.

With these caveats established, I nevertheless do think that these categories can help us think through the process of vernacularisation seen in Ugaritian writing practices. The picture as it appears from surviving evidence – of vernacular written Ugaritic appearing in the record essentially overnight, fully-formed and with an official imprimatur – is not typical of most cases of vernacularisation, especially 'bottom-up'. We can push it slightly, as I've suggested, to allow for a relatively short, smaller-scale existence priorto official adoption, but the absence of any sign of firmly datable non-elite inscriptions before its bureaucratic use argues against a protracted period of grass-roots development, at least in the city of Ugarit.

There are two ways of squaring this circle, I think, and they're are not mutually exclusive. The first is that the official adoption of alphabetic cuneiform and the Ugaritic language didn't result from a bottom-up gradual expansion in the uses of the vernacular, but that it was a top-down project imposed by fiat by one of Ugarit's rulers, the likeliest candidate being ʿAmmiṯtamru II, perhaps during histime as crown prince, towards the end of his father's reign.13 The second possibility, which sits quite nicely alongside the first, is that the early development of this script didn't take place in cuneiform on clay tablets at all, but involved linear writing on perishable materials. ʿAmmiṯtamru's innovation, if it was indeed him, wouldn't have been the invention of the script *per se*, but of a cuneiform variant for writing on clay.

This may require a little more explanation. I'm not arguing here that alphabetic cuneiform is derived from the linear alphabetic writing practised elsewhere in the Levant. We know that to be the case and no one believes otherwise. What I'm suggesting is that the script may not really be a cuneiform one at all, in the sense that it takes the structure of consonantal writing and applies it to newly-created cuneiform signs, and that it is these signs that writers had in mind when they wrote. Rather, I wonder whether it may be a stylistic variant of a linear script that existed in Ugarit – a kind of typeface if you like – for writing on clay, and that the cuneiform signs that are preserved today are merely conventional simplifications of the linear signs their writers had in mind – somewhat like the difference between cursive and lapidary variants seen in other scripts.

There are, of course, no traces of linear alphabetic writing practices at Ugarit. There are, however, a number of peculiarities about the script's 'cuneiform' that would make more sense if the writers had in mind a more linear image of the signs they were

<sup>13</sup> Silvia Ferrara reached similar conclusions about the top-down nature of the emergence of alphabetic cuneiform (2019), for slightly different reasons.

*Fig. 4.1. Curves and strokes in variants of the ṯ-sign. From Ellison (2002, figs 1410–11; 1413–14; 1439). Reprinted with permission.*

writing. Looking at the numerous variations within individual cuneiform sign-forms, as presented by John Lee Ellison (2002), there is sometimes a sense that these result from attempts to render a theoretical sign-shape not primarily envisaged in terms of wedges. Particularly in less common signs – *i.e.*, the ones where habitual ways of rendering a linear prototype in cuneiform could be expected to be less ingrained and conventional – sign variants can include elements beyond the usual cuneiform wedge impressions: namely curves, strokes and circles. This is most apparent among variants of *ṯ* (Fig. 4.1).

In other cases, more typical cuneiform wedges are used, but not always in the usual right- or downward-pointing orientation favoured by Middle Babylonian. See, for example, e *š*.

I readily acknowledge that this is highly speculative, and I'm raising it as a possibility ratherthan a firm proposal. However, as we've already said, itseems very unlikely that linear writing was unknown in Ugarit. The idea that there might be a living, linear version of alphabetic cuneiform (if that isn't a contradiction in terms) being used on other surfaces in parallel to the primarily clay-based and prestige cuneiformised variant is not too far beyond the suggestions that have been made for a long time that the script's cuneiform signs were not arbitrarily chosen, but that at least some of them could have been derived from linear prototypes.14 Correspondences such as *h*̮, *h*, *m* and *b* in Table 4.1 and the apparently rather systematic correspondence between circles in the linear script and Winkelhakens in the cuneiform (see also Table 4.2 and

<sup>14</sup> Stieglitz 1971; Lundin 1987; Dietrich and Loretz 1988; Koller 2018, 4. Pardee (2007, 189) mentions that 'The combinations of wedges forming these signs would have been in imitation of a linear alphabet used in the region' and seems relatively open to the idea of linear writing taking place in Ugarit itself, although he underlines the impossibility of determining for certain whether this was the case: 'For absence of data, it is presently impossible to say whether this linear alphabet was actually used at Ugarit before the invention of the cuneiform alphabet.'


*Table 4.1. Suggested correspondences between the cuneiform and linear alphabets, according to Stieglitz (1971). Several signs are envisaged as rotating or flipping in accordance with the different writing direction.15*

Koller 2018, table 1), do seem plausible.16 In other cases, the parallels are looser or nonexistent and we must conclude that either these signs were so abstracted that they now bear little resemblance to their linear prototypes or that only some of the linear signs passed into the cuneiform repertoire, the gaps being filled by newly-invented, arbitrary signs.

Returning to the question of alphabetic cuneiform's adoption for official purposes, then, the most likely scenario, I think, is that ʿAmmiṯtamru II, or some other senior figure around the same time, decided that henceforth much of Ugarit'sinternal written

<sup>15</sup> Stieglitz is criticised by Millard (1979, 615) for his reliance on later, Phoenician signs for comparison.

<sup>16</sup> Although see Ellison (2002), who points out that the standardised renderings of many alphabetic cuneiform signs found in tablet drawings, transcriptions and sign lists often misrepresent the reality of how signs were actually made; in particular, several conventional Winkelhakens may in fact usually be diagonal wedges (the distinction between these two categories being a fine one and perhaps only important if one wishes to posit a particular development trajectory for one and not the other).


*Table 4.2. Correspondences of signs between cuneiform and linear alphabets, according to Dietrich and Loretz. Based on Dietrich and Loretz (1988, 102).*

(*Continued*)


business should be conducted in the local language, whereupon an existing linear script was adapted for this purpose. Its 'cuneiformisation' for use on clay tablets can be seen as a concession to the familiarity of cuneiform writing practices and the need to maintain a cuneiform bureaucracy anyway for the purpose of international correspondence. It would be considerably more efficient to reconcile the new language and script form to these existing structures and practices rather than to create a second, parallel 'workflow' based on a linear script and perishable materials.17 This functionalist, pragmatic interpretation is not the whole story, however, since the choice of a cuneiformised version of the script would inevitably be informed by, and give rise to, cultural and ideological considerations. The decision to engage with, and allude to, cuneiform writing practices in this way is a powerful statement as to the prestige of these forms of writing among Ugarit's literate elites and the audiences for whom their written material was intended, as well as of their ideas and ambitions regarding Ugarit's position within international 'globalised' networks of culture, scholarship and politics. We will explore these matters in more detail in later chapters.

#### **Standardisation**

The adoption of scripts by elites for official use often goes alongside a process of standardisation. We've already talked about the eclectic nature of the southern Levantine linear alphabetic inscriptions, with an implied contrast to the much more typologically and palaeographically consistent corpus of alphabetic cuneiform. This is not to say that alphabetic cuneiform lacks variation: there were multiple variants including combinations of 30-letter, 27-letter and 22-letter repertoires, *ʾbgḫd* and *halaḥam* sign orders, both left-right and right-left writing directions and significant variation in sign-forms. These are sometimes grouped into two varieties – the 'long' and 'short' alphabets – but the reality is that variation is possible across all these – and other – axes. Coupled with the small number of non-standard inscriptions and the often short length, which can mean it can be hard to tell whether the script used for a given inscription attests a given feature or not, it's often difficult to identify coherent varieties of alphabetic cuneiform with clear sets of consistent characteristics. Instead, once we move outside the large majority of tablets that confirm more or less to an official standard, we're faced with a cluster of essentially unique realisations of the idea of alphabetic cuneiform, which can be more or less closely related to each other, to linear writing practices and to 'standard' writing practices. We should not, then, group 'non-standard' inscriptions into a single 'short alphabet', but on the other hand, the idea of a standard does remain useful, since the vast majority of uses of

<sup>17</sup> The main difference is that she sees alphabetic cuneiform as created as a cuneiform script based on the model of a linear one, and most likely at Ugarit itself. The former might be the case, but I have also explained my reasons for wondering whether the cuneiform appearance might only be a schematised graphic rendering rather than a characteristic built into the script from the start. I see no reason why its development must have occurred at Ugarit.

alphabetic cuneiform do adhere to a certain set of norms – they're written left-to-right, on clay tablets, using – so far as can be determined – a 30-sign alphabet, which is usually written in *ʾbgḫd* order in abecedaries. Similar notions of a standard versus a much more diffuse cloud of variant practices also emerge as important when we consider the development of linear alphabetic writing practices – especially when contrasting the Phoenician of the tenth century and afterwards with the situation in the Late Bronze Age.

And it's not just for the scripts themselves that ideas of standardisation are hard to avoid. As we'll see in the next chapter, they are extremely important when considering the emergence of broader writing practices: cultures, traditions, educational curricula, norms and etiquette for international communication, and so on. In this chapter and the last, we've examined how alphabetic vernacular scripts developed in relation to 'high' globalised scripts, writing practices and languages. But in order to understand how writing practices become global, and thus the nature of the context in which these alphabetic writing practices existed, we need to consider what standards are and how standardisation takes place.

Standardisation in writing has generally been approached using the same methods as linguistic standardisation. The traditional model for the emergence of a standard language or register was formulated by Einar Haugen (1966) and consists of a four-part process of *selection*, *codification*, *acceptance* and *elaboration*. It's questionable, however, how appropriate such linguistic models are for writing. As we've already argued in this book, writing is not an extension of language but a variety of human culture and practice in its own right. There's no reason why it should work in the same way. What's more, by focusing on these abstracted processes, we risk exacerbating the depersonalised and systems-focused tendencies of both writing practice research and linguistics in general. The emergence of standards in writing practice isn't just a matter of agreeing upon universally-recognised character-forms but one encompassing writing materials, education systems, social identities, religious beliefs and practices in which inscribed objects might be offered, and so on. It is, as discussed in Chapter 2, wholly entangled in the great mesh of human social practice and culture.

As such, and particularly because writing culture at Ugarit owed so much to the influence of the cultures and societies to which it was connected, we're better off looking to research into the formation of material culture koines, cultural homogenisation and the emergence of standards and social norms within globalised networks. In particular, Grewal's (2008) concept of 'network power' usefully accounts for processes of cultural convergence, the emergence of standards and the role of power dynamics and agency within these. The key element of this idea is a simple one – that as networks grow, their standards of practice, behaviour and thought can exert power encouraging people to participate and to conform, quite apart from any perceived or real benefits to the practices themselves. Certain kinds of behaviour or standards are simply a requirement if people want to interact with the growing numbers of others already integrated into the network. The outcome is a kind of snowball effect, where

a certain practice can be so widely practised it reaches a tipping-point and becomes essentially mandatory. Crucially, however, this is not merely a simplistic matter of numbers. Grewal pays close attention to questions of agency and the diverse ways in which standards can take hold. Initially, standards can spread through reason, force or sheer chance – or by a combination of several of these. Thus, people might adopt a standard because they discern an advantage from doing so, because they are compelled to, or for reasons that are essentially arbitrary. The meaning of words is an example of the latter: there is no practical reason why the word 'dog', for example, should mean what it does in English; but for whatever reason, people have agreed that it does. Once standards become established, however, they exert power. People can choose to adopt them or not; they can choose whether to use them as a supplement to their existing practices or to replace them. If they do so, they become incorporated into the network of existing users. But the more cultural power a network has, the more widely accepted its standards are, the greater the extent to which it will necessarily constrain and condition the choices people can make, even if they have no particular interest in the practices themselves. If people want to interact with others, they are compelled to adopt the appropriate standards, whether they want to or not. The importance of interaction supersedes consideration of the actual nature of the standard: you do it, because that's what everyone does.

We can see Akkadian and logosyllabic cuneiform in these terms for the ancient Near East. While not a single integrated political entity, the region was nevertheless a single network united by a communications standard involving that language and script. The more places that adopted this standard – whether by force or by choice – the more powerful and attractive it became.

In the case of vernacular alphabetic writing practices, we're dealing with much smaller-scale networks, but because of the close association of alphabetic cuneiform with the Ugaritian elite, also ones with much more local, immediate power. State officials were in a position to mandate the practice of their subordinates and related personnel. The same goes for teachers and their apprentices. Existing power structures and the small numbers of writers involved make the imposition of standard by 'force' relatively easy and practical. Or else, writers could choose to adopt the styles and standards used by favoured people in order to make their own work more acceptable to them, to facilitate their own access to the circles of power. We don't have the evidence to reconstruct specifics, but some combination of imposition and emulation is likely to explain how a standard variety of alphabetic cuneiform became established at Ugarit.

#### **Social questions**

In the final portion of this chapter I want to talk about the social implications of the relationship between cuneiform and linear manifestations of alphabetic writing that I've proposed in this chapter, namely that they were co-present, persistently interrelated and essentially a single socio-scriptal phenomenon. From an epigraphic and palaeographic viewpoint, this makes sense, but we shouldn't lose sight of the pronounced visual difference between the linear and cuneiform script variants. There can be no expectation that people literate in one variant should be able to read the other (though it's probable a few specialists were proficient in these and other variant practices). This mutual inaccessibility was likely even more pronounced when we consider how the scripts were received by non-literates: would the general population who could read neither have perceived the cuneiform and linear alphabets as in any way related? More so than, say, alphabetic cuneiform and its Mesopotamian logosyllabic counterpart?

For all that I'm advocating seeing these not as two separate writing systems but differing manifestations of the same scriptal phenomenon, this isn't necessarily how they would have been viewed at the time, especially by those without detailed specialist knowledge. In this respect, it makes sense to see Levantine alphabetic writing practices as an example of digraphia – that is, the availability of two or more scripts for the same language in the same place.18 Of course, the actual situation is rather more complex because of all the other scripts and languages also in the mix. For now, though, let's restrict our attention to the alphabets.

Digraphia is still a relatively new concept: the written counterpart to the much longer-established and more studied field of diglossia or bilingualism.19 However, as Peter Unseth has shown in a pair of important articles, many of the same concepts apply.20 Often, scholars have treated the choice to use a cuneiform alphabet in Ugarit in an essentially functionalist manner: it was motivated by the need to operate within the Hittite sphere of influence or out of a desire to render the idea of the alphabet as convenient and user-friendly as possible to their existing cadres of Akkadian-trained scribes. But what emerges very clearly from Unseth's work and other studies of digraphia21 is that social and cultural concerns are far more important than the strictly

<sup>18</sup> Essentially since Ugarit was first excavated, there has been an ongoing discussion about the degree to which its culture, religion and language ought to be considered 'Canaanite'. Linguistically, the matter of classification is arguably more important to grammarians and taxonomists than it is to us, since everyone agrees that whether we label it 'Canaanite' or not, Ugaritic is certainly very closely related to the Canaanite dialects and especially the coastal variants from Phoenicia, though naturally it attests certain more archaic featuresthan its first-millennium counterparts. The debate issummarised in Buck 2018. For the opinion that Ugaritic is not Canaanite, see Goetze 1941. For the counter-argument, see Tropper 1994, with further references. Regardless of where we draw the (always essentially arbitrary) dividing-lines between our different labels, the closeness of Ugaritic's relationship with the language of the central and southern coastal Levant justifies treating them as essentially dialects – most likely extremely mutually intelligible – rather than wholly separate languages. More to the point, for our purposes the finer points of modern linguistic taxonomy are less interesting than how the divisions between language were drawn at the time.

<sup>19</sup> On diglossia, see Hudson 2002, with further references.

<sup>20</sup> Unseth 2005; 2008.

<sup>21</sup> Literature is still rather sparse and the contributions to Stéphanie Grivelet's 2001 edited volume of the *International Journal of the Sociology of Language* (Issue 150) remain an important starting-point and collection of case studies (Grivelet 2001 and other contributions to that issue).

utilitarian in determining which scripts people choose to use and in which contexts. Discussing modern digraphia in south Asia, Robert King puts it well:

The script differences in 'typical' cases of digraphia almost always mask profound differences both linguistic and societal: in grammar and vocabulary, in cultural orientation and often in religious orientation as well, in history, in style and preferences for different literary genres, in way of life and sensibility. Digraphia is not unlike the proverbial ten percent of an iceberg that is visible above the water: the part that is visible – the script – is the least of it.22

This is at odds with a general scholarly approach that has tended to stress Ugarit's cultural commonalities with the rest ofthe Levantratherthan the differences. This has its root in the earliest period of Ugaritic studies, when the clear similarities between the newly translated Ugaritic myths and Biblical passages was bringing the site to international attention. Since then, a major strain of Ugaritic scholarship has been interested in its potential as a comparandum for understanding the first-millennium southern Levant.23 I did this myself when studying Late Bronze/Early Iron Age Phoenicia in my doctoral dissertation.24 Ugarit's unique documentary record makes it an irresistible source from which we might draw to help fill some of the lacunae in our understanding of its neighbours. This is fine, and a valid research methodology. But focusing primarily on similarities risks us overlooking the potential for important differences, especially when the material we are comparing is not equivalentin quality, quantity or date.

Rainey showed as early as the 1960s that at least some people in Ugarit apparently didn't consider themselves 'Canaanites'.25 KTU 4.96 is a list of people specified according to either patronymic in the case of natives, or ethnic in the case of foreigners. Alongside a man labelled as *mṣry* – Egyptian – there is one called *knʿny* – a Canaanite. Rainey argues, and most have agreed with him, that this shows that 'Canaanites' were treated asforeignersinUgarit, asthey apparently were too in nearby Alalaḫ. Our understanding of ethnicity and other forms of social identity has of course developed a great deal in the more than half a century since Rainey was writing, and we are now much more awake to identities being situational and shifting, and to the possibilities that groups might define themselves differently from how they are defined by others. In Amarna letter EA 151, for instance, the king of Tyre – writing to the Egyptian pharaoh – seems to use a definition of 'Canaan' that does include Ugarit.26 Evidently, the meaning of the term was as slippery and dependent on the user in antiquity as it is in modern Near Eastern studies.

<sup>22</sup> King 2001, 44.

<sup>23</sup> See, for instance, Smith and Bloch-Smith 1988; Schloen 2001; Amico Wilson 2013; or contributions to Brooke *et al.* 1994.

<sup>24</sup> Boyes 2013.

<sup>25</sup> Rainey 1963.

<sup>26</sup> Lemche 1998.

The point, again, is not whether we decide that this label is or is not appropriate for Ugarit's culture, but to consider the possibility that Ugarit's relationship with its neighbours may have been contested and shifting in ways that we cannot hope to fully retrieve based on the evidence as it currently stands. There is a possibility too, that the situation changed over time, since EA 151 likely predates KTU 4.96 by a century or so. The role of writing practices in this is impossible to pin down with any certainty, except to say that modern examples of digraphia show that the choice of script has extremely strong symbolic value and can provide an important focus for the crystallisation of social identities. Even if there is not initially a strong or straightforward correlation between script use and a given social identity, these can develop – or the perception of such a correlation can. Thus, in his south Asian case study, King highlights how the importance of language and script as axes of the ethnic, political and religious rivalry between Hindus and Muslims is both highly contextual and has become more rigid within observable timescales. At street- or village-level, he describes, spoken Hindi and Urdu can be entirely mutually intelligible, almost indistinguishable; and yet within more urban or socially prestigious contexts this common ground evaporates as specialist vocabulary and intellectual concepts become more marked. The strong sense that Hindi/Devanagari corresponds to Hindu identity and Urdu/Nastaʿliq to Muslim has become increasingly pronounced as the rivalry and antipathy between the groups increased around and following the end of British control. Motilal Nehru, the grandfather of India's first prime minister and a Hindu, spoke and wrote Urdu as a first language, and Urdu was even known as 'Hindi' until around the eighteenth century.27 It's not difficult to imagine a similar situation in the Levant, where the Ugaritic and Phoenician or Hebrew dialects were in many contexts highly interchangeable and weakly differentiated, but where ethnic associations might be much more marked in certain contexts or at certain times. Likewise, the reasons for the adoption of a cuneiform rather than linear alphabet may well have had to do with distinguishing Ugarit – or a certain intellectual, political or religious group within Ugarit – from some other; or this may not have been the original reason, but it may still have become an axis for such differentiation as time went on. Even the short timescale of a few decades, for which alphabetic cuneiform was in use, is more than enough time for important developments in social identities and changes in how people or groups viewed themselves with regard to their neighbours.

Throughout this chapter and the last, we've shown that when it comes to writing, and especially alphabetic writing, the degree of integration and difference between different parts of the Levant is complex and dynamic. At times it has made sense to think in terms of a north–south divide in which, crudely put, the north embraced writing early and looked mainly to Mesopotamia for inspiration, while the south was more reticent and orientated more towards Egypt. At others – especially when it comes to alphabetic writing practices during the Late Bronze Age – there seems

<sup>27</sup> King 2001.

to be a high degree of integration and intermeshing of developments. And yet even then, alphabetic cuneiform looks very different to itslinear counterpart; both visually and in terms of its writing practices it can feel much closer to Mesopotamia than to the southern Levant, and this is likely to have had significant social implications for how literates and non-literates alike from all around this region thought about, developed, used and did not use these scripts.

These chapters have traced the background of writing in the Levant, and the spread of alphabetic writing practices from Egypt to Syria. It's been typified not by rigid taxonomies, formalism and standardisation, but ambiguity, patchiness and heterogeneity. I'd like to suggest that this isn't entirely down to the parlous state of the available evidence, but reflects a Bronze Age writing culture in the Levant that is similarly messy, not just when it comes to the alphabet, but also for what we tend to think of as the more established, standardised and official scripts too.

Neither alphabetic cuneiform nor linear writing constitute coherent singular entities: both exist in multiple variants, some of which can be much closer to variants of the other broad type than to other examples of their own. It's difficult to prove that there was state use of linear alphabet(s) in the Late Bronze Age but we've seen that this is at least a possibility. It's very possible that linear and cuneiform alphabets may have been in use at the same sites at the same time (though by the same people?) and some of the cuneiform alphabetic signs may be simplified versions of the same ones used for linear writing. Certainly it's also possible that none of these things were true, though I think this would be surprising. The key thing is that alphabetic cuneiform should be considered part of the same general mesh of alphabetic practices as the linearinscriptions (as most would agree) and probably differed from them (or certain realisations of them) less than it might at first appear. We should explore its development and use not as an ingenious and idiosyncratic scriptal dead end, but in terms of how its users adapted and manifested the general ideas of alphabetic writing in their own specific social contexts and for particular practical or expressive purposes.

### Chapter 5

### Influence and innovation: networks of writing practice and culture

As we've explored the history of writing in the coastal Levant over the last two chapters, a key theme has been how polities such as Ugarit are integrated into wider networks of practice and culture. We've seen how ideas of writing, as well as specific scripts, have spread through networks of political, commercial and cultural interaction. On several occasions we've talked about 'writing cultures', without really delving into what we mean by this. We've also discussed ideas of standardisation, standards and the existence of variant writing practices that can carry important ideological connotations. In the ancient Near East, one of the most important manifestations of this theme is the interplay between regional standards of writing practice – such as the use of logosyllabic cuneiform and the Akkadian language – and local, often vernacular forms. For our purposes, the most important example of such vernacular writing is alphabetic cuneiform. As we saw in the last chapter, the relationship between this and logosyllabic/Akkadian is complex, and includes endorsements and adaptation of certain writing practices – such as the cuneiform method of making signs or the use of clay tablets – as well as the choosing of an alternative script. In this chapter we'll explore this in more detail, considering the nature of 'cuneiform culture' in the Late Bronze Age Near East and the diverse forms in which it was engaged with and participated in from place to place. We will see that, while the writing practices that resulted in Ugarit incorporated diverse elements drawn from this globalised writing culture as well as innovations in a blend which is highly particular to that site, this broader pattern of adaptation, localisation and hybridisation is by no means unique to Ugarit.

#### **Deconstructing 'cuneiform culture'**

Mesopotamia was not a single political or cultural entity and it didn't have a single monolithic set of writing practices. There was great diversity in cuneiform writing practices and associated culture, across both time and space. This is true even if we confine ourselves just to writing on clay tablets, which makes up the overwhelming majority of surviving Near Eastern Bronze Age written material and has understandably dominated scholarly discussions; but there's a risk that the dominance of the clay tablet can eclipse other forms of writing that we know existed. A dizzying array of objects, materials and media were inscribed in ancient Mesopotamia, necessitating a broad range of techniques and tools, and each having their own lives, afterlives and relationships with people, places and other objects. Some of these are also found at Ugarit; others are not. When we think about how the writers of Ugarit – and those who made use of their services – drew upon, adapted and transformed writing traditions from other parts of the Near East, it's essential that we consider this diversity as fully as we're able to reconstruct it.

The Late Bronze Age (or Middle Babylonian/Assyrian period in Mesopotamian terms) – is relatively poorly attested in terms of Mesopotamian written material compared to those preceding and succeeding it. This is probably not a real indication for any significant decrease in writing during thistime, butratherresultsfrom chance imbalances in survival and lesser interest in excavating and studying Late Bronze Age materials compared to the Old Babylonian and Assyrian periods. Many of the largest and most significant collections of cuneiform-inscribed material at this time come from so-called 'peripheral' sites such as Nuzi, Emar and of course Ugarit itself. This means that the writing culture and wider cultural context are in many ways highly comparable to what we see at Ugarit, and allows us to explore how different polities tackled similar issues in different ways – for example, the impact of incorporation into the sphere of influence of a larger imperial power. But it also makes it difficult to ascertain how the writing practices at these centres might relate to a notional 'core' of practices regarded as more or less standard.

The idea of cores and peripheries can be unhelpful in Assyriology, especially the tendency to lump together anything not from Babylonia or Assyria proper as 'peripheral Akkadian', which has often been seen as in some way substandard or of lesser importance.1 Concomitantly, Near Eastern writing culture has frequently been presented as fundamentally originating in Babylonia,2 an idea summed up by Oppenheim's(1960) notion ofthe 'stream oftradition', an unbroken flow ofscholarship and scribal practice beginning with Sumerian literature and evolving over thousands of years. This privileging of Babylonian literate culture is not without ancient precedent. Accounts from Late Bronze Age Assyria, such as the Epic of Tukulti-Ninurta I, make clear the prestige of Babylonian culture and especially its writings.3 The Epic describes with great pride how, after the conquest of Babylon in the thirteenth century, tablets were brought back to Assyria in large numbers:

<sup>1</sup> Core-periphery interactions are typical of world system frameworks of analysis. See Chapter 2, n. 38 on these, and why relationships of this kind can more usefully be conceptualised in terms of networks.

<sup>2</sup> For discussion, see Foster 2015.

<sup>3</sup> On the treatment of Babylonian literature in Assyria, see also Veldhuis 2012.

Secret[s of …], tablets of […], the corpus of scribal ar[t … ], the corpus of incantation … […], *eršaḫunga*-lamentations according to … the corpus of diviners …, the plans of heaven [and earth …] medical prescriptions according to the lore of the physicians (including) the procedure for band[ages … ] … There was not left any [… ] in the land of Sumer and Akkad (sc. Babylonia).

The abundant profit of the secrets (or: treasures) of the king of the Kassi[tes … ], he filled boats with (rich) yields for (or: …, to) the god Assur [… ].<sup>4</sup>

Dominique Charpin has compared the resulting situation in Assyria to that of Rome after its conquest of Greece – while militarily and politically supreme, the conquest opened the doorto an influx of highly-prized Babylonian culture into Assyria, recalling Horace's famous remark that 'captive Greece captured her savage conqueror and taught the arts to rustic Latium'.5 By the first millennium, this had progressed even further: the eighth-century Assyrian king Sargon II apparently refused to accept letters written on skin in Aramaic, despite that by this point being essentially the vernacular language of his kingdom: the correct form was in Akkadian and on clay: a definitive statement of the entanglement of writing practice, identity and the prestige of the Babylonian tradition.6

It would be helpful to be able to compare writing practices in Babylonia to those in other parts of the Near East to get a better sense of exactly what the similarities and differences were in this period, whether this prestige extended beyond Assyria to other polities in the region, and if, as is likely, it did, then how exactly it played into their own practices. Were historical Old Babylonian traditions more prestigious than contemporary Middle Babylonian practices, for instance? To what extent was this seen as a coherent and fixed tradition, or were changes acknowledged and even welcomed? Instead, in the absence ofsignificant amounts of contemporary Babylonian scribal material, scholars have mostly used Old Babylonian writing culture when trying to assess how so-called 'peripheral' polities responded to and adapted the larger cuneiform tradition. This is clearly still an important comparison to make, and one that can illuminate the processes of cultural interaction and adaptation that were taking place, but it is also problematic in that it flattens out chronological change as well as the potential variation within the many-faceted thing that is Mesopotamian cuneiform culture.

#### *Local variation in cuneiform practices*

When we focus our attention on the cuneiform of the Late Bronze Age, we find that idiosyncrasy is the order of the day. Nuzi, Emar and indeed Ḫattuša (since the Hittite adaptation of cuneiform writing also needs to be considered here despite – and even because of – the language difference) all have very distinctive individual spins on

<sup>4</sup> Machinist 1978, quoted in Fincke 2003–4, 124, n. 108.

<sup>5</sup> Charpin 2010, 213; Horace *Epistles* 2.1.156.

<sup>6</sup> Charpin 2010, 95.

cuneiform practices. As at Ugarit, local linguistic, social and cultural factors motivate the creation of alternative traditions alongside the globalised Akkadian used for international communication.

Nuzi lay east of the Tigris and was a secondary centre in a kingdom centred at Arrapḫe (modern Kirkuk). Although archives have been found at both Arrapḫe itself and the nearby site of Tell al-Faḫḫar, Nuzi has produced by far the most extensive collection of written material, which dates up to its conquest by Assyria around 1325.7 We cease to have writing from Nuzi, then, around the same time it first appears at Ugarit, with only a very small window of overlap. Nevertheless, it's still close enough chronologically for useful comparison, especially since we have every reason to believe that Akkadian and cuneiform are likely to have been used at Ugarit well before the first surviving texts from the reign of Niqmaddu II.<sup>8</sup>

Like at Ugarit, the population of Nuzi was not primarily Akkadian-speaking. Instead, the local vernacular was Hurrian (for which, at Ugarit, see Chapter 9), which had a significant impact on the language of the cuneiform texts, to the extent that they have largely been studied by specialists rather than mainstream Assyriologists.9 This 'Hurro-Akkadian' has been seen by some as a contact language similar to the Canaanite-influenced Akkadian (or Akkadianising Canaanite) ofthe southern Levantine Amarna letters, and parallels similar phenomena for earlier periods in other polities with mainly Hurrian-speaking populations, such as Alalaḫ and Qaṭna.10 Another distinctive feature of the written material at Nuzi is the range of genres: the recovered material is overwhelmingly administrative and utilitarian, with little to no monumental, scholarly or literary writing. Nicholas Postgate compares this situation to that of Middle Assyrian provincial and private archives,11 but it's very different from that of, for example, the Assyrian state archives filled with Old Babylonian literary culture. It differs too from what we see at Ugarit, which, as we've already seen in this volume, has its own generic preferences, including a clear association of certain types of material with either alphabetic or logosyllabic script. We see variation too in the division of written material between state and private archives: at Nuzi, the latter greatly outnumbers the former. The reverse is true at Aššur and the situation at Ugarit appears to be somewhere in between.

Moving west, we turn now to Emar, on the bend of the Euphrates in north-central Syria. Although still considerably further inland than Ugarit, Emar is generally seen as presenting one of the closest comparanda, especially in terms of writing practices. Its surviving corpus of written material is substantially smaller, at around 1000 Akkadian tablets, but is contemporary with Ugarit's. Emar came under Hittite control around the same time as Ugarit, so in this respect too it makes a good case study. We shouldn't underestimate the significant cultural and political differences, however.

<sup>7</sup> For a detailed recent survey of the written material from Nuzi and a comparison with Middle Assyrian writing, see Postgate 2013, 343ff.

<sup>8</sup> Boyes 2019a.

<sup>9</sup> Postgate 2013, 344–345.

<sup>10</sup> Andrason and Vita 2016.

<sup>11</sup> Postgate 2013, 370–371.

There is good evidence that Emar had a rather idiosyncratic political structure prior to the Hittite conquest, with a greater emphasis on its assembly and elders than the king. Although a royal dynasty was established – possibly by the Hittites in an effort to bring the city more into line with the political norms of their other vassals12 – it appears that assemblies and vestigial tribal structures of governance may have continued to play an important role in the administration and culture of Emar.13 Once incorporated into the Hittite Empire, imperial administration of Emar seems to have been markedly less hands-off than that of Ugarit, with Hittite officials and their direct involvement in administration a fixture of the city's political life.

As with Nuzi, written material from Emar is highly idiosyncratic, although in a rather different way. Scholars have identified two distinct traditions of writing at the site, which they term 'Syrian' and 'Syro-Hittite'. These ethnically associated names are unhelpful, but the dichotomy in writing practices they signify is real.

The 'Syrian' tradition at Emar seems to be older: establishing a chronology for the written materials from the city is complex, but it appears that the 'Syrian'-style texts start earlier and remain in use longer.14 This is consistent with the appearance of the cuneiform used, which in general is closer to Old Babylonian models. Tablets tend to be orientated vertically (that is, in 'portrait' orientation). The 'Syro-Hittite' tradition begins later – around 1275 – and lasts until the destruction of the city about a century later. The palaeography is more 'modern' in style and there is a preference for horizontal ('landscape') orientation of tablets, which is quite unusual in the Near East at this time. The two traditions also differ in several stylistic conventions such as how tablets are headed, how 'bullet points' are written in lists, typical sealing practices, and certain choices of vocabulary.15 There are differences in who uses each set of practices and for what purposes: the 'Syrian' mode is preferred by the state and certain civic institutions, as well as many private citizens, whereas 'Syro-Hittite' tends to be used by the prominent family of diviners, the Zu-Baʿlas, certain scribes, and in some texts relating to Hittite imperial functionaries.

From this schematic outline it would be tempting to assume that the advent of 'Syro-Hittite' writing practices at the site was connected with the Hittite takeover, but the reality seems more complex, with certain features of this 'tradition' beginning to appear even in early 'Syrian' texts before the arrival of the Hittites.16 It may be more accurate to say that Hittite influence and preferences may have accelerated and promoted certain changes in writing practices that were already underway, creating a polarised scribal environment in which one's choices in sign-form, formatting and style are likely to have been charged with symbolism regarding one's political, social and cultural identity.

<sup>12</sup> Beckman 1995, 29; Adamthwaite 2001, 201–203; but contra Otto 2014, 39–41, who claims that numerous textual records attest the existence of a king and palace at Emar from at least the fourteenth century, before the Hittite takeover.

<sup>13</sup> Fleming 1992.

<sup>14</sup> Cohen and d'Alfonso 2008.

<sup>15</sup> Cohen and d'Alfonso 2008; Faist 2008; Y. Cohen 2009; C.H. Cohen 2012.

<sup>16</sup> Cohen 2012.

It's very tempting to draw parallels between this dichotomy in writing practices at Emar and the emergence of alphabetic cuneiform at Ugarit. The two phenomena occurred at exactly the same time and in both the relationship with external powers – namely the Hittites – seems to have been an important consideration in which script was chosen. There are clearly significant differences too, however, not least in the relationship between localism, language and tradition. Unlike at Ugarit, this polarisation of writing practices didn't centre around an innovative process of vernacularisation, but instead was markedly conservative. Just like Ugarit, the locally spoken language at Emar seems to have been West Semitic, but there was apparently no desire to write this down, even in logosyllabic cuneiform, let alone in a newlydeveloped script. Instead, the practice preferred by the civic institutions for internal purposes was traditional and 'Mesopotamianist', aligning itself with the Old Babylonian legacy. This points to the markedly different scriptal and cultural contexts of Emar and Ugarit, despite their superficial similarities. The climate of home-grown scriptal innovation and experimentation of the coastal Levant did not exist for inland, Euphratene Emar; instead, the pull of traditional and prestigious writing practices associated with trans-Mesopotamian culture and especially with the Babylonian inheritance stood in opposition to a more innovative and modernising writing culture promoted, or at least preferred, by the conquerors from the west.

This brings us to the Hittites themselves. The history of writing at Ḫattuša and in the subsidiary Hittite centres is of course much longer and more complex than can be dealt with in detail here, but its broad strokes paint a very similar picture to that seen in our other examples. Cuneiform was adapted into a specific local tradition, which existed parallel to practices that fell within the generally recognised regional norms and were used mainly for international diplomacy.17 As at Ugarit and to a lesser degree Nuzi, the most striking feature of this adapted local tradition was linguistic: the use of the local language rather than conventional Akkadian. In the Hittite case, of course, matters are complicated by Ḫattuša's status as an imperial metropolis. As such, its 'vernacularisation' of the cuneiform tradition took place within a very different set of power relations to that seen at non-imperial centres like Ugarit. The massive difference in the scale of the polity also made for a much more complex internal linguistic and scriptal situation, most clearly demonstrated in the existence of a third major set of writing practices – hieroglyphic Luwian. It's difficult to assess the extent of Luwian-speaking within the Hittite territory, but it has been argued that the language gained in popularity throughout the Late Bronze Age until by the end of the period it may even have been an important linguistic influence among the royal elite.18 Certainly Luwian hieroglyphs were adopted by the elite for monumental and sphragistic purposes alongside Hittite and Akkadian cuneiform. Arguably, then,

<sup>17</sup> Mark Weeden (2016) has argued that the Akkadian of Ḫattuša is essentially Syrian in inspiration, drawing parallels with the Middle Bronze Age texts of Alalaḫ. While there are a few fifteenth-century texts from Alalaḫ, no archives survive from the period of Hittite domination in the latter part of the Late Bronze Age (Pedersén 1998).

<sup>18</sup> Yakubovich 2008; Payne 2010.

Anatolia saw two vernacularisations of writing practices, offset chronologically – the first adapting logosyllabic cuneiform for the local language; the second displacing it for the wholly Anatolian script and language of Luwian. Tellingly, it was this latter that survived among the north Syrian states of the former Hittite sphere of influence into the first millennium BC.

#### *Norms and etiquette in writing*

The picture I've been painting so far in this chapter is of a regional 'cuneiform culture', which was not a singular thing with an epicentre in Babylonia. By the Late Bronze Age, it was a network comprised of countless locally specific takes on similar ideas and practices, as well as some of the same key texts, genres and so on. This strikes a chord with the networks of globalisation discussed in Chapter 2. In such terms, cuneiform writing practices constitute a 'global culture' within the interconnected world of the ancient Near East. This perspective allows for a view of local writing practices which takes account not only of the prestige and power of the network as a whole, its standards, and of specific important centres within it – such as Babylonia – but also the agency of local populations and the complex dynamics of their engagement with both the idea and reality of Late Bronze Age connectivity.

When we think about standards in cuneiform culture in the Late Bronze Age, then, it isn't a matter of whether or not the writing practices we observe at a given site conform to those of Babylonia or are in some sense 'peripheral' and thus defective. Rather, the standards we are concerned with are those of globalised networks: the generally accepted protocols or etiquette that facilitated participation, communication and interaction. This is most clearly seen in the field of diplomacy, where there were evident expectations concerning what language and script should be used for international communication, how a letter should be laid out, or one's correspondent should be addressed. These are 'standards' in Grewal's sense,19 facilitating interaction as well as serving as markers of an acceptable level of socialisation within the networks of high-status Near Eastern culture and politics which allowed users to be recognised as part of that world. Nevertheless, variation still existed. As we've already mentioned, the rulers of Phoenicia and the southern Levant wrote their Amarna letters in a peculiar linguistic form, while Alašiya (Cyprus) used either Akkadian or Canaano-Akkadian depending who it was writing to (its Amarna letters are in the latter; its letters to Ugarit in the former). We even have international letters written in local vernacular languages: from Sidon to Ugarit in Ugaritic and alphabetic cuneiform; from Mitanni to Egypt in Hurrian. These are rare, and some may be archival translations, but they do nuance the picture of Akkadian as a regional *lingua franca* or cast-iron standard for participation in international connectivity. It might be more accurate to say that people made efforts towards minimising idiosyncratic local writing practices when communicating with people outside their polity.

<sup>19</sup> Grewal 2008; see also Morley 2014.

Such norms and stylistic expectations would in part have spread through emulation – elites mimicking, or having their secretaries mimic, the practices of those foreign dignitaries they want to impress with their ability to 'play the game'. The spread of practices and material culture in this way is very well known throughout history. Also important, however, is literate education, since that is where writers would be inculcated with ideas about the 'right' and 'wrong' ways of doing things. This is one of the areas in which Babylonian cultural hegemony is often presented as strongest, since across the Near East literate education seems to rely on the same texts – ultimately of Old Babylonian origin. These are mainly lexical lists – usually bilingual or multilingual, organised thematically20 – along with a number of literary and religious compositions such as the Epic of Gilgamesh or the flood narrative of Atraḫasis. Even in regions far from Babylonia, this material is extremely widespread: writing in 1976, Miguel Civil mentioned Late Old and Middle Babylonian-period lexical lists from the main Babylonian series at Nuzi, Ḫattuša, Alalaḫ, Amarna, Ugarit and in Palestine. In fact, Mari was the only site at which a large cuneiform tablet collection had been recovered that did not include this material.21 This pervasiveness of an ostensibly Babylonian-derived educational corpus has also, probably rightly, been taken as indicative of a common, conservative paedagogical method centred initially on the copying and learning of bilingual lists in Sumerian and Akkadian. The body of educational texts or scribal exercises from Ugarit is large – in fact, the fullest of any Near Eastern site for this period – and is very much in line with the tradition of Near Eastern literate education known from other sides and other periods. Canonical lexical lists such as *urra = ḫubullu* are fully present, as are segments of Gilgamesh and a possible fragment of Atrahasis. Unsurprisingly, this material has been widely discussed, most notably by Wilfred van Soldt in two very useful articles.22

The methods of literate education at Ugarit are certainly congruent with what we see elsewhere in the region, but we can be a bit more nuanced than simply regarding this as a straightforward diffusion of a Babylonian cultural and education tradition out to the 'periphery'. Both van Soldt and Civil have noted that several of the Babylonian lexical and literary texts at Ugarit are rather low-quality compared to 'canonical' examples from Babylonia itself, featuring numerous mistakes in copying and transmission. They conclude that this material probably didn't arrive in Ugarit and other 'peripheral' centres directly from Babylonia.23 This contrasts with the view of Huehnergard, who suggests that some of the lexical texts found at Ugarit might be imported from Babylonia due to the small number of 'non-standard' forms, but notes that it is difficult to prove definitively; others, it should be noted, are rich in non-standard forms. He notesthat of the around fifty literary textsin Akkadian found

<sup>20</sup> On the history of lexial lists, see Veldhuis 2014 and especially section 5 for the Late Bronze Age material. They are usually Akkadian and Sumerian in language, but can vary according to the local linguistic situation. At Ugarit lexical lists featuring up to four languages have been found, with Akkadian and Sumerian supplemented by Ugaritic and Hurrian.

<sup>21</sup> Civil 1976, 128–129.

<sup>22</sup> van Soldt 1995; 2011. For Gilgamesh, see George 2007.

<sup>23</sup> Civil 1976, 128; van Soldt 1995, 177.

at Ugarit, few show 'non-standard' or 'peripheral' Akkadian features.24 In a more recent article, van Soldt seems more open to the idea of direct connections, including the movement of specialist personnel.25 At both Ugarit and Emar there are signs that at least one of the writers involved in educating apprentices might have been on Babylonian origin, but this is still a very small proportion of the overall known literate population and while we should never underestimate the ability of people to move around in antiquity, one or two people reaching Syria from south-east Mesopotamia hardly constitutes an ongoing and significant export of educational resources and personnel from Babylonia to the 'periphery'.

Recent work by Eleanor Robson on the geographies of knowledge and scholarship in the cuneiform world is very useful here. Echoing van Soldt and Civil's early comments about the poor quality of educational material from Ugarit, she emphasises the patchiness of access to scholarly material in many Near Eastern centres, pointing out that incomplete or poor-quality texts seem to have been the norm even in Mesopotamia:

Our sobering conclusion must then be that no person or community in the first millennium BC, even the royal scholars of Nineveh, had access to as much of the so-called 'stream of tradition' as we do today.26

Robson also highlights the extremely local networks of trade and information exchange that fed this pattern. Far from Babylonia as a great cultural centre radiating scholarly material out to every corner of the Near Eastern world, she adopts a decentred network model, where would-be scholars sourced such texts as they could from relatively local neighbouring sites, where what was used was based not so much on a universal sense of the prestige and importance of Babylonian scholarly culture as on the simple expediency of what they could get their hands on. Of course, these two factors aren't unrelated: material initially widely reproduced because it was prestigious or politically and culturally desirable is likely to be more readily available, regardless of whether one bought into its ostensible cultural cachet. This model doesn't deny the possibility of direct links with Babylonia, but recognises that these were only one strand – possibly a minor one – in a much more complex web of influences and interconnections. Analysis of geographical terms in the written material from Ugarit ably demonstrates this – although Babylon is mentioned and on the city's cultural radar, it's a minor and distant player in comparison to Syrian centres such as Alalaḫ, Amurru, Karkemiš and Emar (and see also the role of Mari as a mediator of 'Mesopotamian' culture in the Middle Bronze Age, as discussed in Chapter 3).

It's difficult to map these networks. We have no way of knowing where particular documents at Ugarit were originally obtained: surviving examples are mostly likely to be local copies. Arnaud (2007) judges that approximately a third of the 68 texts he discusses were tablets physically imported from elsewhere – mostly ultimately from Babylonia, then Assyria and Ḫatti in descending order. He uses criteria such as

<sup>24</sup> Huehnergard 1989, 13–14.

<sup>25</sup> van Soldt 2011, 211. See also Veldhuis 1997, 71.

<sup>26</sup> Robson 2014, 159.

palaeography and dialect to suggest broad possible routes to Ugarit for much of this material. The rest, he believes, were locally produced copies, predominantly based on Babylonian or Sumerian originals but many showing signs of intermediary copies along the way.

Another approach is to map the spheres of interest or distributions of places that relate to particular genres or scripts, based on the places mentioned in those documents. This does not, of course, mean that the text originated there, nor are the numbers sufficient for true statistical significance – but it does give an idea of the conceptual geographies relating to that class of written material, which may perhaps hint at the networks of knowledge, information and cultural exchange that informed them. Figures 5.1–5.4 illustrate these spheres of interest for four broad genres of written material from Ugarit, in both main scripts.

*Fig. 5.1. Numbers of mentions of sites in alphabetic cuneiform religious and literary texts from Ugarit. In logosyllabic, there is only a literary/religious single text, mentioning Aleppo.*<sup>27</sup>

<sup>27</sup> The data in these maps is largely taken from Belmonte Marín (2001) and so does not include material published more recently. The maps only plot sites that can be relatively precisely located (so sites merely listed as in the kingdom of Mukiš, for example, are mostly excluded) and towns within the Kingdom of Ugarit itself are not included, partly for visual clarity and partly because my interest here is external networks. The use of place-names means that numbers on these maps don't include material which might be seen as ultimately 'Babylonian' in origin but which doesn't include toponyms, such as lexical lists. While these limitations are openly acknowledged, I don't believe they substantially diminish the indicative value of the maps produced.

*Fig. 5.2. Numbers of mentions of sites mentioned in alphabetic cuneiform (top) and logosyllabic (bottom) letters found at Ugarit.*

*Fig. 5.3. Numbers of mentions of sites mentioned in alphabetic (top) and logosyllabic (bottom) treaties and judicial texts from Ugarit.*

*Fig. 5.4. Sites mentioned in alphabetic (top) and logosyllabic (bottom) administrative and economic texts from Ugarit.*

These maps clearly show how different writing systems and different genres manifest different patterns of geographical interest. This can be seen as a rough proxy for the networks of communication and transmission underpinning the movement of textual materials and writing practices into Ugarit. There are limitations to this method – since it's based on a simple numerical count of mentions of place-names, it doesn't account for foreign-derived material which doesn't mention toponym or for more fine-grained analysis of the context or purpose associated with the appearance of a toponym. It also doesn't distinguish between textual material coming from a site and mentions of people. Nevertheless, it can serve as an approximate indication of the key sites within Ugarit's wider world and their correspondence to different kinds of written material. The variations we see between documents in different genres and scripts allow us to roughly outline the multiple overlapping conceptual geographies relating to different branches of Ugarit's writing culture. Above this heterogeneity, there is also a broadly consistent overall pattern attesting to the locations with which Ugarit was most closely interconnected: there is a very strong clustering of places relatively nearby in the coastal Levant, with key centres further afield at Ḫattuša, Karkemiš and Egypt. This is much as we would expect, given what we know about Ugarit's history and political situation. Babylon and Aššur are clearly the most significant of the more distant sites in Mesopotamia, but the amount of material referring to either is small. Given the geography of the region, it would make sense to assume that writings and writers from these places may have travelled up the Euphrates in relatively small numbers, probably coming to Ugarit via intermediaries such as Karkemiš or to a lesser extent Emar.

Even within the predominantly short-range networks of scholarship and knowledge I am suggesting, influences from outside weren't unthinkingly reproduced but were strongly mediated and adapted according to the specific cultural contexts of the sites themselves and the personal and corporate agendas of the people involved in training apprentice literati. This could be relatively small-scale, such as the addition of locally useful extra languages to Sumerian-Akkadian lexical lists (*e.g.* Hurrian, Ugaritic) or the personal preference for more 'standard' orthography by a 'Babylonian' writer whose education tablets were found among Ugarit's Lamaštu tablet store.28 However, it could also be as dramatic an adaptation as the existence of a whole second strand of literate education such as training in alphabetic cuneiform. We must also always bear in mind that education is more than merely the methods and 'teaching resources' that are utilised, but in antiquity as today contains a substantial – if not even more important – oral component in which the beliefs, priorities and agendas of teachers play a substantial role. In accepting that literate education was not education to be literate but education that made use of literacy, we must pay attention to the possibilities that the rest of what was taught might have had quite different content, even if it were based on similar 'textbooks'. The ideological and interpretative framework within

<sup>28</sup> van Soldt 2011, 210.

which this material was presented and taught is likely to have been highly dependent on local culture. The ideological aspects of the oral component of ancient teaching are of course extremely difficult to reconstruct, but there is good evidence that people in Ugarit engaged with and reflected on regionally widespread myths, ideas and norms in distinctive ways. In his recent analysis, for example, Aaron Tugendhaft convincingly arguesthat the Baʿlu Cycle should be seen as a reflection on the nature of kingship and international networks of power and prestige.29 The mythical elements it utilises are doubtless ancient within Ugarit and resonate with parallel myths in other parts of the Near Eastern world, including canonical texts such as the *Enuma Eliš*, but this poetic work remixes and re-presentsthem in a new and distinctly Ugaritian way that is fitted to its experience as a vassal between great powers and to its authorʾIlimilku's possible experience as an official and diplomat. It'slikely that the education offered by ʾIlimilku's colleagues to their apprentices was just as partisan, and culturally mediated, wrapping the standardised Near Eastern methods and canonical texts with their own views, ideologies and experience. This is, after all, what teachers have always and will always do. Literate education in Ugarit might, then, have used the same materials to very different ends than that in Babylonia, Emar or Ḫattuša.<sup>30</sup>

Thus, if we talk of 'norms' or 'standards' in writing practices in the ancient Near East, we shouldn't see this as a straightforward transmission of 'correct' forms from the centre to the periphery and a simplistic decision by peripheral centres to adopt or reject those standards. Instead, it was evidently a much more organic and decentred process. Standards such as those for international diplomacy were likely converged upon in a gradual and organic manner, without having to be associated with a single place of origin – in globalisation terms, they are abstracted and deterritorialised. However, they were implemented or not according to the local specificities of the cultural network within which they existed, based on the particular socio-political situation and agendas of users in given location and the uneven distribution of things like teaching materials or key texts within cultural and exchange networks. What might initially have been driven by prestige or political pressure might, at a greater remove, be mere expediency: it was obtainable, therefore it was used and reproduced, therefore it was obtainable. Conservatism in education was a result, but even so should not be overstated: we shouldn't forget that education wasn't a vehicle by which Babylonian cultural supremacy was unthinkingly reproduced, but a tool that could be turned to specific local scribal agendas.

Within this framework, there is considerable scope for 'norms' to have been imperfectly grasped, adapted, subordinated to other concerns or disregarded entirely. Deviations from what in the past might have been understood as the standard forms of Babylonian culture should not simply be put down to the cultural deficiencies of the Mesopotamian world's western periphery, but understood as a result of a complex

<sup>29</sup> Tugendhaft 2018.

<sup>30</sup> For the diversity of school curricula in Ugarit, Ḫattuša and Emar, see Fincke 2012.

interweaving of networks of trade and scholarship with local agency and agendas. Even where they seem to be adopted relatively unchanged, this may still be due at least as much to expediency in order to participate in political and economic interactions, rather than out of any ideological investment in or perceived emulation of 'Babylonian culture'. When and how people chose to adopt a more generally accessible linguistic and scriptal form, and when they did not, was a deliberate and active decision that tells us a great deal about how the inhabitants of these regions thought about themselves relative to international networks and traditional regional models of prestige, as well as their power to act within them.

#### *Beyond the tyranny of the tablet*

So far this discussion has only focused on clay tablets, but we shouldn't ignore the great diversity of objects upon which cuneiform writing appears across the Near East, which further expands the range of writing practices current in the region. Even within the category of tablets, while clay might have been the most common material, it was not the only one, as Laurie Pearce has amply illustrated.31 Prestigious texts could be written on tablets made of metal or perhaps even precious stone. Although these are less likely to survive intact for obvious reasons, some do, such as the bronze tablet bearing the treaty between the Hittite king Tudḫaliya IV and Kurunta of Tarḫuntašša found at Ḫattuša. A silvertablet bearing the Treaty of Qadešis mentioned in the extant clay version, but doesn't survive. There are textual references to metal tablets, including gold, being placed in foundation deposits from the thirteenth century onwards, beginning under the Middle Assyrian king Šalmaneser I, and these continue well into the first millennium.Uninscribed tablets(ortablet-like objects)in precious metals, lapis lazuli and alabaster have been found at Mari and Uruk, and a number of literary and religious texts refer to lapis lazuli tablets in sacred contexts, especially as belonging to scribe-deities like Nisaba. The Epic of Gilgamesh includes one such reference, and furthermore mentions the lapis tablet being stored in a cedar-wood box with bronze clasps, further enhancing its sense of prestige and value.32

[Find] the tablet-box of cedar, [release] its clasps of bronze! [Open] the lid of its secret, [lift] up the tablet of lapis lazuli and read out all the misfortunes, all that Gilgameš went through!33

While clearly related to the ordinary clay tablet, such objects are evidently a very different proposition in terms of their cultural function and the practical techniques necessary for their production and use. This is hinted at in symbolic terms by

<sup>31</sup> Pearce 2010.

<sup>32</sup> Pearce 2010.

<sup>33</sup> George 2003, 539, II. 24–28, quoted in Pearce 2010, 178.

references to such prestigious tablets being paired with styli of gold, lapis lazuli and carnelian. The impossibility that such tools could actually impress wedges into the solid tablets isn't the point; both key components of the everyday business of writing are elevated to sacred preciosity: conceptually there's a parallel with clay-writing, even if in terms of earthly artisanal craft we know that these items must have required an entirely different production process.

Close to the mundanity of the clay tablet in another way are wax-covered wooden writing boards, which were used widely across the Near East for similar purposes. Although perishable in modern terms and easily reusable, we shouldn't necessarily assume these were considered more ephemeral in antiquity. Doubtless in some circumstances they were, but we also have mentions of libraries of texts on writing boards being stored long-term. Although archaeological finds, textual references and visual depictions all suggest writing boards were widespread, it's possible that they were embraced to different degrees from place to place. Postgate notes, for instance, that no boards or references to boards have been found at Nuzi, whereas the Hittites actively distinguished between '[clay]scribes' (LÚDUB.SAR) and 'wood scribes' (LÚDUB. SAR.GIŠ) – however, it is often thought that the wooden tradition at Ḫattuša is likely to have involved Luwian hieroglyphics rather than cuneiform.34

Moving away from the everyday, there are numerous other classes of objects upon which cuneiform could be written. One broad category can be identified as ritual or religious objects, in which I would include a diverse assortment of items such as clay cones, 'nails' or bricks used in foundation-deposits, objects used in divination, and votive or ex-voto objects. All of these are found at Ugarit to some degree.

Taking the firstsub-set first, foundation-deposits are a well-known feature of Near Eastern building tradition. We've already mentioned that tablets – either conventional clay ones or prestige skeuomorphs in precious materials – could be included in these ritual caches, but there was also a long tradition of inscribed objects specific to this purpose. The best-known are the so-called 'Gudea cones' of the late third millennium BC, especially associated with that eponymous king of Lagaš but certainly not a practice unique to him. By the late second millennium, these clay cones had evolved into clay 'nails', of which one – with a fragmentary and more or less indecipherable inscription in alphabetic cuneiform – has been found at Ugarit. Clay or stone bricks and door-sockets could also be inscribed with foundation inscriptions, although none of these have so far been found at Ugarit.

Inscribed divinatory objects are also known from Ugarit, this time in a rather large corpus of ivory replica livers inscribed in alphabetic cuneiform.35 Similar objects in clay and inscribed in Akkadian have been found at Mari and date to the early second

<sup>34</sup> Waal 2011. Waal also raises the possibility that the Hittite boards may not have been wax-covered, with the writing being made directly on wood with ink, similar to examples from Egypt or Roman Vindolanda.

<sup>35</sup> Gachet and Pardee 2001; Gachet-Bizollon 2007.

millennium BC.36 The change of language, script and material are all testament to the adaptation of this practice to the Ugaritian context, but we are evidently nevertheless dealing with a use of writing with deep Mesopotamian roots; unsurprising given that divination was, in the Babylonian education system, an advanced specialism of scribal training.

Our final ritual subcategory is votive and ex-voto objects. As with foundationdeposits, with which they had much in common, ordinary tablets or styli could serve in this way if dedicated to a scribal deity such as Nabû or Nisaba, but inscriptions can also appear on a diverse array of other material culture as part of its dedication in a religious context, from statues to weapons, thrones to seashells. In first-millennium Assyria, even human beings could be inscribed and dedicated as votive offerings.37 As mentioned in Chapters 9 and 11, many of the inscribed objects from the sanctuaries of Ugarit bore Egyptian hieroglyphic texts, a mark of the overlap between religious devotion and elite display. The most relevant for our purposes is the sole stele to bear an Egyptian hieroglyphic inscription, that of the 'royal scribe and overseer of the royal domain', Mamy. While the individual himself and the appearance of the stele appear to be wholly Egyptian, it's noteworthy that the dedication is addressed not to an Egyptian deity seen as syncretised with a local god, but explicitly to the quintessential Ugaritian deity, Baʿlu of Mt Ṣapanu.38 What we have here then, is a dedication to the local god by a high-ranking foreign functionary who defines himself first and foremost by his position as a scribe. The other two inscribed stelae from Ugarit are in alphabetic cuneiform and come from the precinct of the Temple of Dagan. Other inscribed votive objects include the cache of bronze weapons, tools and other objects found at the house of the High Priest on the acropolis.

A logosyllabic cuneiform tradition that doesn't seem to have been of interest in Ugarit or indeed the wider Levant is the production monumental inscriptions. Once again, this is a class that encompasses a great diversity of objects across time and space, from royal stelae such as the Assyrian military monuments at Kition on Cyprus or the Nahr el-Kelb in Lebanon to the so-called *kudurru* inscriptions of the Late Bronze Age, which preserve high-status royal favours such as donations or exemptions from duties. Statues could also be inscribed with biographical details of the person depicted – as is the case with the famous statue of Idrimi of Alalaḫ in the British Museum – or with apotropaic messages, such as the lions flanking the entrance to the Temple of Ištar in Mari.39

The point of this survey isn't comprehensiveness – there are doubtless many classes of objects I haven't covered, and I've barely scratched the surface when it

<sup>36</sup> Charpin 2010, 50.

<sup>37</sup> A remarkable Neo-Babylonian text describes a slave girl inscribing her own hand with a dedication to the god Nanâ, which is striking as much for its implication of slave literacy as for its elision of the boundaries between human being and votive object (Huehnergard and Liebowitz 2013, 72).

<sup>38</sup> Yon 1991, 284–288.

<sup>39</sup> Charpin 2010.

comes to the physical and cultural contexts in which different kinds of inscriptions are found – but to give a sense of the breadth of the 'cuneiform tradition' beyond the clay tablets it's primarily associated with. It should be very obvious by now that alongside the diversity in how cuneiform and related practices were adopted and adapted from polity to polity and town to town, there must also have been a vast array of skills, tools, professions, techniques, cultural lives and symbolic meanings associated with this incredible range of inscribed material culture. There was not just one 'cuneiform tradition', that of the scribe and scholar working in clay, but a great many, which also brought into their ambit the carpenter who made wooden writing boards; the beekeeper40 and orpiment-miner who supplied the materials for writing-surfaces; the stonecutter and the sculptor who were responsible for stelae, *kudurru* and statues; the hunter who supplied ivory and the carver who created objects which were subsequently inscribed, and countless others besides. We talked in Chapter 2 about the idea of the *chaîne opératoire* – the sequence of technical steps necessary to produce written material – and this is a very helpful notion to bear in mind as we consider this proliferation of cuneiform writing practices and the people involved in them. Equally multiplicitous are the cultural lives and afterlives of so many different kinds of material culture.

There was no single 'cuneiform culture' in the Late Bronze Age, to which the situation in Ugarit can be compared. Robson is surely right in her re-imagining of the world of cuneiform scholarship and tradition along the lines of a Latourian network in which the kinds of texts people have access to were extremely reliant on specific and probably fairly short-range trade links, and where the simple exigencies of what happened to be available probably had far more influence than the fashions, prestige or traditions of faraway centres. Robson is concerned mainly with geographies of knowledge – what scholarship and textual material was accessible – but the same is likely to have been true for the writing practices themselves. In other words, cuneiform culture, such as it was, was exactly the kind of decentred, constantly hybridising phenomenon that we see with alphabetic writing in the Levant,41 and indeed, one beauty of a network-based framework is that these meshes can just as easily be integrated into a larger web of writing practices across the region, without an essentialist division between alphabetic and cuneiform, 'Levantine' or 'Mesopotamian'. This is entirely in line with the model of material culture and agency advanced in Chapter 2.

<sup>40</sup> It's unclear when beekeeping began to be practised in Mesopotamia. The earliest direct reference is an eighth-century stele of Šamaš-reš-uṣur, governor of Suḫi and Mari, who boasts that 'bees which collect honey, which no man had ever seen since the days of my fathers and forefathers, nor had brought to the land of Suhki, I brought them from the mountains of the Khabkha tribe and I put them in the garden of Gabbari-ibni… They collect honey and wax. The preparing of honey and wax I understand and the gardeners understand it' (Ransome 2012, 40; Levey 1957, 159) but references to wax and honey predate this considerably so if we are to take Šamaš-reš-uṣur's claim at face value then we must assume that these products were gathered from wild bees or imported, probably in quite large quantities, at least in some parts of the region.

<sup>41</sup> See Boyes 2019b.

A 'scribal culture', then, is no less problematic than the archaeological cultures that populated nineteenth- and early twentieth-century scholarship. The alphabet or cuneiform are not things in themselves but heuristic labels covering a wide range of practices that varied greatly across time, space, social group and individual practice. And just like other cultural labels, these become particularly ill-fitting at the edges, where they are prone to overlapping, blurring into each other. This is what is often referred to as hybridisation, although that has the unfortunate connotation of suggesting a coming-together of two thingsthat were bounded and distinct in the first place. It is a truism of culture that everything is hybrid, that there is no such thing as the pristine or authentic. As I hope to show, writing practices at Ugarit have to be seen within this conceptual framework, not as a strange offspring of the alphabetic and cuneiform traditions, but as just another local manifestation of writing which reproduces some practices, adapts others, and discards yet more. This is not what makes Ugarit unique; what makes Ugarit unique is the specific way in which these elements are combined as a result of the particular social and cultural circumstances within the city and its kingdom.

#### **Writing practices at Ugarit**

How, then, was writing practised at Ugarit? Where did it stand within this heterogeneous network of local practices? Let's begin by recapitulating the basics. Ugarit has produced around five thousand clay tablets, and a much smaller number of other inscribed objects. The vast majority of these are in either logosyllabic or alphabetic cuneiform. It's generally assumed that these would have been written in the Akkadian and Ugaritic languages respectively, although as Roche (2010) has pointed out, this is by no means a given. In a corpus of 144 logosyllabic administrative texts she found only ten that could confidently be said to be 'in Akkadian', while the rest were either solely personal names, logograms or a mix of logograms and Akkadian syllabic words used as Akkadograms elsewhere in the region – as such, the vast majority are open to having been read in Ugaritic. Hurrian is written in both scripts. When Akkadian is used, it largely follows Babylonian orthography, but a number of examples seem closer to Assyrian norms, and some show a blend of Babylonian and Assyrian features, not to mention influences from various other parts of the Near East that attest to Ugarit's networks of contact and the routes by which Mesopotamian canonical material made it to the city.42 The clay tablets are generally 'portrait' in orientation and follow the usual formatting rules common across the region – for example, tablets are flipped bottom-to-top (rather than right-to-left like modern pages) and overrunning lines continue on the edges. Tablet sizes are very variable and depend on the content of the text to be inscribed, as we would expect. There is no obvious difference between the form or formatting of alphabetic cuneiform tablets and that of logosyllabic ones.

<sup>42</sup> Huehnergard 1989; Arnaud 2007.

According to Ellison, the same styli are used for both scripts – they have a square cross-section and generally (though not always) a bevelled tip.43 This contrasts with the triangular reed stylus believed to have been used for some Mesopotamian cuneiform by Michele Cammarosano,44 but we shouldn't be too surprised by that – as Cammarosano abundantly illustrates, there is a great deal of variation in stylus form and material across the Near East and over time. In a recent publication, Françoise Ernst-Pradal is less convinced than Ellison that triangular styli can be discounted for Ugarit, but nevertheless is still not suggesting that different types of implement were used for alphabetic and logosyllabic cuneiform.45 And this is the important point here: there is no evidence that different styli were used for the two main scripts within Ugarit, despite the different range of marks comprising their signs (for instance, alphabetic cuneiform can include left-pointing wedges and, in some sign variants, twisted wedges, curves, rings and dots, none of which are usual in the logosyllabic script).46 In terms of materials and implements, then, the two varieties of cuneiform at Ugarit as inscribed on clay tablets are not separated from each other, even if they do differ slightly in physical methods of inscription due to the slightly different shapes needing to be produced.47

We should also briefly consider the possible use of wax-covered writing boards at Ugarit. No writing boards have been found at the site but there is a mention of a 'tablet of wax' in RS 19.53, a letter between two scribes thought to have been sent to the city from the Middle Euphrates region.48 Also at Ugarit, a small limestone stele was discovered on the acropolis with an image depicting what Postgate interprets as a treaty-ceremony in which two folded writing boards sit on a table between the participants (Fig. 5.5).49 It's possible that writing on wax might have been facilitated by using a different kind of stylus to that employed for clay.50 In my own practical experiments I found that the square-ended wooden chopstick I use for writing in clay produced soft, hard-to-read impressions in wax. A harder implement with more

<sup>43</sup> Ellison 2002; 2015; Ernst-Pradal 2019, 490.

<sup>44</sup> Cammarosano 2014.

<sup>45</sup> Ernst-Pradal 2019, esp. 23.

<sup>46</sup> Ellison 2002.

<sup>47</sup> The question of how styli were held and manipulated, and whether tablets were also rotated and moved as part of the inscription process, has been fully and convincingly discussed by Ellison (2002; 2015) for alphabetic cuneiform. While he doesn't explore the logosyllabic script in the same way, it's reasonable to assume that the different components making up signs – namely the absence of leftor upwards-pointing wedges and of curved forms – would have led to some minor differences in the movements necessary to inscribe logosyllabic compared to alphabetic cuneiform.

<sup>48</sup> Symington 1991, 121 and n. 74.

<sup>49</sup> Yon 1991, 303–305; Postgate 2013, 401–402.

<sup>50</sup> Seidl (2007, 124) suggests that the stylus with a central groove, visible in several first-millennium Neo-Assyrian reliefs, including the famous one of Aššurbanipal, may have been intended for writing on wax. In her view, the groove – unnecessary for forming the cuneiform wedge shapes – may have contained some substance that could have assisted with writing on wax, either by softening or hardening as necessary, or perhaps reducing stickiness.

*Fig. 5.5. Possible treaty-ceremony involving writing boards. Limestone stele from Ugarit. Drawn by the author after Schaeffer (1936, pl. XIV).*

sharply defined edges (in this case a plastic rod with a square cross-section) worked much better, producing noticeably smaller and better-defined wedges whose orientation was much more easily readable. In this connection, it's noteworthy that bronze styli with square cross-sections were found by Schaeffer in Ugarit's Western Tablet Store, and Ellison reports that replicas were perfect for reproducing the palaeography of alphabetic cuneiform.51 These could, of course, have been used just as easily for clay, if they are indeed styli at all – but they are perhaps additional circumstantial evidence supporting the use of wax writing boards at Ugarit.

The very limited evidence we have from Ugarit links writing boards most readily to the fields of international correspondence and diplomacy. From this we can surmise that

if they were indeed used, then logosyllabic cuneiform was probably written on them, since this is the script most associated with these genres. We can't rule out alphabetic cuneiform as well, however. It would be premature to say anything much more conclusive than that, but as a purpose-made writing surface similar to the clay tablet, it would make sense for writing boards to show a similar pattern of use, with both logosyllabic and alphabetic scripts used, according to genre and intended reader.

There's a much greater difference between the use of scripts at Ugarit once we get beyond the world of purpose-made writing surfaces. Alphabetic cuneiform appears on a wide range of other objects, including hoes, stone stelae, clay vessels, a clay nail and more (Table 5.1). By far the largest category is ivory replica livers thought to have been used in divination. More non-purpose-made inscribed objects occur outside

<sup>51</sup> Schaeffer 1951, 15; Ellison 2015, 167–168. For images, see now Ernst-Pradal 2019, 20, fig. 5.


*Table 5.1. Non-tablet objects inscribed in alphabetic cuneiform from Ugarit, Minet el-Beida and Ras Ibn-Hani.*

Ugarit in a scatter of alphabetic cuneiform-inscribed material across the Levant, Cyprus and as far as Tiryns on the Greek mainland. All this material from outside the Kingdom is written in non-standard variants of the alphabetic cuneiform script, which contrasts with the inscribed objects from Ugarit and its associated secondary centres, which are predominantly in the standard version, or else don't present any diagnostic features and so are assumed to be in the standard alphabet.52

There doesn't seem to be an equivalent corpus of non-tablet items in logosyllabic cuneiform. There are occasional items, such as the Middle Bronze Age royal dynastic seal of Yaqaru and its later copy, but neither the number nor breadth of types of objects seen in the alphabetic script. It's possible this is partly a result of how the logosyllabic-inscribed material has been published – there is no single corpus but a large number of partial publications focusing on different groups of tablets; it's far from inconceivable that non-tablet items might have 'fallen through the cracks' of publication, or else are published outside the main volumes and so are hard to find. Be this as it may, there's reason to believe the preference for alphabetic cuneiform over

<sup>52</sup> For a much fuller discussion, see Boyes 2019b.

logosyllabic on non-purpose-made items is a genuine one. We can note, for instance, that some of these object types find parallels at other Near Eastern sites, but there they are inscribed in the logosyllabic script – inscribed divinatory livers are found at Mari, for example, while the clay nail recalls third-millennium foundation deposits from Mesopotamia.53 This offers a fairly consistent picture in which even objects known in the logosyllabic cuneiform world are adapted to alphabetic cuneiform at Ugarit. This is borne out by the 'genres' this material relates to. Based either on the contents of the inscription itself, or the limited contextual information available, they fall into two main categories: religious items – including votives or commemorations of sacrifices, the dedicatory livers and possible foundation deposits – and mundane items of daily life, often connected with administration: labels, ostraca and inscribed storage vessels. Within the clay tablet corpus, both religious and administrative texts are primarily associated with the local script. The fact that they were for local use would also suggest in favour of the use of alphabetic cuneiform.

This isn't to say that *only* alphabetic cuneiform is found on non-purpose-made inscribed objects at Ugarit, but the second main script isn't logosyllabic cuneiform but Egyptian hieroglyphs. The small but significant number of Egyptian-inscribed objects from Ugarit are overwhelmingly elite display objects and a number are religious dedications, placing them in a similar category to many of the alphabetic cuneiform-inscribed items.

What we have in the non-tablet material at Ugarit, then, is a pattern which in its broad strokes is similar to that seen in other Near Eastern and Mesopotamian polities, featuring similar kinds of objects. There are some signs of something a little unusual in that some items such as the divinatory livers or the clay nail find their closest parallels in much earlier periods, but this might be put down to the spotty chronological coverage of various sites. By and large, the people of Ugarit wrote on a similar range of items and for similar purposes to other societies in the region, at least so far as the evidence permits us to judge. There are, however, clear local preferences in terms of script. Most non-tablet items are inscribed in the local alphabetic cuneiform rather than logosyllabic, and the strong prestige of Egypt in the field of elite display makes objects inscribed with hieroglyphs far more significant than Mesopotamianising ones. This is very much in keeping with what we observe at Byblos, a city which – as we've seen over the last few chapters – has much in common with Ugarit. It also chimes with what we discussed above regarding education, where material of ultimately Babylonian origin is of clear prestige and practical importance in shaping particular areas of practice, but where the strength of direct connections shouldn't be overemphasised; they were tempered and altered in accordance with the specific cultural situation of the Levantine coast.

<sup>53</sup> The Ugaritic inscription is damaged and more or less unintelligible so it's hard to confirm what this object was used for and the significance of the writing it bore. The resemblance to Mesopotamian inscribed cones, pegs and nails may be coincidental, which might make more sense given the significant disparity in date.

If the script preferences in different areas of culture shows both similarities and differencesfrom other polities in the cuneiform world, then the same istrue of actual practices of writing and of the daily lives of literates. Broadly speaking, the structure of Ugarit's literate professions is very familiar from other Near Eastern societies. Education, as we've discussed, is an adapted form of that used elsewhere, utilising similar texts and teaching techniques and visibly differing mainly in the addition of the alphabetic cuneiform training on top – although, as we have also mentioned, it is very difficult to know how similar the ideological, political and cultural spin given to the standard Near Eastern teaching materials would have been to that in other polities and at other times. This system produced literate intellectuals who served in a number of different professions: 'basic' professional writers whose main function was to write what was required of them, international messengers, administrators, politicians, priests, exorcists, diviners, scholars and poets. Some people moved between two or more of these categories, such as ʾIlimilku, who was *ṯ*ʿ*y* (some sort of high official) of the king but also served as a diplomat, accompanying the queen on an overseas trip, and who composed or copied out the main works of Ugaritic literature.54 Even so, there were clearly specialisations and gradations among those who wrote in their professional lives. And at Ugarit, as far as we can tell, those specialisations and gradations look similar to what we see elsewhere.

An important thing to note here is that there's little sign that things worked differently forthose literate in the alphabetic script compared to those working mainly with logosyllabic cuneiform. The jobs undertaken by ʾIlimilku, and histeacher'sstatus as a diviner, are very much in line with what we would expect from someone using the logosyllabic script. Indeed, it's unclear how much overlap there was between the two scripts: did people specialise in one or the other, or was it usual to switch between the two as circumstances demanded? We only know the names of four of the writers who used the alphabetic script: ʾIlimilku, Ṯabʾilu, Bṣmn and Brqn.55 There is no certain overlap between these and the forty-odd names of writers we know from logosyllabic texts. The name ʾIlimilku appears in both, but not all occurrences are likely to be the same individual. The addressee of the Akkadian letter RS 6.198 and the ʾIlimilkus referred to in RS 19.070, 18.20+, 94.2445 and 94.2483 may also be the writer of the alphabetic literary texts.56 If this is correct, then it is clear that ʾIlimilku was proficient in both scripts and their associated languages. It's possible that this ʾIlimilku's teacher, ʾAttenu, could be a fifth alphabet-user, but there's no confirmation that what he taught Ilimilku was the alphabetic script, as opposed to general literacy

<sup>54</sup> van Soldt 1988; Wyatt 1997; 2015; Tugendhaft 2018, 31–35. Del Olmo Lete (2018, 50) takes *ṯʿy* as a title of ʾIlimilku's teacher ʾAttenu rather than ʾIlimilku himself; since he also identifies ʾAttenu with the high priest, he concludes that 'this term has two meanings, one secular, "prime minister, vizier", the other cultic, "offerer, sacrificer". In both cases he acts as the king's deputy, on his behalf, as the actual and only offering priest. In this manner, Attanu represents the fusion of the cultic and magic religious systems.'

<sup>55</sup> Roche-Hawley and Hawley 2013, n. 68.

<sup>56</sup> Tugendhaft 2018, 31–35.

or the content of the religious myths ʾIlimilku was writing down. A specialisation of writers to focus on either the alphabetic or the logosyllabic script would therefore be consistent with the limited data available, and some scholars have evidently assumed this to be the case. Roche-Hawley and Hawley, for instance, posit slightly different traditions and practices for writers working in the two scripts, namely a greater conservatism and emphasis on family tradition among the logosyllabic writers.57 On the other hand, given what we know about education at Ugarit, it seems almost certain that this would start off with logosyllabic writing in Akkadian and Sumerian. The existence of multilingual lexical texts including these languages alongside Ugaritic may simply reflect the first language of the students being taught, but it would also be consistent with an expectation that students should be able to move readily between writing systems and languages, at least at an early stage in their training. If we wanted to reconcile this with the apparent separation of practising writers, we might plausibly reconstruct a situation whereby early education is general and covers both scripts, but in which people later specialised in one and essentially dropped the other, at least for professional purposes. If this were the case, then we shouldn't assume a complete separation between two different systems, but a rather more flexible set of writing practices in which people had preferences and different levels of skill and training in each, but would not be entirely ignorant of the other, even if they had to fall back on hazy memories from their school days. This doesn't preclude particular tendencies or practices arising among writers of one script which may not have been shared by those of the other, but overall, these seem to be just another axis of specialisation within a single professional/educational structure.

In the previous chapter we also discussed the possibility of writing practices outside this official cuneiform bureaucratic structure, either in the form of wider use of alphabetic cuneiform, usually in non-standard variant forms, or the very speculative suggestion that there may have been a linear alphabet in use at Ugarit, of which alphabetic cuneiform would be a 'cuneiformised' variant for use on clay. As we saw then, while quite possible, both of these are extremely hypothetical, based on a relatively small and ambiguous body of evidence. I don't want to push this evidence too far, but both these possibilities would be consistent with the general picture we have presented here in which cuneiform writing practices were more diverse and less concentrated on the world of tablet-based officialdom than traditional approaches would tend to suggest, and in which these practices were at Ugarit deeply intertwined with the emerging linear writing practices of the Levant in a way that is not so much a hybridisation between two distinct traditions as it was characteristic of the interconnected and mutually-entangled nature of writing practices across the Near Eastern world.

That Ugarit is unique is beyond doubt – nowhere else shows patterns of script use or writing culture quite like it. But Ugarit is not entirely *sui generis*. Rather, its people

<sup>57</sup> Roche-Hawley and Hawley 2013.

engaged with the written materials and writing culture that reached them through their networks of contacts: reproducing, altering, adding to or ignoring them as they saw fit. This didn't result in two distinct sets of writing practices, the alphabetic and the logosyllabic, but rather a single system within which people were capable of specialising or altering their practice relatively dynamically, depending on the requirements of any given situation or task. Inevitably the most locally distinctive practices occurred in writing intended for local consumption, while the most extreme of these tended to be avoided for externally focused writing such as letters or diplomacy.

In this, Ugarit's writers were no different from those of any other Near Eastern polity with a strong culture of literacy; something which becomes clear when we look at cuneiform practices across the Near East not in terms of a single 'stream of tradition' or a binary distinction between 'standard' and 'peripheral' Akkadian. The region was a patchwork of local ways of writing and of using written materials, in which adaptations – both scriptal and linguistic – were the norm. The difference between Ugarit and places like Emar, Nuzi or Ḫattuša lies in the specific local scriptal, linguistic, cultural and political circumstances that the writers in each place were reacting to, the particular ideologies and concerns that preoccupied their minds and which together shaped their writing practices into something distinctive. In the next section of this book we will explore this socio-cultural context in detail, charting the social, political and ideological situation of Ugarit in the Late Bronze Age, and how these circumstances – and people's active responses to them – combined to shape the particular writing culture we associate with the polity.

### Part III

Writing and society at Ugarit

## Chapter 6

### The contexts of writing at Ugarit

This third, and longest, part of this book considers the case study of Ugarit in the light of the theoretical and historical contexts laid out in the preceding chapters. We'll consider questions of literacy, diversity, identity, social change and ask what legacy, if any, developments in writing practices at Ugarit had in the wider region. To begin with, however, over this chapter and the next I want to explore the spatial contexts of writing. At its most basic level, this is the simple matter of *where* writing-related materials were found and what we might glean from this distribution about how writing was practised. This survey is the main subject of the present chapter. In Chapter 7 we will then move on to space as a social and conceptual domain, and explore what the distribution of writing in the city and kingdom of Ugarit tells us about its social and cultural significance for different social groups within the city.

#### **Places of writing: archives, libraries and tablet-houses**

Before we begin the location-by-location exploration of where inscribed material is distributed on the site of Ugarit, I want to address the broader question of the terminology we use for places where inscribed material is gathered and what it implies about the practices that took place there and the contexts of writing on the site. The most commonly used term for collections of tablets is 'archive',1 although 'library' is also often found, sometimes in the same sources. Archives and libraries are not the same thing, however, as has been pointed out by several scholars who have considered whether these terms are appropriate for the ancient Near East.2 These writers have sought to delineate differences between libraries and archives, and indeed between ancient archives and modern; and a general sense that such precision is not entirely possible for the ancient Near East weighs against an apparent reluctance to dispense with the traditional terminology. Much of this work draws on the discussion by Ernst

<sup>1</sup> Van Soldt 1991; Bordreuil and Malbran-Labat 1995; Pedersén 1998; del Olmo Lete 2018.

<sup>2</sup> See, for example, Veenhof 1986; Black and Tait 1995; Posner 2003 [1972].

Posner, an archivist first and an ancientscholarsecond. Writing in the early seventies, he drew attention to two modern ideas of what an archive is – the American one, which defines it as a repository of texts no longer in everyday use, and the continental European one where it is used more generally for any collection of records. In his view this distinction is inapplicable to the ancient world, both because it was not a conceptual difference made at the time and because the evidence rarely allows us to distinguish between the two anyway.3 Klaas Veenhof (1986, 7), Jeremy Black and William Tait (1995, 2197, 2202) all essentially follow this line. For them, archives in the assyriological sense are the totality of the documents produced at a given time and place. Black and Tait define a library as a specialised form of archive, namely a 'collection including literary, historical, and perhapsscientific texts, in an institutional building such as a palace or temple or in a private house – where it might be the property of a scholar-scribe or priest' (2197). Pedersén, by contrast, has adopted his own, slightly idiosyncratic, view of the distinction, defining an archive as containing generally functional, often administrative documents while a library, as in Black and Tait's definition, is comprised of literary, scientific, historical and wisdom texts. He adds a secondary criterion based on the number of copies held: an archive generally only has one, or occasionally two copies of a document, while a library might hold multiples for use in different places at the same time.<sup>4</sup> He identifies both libraries and archives at Ugarit, sometimes in the same tablet collection.5

This whole terminological issue ends up feeling rather unsatisfactory. Among those scholars who address definitions directly, the desire for precision and typologisation doesn't fit well with archaeological evidence, which is frequently hard to interpret, even where it's well published. As we've already mentioned, these distinctions seem to be wholly etic – there is little evidence that people in the ancient Near East themselves thought in these terms: the Akkadian term used for the buildings housing such collections means simply 'tablet-house'.6 The use of terminology is also at odds with contemporary uses of the same terms, and thus liable to sow further confusion. Meanwhile, the majority of scholars continue to use 'archives' and 'libraries' loosely and more or less interchangeably. It's hard to escape the sense that we may have been looking in the wrong place for distinctions we can draw between different kinds of places associated with writing. One of the advantages of taking a wider view of writing practices is that we can, at least conceptually, distinguish a wide variety of places associated with writing practices, rather than getting bogged down in the terminological problems of trying to identify different sub-varieties of tablet storage room. Where were the raw materials gathered? Where were blank tablets shaped or other artefacts that could be inscribed manufactured? Where did the actual writing take place? Where were different kinds of objects used, and when? Inscribed items could, of course, have many different functions and places of use at different points

<sup>3</sup> Posner 2003 [1972], 4–5.

<sup>4</sup> Pedersén 1998, 3.

<sup>5</sup> Pedersén 1998, 68–80.

<sup>6</sup> *bῑt ṭuppi*, a calque of the Sumerian *é.dub.ba*.

in their lives, such as a letter read in a throne room, subsequently housed in a tablet store forreference, discarded and used as building material, before finally being excavated and turned into a museum object, its image perhaps appearing on book covers, tea-towels or souvenir mugs. These aren't usually the kinds of questions epigraphers have focused on, and so the published evidence can often be at least as unconducive to answering them as it is to distinguishing an Ugaritian 'library' from an 'archive'. But these kinds of places can at least conceivably be identified by future evidence, and the results will arguably be more useful for understanding the nature of writing practices and culture at a given site than trying to fit tablet stores into modern conceptual categories. As a result, this chapter will largely substitute 'tablet store' or 'tablet collection' for the more usual 'library' or 'archive' in its survey of the places where tablets were discovered in Ugarit. In the next chapter, we will focus in more detail on the other places of writing, including those outside the capital.

#### **Tablet collections in Ugarit**

This section is intended to be read in conjunction with the site map (Fig. 0.1) at the beginning of the volume.

#### *Palace*

The Royal Palace (Fig. 6.1) has suffered particularly badly from the lack of interest in publishing archaeological and contextual information alongside the tablets. It was intended to be covered by Schaeffer in the first volume of Palais Royal d'Ugarit, but that never emerged. More recently, a project to restudy and retrospectively publish material from the palace was begun at the end of the 1980s, but again the promised major publications never saw the light of day.7 From preliminary publications such as Margueron (1995b), it's evident that much of what had previously been understood about the palace needed serious revision: Schaeffer's plans were misleading and important information such as the deliberate blocking-up of doorways had been omitted. Schaeffer also measured depths based on the pre-existing ground surface, which was then removed, rather than on a fixed benchmark. This means that the recording of where objects were found within three-dimensional space as recorded in his documents is wholly unreliable, quite apart from the difficulties of attributing points topographiques with specific objects in the publications (see footnote 13 below).

Perhaps the most serious potential re-evaluations were suggested for the architectural phasing of the building and how this relates to its upper storeys. Schaeffer had always maintained that there was evidence of destruction by fire before the palace's final abandonment, but Margueron has emphasised the severity of the conflagration, which was apparently sufficient to fuse some of the building materials into glass, rendering them nigh-impossible to demolish. Margueron also pointed out

<sup>7</sup> Matoïan 2008.

*Fig. 6.1. Plan of the Royal Palace. Drawn by the author after Saadé (2011, fig. 53).*

that this seemed to affect the western portion of the structure more than the east, which strikes a tantalising chord with EA 151's assertion that 'half of the Palace was destroyed by fire and so half of it has disappeared'. As we will see below, it has often been thought that many of the tablets found in the palace are likely to have fallen from an upper storey, which is generally assumed to have been given over to the royal residential apartments and offices, in contrast to the public and utilitarian function of the ground floor. Margueron, however, thinks it possible that after the palace fire, some of the ground floor may have been abandoned and functions from there moved upstairs to newly built rooms on top of slagged ruins that couldn't easily be cleared.8 This isn't just a striking visual image that could have importantramifications for how we consider the palace as a physical statement and embodiment of royal power and ideology; it also requires a major re-imagining of our already almost totally speculative understanding of the function and spatial organisation of the upper floors, and the contexts for the chancelleries and archives where written materials were made and stored.

Unfortunately, until the palace sees a proper publication in some form, it is very difficult to assess such suggestions, and consequently to reconstruct almost anything of the immediate archaeological and architectural contexts of the palace's written

<sup>8</sup> Margueron 1995b, 191–192.

material.9 Nevertheless, we will summarise the various clusterings of tablets and draw what conclusions we can.

#### *Western Tablet Store*

The Western Tablet Store (Fig. 6.2) comprises four rooms in the north-western section of the Royal Palace, to the left of the main western entrance and opposite the city's outer wall abutting 'Palace Street'. Its positioning invites comparison with the archives of the Mycenaean palace at Pylos, which are also to the left of the main entryway. Initial access to the tablet rooms is through an antechamber, 2, and then into a larger room, 3, which has been interpreted as a 'secretariat' because of a number of bronze styli found there, as well as the lesser density of tablets.10 Two inner rooms, 4 and 5, contained the majority of the tablets – mostly Akkadian administrative texts dealing with lists of personnel and goods, as well as a few royal letters; however, as Figure 6.2 clearly shows, a halo of tablets was also found scattered around outside the archive. Schaeffer attempted to explain this as material that had fallen from an upper storey, but van Soldt (1991, 50) doubts this since joins can be made between tablets in this group and ones inside the archive proper, there are no stairs in this part of the building and there is no explanation for why such debris would have been scattered outside by a collapse rather than falling straight down on to the floor below. Van Soldt is unable to offer a preferable explanation, but we might wonder whether this material was dispersed by human action during the abandonment of the palace or in looting thereafter.

Dating this tablet assemblage is problematic since little in the way of material culture is recorded beyond the tablets themselves. Based on their contents, these seem to cover a spread from the time of Ugarit's incorporation into the Hittite Empire to its destruction, although the majority are from the reign of ʿAmmiṯtamru II: that is, the second half of the thirteenth century.11 Almost all the tablets here seem to relate to external matters to the city – either administration of the hinterland or international letters. This makes good sense given their location at the entrance to the palace, near the city's western gate.

#### *Eastern Tablet Store*

The Eastern Tablet Store (Fig. 6.3) occupies a number of small rooms a little way north of the large open area thought to be the palace garden (Court III). As well as in three small, non-interconnecting rooms, 54–56, tablets were found in a stairwell, 53, and a hallway with two columns, 52. As in the Western Tablet Store, there were two distinct layers of tablets. These were separated by a layer of ash, leading Schaeffer to speculate that after initial destruction, the area was rebuilt and the tablets moved to an upper floor.<sup>12</sup>

<sup>9</sup> Lackenbacher 2008.

<sup>10</sup> van Soldt 1991, 49; Schaeffer XIff. in Nougayrol 1955.

<sup>11</sup> van Soldt 1991, 57–58; Dalix 1997; Pardee 2001.

<sup>12</sup> van Soldt 1991, 61; Yon 2006; Schaeffer XVII–XVIII in Nougayrol 1955.

*Fig. 6.2. The Western Tablet Store of the Royal Palace. Drawn by the author after Nougayrol (1955, pl. III).<sup>13</sup>*

<sup>13</sup> The points on this plan and on Figures 6.3–6.5 represent '*points topographiques*' where small finds were discovered and are taken from the original plans. As mentioned, these are unreliable due to Schaeffer's failure to use a fixed benchmark when recording locations. Furthermore, Schaeffer's plans do not provide keys or lists to allow the reader to identify what specific items are. A complete list of *points topographique*s has not been published and, at this point would probably be impossible to produce even with access to the original excavation notes and objects because of the vagaries of recording and the fact that the numbering system was changed at least once during the 1930s, resulting in several items having multiple numbers (van Soldt 1991, 673–674). The *points topographiques* relating to inscribed material are given in Bordreuil and Pardee (1989) but this does not include uninscribed material, which, of course is the majority of the material and is crucial to understanding the material culture context of the inscriptions. Uninscribed material is included in the catalogue of *points topographiques* in Appendix B of van Soldt (1991), but he doesn't provide details on what most of these items are or their material characteristics. Furthermore, his list only covers 1938–9, which makes it extremely partial. Without a full study attempting to reconstruct and publish the *points topographiques* and their relation to both inscribed and uninscribed material, it is simply impossible to provide detailed and reasonably accurate breakdowns of what objects were found and where. I've included the points on these maps as approximate indications of where material was concentrated, but they should be treated as such, rather than precise maps of the distribution of inscribed (or uninscribed) material. This is unsatisfactory, but the situation is what it is. In cases where contexts have been re-studied in modern times and better details provided, more useful distribution maps are possible – see, for example, Figures 7.2–7.4 in the next chapter.

*Fig. 6.3. The Eastern Tablet Store in the Royal Palace. Drawn by the author after Nougayrol (1955, pl. VI).*

The documents from this assemblage were mostly in Ugaritic, belonging to miscellaneous genres, although a number of Akkadian texts were also present. Van Soldt considers the collection much more chronologically coherent than that from the Western Tablet Store, with almost all datable texts originating in the last half century or so of the city's existence. The lower layer of tablets appears slightly earlier than the upper. In this case van Soldt does consider the existence of an upper floor 'very likely'.14

Unusually, we have some idea of the non-tablet objects found in association with this collection, since Schaeffer provides a doubtless very selective list of other objects recovered from these contexts (see Table 6.1).15

<sup>14</sup> van Soldt 1991, 72.

<sup>15</sup> Schaeffer 1962, 95.


*Table 6.1. Finds from the Eastern Tablet Store, as recorded by Schaeffer.*

This is an assemblage consistent with a high-status building, but it is hard to derive much more from it than that. The weights and moulds perhaps hint at craft activities, as Schaeffer thought, but are little to go on. We have no way of knowing which of these items may have fallen from the floor above or how representative the assortment is of the general scope of finds in the area – Schaeffer would not have thought mundane local pottery worthy of mention, for example.16

#### *Central Tablet Store*

The tablets of the Central Tablet Store (Fig. 6.4) were found in and around an apparently open area known as Court IV, as well as in the adjoining 'Court VI'. While some of the tablets seem to be in their original storage rooms, van Soldt (1991, 74) believes

<sup>16</sup> The material culture from the Eastern Tablet Store is reassessed in Matoïan 2008.

*Fig. 6.4. The Central Tablet Store. Drawn by the author after Nougayrol (1955, pl. X).*

that others probably fell into the courtyard when an upper floor collapsed – since it is unlikely tablets would have been stored in an open area. Yon (2006, 40, 43–44) seems to think they all fell from an upper storey. This unfortunately means we have even less contextual information about this archive than we do for others. The large adjoining area originally known as Court VI is now believed to have been an interior space, possibly a banqueting hall.17 If this is correct, it would indeed be unexpected to find a large archive just next door. If this was a room for high-status feasting it would make sense for there to be large-scale cooking facilities somewhere nearby, but nowhere obviously presents itself – most adjoining rooms are small and many serve as thoroughfares. Perhaps the likeliest possibility is that food preparation was done in the open Courts IV or V (though it is unclear if the latter really was a courtyard); however, to the best of my knowledge no material evidence of such use was recorded by the excavators. Given the way the excavation was conducted and published, this does not necessarily mean none was there, but it is obvious that our understanding of the palace and how its various parts were used is both incomplete and very speculative. For now, all we can say is that *if* Court VI was a banqueting hall, and *if* the banquets were prepared

<sup>17</sup> Yon 2006, 39.

in the outdoor Courts IV or V, then it would be surprising (though not impossible) if these were also hubs of writing or archival activity – at least at the same time.

Van Soldt identifies three separate wings on different sides of the courtyard, each containing different kinds of documents – primarily legal in the north, economic in the east and south, with the latter containing a particular preponderance of texts relating to Queen Ṯariyelli and her relative and agent, the merchant Ṯipṭibaʿlu. The lower layer of tablets date from the later thirteenth century. Because of this, he thinks this area of the first flooris the most likely location forthe queen's private quarters.<sup>18</sup>

#### *Southern and South-western Stores*

These two assemblages of tablets both border the area originally known as Court V, but now thought likely to have been an interior space (Fig. 6.5). Additional tablets were found in the 'court' itself, most notably in the notorious'oven', which Schaefferthought was used to fire tablets and still had its final batch in situ at the time of the palace's destruction. This supposed feature has been subject to considerable re-evaluation

*Fig. 6.5. The Southern and South-western Tablet Stores. Drawn by the author after Schaeffer (1962, fig. 28).*

<sup>18</sup> van Soldt 1991; 2013, 10.

and it now seems clear that it was a product of over-imaginative interpretation from Schaeffer. Most likely it was nothing more than an overturned basket of tablets. Even if there was an oven (which is doubtful), the tablets were not placed within it for firing, but fell on top of it during the collapse of the palace.19

The Southern Tablet Store contained the majority of Ugarit's international legal texts and occupied two rooms, 68 and 69, on the southern edge of the 'court'. Little information is available on its archaeology, even compared to the other tablet assemblages. Van Soldt thinks that 68 may have been a ground-floor part of the store, with 69 housing a staircase leading to an upper section. In his view, the less valuable Ugaritic administrative texts were probably on the ground floor, with the treaties upstairs.20

The South-western Tablet Store was found in the wing west of the 'court', in rooms 80 and 81. It was subject to significant damage during the destruction of the palace. The majority of tablets were apparently found against the eastern wall of room 81, leading van Soldt to speculate that they were shelved along this wall.21 It's worth noting that as well as the tablets, this room was also the source of the largest class of other objects inscribed in alphabetic cuneiform – a set of ivory replica livers, thought to have been used for divination.22

#### *Minor tablet assemblages*

Several other rooms within the Royal Palace contained clusters of tablets, but aren't generally counted as 'archives' (or tablet stores in our terminology) proper.


<sup>19</sup> Margueron 1995a; Coquinot *et al.* 2008.

<sup>20</sup> van Soldt 1991, 91, 107.

<sup>21</sup> Hoftijzer and van Soldt 1991, 114.

<sup>22</sup> Gachet and Pardee 2001.

<sup>23</sup> Schaeffer in Virolleaud 1957, X.

<sup>24</sup> van Soldt 1991, 127.

<sup>25</sup> van Soldt 1991, 128.

whether this should really be counted as an archive, thinking that many of the tablets found there may have originally belonged to the Central Tablet Store, or possibly even were being used here as flooring material.<sup>26</sup>

#### *Tablet stores and the contexts of writing in the Royal Palace*

So what can we conclude from this survey of where tablets were discovered within the Royal Palace? There is, as has often been noted, broad separation of texts by genre: for example, treaties and other international documents tend to be located in the Southern Store, while domestic judicial texts are concentrated in the north wing of the Central Store. This organisational scheme is not absolute, however, and it is not uncommon to find 'out-of-category' documents. Similarly, because there is a rough but not absolute correlation between genre, script and language, particular storage areas tend to favour either Akkadian or alphabetic cuneiform, but not to the complete exclusion of the other.

Beyond this, it's difficult to draw conclusions about how the stores were used or how tablets were produced because of the numerous uncertainties about which tablets were found in situ, and which had fallen from an upper storey. The latter seems more common in most cases, which means we have little information about architectural features, material culture associations or how these rooms related to other spaces and activities within the palace. Our best evidence for how these rooms may have been structured comes from the Western Tablet Store, for which a ground floor location is at least possible. There we seem to have an outer 'secretariat' and an inner 'archive'. The rooms are placed at the palace's main entrance and evoke strong parallels with the placement of Linear B scribal facilities at Mycenaean Pylos. However, comparable placement is not seen in any of the other palace stores or at any of the other Mycenaean palaces; nor does Pylos have additional tablet assemblages from locations other than the entrance like the palace at Ugarit does, so the correspondence is neither very exact nor generalisable. At most we can say that it might point to Pylos and Ugarit adopting similar strategies for administering the hinterland at the palace gates; there's nothing to indicate a direct inspiration in either direction.

The existence of the so-called 'secretariat' in the Western Tablet Store raises questions about the balance between storage and bureaucratic functions – in other words, were these places where tablets were merely deposited for occasional reference, or were they also the places where they were written? And further, were the tablets produced in the same place, as and when they were needed, or was there a stock of pre-made blanks? To address the second question first, a number of blank tablets have been found at sites in Mesopotamia including Nippur, Sippar and Mari, dating from a wide range of periods.27 This implies that at least sometimes, tablets

<sup>26</sup> van Soldt 1991, 131.

<sup>27</sup> Taylor 2011, 8.

were prefabricated and raises the possibility that it may not have been the writers themselves who produced them. However, as Taylor says (2011, 8), there is no direct evidence from Mesopotamia to confirm whether this was the case and if so, how regularly. The working assumption among Assyriologists is that writers produced their own tablets. In the Aegean, there has been a suggestion for a division of labour between tablet-makers and Linear B writers: at the palace of Knossos on Crete, Sjöquist and Åström (1991, esp. 25–28) think children may have shaped the clay. To go further with this question, we can ask what facilities are needed for the manufacture of tablets. The basic answer is merely clay and water, although more elaborate amenities could have existed. At Hammam al-Turkman in the Jazira region of eastern Syria, excavators have identified Middle Bronze Age architectural installations, which they believe are associated with the production of clay tablets. These included a basin fed by water channels with an adjacent kneading table on which fragments of clean clay were found, some still bearing the imprints of fingers and hands from the kneading process. A bronze object tentatively interpreted as a stylus was also discovered.28 Hammam al-Turkman is, to be sure, separated from Ugarit by a significant span of distance and time, but it does point to what a tablet-production workshop might look like. Nothing like this has been identified at Ugarit and the nearest source of water to Ugarit's Western Tablet Store was 30 m away in the south-west corner of Court I. Some of the other tablet stores may have been even less conveniently situated with respect to the availability of water, given their likely locations on upper storeys.

Another interesting consideration is the availability of natural light. This would have been more necessary for reading and writing tablets than for the relatively simple task of shaping tablets. Both alphabetic and logosyllabic cuneiform were often written in very small text and rely on light and shadow playing across a three-dimensional surface to be legible. They would have been difficult to make out by lamplight (eye strain and short-sightedness were likely at least as much an occupational hazard for Bronze Age writers as for medieval monks or modern academics). We can note that most of the tablet assemblages were found in rooms adjoining large open spaces. In part, this is a natural consequence of an architecture based around courtyards, but even so, there are numerous 'interior' rooms where light might have been scarce, and it's not surprising that these tend not to be where tablets are concentrated. Even when the job in hand did not require them to make a house-call, perhaps we should envisage scribes taking their work out into the courtyards or up on to roof terraces?

Taken together, these observations suggest that tablets were probably mostly not produced or inscribed in the rooms where they were found. These were primarily for storage, and if they were associated with 'secretariats', these probably served more as bases of operations orstationery cupboardsthan self-contained offices. The real work of tablet production, writing and reading was probably done elsewhere. Often, this would have been determined by the nature of the writing task itself – whether a certain

<sup>28</sup> Meijer 2004.

person needed a letter writing or a certain place had to be visited for administration. Where it was simply 'office work' that could be done anywhere, the courtyards and probably roof terraces seem the obvious places. A corollary of this is that we should envisage writing practices within the Palace – and by extension in the Kingdom of Ugarit more widely – as mobile and situational rather than tied to specific 'archives' or 'secretariats'. This has implications for thinking about how writers and writing materials moved through space and for the potential for other people to come into contact with them. These are questions we'll revisit in the next chapter.

#### *Tablet assemblages outside the palace*

#### *House of the High Priest – Acropolis*

The large building known as the House of the High Priest is situated on the Acropolis, between the temples of Baʿlu and Dagan (Fig. 6.6). Its identification with the High Priest is based on the colophons of some of the tablets found there and the dedication inscriptions on some of a cache of bronze tools, weapons and other objects found buried under one of the doorways. The tablet collection is most famous for its mythological texts, including the six large tablets containing the Epic of Baʿlu,29 but there were also ritual and lexicaltexts. According to Schaeffer, different genres were found in different locations,30 but the archaeology of this structure and the findspots of specific tablets are not well recorded even by the standards of the site. Based on the published plans, it looks like there might be three separate tablet stores, but whether these were in their original rooms or had fallen from upper storeys is unclear, as is how they were used. Were these a private reference collection belonging to the High Priest, or were they used more widely in religious practice and education?

Del Olmo Lete has recently suggested that the mythological texts found in this collection constitute a distinct 'sacred literature' of Ugarit, tied to royal legitimation and intended for ritual recitation. He sees the texts found in these stores as directly correlating to the functions of the high priest himself – as officiant, diviner and exorcist – rather than a resource more widely available to Ugarit's priestly corps. Administrative, epistolary and other non-ritual documents from this building are interpreted as relating to the wider work of the high priest's 'office' and are likely to have been worked with by bureaucrats in his employ rather than the high priest directly.31 The reasons for del Olmo Lete's analysis are not always easy to unpick, and in my view the assumption that the religious and ritual activities implied by these documents must relate in a fairly direct way to this individual's personal practice rather than to the religious establishment he headed is not entirely certain.

<sup>29</sup> Petersen 1994.

<sup>30</sup> Schaeffer 1939a, 37; van Soldt 1991, 217.

<sup>31</sup> del Olmo Lete 2018, 13–25.

*Fig. 6.6. The House of the High Priest. Drawn by the author after Schaeffer* et al*. (1956, fig. 216).*

#### *Space between Royal Palace and House of Yabninu*

This open square32 seems to have been laid out in the mid-fourteenth century at the latest, since it overlies buildings dated to the fifteenth or fourteenth centuries. Twenty-four tablets – mostly economic – were found here, well dispersed such that van Soldt thinks they could have come from different parts of the surrounding buildings.<sup>33</sup> It certainly doesn't represent an *in situ,* coherent context. Yon (2006, 51) believes the tablets in the north-west of this plaza came from the Southern Tablet Store of the Royal Palace.

<sup>32</sup> See Figure 6.5 above.

<sup>33</sup> van Soldt 1991, 145–147.

#### *House of Yabninu (Yabni-šapšu)*

A structure known by several names, this large building of more than 1000 m2 was originally called the 'Southern Palace' or 'Small Palace' by excavators (Fig. 6.7). More recently it has been known by the name of its most famous occupant, to whom four letters found in and around the house are addressed. This Yabninu – a hypocoristic form of a full name Yabni-šapšu, which also occurs – was a senior official (*šatammu rabu* – high administrator *vel sim*.), merchant and perhaps also an envoy of sorts around the end of the thirteenth century. Unsurprisingly given its initial identification as a palace, Yabninu's residence was not just large but well-constructed and stocked with fine objects, including imports from Cyprus, which is significant considering the contents of the texts. Room 219 was a store room containing a number of large pithoi.34

The house's assemblage of around 82 tablets was found mainly in rooms 203 and 204 in the northern part of the structure. Staircases in rooms 202 (originally an entrance hall but later closed off to the exterior) and 231 attest to the existence of an upper storey, and Yon (2006, 51) believes that the tablets must have fallen from here, on account of their dispersal. This means that we again lack an immediate primary context for the assemblage.

The contents of the tablet store are overwhelmingly Akkadian records of economic activity, much of it relating to overseas trade or the recording of foreigners resident in Ugarit. Yabninu's business interests ranged from Cyprus and Phoenicia down to Palestine and Egypt. It is notable, then, that his collection contained two tablets of broadly Near Eastern type but inscribed in Cypro-Minoan, which, of course, remains undeciphered. Different Cypro-Minoan scholars have taken different views on the likelihood that these were written in Ugarit itself, as we will discuss in more detail in Chapter 9.

#### *House of Rašapʾabu and House of the Scholar*

East of the Royal Palace is a well-to-do residential district where many large homes were found (Fig. 6.8), several of which had tablet assemblages. One insula contained several rather small dwellings, a number of which have produced inscribed material. The exact division of this complex and the assignment of tablets to particular residences has been subject to debate. This has particularly focused on what has traditionally been interpreted as a pair of residences, both with tablet assemblages: the House of Rašapʾabu (Fig. 6.9) and the so-called House ofthe Scholar, named forthe lexicographic, magical, religious and medical tablets discovered there. Both dwellings are small and it has been suggested that the tablets may only belong to one of them, having been scattered during the destruction, or else that they may even constitute only a single house.35

In this connection, we can note that even as it is usually delineated, Rašapʾabu's house is architecturally unusual in that it seems to be divided into two main

<sup>34</sup> Courtois 1990; van Soldt 1991, 149–157; Yon 2006, 51–54.

<sup>35</sup> Yon (2006, 71–72) advocates the former position, while del Olmo Lete (2018, 77–86) advances the latter.

*Fig. 6.7. House of Yabninu, drawn by the author after Saadé (2011, fig. 66).*

interconnected areas, each with its own entrance from the street. The more northerly one seems to have been used mainly for domestic purposes and the southern section principally funerary, with a well-constructed tomb. The location of the tablet store is not recorded but, again, van Soldt thinks it was most likely on the upper storey.36 A complex architectural history for this block, perhaps involving subdivision or merging of units, therefore seems distinctly possible, and would accord well with some of the changes visible in houses elsewhere in the site (see Chapter 10).

<sup>36</sup> van Soldt 1991, 160–163; Yon 2006, 72.

*Fig. 6.8. Division of houses in the Residential Quarter. Drawn by the author after Yon (2006, fig. 36).*

*Fig. 6.9. Plan of House of Rašapʾabu. Drawn by the author after Saadé (2011, fig. 73).*

Rašapʾabu was the *akil kari* – the 'overseer of the quay'. Around 28 tablets were found within what's usually seen as his residence, with a further 18 or 19 in the 'House of the Scholar'. Mostly this material is in Akkadian but there were also a few Ugaritic documents. As well as the scholarly documents already mentioned, there were legal and economic texts and a couple of letters that might be exercises. In consolidating these two contexts into a single one, del Olmo Lete believes it is possible to reconstruct something of the differentiation in how the space was used, with the 'House of the Scholar' as 'the "classroom" and scriptorium where the documents were copied and the apprentices training took place',37 while the 'House of Rašapʾabu' was where the tablets were stored. This suggestion seems to be based entirely on the content of the tablets rather than any archaeological evidence of installations or material culture associated with writing and teaching in the one case and storage in the other.

#### *House of Rapʾanu*

Across 'Merneptah Street' east of the block with the two houses above is a large residence attributed to Rapʾanu (Fig. 6.10). It shared an insula with the descriptively named 'House of Rapʾanu's Neighbour' but it seems that at some point after their construction these two initially separate dwellings were interconnected. Nevertheless, the tablets mostly come from the northern section, the House of Rapʾanu proper. More than two hundred tablets have been published from this building, including international correspondence, most notably that from Cyprus; however, a further 459 tablets remain unpublished.38 As well as this 'state' material, there is also legal documentation relating to Rapʾanu's personal business affairs and an impressive collection of Akkadian scholarly texts: 'working lexicographical tools for scribes in carrying out their profession and at the same time as indispensable instruments in the training of young new scribes'.39 Although almost no exercises have been found, del Olmo Lete has little hesitation in characterising this building as a school for literate training.40 Again, we lack a precise localisation for where the inscribed (or other) material was found, so cannot provide any kind of proper discussion of the building as an archaeological context.

#### *House of the Tablets*

This building was uncovered as part of a large trench through a dense residential area known as the Ville Sud/South City. Because of the concentration of houses attributed

<sup>37</sup> del Olmo Lete 2018, 82–83.

<sup>38</sup> del Olmo Lete 2018, 73. This author notes, however, that it is difficult to establish a precise number for the tablets found in the House of Rapʾanu because of the fragmentary nature of many of them and 'the sometimes confused way of double cataloguing, with descriptive remarks that are not always easy to interpet' (ibid.).

<sup>39</sup> del Olmo Lete 2018, 73.

<sup>40</sup> del Olmo Lete 2018, 74.

*Fig. 6.10. Houses of Rapʾanu and 'Rapʾanu's Neighbour' Drawn by the author after Saadé (2011, fig. 68).*

to artisans in the southern part of this area, it has been considered a 'souq'.41 The so-called House of the Tablets lies at the northern part of this souq, just south of a large square. In Figure 6.11 it can be located in Block X, house B. The northern section of this building contained hundreds of tablets. Van Soldt (1991, 182) notes that access to this area was partially restricted over time, speculating that the intention was 'to set this wing apart from the rest of the structure and to concentrate all scribal activity here'. Nevertheless, within this part of the building the precise find-spots of the written material are again too uncertain to allow for a fine-grained contextual analysis. A large number of tablets were also found in the adjacent square and houses, and may belong to the same store. Yon (2006, 94–95) speculates that some of these may

<sup>41</sup> van Soldt 1991, 182.

*Fig. 6.11. The Division of blocks and houses in the South City. Drawn by the author after Yon (2006, fig. 54).*

#### *House of ʾAgapṯarri and House of the Hurrian Priest*

Another large trench extended in a southerly direction from the Acropolis and produced a similar cross-section of dense residential occupation, although it was poorly preserved and has deteriorated further since excavation. The written material from this part of the site comes from two adjacent residences (although del Olmo Lete (2018, 31) treats them as a single house, with the northern section being residential and the southern devoted to ritual/magical practice) (Fig. 6.12). The first is generally called the House of ʾAgapṯarri (or Agapšarri in Akkadian) because of a lion-headshaped vessel inscribed with an Ugaritic dedication to the god Rašap (Fig. 6.13), which referred to 'the son of ʾAgapṯarri' in an uncertain capacity – the dedicator was another man, Nuranu.

The adjacent building contained two tablet stores. The first, generally known as the Archive of the Hurrian Priest, was found in a so-called 'cella' in the north-eastern part of the house. As van Soldt (1991, 194) notes, this area is

have been older documents that had been discarded, while others may have been scattered from collapsing upper storeys. As well as lexical texts and writing exercises, the store also contained Babylonian literary texts, including the Flood section of Gilgamesh and wisdom texts. For this reason, the house has been seen as a place of literate education: in del Olmo Lete's words, 'a school and notary's office'.<sup>42</sup>

<sup>42</sup> del Olmo Lete 2018, 59.

*Fig. 6.12. House of ʾAgapṯarri and House of Hurrian Priest. Drawn by the author after Yon (2006, fig. 58).*

markedly separate from the rest of the house. Some of the tablets, along with a number of ritual items such as replica livers and a lung for divination (several inscribed), metal bowls, and an Egyptian-style clapper, were found in a 3.5 m deep pit. It is unclear whether they were deliberately deposited here or fell there during the destruction of the building. As del Olmo Lete notes, however, no staircases have been found in this complex, leading him to speculate that this building had only a single storey.43 The tablets, around 86, are almost all in Ugaritic or Hurrian, though a couple of Akkadian ones were also present. They were almost all religious or ritual in nature, including mythological and ritual texts, magic, god-lists and hymns.44

<sup>43</sup> del Olmo Lete 2018, 32.

<sup>44</sup> van Soldt 1991, 194; Yon 2006, 100.

*Fig. 6.13. Lion-head cup inscribed with an alphabetic cuneiform/Ugaritic dedication to Rašap (KTU 6.62). From Bordreuil and Pardee (2009, Text 16). Used by kind permission of Projet PhoTEO, Mission de Ras Shamra.*

The southern part of this building was badly eroded towards the eastern side, but in the south-west corner a second collection of tablets was discovered. These were mostly in Akkadian and mostly lexical in nature, save for a few literary compositions, including a set of Lamaštu incantations for which this 'Lamaštu Archive' is named. Del Olmo Lete characterises this as a 'reference library of classical Akkadian texts of magic performances'.45 As will be dispiritingly familiar by now, detailed archaeological information for this part of the building is not available, in part due to the erosion this part of the tellsuffered.

The differences between the two tablet assemblages in this building – in terms of recoverable context, genre, language and script, are profound and are perhaps the best

evidence we have in Ugarit for the strict organisation of written material by type. Nevertheless, as van Soldt (1991, 209) observes, both collections are scholarly and religious in nature, with no concern for day-to-day administration. There certainly seems little doubt that the occupant (or occupants) of this house was principally interested in cultic and academic matters. Del Olmo Lete considers him to be a magician, diviner, necromancer and cult officiant of high rank, probably second only to the high priest, and likely managing a number of acolytes and apprentices.46 ʾAgapṯarri seems to be a Hurrian name, but whether the owner of this building was a Hurrian, as the name given to it implies, is open to debate and rather depends what we mean by 'Hurrian', something we will discuss in detail in Chapter 9. Literate education also appears to have taken place in this building. Van Soldt notes that the Akkadian script of the Lamaštu Tablet Store is much closer to the Middle Babylonian standard and less prone to 'errors' than most Akkadian from Ugarit; for this reason it has been thought that one of the teachers here may have been an immigrant from Mesopotamia, although this is not confirmed by the colophons, which only attest local Ugaritians.<sup>47</sup>

<sup>45</sup> del Olmo Lete 2018, 53.

<sup>46</sup> del Olmo Lete 2018, 31–32, 39–40, 49.

<sup>47</sup> van Soldt 1999, 45.

#### *House of ʾUrtenu*

The House of ʾUrtenu (Fig. 6.14) in the south-central part of the site isthe mostrecently excavated major archive at Ugarit and has been central to the improved dating of alphabetic cuneiform. This part of the site was for a long time off-limits because of military installations. The first written material came to light in 1973 in a spoil heap produced by the excavation of an army bunker. Schaeffer – then aged seventy-five – was able to obtain passage to war-torn Syria on a cargo ship and visited Ugarit to investigate the finds. In the late 80sthe Syrian army finally demilitarised the tell and afterthe removal of the bunker, archaeologists were able to excavate in 1988 and 1992. They uncovered a large house with an important tablet store containing over five hundred tablets, the vast majority in Akkadian. Because of the modern excavation and the appearance of datable figures such as the pharaoh Merneptah in the texts, this tablet collection can be conclusively dated to the end of the thirteenth century or beginning of the twelfth.

Nevertheless, even with this modern excavation, our contextual information is much less than we might like or expect. While the information has been recorded, the building has only received a preliminary archaeological publication;48 the promised full report – like so many material culture-focused reports across the site – has failed to materialise. Instead, publication efforts have, as ever, focused on editions of the

*Fig. 6.14. House of ʾUrtenu. Drawn by the author after Lackenbacher and Malbran-Labat (2016, fig. 2).*

<sup>48</sup> Lombard 1995.

texts themselves49 – doubtless very significant but frustratingly adding to an already sizeable body of written material from Ugarit, while detailed archaeological information continues to be neglected.

This said, it is recorded that most of the tablets came from the small room 2135 in the south-east of the house. Based on holes for wooden brackets in the south wall of this chamber, Yon believes that most of these tablets were stored in shelves along this wall. Other tablets were found scattered in the northern part of the house, around the area damaged by the installation of the military bunker. Obviously we lack a proper context for this material, but Yon speculates that it may have fallen from an upper storey.50

The contents of the tablet stores include international letters, most notably royal diplomatic correspondence, administrative texts, commercial records and literary fragments. Again we are presented with evidence of a high official who kept both 'state' and 'private' documents in his well-appointed home, suggesting that if such a conceptual distinction even existed in Ugarit, it was not strictly enforced.

#### *Ras Ibn Hani*

Located on a peninsula slightly south of the capital, Ras Ibn Hani is believed to have been a new foundation of the thirteenth century, with comparatively orderly, well-planned rectilinear building alignments where Ras Shamra is a warren of narrow streets and subdivided residences, all crammed higgledy-piggledy within its densely occupied tell. The site has not been fully excavated, in part because of a reluctance to destroy evidence from later strata – including Iron Age occupation levels – in order to reach the Late Bronze Age city. The remains that have been uncovered from our period of interest have been interpreted as palatial (Fig. 6.15). In the south-east of the site is the so-called South Palace, larger but comparatively little investigated, while on the other side of the peninsula to the north-west lies the smaller but much more fully excavated North Palace. A glacis is thought to have run diagonally between these two complexes, marking the eastern boundary of the settlement. What lay between them, or what occupied the rest of the peninsula is currently almost entirely unknown. Erosion and the modern tourist resort situated on the peninsula west of the site mean that answers to these questions are unlikely to be forthcoming.

The South Palace hasn't produced any written material, but 185 tablets or fragmentsfrom roughly 130 documents have been recovered in the final destruction layer of the North Palace, in four main groups (Fig. 6.16).51 Many of these are in poor condition, likely because of disturbance by Iron Age and Hellenistic activity on the site. Because some of the texts found in the building concern the queen, the excavators

<sup>49</sup> Bordreuil and Malbran-Labat 1995; Bordreuil and Pardee 1999–2000; Arnaud 2007; Malbran-Labat 2008; Bordreuil *et al.* 2012; Lackenbacher and Malbran-Labat 2016.

<sup>50</sup> Yon 1995, 439; 2006, 86–88.

<sup>51</sup> Bounni *et al.* 1998, 91; Bordreuil *et al.* 2019, 14–16.

*Fig. 6.15. The site of Ras Ibn Hani. Drawn by the author after Bounni et al. (1998, fig. 1).*

have been keen to associate the North Palace with her in particular (specifically the queen mother rather than the wife of the king).52 However, other recent work, such as that by van Soldt (2013), has questioned whether we can narrow down the owner of the building so precisely, rather than merely seeing it as a residence controlled by the royal family in general.

Architecturally the North Palace at Ras Ibn Hani is divided into two main zones. The principal one, situated to the south-west, is organised around the central Court II and is characterised by well-planned, rectilinear architecture. The excavators contrast this with what they see as something of an annexe to the north-east, separated from the main area by a long north-west to south-east wall in which few if any openings have been found. The architecture of this 'annexe' is less well-planned in their view,

<sup>52</sup> Bounni *et al.* 1998, and see recently Jacques Lagarce's preface to Bordreuil *et al.* 2019, 18.

*Fig. 6.16. The North Palace at Ras Ibn Hani, with tablet locations marked. Drawn by the author after Bordreuil et al. (2019, fig. 1).*

and points to it being something of an afterthought, perhaps for more utilitarian purposes. Jacques Lagarce suggests that the period of royal occupation was mainly confined to the mid-thirteenth century, when the Royal Palace at Ugarit was being repaired after earthquake damage (see Chapter 10 of this volume) and that after this the North Palace at Ras Ibn Hani was given over to artisanal workshops, during which time the tablets sat more or less forgotten in their store rooms until the building's final destruction. More precisely, he dates Groups 1, 1 bis, 2 and 4 to the reign of ʿAmmiṯtamru II, with Group 3 perhaps slightly later.

Lagarce gathers the four tablet assemblages into two principal larger 'archives', the West (Groups 1, 1 bis and 2, concentrated on room VII, with some spillover into Court II) and East (Group 4, concentrated in rooms XXIX and XXXV). He notes a striking difference in the character of the material stored in these two locations, with the former containing mixed genres in both the principal languages and scripts (though the corpus for the site as a whole is overwhelmingly weighted towards Ugaritic and alphabetic cuneiform), while the 'East Archive' consisted only of 16 economic texts in alphabetic cuneiform. Lagarce thinks these latter may have been a private 'archive' relating to the property of one of the Palace's occupants; possibly the queen mother.53 Both tablet stores seem likely to have been situated on an upper storey, like their counterparts at Ras Shamra.54

The North Palace was excavated between the 1970s and the early 2000s, so under considerably more modern conditions than the Royal Palace at Ugarit. We might expect, then, that it could provide some of the material culture associations that are largely absent for the capital. However, the excavators note that the number of finds recovered was relatively low, leading them to speculate that the complex had been largely emptied prior to its destruction by fire, with only the largest, least valuable, or particularly specialised items remaining. Notably, everyday ceramics are among the under-represented categories, which points more towards the inhabitants taking their most useful and most valuable items with them as they abandoned the site, rather than a wholesale plundering of the palace by interlopers.55 The remaining material culture has not been systematically published, but the excavators do highlight a large sample of material in the first volume of the site report (Table 6.2). Of this, a good deal comes from rooms where tablets were found, although precise find-spots are not available.


*Table 6.2. Finds from archives at Ras Ibn Hani.*

<sup>53</sup> Lagarce in his preface to Bordreuil *et al.* 2019, 18–21.

<sup>54</sup> Bounni *et al.* 1998, 91.

<sup>55</sup> Bounni *et al.* 1998, 53.

<sup>56</sup> One inscribed in Ugaritic as seal of king ʿAmmiṯtamru. The same seal was used on KTU 6.23, found in the Royal Palace at Ras Shamra

Court II and Room XXIX stand out clearly as containing the greatest concentrations of finds, unsurprisingly given that they're fairly large and that the former is the central space around which the palace was organised. We shouldn't read too much into the fact that the material listed tends towards the moderately high-status and imported, since mundane ceramics are not listed in full and are anyway thought likely to have been removed. The excavators note that Mediterranean imports accounted for less than 1% of the total ceramics overall. A key question, then, is which of these finds belonged to the upper storey and should be associated with the tablets and which were found in the ground floor rooms into which the tablet stores collapsed. The items that can be linked to the tablet assemblage with the highest degree of probability seem to be the seals and sealings from XXIX and adjacent rooms. The fact that these were scattered is consistent with their falling, and sealing practices are of course connected with writing and administration. Significantly, this group accounts for around half the seals and sealings found in the palace. On the other hand, the beads and precious stones found in XXIX seem to point to artisanal craft production, which may well have occurred on the ground floor. We can draw parallels here with the comparable assemblage reported in the Eastern Tablet Store of the capital's Royal Palace (see above). Given our lack of other information with which to compare it, this may simply be coincidence, or it may indicate that there was a tendency for tablet stores or secretariats to be situated on upper storeys in similar parts of the palace to craft activities. This raises an interesting question of whether writing was considered fundamentally an artisanal or craft activity. This would make a lot of sense theoretically (see Chapter 2) but is somewhat difficult to square with the very high status and political offices of a lot of writers at Ugarit. This, obviously, is to assume that taking part in craft activities was in some way incompatible with high office, which may not be the case at Ugarit. Or else, there could be differences in how writing was viewed depending on who did it and for what reason.

#### **People, place and the practice of writing**

This brief survey of the contexts where tablets have been found at Ugarit and Ras Ibn Hani has highlighted the difficulty of the available data. On the one hand, the information we have is quite extensive – we can identify multiple assemblages of inscribed material and break these down by script, language, genre and more. In many cases we can locate find-spots reasonably closely. On the other hand, our ability to reconstruct the contexts of writing practices is plagued by ambiguities and gaps. Most frustratingly, we have very little information on stratigraphy or on the rest of the material culture assemblages of which these tablets were a part. We know enough to be confident that in most cases tablets were stored on upper storeys, but nothing of these floors survives. As a result, it is nigh impossible to accurately reconstruct the relationship between writing practices and Ugarit's architectural spaces. How did people use the rooms that I have called 'tablet stores' and which are more often termed 'archives'? How did writers and writing-related material culture move through other parts of these buildings? What changes occurred over time?

Despite the undoubted problems with interpreting the data we have, we shouldn't allow ourselves to be drawn into complete aporia. There are conclusions we can tentatively draw and speculations we can reasonably make. For instance, we can observe the potential implications of the vertical and horizontal associations of the locations of the tablet stores. As we have noted, the majority were located on upper storeys. This is also often thought to be where the private residential apartments would have been situated.57 This would have been convenient for the members of the elite who owned the tablets and may imply that they interacted with the tablets on a fairly regular basis – either by writing and reading them themselves, or by keeping employees who did that on their behalf close at hand. It might also have implications for how writing was viewed: as an 'above-stairs' activity an important person might carry out in their private quarters and offices, or that employees would manage near at hand? On the other hand, writing may never quite have escaped an association with other craft activities – while horizontally tablet stores may have shared a floor with the great and the good, there are some indications that vertically they were placed above workshops specialising in things like jewellery or gem-working. It's tempting to wonder if this had something to do with the production of blank tablets – this could have been carried out alongside other fairly low-status craft activities and then they could have been easily moved upstairs to the 'stationery cupboards' for collection and use by higher-status writers. Other possibilities can also be considered: perhaps tablets were located upstairs mainly for security reasons – as has often been suggested – or even just because there was more natural light (windows were likely more common on upper storeys in Ugaritian architecture, they were less likely to be shaded by nearby buildings, and roof terraces were closer).

Another question we can ask – if not necessarily answer yet – is the relationship between spatial organisation and social groups. Is it possible to identify specific people with particular places within the city, and then use this to begin understanding something of how they engaged with writing? Obviously this can be done to a certain extent, as the existence of buildings known as the 'House of ʾUrtenu', the 'House of the High Priest' and so on clearly attests. For a long time we've been able to observe certain individuals who seem to be associated with particular assemblages of tablets. Even within a single building like the Royal Palace, we can see that tablets belonging to some writers only appear in a single tablet assemblage. So, for example, Munaḥimmu is named as the writer of nine tablets, all of which were found in the Central Tablet Store.58 In other cases, however, we find the work of a single writer scattered across a wide range of locations across the city, as is the case with Nuʿmerašapu, whose tablets crop up in at least five different assemblages. This

<sup>57</sup> Yon 2006, 29.

<sup>58</sup> We know from patronymics this is a different Munaḥimmu to the one responsible for RS 17.22+ and RS 17.149, whom del Olmo Lete regards as Rašapʾabu's chief 'scribe' (2018, 82).

disparity is interesting, and points to the existence of quite different working patters within the ranks of Ugarit's writers.

Where people work in a particularlocation, especially within groups, often affects the social dynamics of their practice. It's useful here to think about Lave and Wenger's concept of 'communities of practice', which is where informal groups form in which people share skills, expertise and advice regarding practice. These need not necessarily require working in close proximity, but this can obviously facilitate their arising. Access to such communities can be an important way for people to gain education in a particular skill, or for particular ways of doing things to be passed on and proliferated, as well as for the formation of distinct social identities.59 Identifying such groups is difficult: although we can list writers whose work is collected together, it is more problematic to identify those who certainly knew each other or interacted on a regular basis. Our understanding of the prosopography of Ugarit is still at a relatively early stage. The most recent catalogue of writers named in the tablets is still the one that appears in van Soldt (1991, 19–32), which omits the many developments made in the three decades since. There has not been a broader study of onomastics and the people of Ugarit since Gröndahl (1967), while investigations into the possibility of identifying particular people's handwriting are even more in their early stages (Ernst-Pradal 2019). What is arguably needed is an up-to-date prosopographical study linked to social network analysis.60 Until that is done, our best way of identifying writers who certainly knew each other and interacted is through multi-generational literate lineages. There are several known families where multiple members have written tablets. Usually these consist of fathers and sons, but the most famous example is the aforementioned Nuʿmerašapu, who had at least four children and one grandchild who also wrote tablets.61 Is there any sign of commonality in the practice of these families, and in particular a link with particular locations?

Nuʿmerašapu, as we noted above, cannot be identified with a particular location even himself, and his progeny are similarly widely distributed around the tell. Whether their writing practice has anything in common is impossible to determine without access to the tablets themselves. In other cases, more of a correlation is apparent. If we return to the other example we cited above, Munaḥimmu, we can find that his father, Yarimmu, was also a writer. As with Munaḥimmu, all Yarimmu's tablets were found in the Central Tablet Store. However, since the latter only put his name to two documents, it's questionable how much significance we should place in this. Across the other families, the results are similarly ambiguous: Huṣanu and his son Yaʿḏianu are both concentrated on the Central Tablet Store; Karranu and his son Iltaḥmu both produced tablets scattered across several places; Šapšu-malku has a large number of tablets, almost all found in the Central Tablet Store; his son Iliramu has only one,

<sup>59</sup> Lave and Wenger 1991; Wenger 1998; and see also Wendrich 2012 for an archaeological application of these ideas.

<sup>60</sup> On social network analysis and its archaeological applications, see Knappett 2013; Mills 2017.

<sup>61</sup> van Soldt 1991, 22–23; Roche-Hawley and Hawley 2013.

seemingly found in the East Store.62 There seems to be a relationship with genre: where parents' and sons' documents are found in the same place, they tend to be of the same genre; where they are scattered, they are generally of different genres. This is not unexpected given that tablet stores seem to be broadly organised along genre lines. The reasons why some children work on similar things in similar places to their parents while others do not are not reconstructable at the moment. It is certainly not the case that a son would automatically follow his father into working on documents that would be gathered in the same location. We can't make any assumptions about familial relationship directly mapping on to involvement in the same communities of practice.

Ultimately, our understanding of space within Ugaritian buildings – especially regarding their upper floors – remains extremely speculative and more work is urgently needed, not just to delineate the specifics of architectural plans, phasing and the relationship to both inscribed and non-inscribed material, but also to think about how space and architecture worked socially and culturally, as arenas for, determinants of and results of, human practice and ideas. We will begin to address this aspect in the next chapter.

<sup>62</sup> Bordreuil and Pardee (1989, 169) place it (RS 18.285) here; van Soldt (1991, 21) has only a question mark for its location.

## Chapter 7

### Writing and the social construction of place

Archaeology has always been concerned with questions of space, geography and topology, whether landscapes, cityscapes or the microcosms of individual structures or stratigraphic context. Since the advent of post-processual approaches in the 1980s, however, archaeologists have increasingly realised that space isn't simply a backdrop, a three-dimensional matrix within which the archaeological record is located and organised, but is itself a product of, and constitutive of, society. As Henri Lefebvre argued,

social space 'incorporates' social actions, the actions of subjects both individual and collective who are born and who die, who suffer and who act. From the point of view of these subjects, the behaviour of their space is at once vital and mortal: within it they develop, give expression to themselves, and encounter prohibitions; then they perish, and that same space contains their graves. From the point of view of knowing (*connaissance*), social space works (along with its concept) as a tool for the analysis of society.1

Post-processual archaeology has been less interested in 'space' as an undifferentiated abstract and more in 'spaces' and 'places', specific and particularised. It has embraced this idea with a multiplicity of approaches, exploring how spaces and places are created and maintained, imagined, experienced and reproduced; their relationship with the body and the senses, with gender, status and, above all, how all these relate to material culture.2 This isn't a one-way street (to use a metaphor from the organisation of urban space) in which 'context' imparts meaning to material culture, or even a simple two-way one in which the artefacts in turn help us identify the purpose and meanings of excavated structures. Rather, material culture helps create, transform, contest, appropriate and symbolically charge spaces, imbuing them with social, ideological, ritual and other meanings. There's a sense, then, in which the distinction

<sup>1</sup> Lefebvre 1991, 33–34 [French ed. 1974].

<sup>2</sup> Parker Pearson and Richards 1994a; 1994b; Tilley 1994; Ashmore and Knapp 1999; Ashmore 2004. This approach is ultimately Heideggerian, drawing, as Tilley puts it, on the distinction between 'being' and 'being-in-the-world'.

between object and context is evanescent; for many scholars who've explored these issues, landscapes, cityscapes3 and other places are themselves material culture.4 And very often they're conceptualised as *inscribed* objects: palimpsests of human meaning and symbol that can be 'read'.5

If writing solidifes or objectifies speech into a material medium, a text, which can be read and interpreted, an analogy can be drawn between a pedestrian speech act and its inscription or writing on the ground in the form of a path or track. Both are sedimented traces of activity, and both provide ways to be followed. A strong path is inscribed through a forest or across a tract of heathland through a multitude of pedestrian speech acts that keep it open; a strong text is also one that is kept open, read many times. Just as the writing of a text is dependent on previous texts (it has the characteristic of intertextuality), the creation or maintenance of a path is dependent on a previous network of movements in particular, and reiterated directions through a landscape; it works in relation to a previous set of precedents.6

Writing is thus closely bound up in how we think about social spaces; it is, after all, the 'graph' part of 'geography' and 'topography'. The spaces people inhabit are also things made, inscribed with meaning just as much as clay tablets or papyrus sheets. And the inscription of place isn't just metaphor: it's long been recognised that writing is an important tool with which people create, claim and order the spaces of their worlds. At a grand level, this might mean a ruler literally writing on the land- or cityscape with a monumental inscription. This is a familiar phenomenon in the ancient Near East where pharaonic stelae or rock-cut inscriptions from Anatolian, Syrian or Mesopotamian rulers mark out important ritual sites or memorialise military achievements.7 Within the Levant, this is nowhere clearer than at the mouth of the Nahr el-Kelb, just north of Beirut and now alongside Lebanon's main north–south road. The rocks there play host to nearly two dozen monumental inscriptions spanning more than four thousand years, from Ramesses II to Nebuchadnezzar, Napoléon III and the European imperial powers of the nineteenth and early twentieth centuries to the nationalist presidents of the independent Lebanon. The most recent official inscription was added as recently as the year 2000.8

<sup>8</sup> Volk 2008.

<sup>3</sup> There's often a tendency, reinforced by this terminology, to present the city and the countryside as dichotomous. This is, of course, largely a product of modern times and the social changes wrought by the Industrial Revolution. As the overlap in names renders vivid, the boundaries between the city and kingdom of Ugarit were permeable, conceptually, practically and economically. Schloen (2001), for example, thinks it likely many city-dwellers worked fields out in the hinterland, while 'rural' features such as livestock were probably ubiquitous even in the cramped defiles of a city like Ugarit.

<sup>4</sup> Moskowitz 2009, 71; Harmanşah 2013.

<sup>5</sup> The origin of this metaphor has been attributed to Mae Theilgaard Watts in her 1957 book *Reading the Landscape: An Adventure in Ecology*, and was picked up in an influential treatment by Peirce Lewis in his 1976 essay 'Axioms for reading the landscape: some guides to the American scene'. Moskowitz 2009, 77.

<sup>6</sup> Tilley 1994, 29–30.

<sup>7</sup> Shafer 1998; 2007; Harmanşah 2013, 48ff., 93–99.

But the inscription of place isn't just a matter of elite monumentality, nor even of the graffiti and other informal inscriptions that fulfil a similar purpose at less elevated levels of prestige. Writing exists within and inscribes the land- and cityscape in countless forms. Signs and labels, shop-fronts and warnings; receipts and discarded shopping lists blowing along the pavements.9 Writing can even be part of the physical fabric of the spaces: in Mesopotamia, inscribed clay pegs could be inserted into walls with messages for the gods and for future remodellers, bricks could be stamped with cuneiform messages and old clay tablets were used aslevelling and fill materials.10 In a more recent parallel, we might consider England's M6 Toll motorway, which included a layer comprised of two and a half million pulped paperbacks. A BBC news report quotes the project manager as saying 'We use copies of Mills & Boon books, not as a statement about what we think of the writing, but because it is so absorbent. They may be slushy to many people, but it's their "no-slushiness" that is their attraction as far as we are concerned.'11 There's a lot going on in this light-hearted remark, which entangles materiality, gender, social status, reception of particular genres and the practicalities of civil engineering. Pulped the books may have been, but the cultural meanings attached to romance novels are still remembered and cited alongside their practical material qualities.

Not all writing is fixed as a physical component of the spaces it helps construct; it can also be mobile, transient, but nevertheless associated with certain places. Inscribed objects also moved around. The most common surviving examples are pottery – such as Mycenaean inscribed stirrup jars12 or Levantine vessels with cuneiform or linear alphabetic inscriptions13 – and seals, labels and sealings;14 but these almost certainly wouldn't have been the only portable objects to bear writing. Writing might also be associated with people, either worn – for example embroidered into the patterns of clothes, or on a seal around the neck15 – or made part of the body itself, such as through tattooing or branding. While mobile, the patterns of movement, the habitual flows of inscribed objects can still define places and their relations to each other: this is familiar from our own world, for example in the public square or sports ground where banners, placards and badges of affiliation are regularly brought together, or in the contrast between a predominantly young, trendy area frequented by people whose clothing bears slogans, quips or branding and a more conservative district whose older, perhaps wealthier, inhabitants favour

<sup>9</sup> Warnke 2013.

<sup>10</sup> Tsouparopoulou 2016.

<sup>11</sup> 'M6 Toll built with pulped fiction'. http://news.bbc.co.uk/1/hi/england/west\_midlands/3330245.stm, accessed 17 October 2018.

<sup>12</sup> Judson 2013.

<sup>13</sup> Boyes 2019b.

<sup>14</sup> Finlayson 2018.

<sup>15</sup> Inscribed seals in both logosyllabic and alphabetic cuneiform have been found at Ugarit, and a clay 'bead' pierced with a hole and inscribed in alphabetic cuneiform was found at Ras Ibn Hani (RIH 86/ [03], Bordreuil *et al.* 2019, 279).

plain clothing, anepigraphic apart from, possibly, a discreet designer label inside the collar or hem.

Particularly when we study early writing in ancient contexts, there's a tendency for us to think about writing only in those places where it has persisted for us to rediscover it in modern archaeological work, the aforementioned monumental inscriptions or the tablet stores in palaces and metropolitan homes. Anything else can be understandably hard to detect and to factor into our reconstructions. But consider them – or at least the possibility of them – we must. When we think about the economic records of the Ugaritian or Mycenaean countryside, for example, we need to consider how that information reached the centre – the movement of writers or written material through the landscape and the implications this had for how rural populations interacted with writing and the literate bureaucracies (and vice versa). We need to think about where inscribed pottery was produced or imported and the routes it took to the places it was sold, and thereafter to its places of use and deposition. We need to think about whether epigraphic clothing or jewellery such as seals might have been more present in some areas than others, and whether this may have contributed to these places' identities. This can be very difficult – often it'll be impossible to arrive at any concrete answers – but this chapter will demonstrate that these evanescent, transitory interactions of writing and place may be among the most important for wider popular engagement with writing practices.

There has been some research into the territory of Ugarit beyond the capital. In particular, there have been very useful contributions such as van Soldt (2005), Calvet (2012) and Yon (2008). However, this has been conducted entirely from the written sources of the tablets, with very little survey or excavation outside the coastal towns. The ideological, social and symbolic character of place within the Kingdom of Ugarit has been almost totally ignored. Apart from the obvious points of Mt Ṣapanu, the temples and the Royal Palace, we don't know which places were sacred, which were charged with symbolism. We don't know which spaces were haunted by historical meaning or accumulated folklore and which defined by their trendyness, dynamism and sense of fun. We've drawn maps of Ugarit, but we've barely begun to think about how the people that inhabited them *made* these spaces. A social geography of Ugarit has not been written, and this is not the place to provide one. What I do hope to do here is to begin to explore one small aspect of the social construction and negotiation of Ugarit's spaces, namely how they were inscribed. In this I'm interested not just in what writing tells us about how people construed land- and cityscapes, but also how the presence and absence of writing in these places, and its passage through them, helped shape how people construed the practice(s) of writing itself.

#### **Writing in the city**

My intention here is to build on the survey of the locations where inscribed material was found at Ugarit presented in the last chapter and think about how that distribution may or may not represent the real ancient situation, and how it shaped Ugarit's urbanism and relationship with writing.

The immediate impression is one of decentralism and diversity; it is commonly remarked how Ugarit's tablet assemblages and writing schools are distributed between the residences of various notable personages rather than being confined to the main 'public' buildings, the Royal Palace and the principal temples. This is true, but it doesn't necessarily translate into a widespread presence of writing across the city. A real question remains over how visible and 'present' writing in these residences was for anyone outside the highest elite or the owners of these tablet collections and those working for them. These tablet stores were in the homes of the city's most prominent and powerful; there is good reason to think many were on the less public upper storeys. There's every reason to expect these collections of writing would have been highly restricted.

That's not to say they must have been entirely invisible to people outside these households or the community of those involved in writing practices. We already speculated about the possibility – perhaps even probability – that writers worked outdoors rather than in the rooms where the tablets themselves were found. This is unlikely to have been in the streets, which were narrow and presumably often raked by steep contrasts of light and shadow, but more open spaces are a possibility. At Nuzi and Arrapḫe in Mesopotamia, city gates were important areas of writing practice, and this has obvious utility for administering traffic of people and goods in and out of the city.16 The only gate of Ugarit that has been investigated archaeologically to a significant degree is the fortified western entrance to the royal zone.17 This hasn't produced tablets, although it is close to the palace's Western Tablet Store. It may not have been typical of the city's other entrances, given that it opened directly into the heart of the city's most high-status area. It's assumed there was a southern entrance to the city through which the path from the bridge/dam across the Nahr ed-Delbeh connected to 'Main Street', but this has not been located archaeologically as it is thought to underlie modern orange groves.18

Within the city, courtyards are another possible location for writing, but flat roof terraces are also a distinct possibility, assuming water could be carried up there. If that were the case, such locations would most likely have been favoured more at particular times of day and in certain seasons, since the full glare of the Syrian summer sun at the height of the day would not have made for comfortable working conditions.19 Mornings or evenings would have been more tolerable, as well as offering lower angled light, which would have better shown up the inscribed cuneiform wedges. Relatively high rainfall between December and February (at least today – see Chapter 10 for the likely differences in the ancient climate) may have made rooftop writing impractical during the winter.

<sup>16</sup> Postgate 2013, 349.

<sup>17</sup> Yon 2006, 31–34.

<sup>18</sup> Yon 2006, 84–85.

<sup>19</sup> I am grateful to Christopher Rollston for this point.

*Fig. 7.1. The House of the Tablets. Drawn by the author, based on Saadé (2011, fig. 82).*

If this suggestion is correct, then especially in the spring and autumn, and especially in the mornings and evenings, the people whose residences jostled around these elite dwellings may have been able to see writing practices going on when they were up on their own rooftop terraces. This visibility is likely to have been enhanced for those houses where writing was also taught. If room was needed for teaching apprentices, the broad space of the rooftop seems a good candidate. We should also consider comings and goings – water was usually drawn from wells within houses but clay would have had to be brought in. Finished tablets evidently arrived and departed – especially letters

but also administrative records. Forthose houses where literate education was offered, we might imagine young trainees filing in of a morning and dispersing again at the end of their working day. All of these arrivals and departures would have spilled elements of the business of writing out into the surrounding streets. As an example, we can look at the 'House of the Tablets' in the South City (Fig. 7.1). As we saw last chapter, this was a densely occupied part of the city sometimes described as a 'souq'. The house in question, thought to have served as a centre for literate education, stood at the intersection of two open public squares and a narrow street. It was a large, rich residence with signs of having been extended in the thirteenth century.20

Despite its relative grandeur, this building was heavily overlooked by neighbours on three sides. If it had a rooftop terrace where writing took place, it is likely that it would have been very visible from the adjacent structures. The public squares to the east and especially to the north would have provided further scope for encounters between the wider public and the writing practices taking place within the house. The tablet scatter was concentrated in the northern part of the building and van Soldt (1991, 182) believes that over time efforts were made to segregate this part of the house from the rest, as if this were a dedicated area for writing and literate education. A northern entrance giving out on to the large square is conveniently located near

<sup>20</sup> See Callot 1994, 60–61 and also Chapter 10 below.

a staircase leading to the upper floor where the tablet collection is believed to have been located. It stands to reason that deliveries of clay and other writing materials, as well as the arrival and departure of apprentices, are likely to have taken place through this northern entrance, and consequently from the large public square. These comings and goings would have been visible to others using this space. The very act of separating these writing-related parts of the structure from the rest of the residence perhaps points to them as more public activities that the occupants wished to keep separate from their private domestic arrangements.

This said, we should stress that it's likely to have been writing *practices* rather than the writing itself that would have made the greatest mark on urban spaces. No one would have been able to make out cuneiform signs from even an adjacent rooftop; arriving students probably weren't waving their homework around, showing and explaining it to anyone who happened to be passing by. Those not initiated into the world of Ugaritian literacy might recognise that something went on in these houses involving the impression of clay with styli, perhaps even have a general understanding of what its purpose was, but it seems very unlikely that they would have had much contact with writing itself from this source.

So was Ugarit an 'inscribed city' at all, then? Or was it something else, where writing practices and their paraphernalia were occasionally publicly visible but all the actual writing was sequestered away behind high walls in these secret bastions of literacy? Certainly this is how it appears from the surviving evidence, but there are many categories of writing whose existence can be plausibly guessed at, but which can't reasonably be evaluated. This includes potential worn or embodied writing. We don't know whethertattooing, branding orscarification were practised at Ugarit. Still less are we able to say how common they would have been, which segments of the population may have practised them and whether or not writing of this kind would have been publicly visible or hidden by clothes. Likewise, we can't guess at the likelihood that clothes themselves may have incorporated writing into their patterns. Text being incorporated into textiles (no pun intended) is attested from other parts of the Bronze Age east Mediterranean: fragments of linen with woven hieroglyphs were found in the tombs of Tuthmosis IV, Tutankhamum and other high-ranking New Kingdom Egyptians.21 These are extremely high-status objects, but so little textile of the period survives, especially outside Egypt, that it's impossible to guess whether such fabrics might have been visible around a city like Ugarit and if so, how commonly and in which places. Other portable objects such as inscribed ceramics are more likely to have presented an opportunity for writing to be encountered in Ras Shamra's streets and homes: inscriptions labelling objects with their makers' and/or recipients' names aren't common, but they do occur on storage jars both inside the city and outside. It's not unreasonable to assume writing being used in the workshops, warehouses and distribution channels for such items and the commodities they contained. In

<sup>21</sup> Barber 1982; Janssen 1992.

this respect it might perhaps have been more prominent in harbour towns like Minet el-Beida, but we can also reasonably envisage the occasional inscribed vessel in Ras Shamra's markets. Like the wall-pegs from Mesopotamia, there was also unseen writing deep within the fabric of the city – either foundation deposits buried as part of their dedication, or treasure buried for safekeeping. An example of this can be seen in the cache of bronzes – some inscribed – found under a step in the House of the High Priest on the Acropolis.22 Whether this was a treasure horde or a foundation deposit remains rather unclear, but either way it serves to demonstrate how writing could contribute to an unseen world of meanings beneath the surface level of the urban environment.

#### *Writing and sacred space – the Acropolis temples*

Our best opportunity to assess the visible presence of writing in creating Ugarit's urban spaces comes in the sphere of religion. Rituals, magical spells and other supernaturally-charged practices were written down, but we know little about how they were performed: whether they were publicly read out from the tablets, which could have made writing part of the cultural and ritual associations of a certain place,23 or whether the written versions were confined to private collections to serve as personal *aides-memoires*, records or training devices for practitioners. What we do know is that inscribed objects were on display in the major temples (or, more properly, on the *sites of* the two main temples, since the matter is more complex than it first appears).

The Acropolis temples were imposing buildings of a type known as tower-temples, fortress-temples or migdal temples, found across much of Syria and Israel/Palestine in the Middle and Late Bronze Age.24 Their thick walls supported a tall structure, reconstructed at around 20 m based on the thickness of the masonry. Added to this, they stood on the highest part of the tell, raising their pinnacles to perhaps around 50 m above sea level, making them significant landmarks not just for the city but for the surrounding plain (perhaps more so from outside the city than within, given that streets were narrow and buildings tall). They are often depicted in reconstructions with fires burning on their roof terraces, raising their visibility still further; there's even been some speculation that they may have served as lighthouses or at least navigational landmarks for ships,25 though Olivier Callot is doubtful.26 Be this as it may, they could easily have pulled dual service as watchtowers. Much of the architecture of the temples is conjectural so we can't be very certain of what their interiors were like, but it seems likely that they were rather dark. If the Temple of Baʿlu can be seen

<sup>22</sup> Schaeffer and Dussaud 1929.

<sup>23</sup> A Hittite text from Ḫattuša may imply this was the case there, at least sometimes: KBo IV 2 IV 42 f refers to a ritual being performed according to instructions 'inscribed on an old writing-board' (Symington 1991, 116–117).

<sup>24</sup> Mazar 1987; Buck 2018, 157–169.

<sup>25</sup> *e.g.* Yon 2008, 40.

<sup>26</sup> Callot 2011.

as an urban, worldly analogue of the new palace that he builds in his epic, the detailed description may imply that it was not burdened with an overabundance of windows – their number and placement is presented as significant and subject to restrictions.<sup>27</sup>

These two Acropolis temples were built early in Ugarit's urban period, in the Middle Bronze Age, and remained in use for several centuries until they were largely destroyed in an earthquake thought to have occurred in the mid-thirteenth century (see Chapter 10). While the Temple of Baʿlu had been restored by the time of Ugarit's destruction, the same was not true of the Temple of Dagan, possibly because of a decline in that deity's importance. Instead, the site seems to have been levelled, perhaps in preparation for a rebuilding which never occurred. Despite this, the location continued to host religious activity centred around Dagan.

Inscribed objects were displayed in both phases of the Temple of Baʿlu and on the site of the destroyed Temple of Dagan. While all essentially votive and in sanctuary contexts, great diversity is apparent in the types of objects, their religious function and the social meanings likely to have been attached to them and the places where they could be seen and interacted with. This hasn't really been highlighted by past literature, which has focused mainly on the temples as architectural structures and on their inscribed contents as indices of Ugarit's relationship with Egypt, since this is where many of them originate.

All the inscribed items that can be assigned to an actual temple come from the Temple of Baʿlu (Figs 7.2 and 7.3). It's the Egyptian material that attracts the most attention, but the two Akkadian-inscribed items found in the courtyard are both unusual and worthy of discussion. One, RS 4.458, is a fragment of green stone, possibly the same stone from which the sphinx(es) of Amenemhat III (1849–1801) were made (see below). The cuneiform inscription is Akkadian but is too fragmentary to decipher. Nevertheless, based on the style and the possible connection with the Middle Kingdom sphinx, it may be Old Babylonian, and so a contender for the oldest example of logosyllabic cuneiform at Ugarit.28 Also found in the courtyard – but not included in the tables and plans of finds within the temples Callot was able to publish from Schaeffer's notes – was a clay tablet inscribed with a letter from king Niqmepaʿ of Alalaḫ to ʾIbira[nu] of Ugarit. Based on the former king, this seems to be an example of fifteenth-century diplomatic correspondence, rather than belonging to the thirteenth-century ʾIbiranu known from other documents. It deals with the extradition of a thief, so its presence in the temple is unexplained.29

Apart from these items, all the inscribed objects from the Temple of Baʿlu are Egyptian sculptures bearing hieroglyphic inscriptions – likely due in part to the pre-eminence of Baʿlu in the local cult, which made his temple a focus for the dedication of prestigious diplomatic gifts, and in part due to Baʿlu's incorporation

<sup>27</sup> Callot 2011.

<sup>28</sup> Schaeffer 1933, 120; Arnaud 1996, 47–48, n. 6.

<sup>29</sup> Arnaud 1996; Singer 1999, 620. Another early example of Ugarit-Alalaḫ diplomacy is AT 4 found at Alalaḫ, which is also concerned with matters of extradition.

*Fig. 7.2. Distribution of inscribed and other objects in the Temple of Baʿlu. Drawn by the author based on Callot (2011, 50–54 and fig. 38).*

*Fig 7.3. Distribution of objects by type in the Temple of Baʿlu. Drawn by the author based on Callot (2011, 50–54 and fig. 38).*

into Egyptian cult. Often this was through syncretism with the god Set, but one find in the temple, a funerary stele of the official Mamy ('royal scribe, overseer of the royal domain'), mentions him in his specifically Ugaritian aspect, as *bʿr ḏ*ꜣ*pwn*ꜣ – 'Baʿlu of Mt Ṣapanu', although the iconography of the (damaged) main image is typical of Egyptian depictions of Set.30 The Mamy stele is New Kingdom – that is, Late Bronze Age – but most of the otherinscribed Egyptian itemsfrom the Temple of Baʿlu (mostly sculptures of various kinds, including sphinxes bearing pharaonic cartouches) belong to the Middle Kingdom and thus predate our period of interest by several centuries. All these objects are dedicatory, although there is a contrast with the Ugaritic inscriptions in that these Aegyptiaca were prestige objects in their own rights – they *are* the dedications, rather than simply memorials of them. A letter from Merneptah to the king of Ugarit from the end of the thirteenth century gives insight into the diplomatic processes behind the presence of Egyptian royal objects in the temple:

And you wrote thus: 'May the king grant that a sculptor should come out to me to make an image of Merneptah Hatpamua in front of the image of Baʿlu which will be in the house of that god, the new one I am in the process of making for Baʿlu of the land of Ugarit.' You have expressed yourself this way. The sculptors who work here in Egypt are engaged in fulfilling their duty for the great gods of Egypt. Behold, since the king has taken his seat on the throne of Ra, these have worked for the great gods of Egypt. But as soon as they finish, the king will send to you the carpenters that you have asked for in order that they may perform all the tasks that you will command them (by saying): 'Do them!'.31

The archaeological excavation and recording of the temples took place early in the study of Ugarit and is particularly poor, but Callot's much more recent re-examination of Schaeffer's unpublished record allows us to understand something of the location of these inscribed objects in relation to the temple structures, if not the full details of the stratigraphy and complete material culture assemblage (the Temple of Dagan is particularly short on recorded artefacts, which can't entirely be due to its ancient destruction).

There are four main clusters of Egyptian material in the Temple of Baʿlu – the entrance to the holy-of-holies, the vestibule, the entrance steps and the courtyard. Most of the items in the first two categories are fragments of the Mamy stele, which points to it most likely having been displayed inside the temple. The remaining fragments are mostly in the holy-of-holies. The depths recorded for all these objects vary considerably, even for the various fragments of the Mamy stele, which indicate the inadequacy of Schaeffer's recording, the disturbance of the site or both.<sup>32</sup>

<sup>30</sup> Levy 2014.

<sup>31</sup> RS 88.2158, 10'–25'. Translation 10'–16' by the author, based on the French of Lackenbacher 1997. Translation 17'–25' from Singer 1999, 709.

<sup>32</sup> Schaeffer didn't always give enough thought to the potential for multi-storey buildings and some cases where he sees layers as resulting from construction phasing may in fact be due to the collapse of upper storeys. I am grateful to Kevin McGeough for reminding me of this.

The next main group of objects were found beneath and around the main entrance steps. These include fragments of one or more sphinxes of Amenemhat III (RS 4.416; Schaeffer mentions two, but later scholars have been less sure).<sup>33</sup> Schaeffer also mentions a statue of the Middle Kingdom vizier Senwosret-Ankh and his wife and daughter (RS 4.466+) as having been found near these, but the precise location is unknown.34 It may not have been actually within the temple and does not seem to be one of the items recorded in Callot's tables or, as a result, the plans here. The courtyard items also seem to be fragments of sphinxes or other sculptures. Other Middle Bronze Age objects were also found in these areas. From the surviving records it seems that the Middle Bronze Age material was mostly found in the collapsed rubble of the earlier temple, and therefore, while it may have been part of the temple furniture for several centuries prior to the earthquake, it didn't survive the destruction and wasn't transferred to the rebuilt structure. In Callot's view, the Mamy stele hails from this second, thirteenth-century Temple of Baʿlu.<sup>35</sup>

If this diachronic information is correct, it means that the same inscribed objects are likely to have been visible in the Temple of Baʿlu for a very long time, giving them ample opportunity to become thoroughly intertwined with how the place was conceived; however, it's unlikely that the writing was as important in the cocktail of meanings attached to these objects as other aspects – their foreignness, their antiquity, the memorialising of prestigious diplomatic relations with Egypt. Much of this comes down to the little we know about the character of the temples' interior space and how people moved through it. The temple we know archaeologically is the second, thirteenth-century one, and even that reconstruction involves a good degree of conjecture; very little is known about the architecture of the earlier, Middle Bronze Age temple, although Callot thinks it was most likely similar in form and footprint. Even moving beyond Ugarit, we don't know a great deal about how these tower temples were used. The relationship between architecture, spatial organisation and affect in Near Eastern temples is discussed in broad regional synthesis by Michael Hundley (2013), but information is limited. We can reasonably surmise that access to temples was restricted, either by social factors such as status, gender or profession, or to specific times – or, very probably, both. The interiors were probably fairly dark, based on architectural reconstructions, which could be used to enhance the sense of sacrednessthrough effectssuch asshadow orthe flickering of lamp- ortorchlight. It's probable that when most people moved through these spaces, they were engaged in ritual activity and so may have been moving in prescribed ways or engaged in ritual activities; they may not have had opportunity to linger and inspect inscriptions. For these reasons, it seems most likely that the principal effect of dedicated objects such as these would have been to add to the general splendour of the setting rather than to serve as documents intended to be read – at least by humans. The letter RS 88.2158 hints at a further dimension – that objects dedicated in the temple might be placed in

<sup>33</sup> Schaeffer 1962, 223; Helck 1976, 104; Giveon 1981, 57.

<sup>34</sup> Schaeffer 1962, 217–219.

<sup>35</sup> Callot 2011, 64.

a deliberate relationship with the cult statue: 'in front of the image of Baʿlu', perhaps standing as proxies for their dedicators, engaged in permanent devotion when the dedicator had to be away doing other things.36

We also need to take into account the effects of the earthquake and the rebuilding of the temple for understanding its social meanings in the later thirteenth century. At the practical level, we have to ask how long the rebuilding took: for how long was the temple not there, or merely a building site? For how much of our period of interest was it shrouded in scaffolding, emptied of its treasures or not fully accessible? The socio-cultural repercussions of the destruction are even harder to get a handle on. What did it mean to the various people and social groups of Ugarit to lose such an iconic and sacred structure, which had crowned the acropolis of the city for over half a millennium by the time of its destruction? What did they think about the rebuilding and what effect, if any, did it have on their engagement with the writings contained in both the old temple and the new? As well as destruction, rebuildings often afford opportunities for the rediscovery of inscribed materials from the past, and were an important way in which people of the ancient Near East engaged with the literacy practices of their distant predecessors.37

RS 88.2158 above is tantalisingly close to giving us a status update on the repairs of this building soon after the succession of Merneptah – that is, around 13–10 years before the end of the thirteenth century. However, the grammatical nature of Akkadian makes it more ambiguous than it appears in the translation cited above. The key word here is *ēteneppuššu*, a Gtn stative of *epēšu*, the verb 'to make'. Akkadian's verbal system is primarily aspectual rather than tense-based, so while this verb emphasises the step-by-step nature of the action – well captured by Lackenbacher's 'en train de faire',38 which I have rendered in English as 'in the process of making' – the actual temporal positioning of this action is ambiguous. A present-tense meaning is possible, but it could also refer to the future 'I will make, bit by bit' or the past 'I spent a long time making'.39 So we know that the construction of the Temple of Baʿlu was a piecemeal and gradual undertaking, but not precisely when it occurred in relation to the correspondence between Merneptah and his Ugaritian counterpart – probably ʿAmmurapi. We might, however, draw the reasonable inference that if the Ugaritian king was at the stage of turning his attention to the statuary, then it's more likely that rebuilding work was well in hand or perhaps nearing completion, rather than

<sup>36</sup> This is a feature of Mesopotamian temple dedications as early as the Sumerian period: see Highcock forthcoming.

<sup>37</sup> On the antiquarian/archaeological interests of especially the Neo-Babylonian kings, see Winter 2000. Of these rulers, probably the most famous for his interest in antiquity (and his professed literacy) was Nabonidus, who boasts of excavating and restoring various ancient structures. This included reading inscriptions by earlier kings, adding his own inscriptions and returning them to their original context. Evidence of the making of clay casts of ancient inscriptions has been discovered (Winter 2000, 1788–1790) and we even know of specialist ancient 'epigraphists' attached to archaeological excavations (1787).

<sup>38</sup> Lackenbacher 1997, 31.

<sup>39</sup> I am grateful to Martin Worthington for clarifying my understanding of the grammar here.

still to begin. This would also fit with the several decades that had passed since the likely time of the earthquake. In this case, for much of our period of interest, the temple was probably a building site – partially reconstructed, perhaps obscured by scaffolding and lacking its full interior trappings – a far cry from the idealised reconstructions seen in books.

As to the social dimension, our best hope is to approach the issue through modern and recent historical analogies. While I was writing this chapter,the Notre Dame fire of 15 April 2019 offered a vivid demonstration of the short-term responses of people and institutions to the partial destruction of a historic and iconic sacred structure, and how its reconstruction can become contested politically and socially. Accounts in the wake of the destruction of the cathedral's roof and spire present both secular and religious reactions to the sudden absence (or partial absence) of an old and beloved civic icon. An article in the Guardian the day after the blaze quoted a long-time resident of the Le Marais district near the cathedral as saying, 'There's something empty, missing. It was the face of Paris, now it's a face missing its teeth.'40 In the days that followed, the outpouring of sorrow and offerings of financial support from around the world received both positive approval and a measure of condemnation when compared to the lack of aid for smaller, poorer and less iconic structures and causes in France and globally. Wealthy corporations and individuals have been accused of pledging funds for restoration in the interests of publicity and associating themselves with the famous building, and then being tardy about actually paying up once media attention had moved away. There has also been controversy over how exactly the structure ought to be restored, with a vigorous debate between those who would prefer a faithful recreation of the cathedral roof as it was and modernisers spearheaded by the French President Emmanuel Macron who would like to take the opportunity to leave their mark on the structure – and the city of Paris – with a 'contemporary gesture', 'more beautiful than before'.41 Amid all this secular wrangling, we also observed a desire from the cathedral's religious community to reclaim the space as quickly as possible for religious practice: a small party of priests in hard hats returned to practise mass in the fire-damaged Notre Dame within weeks of the conflagration. This is evidence of how cult activity can continue on a site in spite of its destruction, even with only a short interruption, but no one would argue this marked the resumption of ordinary religious practice on the site: this was clearly a deliberate and symbolic act in deliberate response to what had occurred.42

<sup>40</sup> '"On our watch we let it burn": Notre Dame fire leaves hole in heart of Paris' by Angelique Chrisafis, Guardian, 16 April 2019: https://www.theguardian.com/world/2019/apr/16/like-looking-at-bombingnotre-dame-fire-leaves-hole-heart-of-paris, accessed 25 June 2019.

<sup>41</sup> The quotes are from Macron, cited in another Guardian report: 'Rooftop pool? Notre Dame proposals defy traditionalists' by Kim Willsher, Guardian, 15 May 2019: https://www.theguardian.com/ world/2019/may/15/swimming-pool-roof-notre-dame-architect-proposals-shock-traditionalists, accessed 25 June 2019.

<sup>42</sup> 'Priests in hard hats to attend Notre Dame's first mass since fire' by Angelique Chrisafis, Guardian, 14 June 2019. https://www.theguardian.com/world/2019/jun/14/notre-dame-cathedral-paris-firstmass-fire-priests, accessed 26 June 2019.

The Temple of Dagan presents a very different picture in the kind of inscribed material present. Although they are, as we would expect, votive, they differ from the objectsin the Baʿlu temple in type,script, language, origin, date and relationship with the built structure. The objects we're concerned with are two aniconic43 stelae with alphabetic cuneiform inscriptions recording acts of sacrifice and dedication, which were found on the site of the courtyard of the temple's courtyard (Fig. 7.4). These are noteworthy for several reasons and we've already referred to them on multiple occasions in this volume. Unlike the inscribed objects in the Temple of Baʿlu, they are not prestige items in their own rights, and commemorate dedications rather than being votive objects themselves. As such, they're our only examples of public writing in the city of Ugarit. This said, the inscriptions don't conform to our preconceptions of 'public writing' – they're fairly small in relation to the stones, which are themselves very small. They're also haphazardly written and positioned seemingly randomly on the stone, with KTU 6.13 not even maintaining a horizontal line (see Fig. 8.1 next chapter). The apparently rather careless utilisation of the space is in stark contrast to the efficient space-filling we see on the clay tablets, perhaps pointing to a much lower degree of experience in the medium of stone, or else a significantly different set of stylistic and aesthetic norms. Nevertheless, these were made on behalf of important citizens – the famous Queen Ṯaryelli, in the case of KTU 6.13 – and placed in the prominent location of one of the Acropolis's two principal sanctuaries.

There is, however, an additional wrinkle to all this, which is that these inscriptions must post-date the earthquake, and thus the destruction of the temple and clearing of its site. Their setting is not a historic, or rebuilt, temple akin to the two successive Temples of Baʿlu, but an open sacred precinct. This hasimportantrepercussions both for accessibility and for the social meanings of the place and the interactions people had with it. An open space – even a possibly walled and still sacred one – could offer more opportunity for people to see and engage with the writing on display. Indeed, we might wonder whether this had something to do with the choice to create the apparently new kinds of objects the inscribed stelae represent (though we shouldn't forget their small size).

The social dimension of thisformertemple site is much more difficult to approach, since the Ugaritians themselves have left us no written records of what they thought or felt about the absence of this prominent and sacred landmark, which was more than half a millennium old at the time of its destruction. It's likely that, in a sense, it was this absence that defined the site, as we can observe in contemporary places such as the Western Wall in modern Jerusalem44 or the former site of the World Trade Center in New York. Two very different sites, but both defined by – haunted by

<sup>43</sup> Pictorial stelae also existed in or around the temples, such as the very famous example showing Baʿal holding a thunderbolt or perhaps a piece of vegetation, now in the Louvre (RS 4.427). This did not have an inscription, but we have no way of knowing whether the mutual exclusivity of image and inscription across surviving stelae is mere chance or indicative of a more systematic trend in the production of monuments.

<sup>44</sup> I am grateful to my colleague Robert Crellin for this analogy. The idea of present, palpable absence and its effect on the social construction of space has been explored through Derrida's concept of hauntology and the numerous receptions of it. See, for example, Bell 1997 or Shepherd 2013.

*Fig. 7.4. Distribution of inscribed and other items from Temple of Dagan. Drawn by the author based on Callot (2011, 80 and fig. 68).*

– structures that are no longer there. In the former case, the site's sacredness derives from that of the missing building and all that it has in addition come to symbolise over two thousand years; in the latter case, the destroyed buildings were secular, but the traumatic act of their destruction lent the site a quality of sacrality. This is particularly apparent during the annual memorial services held on the anniversary of the destruction.

These modern cases can offer us a way in, a set of guides that can help us imagine the rich and deeply-felt possible meanings that may have been attached to the site of the Temple of Dagan at the end of the thirteenth century and beginning of the twelfth. Of course, context matters and no two cases are the same: they cannot tell us what those meanings actually *were*. At Ugarit itself, we know almost nothing specific – while the Temple of Baʿlu is mentioned in texts, that of Dagan is not. In fact, the rarity of mentions of Dagan has led to the suggestion that by the late thirteenth century he was a waning god, a relic of ancient times.45 That in itself would be an important ingredient in a potentially very heady symbolic and semiotic brew.

For two such doughty and significant buildings, the temples of Ugarit's Acropolis are thus remarkably shadowy beasts, not just because of the deficiencies of their excavation and original publication but because their specific histories render them spaces defined by absence, obscuredness, partialness and missingness – especially in the crucial period of the second half of the thirteenth century. It is noteworthy that the one place in the city in which writing is known to have been displayed – even to a relatively limited degree – is the prestigious 'sanctuary district' of the Acropolis. It's possible that, inasmuch as the ordinary Ugaritian thought about writing at all, they might have associated it with these sacred spaces (although a real question exists as to how accessible these spaces were and how much idea – if any – a typical Ugaritian would have had about what was inside any temple, especially the main ones on the Acropolis). This religious association may have been further strengthened if written liturgies were visibly a part of ritual performances or if magical practitioners did consult their texts when casting a spell or incantation. The prominence of the temples of Baʿlu and Dagan could then have been significant when thinking about writing in the Ugaritian cityscape – and indeed the wider landscape. If the temples – or at least the Baʿlu temple – were visible some distance beyond the city, then there is potential for an even richer blend of meanings and significances. They could quite easily have come to stand as icons of the city and of the urban cult; it's easy to envisage the house of Baʿlu surmounting the city's tell being seen as echoing or reproducing the greater Mt Ṣapanu to the north, where Baʿlu was thought to have his palace. Writing may have been fairly low down in the mix of what these landmarks meant to people looking towards the capital from the countryside, but it was perhaps there; a complex of urban-sacred-literate bound up in the icon of the tower-temple and its rising smoke.

<sup>45</sup> Feliu (2003, 64) calls him 'secondary' and 'marginal' at Ugarit.

On the other hand, I think it's easy to overestimate the importance and visibility of inscriptions in the construction of these sacred places. They were not large, prominent or positioned such that large audiences would have been able to observe them. On the contrary, the environment I have proposed is one of restrictedness and obscuredness within places that are themselves hard to get a handle on. It seems unlikely that writing was a primary attribute of the old and exotic treasures displayed in the Temple of Baʿlu, and while it seems more important on the alphabetic cuneiforminscribed stelae on the site of the former Temple of Dagan, these were still very small and unprepossessing objects, which served primarily as reminders of sacrificial practices. I doubt that many of Ugarit's general population would have associated either of these temple sites primarily with writing, though it's conceivable they may have associated writing to some extent with these places. Even if they did, this association with literacy was only one element of a complex mix of sacredness, reverence, awe, fear, resentment, envy, love, loss, regret and countless other meanings likely to have been bound up in these places – both as flourishing living buildings and ruined but still-respected remnants. Writing, through the labelling and memorialising of dedicative acts, helped construct these places as sacred, rendering ephemeral religious acts permanent and identifiable, tying them to the elite culture and specific high-status individuals responsible. It also potentially imbued the act and product of writing with a sense of sacrality they might not otherwise have had. But these were not museums of writing or places where people are likely to have done a great deal of reading.

#### **Writing and place-making in the Ugaritian landscape**

Landscapes are constituted by an intricate web ofsocially significant placesthat become represented in various ways; they are *imagined, mythologized, marginalized*, or *contested*. 46

To this quote we might also add *inscribed*. Writing is an important means for people to claim, define and represent the landscapes they inhabit, as well as landscapes with which they might only have an outsider's acquaintance. In this respect, the rural world is no different from the urban one discussed above, although the two might be constructed very differently within any given cultural frame of reference. In keeping with the urban discussion above, my goal in this section is to explore the relationship between writing practices and the social construction of the Kingdom of Ugarit's landscape.

As I hope to demonstrate, writing practices likely existed in various forms – whether permanent or transient – in the Ugaritian hinterland, but the exact nature of these practices is very difficult to specify from presently available evidence, which, even if it may in some cases have been produced by rurally situated literates, is nevertheless wholly orientated towards the needs of the urban bureaucracy. The voices

<sup>46</sup> Harmanşah 2013, 30. Italics original.

of rural Ugarit are consequently lost to us: we know next to nothing about how rural people perceived and interacted with their land, about their folklore and beliefs or about whether writing practices were part of these. As a result, the two areas where we can say something meaningful about the interaction of writing practices and the construction of the rural landscape both centre on urban elites.

Firstly, we can reflect on the ways in which writing shaped urban interaction with and perception of the countryside. The gathering of administrative information on rural populations and their economic activities has implications for the presence of writing and literacy outside the capital, as we've seen, but these kinds of administrative documents also shape how the people who use them conceptualise and engage with the places being administered. They aren'tsimply a passive reflection of how the landscape *is*, but represent a framework for understanding and interacting with it. They order, define and commoditise that landscape through the written word. They rival and potentially supplant other forms of knowledge and ways of experiencing the landscape through myth, memory and lived experience. They are the geography of the Ordnance Survey and the tax-collector, not of the herdsman, the farmer or the teller of folk tales. This landscape of rural administration existed for the administered too, but its strongest impact is likely to have been on those dealing with the tablets and the information contained on a daily basis – the urban elite world of bureaucrats, officials and landowners.

It would be interesting to know how well-established this kind of administration was in Ugarit – whether we should be thinking in terms of a new shift in urban elite attitudes to the countryside or a much more ingrained set of relationships. In the extant corpus, rural administration is strongly associated with alphabetic cuneiform, but one assumes that administrative documents probably existed before the advent of the new script, presumably in logosyllabic cuneiform. But this returns us to the currently intractable problem of whether Akkadian was used at Ugarit before the Hittite takeover, and if so why none of this documentation survives.

The second area where we can make some progress concerns monumental inscription in the landscape – or rather, the lack of it. The absence of such public writing that we observed above for the urban spaces of Ugarit extends out into the hinterland. No monumental inscriptions or reliefs of any kind are known from the territory, in stark contrast to much of the surrounding region. As mentioned above, the practice of rulers or other elite personages erecting monuments – inscribed, pictorial or a combination of both – at significant locations in the countryside is extremely well-attested in the ancient Near East. In Phoenicia, as we've seen, there are the inscriptions at the Nahr el-Kelb, among others; in Anatolia and the Hittite region of northern Syria, the hieroglyphic Luwian writing system is particularly associated with rock-cut inscriptions and monumental orthostats, which continued to be produced well into the Iron Age and survived the transition into the linear Phoenician script at sites like Karatepe. In Mesopotamia, rulers frequently erected stelae or had inscriptions carved in rock to commemorate victories or at important ritual sites.

Egypt's proclivity for monumental writing is so well known it hardly warrants mentioning.

The curious absence of such material from the territory of Ugarit can best be explained through comparison with the situation in Phoenicia. The regions have similar climates and cultures, and comparable Late Bronze Age political situations, whereby polities occupied a niche at the edge of imperial control, walking a line between vassalage and autonomy.47 The monumental inscriptions of Phoenicia were, for the most part, made by great imperial powers during transitory incursions into the region. In the Late Bronze Age this was mainly Egypt; in the early first millennium it was Assyria. They were erected on important routes, especially at choke-points such as the Nahr el-Kelb river crossing. These monuments to imperial reach and military campaigning are, of course, all about claiming the land and legitimising their presence there. As Ömür Harmanşah notes,

foundation of new cities and carving of rock reliefs as commemorative monuments, seem to be correlated in interesting ways as two significant colonizing gestures of taking hold of new territories of settlement. They share a powerful rhetoric of legitimation through inscribing places by royal interventions with the implication of previously 'untouched' landscapes, be it an inviting rock surface or an agriculturally fertile yet uncultivated territory (*terra nullius*).48

If we exclude such foreign 'imperial' inscriptions by Egyptians or Assyrians, then monumental inscriptions in Phoenicia are notably sparse. The only potential candidates from this period are a number of stele fragments in the notoriously inscrutable Byblos syllabary, which have essentially no contextual information.49 One of these, Stele g, is divided into vertical columns in a way that recalls Egyptian hieroglyphic practice, while others are ruled horizontally and more closely resemble the early Phoenician inscriptions found at Byblos and dated to the early tenth century BC. Indeed, a direct connection between the syllabic and linear alphabetic inscriptions may be suggested by the fact that the alphabetic inscription of the tenth-century king Yehimilk seems to be a palimpsest, with traces of an earlier syllabic inscription underneath (Fig. 7.5).50

It's tempting to interpret these Byblian examples in relation to Egyptian imperial monuments, as a localised Phoenician experimentation with, and transformation of, an Egyptian-inspired practice of publicly displayed writing. Byblos, after all, had extremely deep-rooted connections with Egypt and the importance of Egyptian stimuli for local writing practices is apparent in the hieroglyphic inscriptions on Middle Bronze Age Byblian kings' tombs and perhaps in the 'pseudo-hieroglyphic'

<sup>47</sup> There is, of course, a key difference in that Ugarit was nominally a Hittite vassal while the Phoenician polities were independent. In practice, however, all had a significant degree of autonomy which was contingent on keeping on the good side of whichever great power exerted the most influence. See Boyes 2013.

<sup>48</sup> Harmanşah 2013, 49.

<sup>49</sup> Vita and Zamora 2018; Sass 2019.

<sup>50</sup> Vita and Zamora 2018.

*Fig. 7.5. The Yehimilk Inscription (KAI 4), Byblos Museum. Photo by the author.*

script itself. Even so, this is an extremely limited local engagement with the production of this kind of public writing – as far as we know, it's only confined to Byblos, and these seem to be urban objects recovered from the city itself, not from its hinterland. They're not indicative of a practice of monumental inscription of the wider landscape. We can't read the syllabic inscriptions, but the alphabetic Yehimilk inscription is a religious dedication relating to the building of a temple. It doesn't necessarily follow that the syllabic inscriptions should also be dedications, but it wouldn't be at all surprising if they were. If so, they would be broadly comparable objects to the Ugaritian dedicatory stelae from the Temple of Dagan and represent a Late Bronze Age(?)51 experimentation with public writing in religious contexts, but not any kind of large-scale and military-imperial inscription of the landscape.

We can tentatively suggest, then, that elite culture in the northern, coastal Levant had little interest in writing for display,52 and at the end of the Bronze Age

<sup>51</sup> Dating the Byblos syllabary is notoriously uncertain. Sass (2019) offers a recent attempt at a radical redating to the ninth century, in line with his similar reconfiguration of the linear alphabet (see Chapter 3 above). This convincingly points out the evidence in favour of the chronological proximity of the syllabary and linear alphabet, but is ultimately unpersuasive in its main goal. Well aware of the lack of solid evidence on which to date the script, Sass is reliant on a single arrowhead, broadly datable on typological grounds to the thirteenth to ninth centuries – *i.e.*, fully in keeping with dating the Byblos syllabary in the Late Bronze/Early Iron Age transition, with a potential overlap with the linear alphabet towards the end. While acknowledging that an earlier date can't be ruled out, Sass prefers to locate the entire usage of the script within a very narrow window at the very end of this chronological bracket, essentially to support his redating to the linear alphabet.

<sup>52</sup> There are two more possibilities, but neither is very convincing. The first is that monumental inscriptions exist (or existed) in Ugarit's territory but we simply haven't found them becase of the lack of

was only just beginning to experiment with such forms, particularly in cities with close and long-lasting relations with Egypt, such as Byblos and Ugarit. The key difference between Ugarit and Phoenicia, which explains the absence of monumental rural inscriptions in the former, must then be their specific relationships with the imperial powers that were the main producers of such writing. Ugarit's relationship with the Hittites was on a more formal footing than were the Phoenician polities' with Egypt. Ugarit was a Hittite vassal, of course, but it was not conquered; it entered the Hittite sphere of its own accord. This was something that evidently mattered to the Hittites and contributed to the degree of autonomy Ugarit was afforded and the relative lightness of Hittite presence in the Kingdom. A reluctance by the Hittites to tramp about putting their stamp on the Ugaritian landscape would be very much in keeping with the general light touch of their relationship. On the Ugaritians' part, the lack of emulation of imperial practices such as the production of monumental rural inscriptions marks a difference from some of its neighbours in north Syria, specifically those that would go on to form the 'Syro-Hittite states' of the Early Iron Age. It seems part of a general coolness towards Anatolian elite culture, and perhaps globalised traditions of prestige in general, which we will discuss in much more detail over the coming chapters.

#### **Writing and place at Ugarit**

Writing is an important contributing factor in how humans construct landscapes and cityscapes. The fact that we have such fragmentary evidence from Ugarit doesn't change that. This isn't to say that it was necessarily the most important factor in creating Ugarit's social spaces. On the contrary, it seems to have been rather restricted. As far as we can tell, monumental and other public inscriptions were not something the people of Ugarit were greatly interested in, either in the city or in the hinterland. The kingdom's location and political status meant there was little reason for the imperial

archaeological survey of the kingdom's countryside. While one or two inscriptions may lie undiscovered in remote places, however, a lack of archaeological investigation is not the same as the region being *terra incognita*. Ugarit's hinterland has been inhabited, explored, exploited and transformed for millennia, and the clear identity of modern village names with locations in the Ugaritic tablets attests to the continuity of population and memory. Moreover, the whole point of a monument is to be visible, to be remembered. It seems unlikely that if Ugarit had an ample supply of rural monumental writing, nobody should have remembered or mentioned it at any point over the intervening three millennia.

Another possible explanation that can only go so faris the nature of the region's climate, flora and terrain. Many of the neighbouring regions where such inscriptions are found are flat, arid and rocky, presenting relatively clear visibility for monuments. By contrast, Ugarit is hilly, has wet winters, and in antiquity was heavily forested. This has an impact for the modern discoverability of inscriptions, but could be argued to have discouraged their creation in the first place. Arguably, Ugarit's countryside presented a less attractive surface for elite inscription than did the regions to the north or east, unless you wanted your timeless memorial to be obscured by trees and eroded by the weather. However, much the same climate and topology exists in Lebanon too, and that doesn't seem to have deterred the creation of rural monuments there.

great powers to make their own textual marks on the environment. The only places where we know inscribed objects were displayed were the acropolis temples, and the writing was probably not the most salient feature of those objects. Access to it was probably restricted both in absolute terms – who was allowed to enter the sacred precincts – and by the features of the temple setting – its likely darkness, prescribed ways of moving through and interacting with the space, and so on. On top of this, the temples themselves by the late thirteenth century existed in a rather liminal state – under (re)construction or absent but nevertheless the site of ritual practice. These were certainly very meaningful spaces, but what did writing contribute to those meanings, and for whom?

The only other places where we know for certain inscribed objects were gathered were also restricted – private residences belonging to elite members of urban society. It's likely the *practice* of writing and the coming and going of materials and personnel would have been visible to locals, at least at certain times, but the writing itself is rather unlikely to have been widely accessible at these locations. As such, we're faced with a similar quandary as for the temples: how much did writing figure in how ordinary members of Ugaritian society conceived of and interacted with these places? How did it differ according to social factors such as status, profession, gender and so on? These are not questions we're presently well-equipped to answer.

The places with the greatest potential to allow ordinary Ugaritians to encounter writing and writing practice are the ones we're least able to define with any precision – gates, market stalls, workshops and rural villages – the places of routine administration and data collection. We can infer with varying degrees of certainty that these kinds of encounters with literacy probably took place, but archaeological investigation to date has not focused on these kinds of locations. Some are potentially locatable if we're fortunate and look in the right places: we might hope to find tablet collections or remains of writing paraphernalia at the city gates – as we may have done near the entrance to the Royal Palace in the shape of the Western Tablet Store. More optimistically, we could envisage a future rural village excavation uncovering evidence of a notary or local literate official's office, or a workshop with inscribed material. The nature of rural literacies will be discussed in more detail in the next chapter. Other associations of place and writing would by their natures have been more evanescent: the inscribed pot on the market stall, the peripatetic administrator on an accounting tour of the hinterland, the businessman travelling with his secretary. In these cases, the relationship is not a fixed and precise one, but rather the potential that repeated actions of these kinds might accumulate associations with particular places. If you hardly ever encounter writing, but when you do it tends to be on market stalls, or in the fields when the government representative comes calling, or in the kinds of places where travellers like to dictate their letters, then writing can become part of the meanings bound up in these places – and vice versa. This is not something we can pin down archaeologically – at least, not easily. But nor is it a dimension we can afford to ignore when we think about the place of writing in an ancient society.

The challenges inherent in the kind of social approach to writing and place attempted in this chapter, and the many ambiguities of the discussion I have offered, demonstrate the need to refocus our approach to the archaeology of writing at Ugarit and elsewhere. The time has come to look beyond the great tablet-houses of the capital and to redouble our efforts on the less prestigious places – the rural, the industrial and the workaday. This shouldn't primarily be a search for tablets – anyone attempting such a thing would be likely to come away disappointed – but for a better understanding of how place, landscape and cityscape were constructed in general. If, as it seems, writing was probably a rather minor part of the way people created and negotiated the places of their lives, then it's only by focusing on the other factors too that we'll be able to define the role it did play in any detail. This is, of course, a restatement of the main theme of this book – that social practices and material culture are all fundamentally entangled: to privilege one at the expense of others prevents us from adequately understanding them all. I hope in this chapter I've begun to sketch out the beginnings of an understanding of the relationship between writing and social space, but to fill in the details we will need a new approach to the archaeology of the Kingdom of Ugarit.

### Chapter 8

### Who wrote? Literacy in Ugarit

This chapter is broadly concerned with the question of 'literacy', in the sense of how many and what kinds of people were writing in the Late Bronze Age Kingdom of Ugarit, and what kinds of things they were writing. However, 'literacy' is a term that has acquired a great deal of baggage over the years – and indeed, different kinds of baggage in different disciplinary and socio-political contexts. In anthropology, questions of literacy call to mind the long-running debates surrounding the so-called 'literacy thesis' and the 'great divide' twentieth-century writers such as Jack Goody or Walter Ong saw between 'literate' and 'oral' societies,1 or the numerous models that have supplanted them since they fell from favour in the 80s and 90s.2 In contemporary education, however, the idea of literacy is approached very differently, as a question of basic skills and paedagogy.3 In developmentstudies, the discussion has been different again. All these fields overlap to some degree, of course, and all have something to contribute to our understanding of the culture surrounding writing at Ugarit, but it's important that we recognise and unpack the complexities inherent in terms like 'literacy' before we begin. At the same time, we must also understand that together with literacy come the ideas of illiteracy or non-literacy (apparently similar terms, but with slightly different connotations). When we think about who was literate in Ugarit – whatever we mean by that – we can't ignore the people who were not. Their interaction with writing culture should be just as important to us as that of their literate counterparts, although it can be even more difficult to reconstruct, for obvious reasons. This is not to revive Goody and Ong's dichotomy between the 'literate' and the 'oral' – as we will see, it's now clear that those are richly intertwined – but to understand that writing can affect members of society beyond just those who are able to take advantage of it directly. From changing administration practices to its effects

<sup>1</sup> Goody and Watt 1963; Goody 1968; 1987; Ong 2012.

<sup>2</sup> For a summary of developments in the anthropology of literacy, see, for example, Bartlett *et al.* 2011.

<sup>3</sup> *e.g.* Martin‐Jones and Jones 2000.

on literature, myth and public ritual, writing has effects that ramify out well beyond the perhaps relatively small number of actual writers.

#### **Types of literacies**

Most scholars who write on literacy in the ancient world begin their discussions by highlighting that it isn't a simple binary: one is not simply 'literate' or not.4 There are degrees of literacy, types. One can be literate to different extents in different genres of writing, or the term can be broadened out and acquire a more-or-less metaphorical quality when applied to other kinds of knowledge, which can be very much dependent on the specific social context. Within our own society, for example, we're accustomed to hearing terms such as 'computer literacy' or 'financial literacy', but it could also be applied to the ability to read the landscape, the weather, to divine the hidden meaning in the flights of birds orto decipherthe will of the gods from the entrails of a sacrificed animal.5 It is people's relationship with writing that is the subject of this book and so our primary interest here, but we must also recognise that the value we place on being able to read and write as the most fundamental form of knowledge is very much a modern perspective and is unlikely to have been shared by the people of Late Bronze Age Ugarit.

Most scholars who've discussed the matter recently have agreed on the need to distinguish multiple literacies in the ancient world, but the categories they have identified have differed according to each of their areas of interest and on the societies they have been focused on. For Mesopotamia and its logosyllabic writing system, Niek Veldhuis limits his discussion to three 'levels' – functional literacy, technical literacy and scholarly literacy. He defines the first as the ability to 'write or read a letter or an ordinary business document';6 the second as the ability to work with technical jargon and specialist orthographies such as those that appear in divination reports, mathematical texts and so on,7 and the third as involving 'knowledge of all the ins and outs of the cuneiform writing system and its history'.8

Those working on literacy in the classical world have often been more open to larger pluralities of types of literacy. Among the many kinds that have been identified are artisan's or craftsman's literacy – the ability to write short, formulaic texts related to craft production or distribution;9 the ability to recognise one's name for purposes

<sup>4</sup> *e.g.* Thomas 2009; Veldhuis 2011. This approach owes much to the so-called 'New Literacy Studies' spearheaded by Brian Street (1993), which supplanted Goody and Ong's approach and sought to examine literacy not as an autonomous thing in itself, but as a set of social practices embedded in socio-cultural context. See also Bartlett *et al.* 2011.

<sup>5</sup> Sasson 2005, 218.

<sup>6</sup> Veldhuis 2011, 71. As Thomas (2009, 16) points out, the term 'functional literacy', while ubiquitous, lacks any consistent meaning and is entirely dependent on the specific social context for what sort of ability to read and write is necessary for someone to get by.

<sup>7</sup> Veldhuis 2011, 73–74.

<sup>8</sup> Veldhuis 2011, 74.

<sup>9</sup> Harris(1989, 8) definesthisin a rather more totalising way ('not the literacy of an individual craftsman but the condition in which the majority, or a near-majority, of skilled craftsmen are literate, while

of democratic or legal participation; the ability to recognise a legal document or contract (even if not entirely able to draft one); the ability to scratch a graffito; the ability to write an epic poem, and so on.10 These categories are, to an extent, arbitrary; the point is not so much to create a definitive, cross-culturally applicable, taxonomy as to illustrate and explore the diversity of practices.

Along with this proliferation of forms of literacy, a key question is the degree to which they are integrated – how easily do skills transfer from one category of practice to another, or how specialised are particular areas of writing? Do people tend to focus on one, or a small number of, types rather than having general writing and reading abilities that are applicable across a wide range of fields and functions. Greg Woolf discusses this question in the context of the Roman Empire, concluding that:

New forms of document emerged, along with new kinds of readers well equipped to use them. A few – such as legal formulae or the labels on Dressel 20 amphorae – were highly specialized. But the peculiar conditions of Roman alphabetic literacy, and in particular the centrality of the aristocratic slave household in most of these webs of exchange, held Roman literacies together. There was no real fragmentation of writing practices, no specialized literacies, and the practices of writing – in particular the use of complex formats, of a set of graphic symbols, and of particular resonances associated with personal names – moved easily between different genres of text. Roman writing practices, in brief, were joined up.11

For the Near East, this question of integration brings us back to the discussion about the role of the 'scribe'. If we see writing culture in the Near East as fundamentally 'scribal' – that is, essentially concentrated in the hands of a relatively homogeneous group of professionals who can be usefully labelled scribes – then almost all writers are assumed to have passed through more or less the same educational process, even if some may have taken it further than others. Diversity of literacy is consequently essentially one of level achieved, with a small amount of horizontal variation due to specialisation. Fundamentally, the writing practices are integrated through the scribal education system and the group culture and identity that result from it. In practice, even with the dominance of the idea of scribes in Near Eastern studies, it's rare for anyone to argue that literacy was solely restricted to this narrow elite; nor is it assumed that every 'scribe' was able to read every text. However, the vast majority of our data relates to this 'scribal' world and so with the best of intentions, discussions inevitably tend to focus on them.

women and unskilled laborers and peasants are mainly not, this being the situation that prevailed in most of the educationally more advanced regions of Europe and north America from the sixteenth to the eighteenth century') than I am using it here. When I talk of 'craftsman's' or 'artisan' literacy, I simply mean the ability of some people to write inasmuch as is necessary to produce or distribute objects that require specialist skills to manufacture, such as inscribed vessels, votive objects, lapidary inscriptions and so on.

<sup>10</sup> Thomas 2009; Woolf 2009.

<sup>11</sup> Woolf 2009, 61.

Given the scepticism I have voiced towards the term 'scribe' and the idea of a homogeneous 'cuneiform culture', it will come as no surprise that I think we should beware of making the scribe our central figure when thinking about types of literacy. A key aim of this chapter and the next is to explore the possibilities of diversity within the backgrounds of those who wrote at Ugarit, how they acquired these skills and how they put them to use. Even when they are formally trained literates who would be covered by the conventional label 'scribe', there remains considerable scope for differences in terms of background, identity and outlook. As well as the purely demographic, there is also a geographical component to this: the restriction of literacy to a conservative and metropolitan bureaucracy implies physical as well as social centralisation, while the possibility of writing outside this structure adds urgency to the underexplored question of whether there was literacy in the Kingdom of Ugarit beyond the major palatial centres of Ras Shamra and Ras Ibn Hani. As we'll see, this approach often involves asking questions that the available data isn't always able to answer. Nevertheless, it's important to broaden the discussion regarding Ugarit's writing practices as much as possible beyond the stereotype of high-status men working in service of the religious or political administrations.

#### **Women's literacy**

Probably the most overlooked group within discussions of literacy in Ugarit – if not in Ugaritian studies in general – comprised around half the kingdom's population. Rather little attention has been given to whether any of Ugarit's women were able to read and write. The evidence here is sparse,12 but there are a few reasons why we might expect to find evidence of female literacies at Ugarit.

The first is comparative. There is good evidence for women writing in the ancient Near East at various points in its history and in different places. The best evidence for this comes in the earlier periods of Mesopotamian history, though we should add the caveat that these are also the times for which we have the fullest data regarding writing practices in general. Sumerian doesn't employ gender markers to distinguish male and female scribes,13 but we know of high-status Sumerian women, such as the princess and priestess Enheduanna, who seem to offer prominent examples of female literacy. However, it can be difficult to disentangle the acts of composition and actual writing. Some attribute the physical practice of writing down Enheduanna's hymns to a professional writer working on her behalf, and others even argue that the compositions themselves were only attributed to her as flattery or to garner additional prestige from association with her name and that they were really the work of others – presumably men.14 Divine comparisons aren't always reliable guides

<sup>12</sup> On the lives of women at Ugarit, see Schloen 2001; Marsman 2003; Amico Wilson 2013; McGeough 2016; Yon 2016, although none of these directly address questions of literacy in any detail.

<sup>13</sup> Meier 1991, 541.

<sup>14</sup> Lambert 2001; Black 2002.

to everyday earthly practice, but we can also note that the Sumerian deity of writing was a woman, Nisaba. She continued to feature in the cult practice of literates right up until the end of the Bronze Age – she is mentioned in several colophons from Ugarit, for example – but like many Sumerian goddesses, over time she was joined and more or less supplanted by a male equivalent: Nabû, adding to a widespread narrative in which Sumerian openness towards women was gradually rolled back in later periods.

Literate women are also attested in the Old Babylonian period, such as fourteen women in the cloister (effectively a kind of Mesopotamian nunnery)15 of *nadītu*-priestesses at Sippar, or nine from Mari.16 Probably the most interesting examples from this period, however, are the correspondence between Old Assyrian merchants at the Anatolian trading emporium of Kaneš and their families back home.17 These deviate markedly from accepted standards of professional writing: judged according to these the handwriting is poor and the syntax and grammar strewn with errors (Mogens Trolle Larsen goes so far as to call many of them 'outrageously hideous'),18 which argues in favour of them having been written by the correspondents themselves, rather than their employing a freelance professional writer for the purpose, as Pearce has suggested.19 Letters from women left behind in Aššur feature prominently in this corpus, and there's every reason to assume that these middle class women engaged in commerce wrote them themselves.

The evidence for the later second millennium is patchier and there is less material directly concerned with the careers and education of writers themselves. However, material from the first millennium may point to a continuation of female literacy in Mesopotamia. For example, female writers are recorded in the queens' palaces at Nineveh and and Kalah.20 This raises a couple of interesting questions. The first is whether it represents a shift from the primarily sacral context of female literacy in the temples and cloisters of the third and earlier second millennia to a more secular world of royal households. This is hard to address given the very limited evidence available, but is potentially significant. The second question is the relationship between female writers and the households of powerful women. To a degree, the literate *nadītu*s of the Old Babylonian period can already be seen in these terms, since the cloister functioned as a prestigious, female-only household. The Old Assyrian commercial families whose menfolk were away in Kaneš can also be seen as effectively female-headed households. Samuel Meier (1991) cites a number of examples of Mesopotamian women expressing a preference for female writers and female messengers, if possible. There is also a tendency for other servants of elite Near Eastern women to often employ other women as servants if possible. It is reasonable to assume – though by no means an absolute


<sup>15</sup> Harris 1963.

<sup>16</sup> Meier 1991, 542. On gendered representations of literacy in Old Babylonian-era Sumerian literature, see also Robson 2007.

<sup>17</sup> Larsen 2015, 54ff.

<sup>18</sup> Larsen 1976, 305.

correlation – that female writers who were not members of the elite themselves were most likely to have been employed by high-ranking women.

This brings us back to Ugarit and the second reason why we might expect to find signs of female literacy at the city – the association of writing with the queen. Ugarit's queens were evidently people of profound political and economic significance.<sup>21</sup> They had their own households that were distinct from the king's: one legal text refers to the *sākin bīt šarrati* – 'governor of the house of the queen' (RS 17.325).22 This has sometimes been interpreted as meaning the queen had her own palace, which various commentators have located in the so-called Queen Mother's Residence, or at the Northern Palace of Ras Ibn Hani. A number of documents mentioning the queen were found at the latter but van Soldt (2013) has convincingly argued against either of these structures being securely associated with the queen. Both he and Juan-Pablo Vita believe that the reference in RS 17.325 is to the queen's *household* rather than a physical house.23 This aside, numerous documents show Ugarit's queens to have been involved in diplomacy, on their own accounts (*e.g.* KTU 2.3) and in conjunction with the king (*e.g.* RS 19.70). Domestically, they had an important role as intermediaries between petitioners and the king (to which we might adduce the mythological analogues of ʾAṯirat and to a lesser degree ʿAnat as intercessors between Baʿlu and ʾIlu in the Baʿal cycle), and legal documents attest to their management of private estates, dependents and business interests. They were also involved in cult practice, as shown by the inscribed stele KTU 6.13 (Fig. 8.1), which records the sacrifice of a bull by the long-lived and powerful Queen Ṯariyelli on the site of the former temple of Dagan.

From diplomacy and business to religion, all these functions involved writing, and we have a significant corpus of documents relating to the queen's household. These primarily include letters in Akkadian and Ugaritic – from both male and female correspondents – and legal texts, as well as otherforms of writing such asthe sacrifice stele mentioned above. Many are found in the Royal Palace, but a good proportion are also scattered through the private archives across the tell.24 It's often impossible to determine exactly which queen each text relates to since names are used less commonly than titles, but there's no reason to assume that any of this material was written by the queen herself, any more than we think the king sat down and wrote out his own diplomatic letters and accounts (indeed, most letters are written *to* the queen, rather than from her, for obvious reasons; but a number allude to outbound

<sup>21</sup> There is some discussion over the nature of queenship at Ugarit. Most agree that it was a position held for life, but opinions differ as to whether the wife of the king became queen immediately or only upon the death of the previous queen; in other words, could there be multiple queens simultaneously, or only one? Ugaritic certainly does not distinguish in title between a queen and queen-mother, and some queens such as Ṯariyelli appear to have been exceedingly long-lived, with the result that it is not always easy to determine how many individuals our queenly correspondence relates to, or their exact relationship with the kings who appear in the texts.

<sup>22</sup> van Soldt 2013, 9.

<sup>23</sup> Vita 1999, 469–470; van Soldt 2013, 9.

<sup>24</sup> For a thorough discussion, see van Soldt 2013.

*Fig. 8.1. Stele recording a sacrifice to Dagan by Queen Ṯariyelli (KTU 6.13). From Bordreuil and Pardee (2009), Text 14. Reproduced by kind permission of Projet PhoTEO, Mission de Ras Shamra.*

letters). In fact, we have a couple of instances of colophons in which writers refer to themselves as working in the queen's household. It's not unreasonable, then, to wonder whether some of these literates employed by Ugarit's queens to take down their letters and conduct their business could have been women.

A third reason why we might expect female literacy at Ugarit is related to this: even outside the royal sphere, it has previously been asserted that Ugarit's was a relatively progressive society in terms of its treatment of women – at least in relation to the first-millennium southern Levant, which is the usual comparison made.25 Eleanor Amico Wilson concluded that '[t]here is nothing whatever of misogyny in the culture. If the queen and goddesses were models, women's opinions were respected, and women were expected to take part fully in Ugaritic life.'26 As we've seen, high-status women at least were able to participate in a wide variety of public life, including the wielding of political power and the ability to own property and conduct business on their own behalfs.27 Hennie Marsman largely agrees with Amico Wilson on the relative liberty afforded to women of the highest rank, but is considerably more circumspect when it comes to the general population, given their almost total absence from Ugarit's economic administration texts: 'Ordinary women were probably worse off. Their contribution did not count, not even if they

participated in the economic life of the kingdom. To a large extent they were invisible'.28

It's hard to argue with this assessment. The very lack of data points to the marginalisation of Ugarit's women, even if they were not marginalised quite as comprehensively as in some neighbouring regions. Archaeological indications are also suggestive in this regard. In the city's urban planning and architecture, for example,

<sup>25</sup> Marsman 2003; Amico Wilson 2013.

<sup>26</sup> Amico Wilson 2013, 204.

<sup>27</sup> On the legal rights of women, see McGeough 2016, 481–484.

<sup>28</sup> Marsman 2003, 680. The position of even royal women was precarious, however, as ʿAmmiṯtamru's divorce and presumed execution of his wife makes abundantly clear. This will be dealt with in more detail in Chapter 10, and see also Thomas 2019.

Schloen (2001, chapter 6) discerns a strong concern with privacy and restriction of access to domestic space by those outside the family unit, which cross-culturally often goes together with practices limiting the visibility, freedom and agency of women. However, there have been no major studies of what archaeology can tell us about non-elite women's lives at Ugarit. Their invisibility, both then and now, represents a major gap in our understanding of the city and its culture; one that goes unremarked far too often.

One possible area of significance that we're missing is the role of women in craft activities, and the relationship between these and writing. We will discuss 'craftsman's literacy' below, and especially the idea that this may have been a locus of non-elite literacy and the use of variants of the alphabetic cuneiform writing system outside the world of state-aligned bureaucrats and intellectuals. What's important to note for the present is that craft activities are often extremely gendered,29 and that we don't really have a good understanding of how this worked in the Near East at this time. Some work has been done on the intersection of gender, craft practice and identity in Mesopotamia, but this has tended to focus on the south-east in the third millennium and is consequently distant from Ugarit in terms of both time and space.30 Pottery production is of particular interest here, given the ability of vessels to be inscribed with writing or writing-like signs. However, no female potters are mentioned in documents from Ugarit (though since no women workers of any kind are thought to be worth mentioning, we perhaps shouldn't read too much into this).31 But we shouldn't overlook the possibility for writing to be incorporated into other crafts, such as textiles, or the role of writing in activities surrounding craft production and distribution, such as labelling practices, accounting, rations and taxation, and so on. Joanna Smith has noted that pottery manufacture on Cyprus during the Late Bronze– Early Iron Age transition is likely to have been heavily female-dominated, although it's not entirely clear what she bases this conclusion on, apart from the household context and a close relationship with textile production – neither of which we should automatically assume must stand as proxies for women.32

We should of course remember that writing itself was a form of craft production, involving a full *chaîne opératoire*. Even if women were not involved in the actual business of inscription, there's every possibility they may have been involved in other parts of the production process, from gathering raw materials such as clay, water, wax, wood, reeds or orpiment, to shaping blank tablets or producing writing boards. As mentioned in Chapter 6, there's no reason why these 'preparatory' activities need to have been undertaken by the writer themselves.

The social archaeology of craft activities at Ugarit is an area of research whose importance reaches well beyond its relationship with writing, but it has been

<sup>29</sup> See, for example, Costin 1996.

<sup>30</sup> *e.g.* Wright 1996; 1998; Pollock and Bernbeck 2000.

<sup>31</sup> Marsman 2003, 688.

<sup>32</sup> Smith 2009, 240–242.

essentially untouched so far – in part due to the neglect of the polity's material culture in comparison to its documentary record. A fuller understanding of this area would go some way toward filling the lacuna in our knowledge surrounding non-elite people in general and women in particular. I suspect it may also open new directions for our investigation of writing practices beyond the political and religious bureaucracies. At present, however, there is very little to be said regarding this *terra incognita*.

We have, then, a number of reasons why it would be reasonable to think women *could* have written at Ugarit. However, there is currently no evidence whatsoever that they actually *did*. None of the writers' names we know are obviously female. Nor do terms for female writers – either using gendered determinatives or feminine forms such as Akkadian *ṭupšarratu* or the unattested Ugaritic \**sprt* – occur in any texts found at Ugarit.33 Even in the household of the queen, the only gender-specificity we find points towards men. The colophons of the writers employed by the queen use male determinatives, as in this example from a legal text found at Ras Ibn Hani:

[šu-t]i m XXX LÚ A. BA [xxx SA]L. LUGAL-ti KUR Ú-ga-ri-it[34 […] MrXXX Mrscribe […] of the queen of Ugarit.

or this, from Ugarit itself:

[šu PN] lúdub.sar lúka[b-zu-zu ša…]/[….]x-di lúsukkal munus.lugal […]/[…èr d ]AD ù d nisaba.35 [of PN] Mrscribe, Mrpupil of … Mrofficial of the queen … servant of Nabû and Nisaba.

The question, then, is how we should interpret this situation. Certainly there's no basis for saying with any certainty that women wrote at Ugarit; but this isn't the same as saying that they definitely did not. It's equally possible that women wrote in the same invisible way that they did their other day-to-day activities, that for whatever reason they didn't draw attention to themselves in colophons. Given the small numbers of female writers known across the whole span of ancient Near Eastern history, we shouldn't be too surprised that none are visible in the quite brief period for which writing is attested in the small kingdom of Ugarit.

Nevertheless, the evidence at present does point one way, and this is the same direction as later, male-dominated, writing culture in Israel-Palestine. So let's assume, for the sake of argument, that the absence of women writing at Ugarit is real. This would mark a difference from Babylonian and Assyrian writing practices, one all the

<sup>33</sup> It is debatable whether it occurs in ancient Hebrew. The only possible example is in Ezra 2.55 and Neh 7.57, where it seems to be a name – usually taken as a man's name. In one of these cases, it incorporates a definite article as if it is a title 'the female scribe', making this occurrence hard to interpret – though the use of the feminine participle *qhlt* as a masculine title may provide an equivalent. I am grateful to Stephen Bigger and Ola Wikander for this information (pers. comm.). A different word *sprt* does occur at Ugarit, meaning a type of document – see del Olmo Lete and Sanmartín 2015, 758.

<sup>34</sup> RIH 77/9. Arnaud and Kennedy 1979, 321. On the sumerogram A.BA, see Roche-Hawley and Hawley 2013, n. 30.

<sup>35</sup> RS 22.437B. van Soldt 1988, 315.

more striking given the measure of economic and social freedom afforded to Ugarit's higher-status women and the strong evidence for their use of written documents to conduct their affairs. Ifthisis a real feature of writing practices at Ugarit, then it would seem to be part of a specifically Levantine set of attitudes to the role of women and their relation to writing, and would be one of the ways Ugarit's people adapted and moulded the practices they borrowed from the cuneiform world to suit local culture.

#### **Writing outside the establishment**

Discussions of writing in the Kingdom of Ugarit – this one included – focus overwhelmingly on the capital, its ancillary sites at Minet el-Beida and Ras Ibn Hani, and on the elite bureaucracy comprised of literate officials, secretaries, ritual practitioners and scholars. This is understandable given that almost all our evidence relates to these groups. However, it's also important that we consider the possibility of writing beyond this elite, metropolitan world. This takes us into a number of interrelated discussions about factors such as status, landscape, economic organisation and social mobility.

### *Craft, literacy and variation*

To begin, let's return to the question of so-called craftsman's – or perhaps better, artisan's – literacy, since this is the closest we come to directly attested writing practices outside the bureaucracy. There is a small but diverse assortment of inscriptions from Ugarit and the surrounding region on objects other than clay tablets. Generally, discussion of this material has focused on the script used, which is often (but certainly not always) one of the multiple non-standard variants of alphabetic cuneiform (see Chapters 1 and 5). Alongside the important questions regarding script, however, we shouldn't overlook the matter of who created these objects and how. A number of them attest to production processes and contexts quite different from the usual clay tablets, something that only reinforces the sense of difference sometimes engendered by the use of script variants.

Within the material listed in Tables 8.1 and 8.2 we have ceramic objects very different in production from the familiar clay tablets, including large storage vessels, a Cypro-Minoan-style clay ball, labels, a clay nail reminiscent of Sumerian dedicatory cones, and zoomorphic sculpture such as a lion-head cup (see Fig. 6.13 above).

There are also examples of inscribed metalwork, such as a silver bowl inscribed in the short alphabet found at Hala Sultan Tekke on Cyprus36 or the bronze adzes or hoes found in a cache of bronzes at the House of the High Priest. There's ivory-work, such as the numerous replica livers thought to have been used for divination,37 or

<sup>36</sup> Åström and Masson 1982; Masson 1982; Bordreuil 1983.

<sup>37</sup> Gachet and Pardee 2001; Gachet-Bizollon 2007; 2008.


*Table 8.1. Objects inscribed with non-standard varieties of alphabetic cuneiform.*

*Table 8.2. Non-tablet objects inscribed in the standard official form of alphabetic cuneiform (AKA the 'long alphabet'), or for which the script variant used is unspecified in KTU.*


the rod found at Tiryns on the Greek mainland.38 There's glyptic material, including a number of seals and seal-impressions.39

I've listed this material quite summarily so I want to emphasise that not all of it comes from Ugarit, and some of it may not even be inscribed in Ugaritic. One short alphabetic inscription – a storage jar handle from Sarepta in Lebanon – is certainly in Phoenician, and it may well be that others were also produced by Phoenician-speakers even though they don't exhibit any diagnostic features that would allow us to definitively distinguish that language from Ugaritic. However, what this survey does underline very clearly, I hope, is that the use of alphabetic cuneiform reached well beyond the geographical core of the city of Ugarit and the material core of purpose-made clay tablets. The production of this material must have involved specialist metalworkers, ivory-carvers, sculptors, seal-makers, stone-carvers and so on. It seems unlikely that all these people would have been 'scribes' as we typically think of them.

There is, of course, a major, and very hard to answer, question about the production practices involved in the addition of writing to these objects. Were the inscriptions made by the same craftspeople responsible for the production of the objects themselves? This likely varied from object to object. For example, it's easy to imagine a stele being produced – at least roughed out – by one carver before the decoration – whether that be an inscription or an image – was added by a different specialist. On the other hand, if the text on seals was seen as just another example of fine carving, to be copied from a prototype and incorporated into the figural decoration, it could be quite possible that one person might have carried out all the carving. But even if we attribute an item to a single craftsperson, it's still very hard to know whether that artisan understood the writing they were inscribing, or whether they were perhaps merely copying it from a text provided to them. Such a scenario is certainly plausible, though Ockham's razor suggests that the most economical explanation is that they knew what they were doing. There's also the possibility for a middle course: a craftsperson without any formal training in reading or writing who is nevertheless called upon day after day to inscribe dedications or gift-giving formulae might before too long get a fairly good understanding of how to produce and interpret those formulae without help. This, after all, is the essence of artisan's literacy – the ability to write a limited range of text-types relating to craft production or distribution, which doesn't necessarily translate into a more general command of the script for all purposes.

One of the best indications that this kind of literacy may have existed in the region in the Late Bronze Age is the aforementioned Phoenician-inscribed handle from Sarepta, Lebanon (Fig. 8.2). Unlike a lot of this material, this was excavated using relatively modern methods (it was discovered in 1972) in a fairly well-handled and well-published excavation. It has a known archaeological context: near a wall in a late thirteenth-century workshop, near a kiln, in an area of the site with a high concentration of such kilns. Sarepta is particularly known for its ceramic production, with

<sup>38</sup> Cohen *et al.* 2010.

<sup>39</sup> Amiet 1992.

*Fig. 8.2. Inscribed jar handle from Sarepta (KTU 6.70). Drawn by the author after Bordreuil (1979, pl. I).*

its potteries apparently in use without interruption from the Late Bronze Age to the Persian period. There is also evidence of other craft activities in the area, including textile-manufacture and metallurgy.40 The inscription was published in 1975 and was shown to be in Phoenician by Edward Greenstein in 1979.41 It is read by Dietrich and Loretz (1988, 234–235) as:

ʾgnn z pʿl yd\* n\* bʿl z l ḫdšbʿl

Something like 'The jar which Ydnbʿl made, which is for Ḫdšbʿl'.

We have a pot, then, inscribed in Phoenician, found in a pottery workshop in Phoenicia and bearing an inscription saying who made it and for

whom. We don't know for a fact that this pot was made in the workshop where it was found, but this does seem the simplest interpretation. The inscription is interesting for a number of reasons apart from the language and its use of the cuneiform short alphabet. Structurally it is very similar to a lot of other alphabetic cuneiform inscriptions on non-tablet objects, which mostly name the maker or recipient. As Greenstein pointed out, it also closely resemblesthe typical north-west Semitic 'offering formula' seen in textssuch asthe Aḥiram sarcophagusinscription from Byblos(probably tenth century – see Chapter 12).42 As I've argued previously,43 I don't think we need to interpret these inscriptions as necessarily religious or dedicatory in nature (though some most likely are). Most of the parallels for inscriptions on the handles or shoulders of ceramic vessels are usually thought to be concerned with labelling or distribution – for instance, Linear B signs painted on Aegean stirrup jars,44 or Cypriot potmarks incorporating Cypro-Minoan symbols.45 Although there are some significant differences – such as the addition of the marks post- rather than pre-firing – the latter are a particularly close parallel to alphabetic cuneiform-inscribed pottery due to features such as variation in writing-direction, a preference for placement on the shoulders

<sup>40</sup> Pritchard 1975; 1988; Anderson 1987; 1988; Khalifeh 1988.

<sup>41</sup> Teixidor and Owen 1975; Greenstein 1976.

<sup>42</sup> Greenstein 1976, 53.

<sup>43</sup> Boyes 2019b.

<sup>44</sup> Olivier 1996–1997; Van Alfen 1996–1997; Duhoux 2011; Judson 2013.

<sup>45</sup> Hirschfeld 2000; 2004.

and handles of storage vessels and their distribution in Cyprus and the Levant. Over a hundred examples are known from Ugarit itself.

It makes reasonable sense, I think, to interpret the Sarepta handle as part of a large vessel produced at the site by a craftsman named Ydnbʿl and labelled for shipment to a man named Ḫdšbʿl but, for whatever reason, never sent.46 Alternatively, it might have been made elsewhere and Ḫdšbʿl might have been the recipient in Sarepta. Since we know very little about what exact kind of vessel this was and there has been no fabric analysis, we cannot say. But its presence in a ceramic workshop is extremely suggestive. The object itself is now believed to be lost, making further progress on this matter difficult.

If we're correct and this inscribed object was produced in a ceramic workshop in Sarepta, far from the Ugaritian metropolis and not even in a major Phoenician centre of politics or administration, then the implications for our understanding of literacy are significant. Taken alongside the other non-tablet objects inscribed in alphabetic cuneiform and found scattered across the central Levant and Cyprus, but rarely in major administrative centres, it's suggestive of a much more diffuse and varied network of writing practices than our focus on tablets and elite culture often leads us to suppose. This network may well have included craftspeople such as potters, metalworkers and ivory-workers. Even if we assume that the inscriptions were not made by the same people responsible for the objects themselves but by professional literates on their behalfs, then we still need to assume the presence of such writers in places like Sarepta, serving commercial artisans rather than elite bureaucracies. This is most obvious in the short alphabet inscriptions, because they come from places without bureaucratic use of alphabetic cuneiform, but we shouldn't exclude Ugarit itself from this phenomenon. It too will have had its artisans and workshops, and while these might have been more likely to use the long alphabet and so stand out lessfrom the background of official writing practices, we need not necessarily assume that all writing in Ugarit was part of this same elite, literate, intellectual culture.

#### *Social mobility and writing in the hinterland*

Following on from the last chapter's discussion of the relationship between writing and place in the Ugaritian countryside, in this section I explore exactly what forms of literacy and writing practice we might be able to reconstruct for this hinterland. As will be a familiar theme by now, the evidence is indirect and open to interpretation, but taken together can be seen to support the idea that writing wasn't entirely confined to the capital, Minet el-Beida and Ras Ibn Hani. A central figure in this discussion is well known to us by this point – ʾIlimilku, the senior official who wrote Ugarit's principal mythological texts. Of particular interest here is how ʾIlimilku identifies himself in his colophons – not using a patronymic, as writers using the logosyllabic writing system and Akkadian language tend to do – but with the name of his home village, Šubbanu. Some scholars have proposed the intriguing idea that this might relate

<sup>46</sup> On the question of whether Ḫdšbʿl is an anthroponym or theonym, see Boyes 2019b.

to a cultural difference between conservative Akkadian-using writers, who are keen to stress their belonging to literate, elite lineages, and potentially more open writers in the alphabetic tradition who are less interested in matters of heredity.47 Enticing though this hypothesis is, it's doubtful whether alphabetic and logosyllabic users constituted two distinct groups with their own writing cultures. It seems more likely that many writers – at least at an elite level – would be able to move back and forth between the two, though they may prefer one or the other based on their interests or the genres they tend to work in. The other problem is that ʾIlimilku's are our only good examples of alphabetic colophons of this kind, and one man's practice is not a good basis from which to extrapolate.

Whatisimportantis whatʾIlimilku's origin in Šubbanu meansfor our understanding of the presence of writing in the villages, and possibly also for matters of social mobility. To put it simply, what route a boy takes from a small rural village (perhaps having only fifteen households if administrative text KTU 4.810 is to be believed)48 to the highest political levels in the capital and the ability to write epic mythological poems in the alphabetic cuneiform script (documents seemingly without parallel from any other writer at Ugarit)? This is another question it's not possible to answer. We know nothing of ʾIlimilku's background or what took him to the capital. We don't know whether he was from a privileged family and sent to Ugarit for education or whether he was a poor boy who somehow showed some talent (in writing?) that was recognised by people in a position to advance him. We can concoct any number of stories. The only solid fact is that probably the most important literate known from Ugarit identifies himself with a small village in the hinterland.

Other documents also offer glimpses of a possible presence of writing beyond the capital, though without the same potential to conjur romantic fictions. Most obvious are the administrative documents concerned with the households, farms and estates of the countryside. These include lists of villages, for purposes of organising tribute, corvée labour and military service and the distribution of food and other rations; documents relating to the management of royal estates (*gt*); texts relating to professional groups, and lists of households for purposes of taxation.49

Some of this information – namely records for the distribution or receipt of goods by the central authority – would probably have been available to hand in the city, but in many other cases it is clear that data would have needed to have been supplied from the countryside itself. There seem two main possibilities for how this could have been done: either representatives from the location in question had to report to officials in the capital, where the data was then written down, or – more

<sup>47</sup> Roche-Hawley and Hawley 2013; Hawley *et al.* 2015.

<sup>48</sup> van Soldt 2005; Hawley 2008.

<sup>49</sup> Heltzer 1999, 423–431. These categories are cited to give a sense of the scope of the administrative documents relating to rural Ugarit, rather than to endorse Heltzer's specific reconstruction of how its economy was organised, which is a rather old-fashioned two-sector model proposing a dichotomy between royal dependents and 'free peasantry' and has been subject to criticism – see Heltzer 1976; 1988; Liverani 1989; 1987; and, for critiques, Schloen 2001, esp. ch. 11; Monroe 2009.

likely – written records were made on-site, either by locally based literate officials or peripatetic ones dispatched from the capital. There's some evidence that something like this occurred in the Mycenaean world, whose Linear B economic records form a good analogue for the Ugaritic administrative tablets. One of the Pylos tablets (Eq 213) records an official (perhaps also the writer?) *a-ko-so-ta* (probably Alksoitās) carrying out a tour of inspection of five sites in the hinterland.50 The use of tokens, nodules and sealings is also thought to have contributed to the administration of the Mycenaean countryside. Elsewhere in the Late Bronze Age Near East, even quite small villages have literate personnel stationed there, as is seen in the territory of Nuzi.51 There is also the possibility that literate state-affiliated officials were only stationed at larger rural settlements, which would bring with it an additional set of social implications for those required to travel to interact with them. As ethnographic research has shown, in such situations factors such as the quality of roads and the necessity of people to take time off from other tasks to make the journey can become very important for rural communities, leading to potential isolation, disruption and other social effects.<sup>52</sup>

List making itself is not a neutral exercise. A number of letters from Mari record acts of resistance to a census, including refusal to be registered and people being hidden by their families. These were met with threats and penalties from the authorities.53 List making practices also have the potential to be appropriated by individuals or groups for purposes likely not envisaged by the central authority, and in particular as tools for negotiating their own positions, power dynamics and social identities within their communities. This can be particularly pronounced when the literate list makers are part of the community they are administering, as we can see in Kulick and Stroud's description of list making – albeit of a very different kind to that of the ancient Near East – in the Papua New Guinean village of Gapun:

Other uses of writing include the habit of a few villagers of recording the dates of deaths in the village, and sometimes the writing of lists of villagers' names by men elected to positions of nominal importance instituted by the national government

<sup>50</sup> Nightingale 2008; Finlayson 2018, 5.

<sup>51</sup> Postgate 2013, 349. There have even been suggestions that a word for a village writer appears in Ugaritic documents: the word *md* has been suggested to mean 'scribe' by Sanmartín (1989) and Vita (1999, 465), which would be significant since its use with the names of villages might imply the existence of village scribes. The reading has been considered unlikely by Pardee (2000, 75) and McGeough (2007, 110; 2011, 28). The word remains problematic and there is no general consensus on its translation, but most believe that it denotes some kind of mid-ranking official or expert (cf. Akkadian *mudu*).

<sup>52</sup> For a more or less contemporary example, see Bloch's (1993) description of how settlement hierarchy and literacy related in Madagascar in the 1970s and 80s. In the Levantine context, the most famous illustration of the social disruption necessitated by the need to travel to interact with local bureaucrats is undoubtedly the New Testament's Nativity story, which has made countless people deeply familiar with the inconveniences of travel and accommodation inherent in interacting with the Roman bureaucracy in first-century Palestine.

<sup>53</sup> von Dassow 2008, 362, with further references.

(the listing of these names serves no other purpose than giving these men occasion to tell the villagers, in dark tones, that their names have been recorded 'in the book').54

Regardless of whether Ugaritian list making served a purpose for the central administration (and we have to assume that it did, given that these lists were found at the capital), the New Guinean example demonstrates how writing practices of these kinds can have social meaning quite independent from any actual administrative function the document served. These are likely to be irrecoverable for ancient rural contexts such as the Kingdom of Ugarit, but ethnographic comparisons at least highlight the kinds of social repercussions that may have existed.

Letters also indicate probable writing practices beyond the capital at Ras Shamra. It wasn't general practice to record where a letter came from, so outside the international diplomatic correspondence we're generally unable to identify the precise location where any of this material was sent from. However, we know that Akkadian and logosyllabic cuneiform were usually used for international correspondence, while Ugaritic and alphabetic cuneiform generally pertain to internal matters. We can assume, then, that the Ugaritic letters were written by people from the Kingdom. Some may have been travelling overseas at the time, as the unnamed king was when he sent KTU 2.30 from the Hittite court, but the majority probably originated within the Kingdom of Ugarit. A good number can probably be expected to have come from outside the capital. Some letters were probably written by, or on behalf of, travelling businessmen and other agents from Ras Shamra, but others don't seem to be from people on their way to somewhere else.

To these we can add the likelihood of local craftsmen with some grasp of writing, as discussed above. The overall picture is shadowy and gives us little definite idea as to the nature of rural literacy in the Kingdom of Ugarit, but it does seem to suggest that it did exist – sometimes in transitory form as literate people travelled and conducted administration – but perhaps also in the form of local writers.

#### **Orality, literacy and non-literacy**

Despite our interest in writing practices and the prominence they receive in the surviving record due to the durability of clay documents, there can be little doubt about the overwhelming importance of oral culture in the ancient Near East. To understand the nature of literacy in a place like Ugarit, it's vital that we think carefully about the interaction between writing and orally-transmitted forms of knowledge, together with the practices that go along with these. As we mentioned at the outset of this chapter, for a long time scholarly discussion of the relationship between orality and literacy was unnecessarily dichotomous. Influential writers like Jack Goody, Walter Ong and Eric Havelock promoted the idea of a 'great divide' between 'oral' and 'literate' societies, and even the notion that the advent of the Greek alphabet (which was often

<sup>54</sup> Kulick and Stroud 1993, 33.

presented as more or less synonymous with the introduction of literacy) was responsible for significant cognitive changes in the people who used it, allowing for greater rationality, complex thought and scientific enquiry.55 As late as 1986, Havelock wondered, 'may not all logical thinking as commonly understood be a product of Greek alphabetic literacy?'56

We can now recognise the extreme Eurocentrism inherent in these ideas, and their patronising approach to both non-western societies of the present day and recent past and those societies of the ancient world who developed complex literacy before the transmission of the alphabet to the Greeks. For anyone who has read this far, it will, I trust, seem absurd to think that even if there were any hypothetical advances brought about by alphabetic writing, credit should be given uniquely to the Greeks rather than the consonantal writing system's developers and elaborators in Egypt and the Levant. Accordingly, anthropologists have now almost universally rejected the so-called 'literacy hypothesis'.57 It's now clear that not only are these supposed cognitive boons from alphabetic or other literacy hard to identify (although it does seem likely that reading and writing can effect changes in the brain, as any often-repeated and routine action will);58 the posited dichotomy between orality and literacy is itself spurious.59 Rather than the latter supplanting the former in a linear, evolutionary way, these are two elements of knowledge and practice that coexist and intertwine in any society that uses writing. Even in modern western society, where the written word has become ubiquitous and utterly quotidian, oral culture is alive and flourishing, having adapted itselfsuccessfully to mass media and the internet age. The signs of this should have been evident when Havelock and Goody were writing, with the rise first of radio, then of television. John Postill mentions the example of Thomas T. Laka, a Sarawak storyteller, who gathered oral tales, wrote them down, edited, proofread, then broadcast them back into the oral domain via radio.60 Now, the internet allows almost anyone to have a YouTube channel or a podcast. The reach of the spoken word has never been greater. It might be argued that since some of this material is scripted, it doesn't really count as 'oral'; but that is exactly the point: spoken and written culture interact and adapt to each other, giving rise to new forms of knowledge and new practices surrounding them. Modern mass media couldn't exist without writing, but their existence offers new avenues for oral communication and transmission of knowledge.

This change in perspectives, from the opposition of literacy and orality to recognising their interrelationship, has been apparent in Near Eastern studies. There's little doubt

<sup>55</sup> Goody and Watt 1963; Havelock 1963; 1982; 1986; Goody 1968; 1986; 1987; Ong 1986; 2012.

<sup>56</sup> Havelock 1986, 39.

<sup>57</sup> Halverson 1992a; 1992b; Bartlett *et al.* 2011, 156.

<sup>58</sup> See, for instance, Dehaene *et al.* 2015.

<sup>59</sup> The 2012 reissue of Ong's major work on the subject includes a useful retrospective final chapter by John Hartley, exploring some of these issues.

<sup>60</sup> Postill 2003, 91–92.

that the majority of Near Eastern documents, in the forms in which they survive, are the products of a fundamentally textual tradition of copying, editing and adaptation. This is especially true of literary compositions. Many such texts exist in multiple versions, and where these come from different chronological periods, as is the case with, for example, Gilgamesh, it's possible to produce histories of the development of the text over time. What's more, many of these literary works draw on Sumerian material. Since Sumerian was a dead language by the time the extant versions were written down, these elements must have been preserved through writing. For these reasons, some scholars, such as Piotr Michalowski, have sought to place these works within a quite restricted social context, arguing that they were preserved and propagated within highly learned, literate textual communities61 but had little currency outside them. Even the language used, it is suggested, was significantly different to the vernacular. For Michalowski, the great Mesopotamian epics such as Gilgamesh existed only in the heads of scribes and their teachers and were 'performed' or 'realised' only through the act of writing them down in scribal exercises.62

Others have seen greater scope for an oral aspect in Near Eastern literature. Joan Westenholz criticised Michalowski's 'purely literary' approach and highlighted the oral element in the development of early pseudo-autobiographical epics such as those of Sargon or Narām-Sîn,63 while Karel van der Toorn has argued that throughout Mesopotamian history, literary works were designed for oral performance: the story of Atraḫasis is described as a 'song' that is 'sung to all peoples' and the much later Song of Erra makes direct reference to being sung and performed – and the blessings to be gained by a singer for doing so – in its epilogue.64 This makes sense, since every indication is that mythological stories of Gilgamesh, Atraḫasis and the gods were current in Mesopotamian and wider Near Eastern society well beyond narrow textfocused scholarly communities. Indeed, although we have little knowledge of the history of these stories before they were first written down, such a wider currency is the best explanation for how they arose and became important enough to write down in the first place. In fairness, Michalowski doesn't exactly deny this, though for him the popular, oral versions of the tales and the literary, written ones are quite separate. For van der Toorn, they are one and the same – at least for successful compositions that were accepted into the repertoires of bards.

[I]n Babylonia the products of creative writing could reach an audience only if a singer was willing to include them in his repertoire. Who but a few would have read the

<sup>61</sup> The idea of 'textual communities' was developed by Brian Stock (1983) for understanding religion in the Middle Ages but has since been extended to other social contexts with some success (*e.g.* Briggs 2000). They can be defined as social groups organised around texts, but where not all members need be literate themselves. Instead, small numbers of literates can serve as intermediaries disseminating the contents of the central text(s).

<sup>62</sup> Michalowski 1992; 2010.

<sup>63</sup> Westenholz 2010, 30–31.

<sup>64</sup> van der Toorn 2007, 13. See also Pearce 1993.

Gilgamesh Epic? The common people knew this work because they had heard it from the mouths of bards and singers.65

For the Levant, there has been a great deal of research into the oral aspects of the production and transmission of the Hebrew Bible. These have often emphasised the fundamentally oral nature of first-millennium Israel/Palestine and recognised the importance of interplay between the written word and this non-literate context. As Susan Niditch puts it:

To study Israelite literature is to examine the place of written words in an essentially oral world and to explore the ways in which the capacity to read and write in turn informs and shapes orally rooted products of the imagination.66

The place of the written word in Ugarit bears comparison with both the Mesopotamian and biblical situations. Some literary texts refer to singing and harp-playing during the narrative and it has been suggested that the use of short stichoi and lack of a defined metre make for a 'free rhythm' considered to be well suited to musical accompaniment.67 The use of repetition and stock formulaic epithets for characters, while not directly equivalent to their widely discussed use in the oral epic poetry of the Aegean in that there is no need for a particular formula to fit a given metrical environment, nevertheless point to an oral environment of composition and transmission where ease of memorisation is important. Of clear relevance here is the debate as to whether ʾIlimilku composed the literary works that he wrote down, or whether he was merely copying an older version. There's no doubt that the poems incorporate ancient elements – the weapons that Baʿlu uses in his battle against Yam, for example, were already mentioned in two eighteenth-century letters from Mari68 – but there's no clear evidence for whether ʾIlimilku's sources were written, oral or a combination of the two. Many scholars today tend to assume that at the very least ʾIlimilku significantly transfomed the ancient tales he told, and that the written versions we have are not verbatim reproductions of originals composed much earlier,69 but there are still prominent advocates of the view that he was little more than a copyist at the culmination of an existing tradition.70 My own inclination is to see ʾIlimilku's work astransformative and essentially authorial, but whichever position we take, there seems little doubt that he was not working in a vacuum but adapting stories that had long existed in other forms. Whether there were earlier written adaptations we cannot say, but it seems plausible that just like Mesopotamian epics, the myths of Ugarit would have been sung and passed on within an oral tradition.

<sup>65</sup> van der Toorn 2007, 13.

<sup>66</sup> Niditch 1996, 134.

<sup>67</sup> de Moor 1978, 132, cited by Watson 1999, 168.

<sup>68</sup> Töyräänvuori 2012.

<sup>69</sup> For a summary of the discussion see Tugendhaft 2018, 29–30, as well as Smith 2014, 38 for criticism of the oral poetry interpretation of the Baʿlu Cycle.

<sup>70</sup> See, recently, Pitard 2009; Greenstein 2014.

We should probably envisage the composition of literary texts at Ugarit not as a purely textual act but as a weaving-together of a range of sources, sung, spoken, remembered and written. Even the act of writing is likely to have had an oral component: silent reading was uncommon in ancient times and we might imagine that the same was true of silent writing, especially when one was composing for performance. Donald Redford (2000) envisages oral and written composition as happening simultaneously: the writer trying out phrases and working them out verbally as they commit them to writing. The end products – mythological epics like the Baʿlu cycle – were probably not solely intended to be shut away in the shadowy tablet collections of the elite, but to contribute to a living tradition of performance and storytelling.

In discussing the oral dimension of writing practices we've focused mainly on literary, narrative material and especially on epics. In part because of the influence of the flourishing discussion regarding the oral element in the production of Aegean epic poetry,71 and in part because of interest in how Ugaritic literary material can be used to support discussions of biblical composition, it is this narrative material that has been the main focus of most discussions of orality at Ugarit. However, we shouldn't overlook the significant degree to which orality shaped the production and use of documents in other genres. This is most obvious in things like letters, which we assume to be the products of dictation and would in most cases have been read aloud to their recipient; but there are also ritual and magical texts that contain a strong performative element. These texts are notoriously difficult to interpret in Ugaritic, but consider, for example, the so-called 'Ritual of National Unity', which we will be discussing in more detail in Chapter 9 (see p. 202 for extract). The ritual text shifts ambiguously between a script (in the theatrical sense), which is to be recited out loud and 'stage directions', which are not. It also includes numerous references to 'statements' by various social groups (*p* – 'mouth' in the original Ugaritic, and admittedly open to multiple interpretations, as we will see in Chapter 9). The refrain 'Here is the donkey!' recurs throughout the ritual and is obviously performative. We can safely assume that the sacrificial donkey would have been indicated or proffered in some way as these words were spoken.

There was also an oral element – as well as an important material one – to legal texts. It can readily be assumed that a great many of the parties to any legal agreement, deed or contract would not have been literate. Tablets recorded a transaction carried out orally, in the presence of witnesses, and if the tablet actually needed to be read, then most people would have had to find a literate person able to do so for them. But Ignacio Márquez Rowe also points out that the physical appearance of the tablet itself would have been important:

It obviously did not matter whether or not he or she could decipher the text. What really mattered is that the holder of the title deed knew the content of the text(s) he kept. […] What I mean to stress is that the physical appearance of the deed, that is,

<sup>71</sup> This discussion was kicked off by Lord and Parry in the mid-twentieth century and the subsequent bibliography is too vast to list.

the nature and position of the seal impression on the tablet, the shape of the tablet itself, and the arrangement of the inscription conveyed a message of their own; a message that even the larger body of illiterates of a society such as Late Bronze Age Ugarit could read.72

This brings us to why the discussion of orality and writing in Ugarit is so important for a chapter centred on literacy. It's not just that exploring the relationship between oral traditions and textual composition gives us better insight into how writing practices worked – although this is true and important; it also gives us our best way into understanding how non-literate or barely-literate people interacted with writing and writing culture. Almost nobody in Ugarit is likely to have been able to read the tablets of the Baʿlu Cycle. Even if they had the skills, there's no guarantee they would have been permitted access to the House of the High Priest or to the tablets. But derivatives of these literary texts in the form of songs or performed poems may well have reached much larger audiences. Likely these differed in significant waysfrom the written texts – it is typical of oral performance that each performance is unique, even assuming the singer in question had at any time had occasion to personally see or hear the original. Rather, we might have second-, third- or more-hand bootleg versions of ʾIlimilku and other writers' compositions, circulating and mingling with pre-existing and alternative traditions. The written version may thus have fed into wider 'popular culture', even if it itself remained restricted to quite limited elite textual communities.73 This seems quite plausible, but of course, we can't prove it. Even so, if there wasn't this kind of two-way inspiration between written literature and oral narratives then at least a one-way relationship – with popular songs and performances of traditional stories feeding into elite literary culture – does seem probable.

And other written documents would certainly have shaped the lives of even the non-literate members of society. We have discussed the oral elements of ritual, magic, medicine, letters and legal documents. There were likely differences in people's abilities to access these depending on their wealth, profession and social circle, but the deciding factor is unlikely to have been the ability to read and write. As we have seen, even non-literate members of Ugaritian society could have participated in a large-scale civic ritual such as that seen in KTU 1.40, or called upon the services of doctors, exorcists or diviners – assuming they had the resources to do so. The form these practices took and the way they were carried out were shaped by the existence of writing, even if we can never know whether literacy was a direct part of their everyday performance: did the priests read out their liturgy from tablets or were the

<sup>72</sup> Márquez Rowe 2006, 105.

<sup>73</sup> For a modern analogy we might look at something like the 1982 Turkish film *Dünyayı Kurtaran Adam* – 'The Man Who Saved the World', perhaps better known as 'Turkish Star Wars' because of its use of fragments of pirated footage from *Star Wars*, as well as clips and music from other films, to tell a *Star Wars*-inspired science fiction tale. Lest we make the mistake of assuming this is solely a phenomenon of non-Western markets, we could also note the wealth of straight-to-DVD cash-ins that appear with similar titles and cover art when Hollywood blockbusters are released.

written versions merely aides-memoires for private consultation? If you consulted a doctor regarding a complaint, would they rummage through their tablet collection for information on diagnosis and treatment, or was that knowledge all already committed to memory?

For this reason, the likeliest areas in which non-literates may have come into direct contact with writing practices would not be those with significant performative or oral dimensions, where memorisation and traditions of knowledge had always been important. Rather, we should probably look to where new information was created and needed to be recorded: legal transactions and economic administration. Even relatively lowly households would have been included in the numerous lists and records with which the bureaucracy managed the city and its hinterland. As we have already discussed, this necessarily would have brought at least a representative of each household into contact with a literate functionary of some kind – whether that be a roving agent of the palace, a locally-based notary-cum-tax official, or an official to whom they had to report periodically in the capital.

#### **Literacy, non-literacy and encounters with writing practices in Ugarit**

Despite the prestige attached to writing and education within the higher levels of Near Eastern societies, it would be a mistake to assume that this extended to the wider population. The idea that non-literate peoples should attribute any mystique to writing, or view it with a kind of awe is a topos with a very long history, but has been shown to be rooted more in Western cultural values and ultimately probably in the central role of scripture in the Judaeo-Christian tradition than in any real understanding of how non-literates in different contexts view(ed) writing practices. Peter Wogan explored this for Native American encounters with European missionaries in the Early Modern period.74 Modern cases where scripts have been developed for previously unwritten vernacular languages in social contexts with low levels of literacy show that interest in, and prestige of, any given script often relate to its perceived usefulness for these communities, and this is strongly dependent on cultural factors and the socio-economic situation. So, global writing systems such as the Roman alphabet can be seen as relatively prestigious because of the economic opportunities they seem to offer or their colonial heritages, while scripts associated with vernacular languages might be used only in relatively restricted contexts – such as religion – and by small numbers of people.75 There are, of course, many major differences between modern ethnographic comparanda and ancient Ugarit – the role of Christian missionaries, the impact of educational systems established and still influenced by colonial powers, the idea of mass literacy as any kind of desideratum. But such comparisons are probably still our best hope for beginning to approach the reception of writing among the general population at

<sup>74</sup> Wogan 1994.

<sup>75</sup> See Kulick and Stroud 1993 as well as other contributions to Street 1993, and also Martin‐Jones and Jones 2000; Postill 2003; Moseley 2017.

Ugarit. Furtherresearch isrequired, but for now we are probably justified in envisaging a high degree of indifference, or at least a sense that writing might be important for particular tasks or areas of practice which had little direct bearing on the vast majority of Ugaritians.

This chapter hasn't attempted to guess at what proportion of Ugarit's population may have been able to read and/or write. We have almost no basis on which to make such a determination, even if literacy was a straightforward binary of this kind. Instead I've aimed to survey some of the broad types of literacy that we either know, or can reasonably surmise, existed in the Kingdom, especially those outside the stereotypical demographic of highly educated, relatively high-status male writers working within, or for, the state or religious establishment. While there is very little that is certain, I hope I have shown that there is at least ample scope for a rather broader and more diverse tapestry of literacies. In Chapter 9, we will continue with the theme of demography and examine how writing practices relate to the existence of possible communities based on ethnic and other forms of social identity.

### Chapter 9

### Writing practices and minority communities

Much has rightly been made of Ugarit's social diversity. Its texts and material culture portray it as a place where people from around the eastern Mediterranean and Near East mingled and exchanged ideas and objects.1 This plurality is reflected in the broad range of scripts and languages attested at the city. Some of these, such as Hittite and Luwian, probably only reached the city on documents or sealings sent from elsewhere. But even if we just look at the documents and inscriptions probably written in the Kingdom of Ugarit itself, then as well as the usual alphabetic and logosyllabic cuneiform, we must contend with writing in Cypro-Minoan, Hurrian (in two scripts) and possibly Egyptian. In short, we have a remarkably polyglot and multiscript environment. In this chapter I will explore the extent to which this linguistic and scriptal diversity reflected social distinctions within Ugaritian society – what might be termed 'ethnic minorities', and in particular how writing was used to define and engage with ideas of ethnicity, identity and community within the kingdom.2

In practice, this means looking at two main identities and their potential associated communities: Hurrian and Cypriot. The exact number of languages and scripts pertaining to these groups at Ugarit is hard to quantify. The traditionally accepted count of texts in the Hurrian language from Ugarit is one hundred, of which twenty-eight are in alphabetic cuneiform and seventy-two in logosyllabic cuneiform.3 These broad linguistic categories mask a more nuanced reality: five ofthe 'Hurrian' alphabetic texts are Ugaritic-Hurrian bilinguals, one of the logosyllabic ones is an Akkadian-Hurrian bilingual and many more are lexical lists in up to four languages. Finally, one text is

<sup>1</sup> Schniedewind and Hunt (2007, 8) suggest that up to 16% of the population may have been foreigners, though it's not clear how they come by this figure. In terms of elite culture, Feldman (2002b; 2006) sees at Ugarit key examples of her 'international style' of art current across much of the east Mediterranean, alongside more specifically local artistic traditions.

<sup>2</sup> Parts of this chapter are adapted from Boyes forthcoming.

<sup>3</sup> Pardee 1996, 64–65. For the original publication of most of these tablets, see Laroche in Nougayrol *et al.* 1968, 447–544.

in a hybrid language combining elements of Hurrian and Akkadian. As for the much smaller Cypriot corpus, we have only one script – Cypro-Minoan – but since it's undeciphered, we can't say with any certainty how many languages are reflected in the nine examples from Ugarit. In the Iron Age and thereafter, Cyprus was characterised by a significant degree of linguistic (and scriptal) diversity for a relatively small island,4 so it would not be wise to take for granted that all Cypro-Minoan texts are necessarily in the same language. As we'll see, there is also a good case to be made that one tablet may be in Cypro-Minoan script but Ugaritic language.

There has often been an assumption that the use at Ugarit of these languages and, in the Cypriot case, script indicates that there were 'Hurrians' or 'Alašiyans' living in the city. This may seem like an obvious and reasonable conclusion since both these gentilic terms occur with some frequency in the texts. The documents also refer to Ašdodians, various kinds of Mesopotamians and perhaps also Hittites, Egyptians and Cretans, although generally at the level of individuals rather than groups, meaning that scholars tend to discuss these in terms of single travellers rather than long-term communities. However, the situation is much less clear than it at first appears. The meaning of the term 'Alašiyan' has been subject to a long-running debate. It's now almost universally associated with Cyprus, but whether it refers to the whole island or a specific polity (and in the latter case, its nature and location) remains a matter of much discussion. 'Hurrian' is not usually subject to the same explicit questioning, but should be. Despite being extremely widely used within Near Eastern scholarship, the term is rarely defined and, when probed, becomes increasingly hard to pin down. Unlike many of the ethnic labels used for the ancient Near East, it's not associated with an easily fixed geographical location, nor with a particularset of material culture. The 'Hurrian' religion is not homogeneous, but more a series of overlapping local cults, similar to what we see with 'Canaanite' religion.5 In fact, the definition of Hurrian seems to be almost entirely ethnolinguistic:6 when scholars talk about Hurrians, more often than not what they mean is 'speakers of the Hurrian language'. But a common language has never been a sure indication of a shared identity, and certainly not in all contexts and all times (just call a proud Scot an Englishman and see what happens). Von Dassow (2008, 68–76) at least acknowledges this, but continuing to use 'Hurrians' to mean 'Hurrian-speakers', as she does, is potentially misleading and cannot help but imply a common identity even when it is stated to merely be a linguistic term.

And yet, 'Hurrian' is a term used in texts from Ugarit and elsewhere in the ancient Near East; it isn't entirely a modern scholarly figment. If we're to understand the relationship between 'minority' languages, writing practices and identity, we will have to tease out what we can of how words like 'Hurrian' were understood; both by the people doing the labelling and the people being labelled. We can't assume an automatic correlation with our modern linguistic category of the same name.

<sup>4</sup> Steele 2018.

<sup>5</sup> Trémouille 2000.

<sup>6</sup> Salvini 2000, 26; von Dassow 2008, 68.

#### **Defining communities**

To begin with, a brief excursus is necessary on the conceptual framework underpinning the approach I'll take here and what we mean by terms like 'community', 'ethnicity' or 'identity'. The old-fashioned approach – still current in some quarters – was to assume a congruence between identity, political organisation, language, territory and material culture. In other words, groups such as 'the Hurrians' or 'the Philistines', and so on could be identified, who were associated with given political polities and a particular territory, spoke a particular language and could be recognised through the presence or absences of particular artefacts or artistic styles. These are assumed to be natural communities.

For several decades now, work on ethnicity and other forms of social identity has largely rejected these sorts of definitions.<sup>7</sup> Instead, it's now recognised that community is a matter of common identity and as such is largely ascribed subjectively, both by people themselves and by the others with whom they come into contact. As such, it is performative, participatory and contextual rather than being defined positivistically by a fixed set of material or practical indices. Indeed, very often identities and communities crystallise in the face of a perceived Other: they are defined in the negative, through contrasts with what they're seen *not* to be.8 They're not enduring, bounded, homogeneous entities in themselves but rather properties of – or dispositions arising from – the interrelations between different people and groups.

These general principles are relevant to all forms of social and cultural identity. What's more, the various subcategories often used – such as ethnicity, gender, status, political identity and so on – are themselves not conceptually distinct but riddled with overlap. Ethnicity can be gendered, both can be political, all are bound up in status. To a degree, then, any subdivision of cultural or social identity is merely an exercise in arbitrary taxonomy. On the other hand, taxonomy is often a helpful tool in understanding and analysing complex situations, so long as we do not over-reify the category labels we assign, and especially when those categories coincide with those understood by the people we are studying. Some subcategories – such as gender or age – are of obvious usefulness. Others can be so nebulous as to be essentially redundant. The category of 'ethnicity' is one of the more persistent, treasured by both

<sup>7</sup> See, originally, Barth's seminal introduction to his 1969 edited volume 'Ethnic Groups and Boundaries' and also Jones 1997; Hall 2002; Jenkins 2008. For an up-to-date discussion of the archaeology of communities, see Mac Sweeney 2011. She sees 'community' as a form of social identity relating to shared local space. I would nuance this by adding that the space might be either real or virtual – accounting for communities in the modern world centred around particular social media sites or internet fora – but for antiquity her definition works well. Following this, there is obviously considerable scope for overlap between the localism-based 'community' and the origin/descent-based 'ethnicity'. In this chapter I use 'community' essentially to mean a group collected in a particular locality and sharing a sense of common identity, usually based around common origin or descent.

<sup>8</sup> A classic example is the formation of Tswana identity and a notion of Setswana – Tswana ways – as a consequence of interaction with European missionaries (Comaroff and Comaroff 1992, 235–263, cited in Jones 1997, 95).

popular and specialist discourse alike; however, its meaning is much debated and highly inconsistent. Within anthropological and sociological works it tends now to be seen as a form of social identity based on real or fictive common origin or descent,<sup>9</sup> but outside these disciplines it is still widely used with meanings other than this – or without a precisely defined meaning at all. In the Near Eastern context, where it is still often used in the old-fashioned sense as denoting an eternal and authentic origin and national character, it is tempting to avoid the term altogether, but in practice this is difficult and cumbersome. So long as we are clear what we mean, terms like 'ethnicity' have heuristic usefulness; but we should not assume they have any intrinsic meaningfulness, for ancient cultures or even modern ones.

Within the shifting web of interactions, practices and symbols which constitute cultural identities, features like territory, language or material culture might all be important, but they need not necessarily be so, and not necessarily in all contexts. This was amply illustrated as early as 1965 in Michael Moerman's discussion of the Lue people of northern Thailand. Moerman showed that Lue identity could not be tied straightforwardly to territory, language, dress and so on, which were all shared to some extent by other groups inhabiting the same region. Nevertheless, the identity was real and important to both the people who subscribed to it and to their neighbours. Moerman concluded that 'Someone is a Lue by virtue of believing and calling himself Lue and of acting in ways that validate his Lueness.'10 Those 'ways' will of course be culturally determined, and context-dependent. They may include using particular types of pottery, speaking a given language or dressing a certain way, but we cannot assume *a priori* that any will be symbolically significant in a given social context.

This throws up obvious difficulties for the archaeologist. If social identity is ever-changing, contextually contingent and not straightforwardly tied to material culture, how are we to detect it except where we have explicit written sources giving insight into ancient people's minds? Jonathan Hall, in his discussion of Hellenic identity in ancient Greece, largely despairs, arguing that it's impossible to distinguish different forms of social identity without written texts.11 Others are less pessimistic. Responses have generally focused on the importance of context, practice and agency, in the same way as we already discussed in Chapter 2 with regard to human practice more broadly. As another example of the interaction of the habitus and practice, identities are susceptible to the same kinds of approaches.12 By examining closely the intersection between context, material culture and the use of writing with regard to minority groups such as the Hurrians, in theory it ought to be possible to explore key questions.

<sup>9</sup> Jones 1997; Hall 2002; Jenkins 2008.

<sup>10</sup> Moerman 1965, 1222.

<sup>11</sup> Hall 2002, 23–24.

<sup>12</sup> Jones 1997, 125–126; Fowler 2010, 360.

#### **Identity terms in writing from Ugarit**

As I've already mentioned, the terms 'Hurrian' (*ḫry*) and 'Alašiyan' (ʾ*alṯy*) both occur at Ugarit numerous times, as do many other such ethnic labels (often called gentilics). What did these mean to the people who wrote the texts? What (if anything) can we reconstruct from these texts about how identities were conceptualised and constructed in society more broadly? We must remember, after all, that the texts that bear on this topic are relatively few and most are not concerned with identity directly; and the views we're being supplied with are those of various facets of the controlling elite (either writing directly or mediated by literate secretaries): the voices that remain to speak to us come from a relatively small segment of the population, and most likely not the people and communities whose identities we are concerned with.

There are clear differences between genres, reflecting how social identities came to the fore or receded in different social contexts. Within the alphabetic cuneiform corpus, gentilics occur in two genres: administrative texts and cult ones.13

In administrative documents, gentilics often appear in place of patronymics or even personal names. As Christopher Monroe (2009, 219) points out, the distinction seems to be between people with known lineages, meaningful to the Ugaritian authorities, and those without; a case of 'Are they local? Do we know their parents?' There doesn't appear to be any concept of degrees of 'foreignness': there is no distinction between, say, people of Levantine origin and those from further afield. Notably, however, *ḫry* – Hurrian – does not appear in this context. 'Hurrian' is not an administratively significant category at Ugarit and people with Hurrian names are administered just as are those with 'local' West Semitic names: they are local, their parents are known.

The administrative documents are terse and provide no more information than they need to. They weren't written with questions like these in mind and don't trouble themselves to actually define the basis for someone being labelled with a given gentilic. Is someone a Canaanite or an Alašiyan if they were born there? If their parents were? If they speak the language, dialect or have the accent? If they look, dress or act a particular way? There's little in these texts to help us with this, although one example does suggest the matter is not straightforward and that the definitions of the Ugaritian scribes and bureaucrats may not coincide with what seems obvious to us. KTU 4.775, a tablet found at Ras Ibn Hani that lists people and associated numbers of sheep and goats, includes mention of *ṯpṭbʿl mṣr[y]* – 'Ṯipṭibaʿlu (or Šipṭibaʿlu) the Egyptian'. On the face of it, this is a typical case of a gentilic being used in place of a patronymic. However, Ṯipṭibaʿlu is not an Egyptian name but a West Semitic one, extremely well attested at Ugarit. Moreover, there's every chance this is the same Ṯipṭibaʿlu who wasthe son-in-law and agent of Queen Ṯariyelli: a prominent merchant known to have had business links with Egypt. Three Akkadian documents dealing

<sup>13</sup> An apparent exception, the appearance of *ʾalṯy* in letter KTU 2.42, is still a religious reference (*kl. ʾil. ʾalṯy* – all the gods of Alašiya).

with the Queen's business affairs and found in the Royal Palace's Central Archives were impressed with an Egyptianising seal bearing Ṯipṭibaʿlu's name in hieroglyphs.<sup>14</sup>

If these refer to the same man (and, though probable, this is admittedly not completely certain), then 'Egyptian' here is probably not a marker of where this man came from in the literal sense, but more of a nickname based on his practices, affinities and interests. Together with the seal-ring, it may point to a man who was identified with Egypt and chose to embrace these associations, despite being what we might call 'Ugaritian' by birth.15 In other words, the literate administrators of Ugarit may have recognised ethnicity as performative as well as innate.

Let's turn now to the sphere of ritual and cult. Within this genre, by far the most significant document relating to identity and community at Ugarit is the so-called 'Ritual of National Unity' (Fig. 9.1).16 This lengthy text in Ugaritic language and the alphabetic cuneiform writing system has been found in two fairly well preserved copies (KTU 1.40 and KTU 1.84) as well as four other fragments. These different versions come from widely dispersed find-spots (the tablet collection of the High Priest, Room 90 of the Royal Palace, the House of the Divination Priest) and were in different handwriting. No other text from Ugarit is known from such a diverse set of copies, pointing to its importance.17 The ritual itself concerns the well-being and rectitude of the people of Ugarit, with animal sacrifices offered in assurance of this. Scholars have found it hard to resist comparison with the later Jewish Yom Kippur and its similar use of a scapegoat as a vessel for human sin.18 What is notable for our purposes is how the ritual enumerates the various communities and classes of people within Ugarit's society.

(26') Bring near the donkey of 'rectitude'; rectitude of the son of Ugarit: and[ well-being of the foreigner within the walls] of Ugar<it>,

(27') and well-being of YMʾAN

and well-being of ʿRMT,

and well-being of[ ]

(28') and well-being of *Niqmaddu*;

whether your 'beauty' be altered;

be it according to the statement of the Qa[ṭien,

be it according to the statement of DDM]Y,

(29') be it according to the statement of the Hurrian,

be it according to the statement of the Hittite,

be it according to the statement of the ʾAlashian,

be it according to the sta[tement of ǴBR,]

be it according to the statement of (30') your oppressed ones,

be it according to the statement of your im[pov]erished ones,

be it according to the statement of QRZBL;

<sup>14</sup> Schaeffer *et al.* 1956, 85; Nougayrol *et al.* 1968, 261; Singer 1999, 696–697.

<sup>15</sup> Vita and Galán 1997; Singer 1999, 696–697.

<sup>16</sup> Pardee 2002, 77ff.; Sanders 2004, 51.

<sup>17</sup> Pardee 2002, 78.

<sup>18</sup> de Moor and Sanders 1991; Pardee 2002, 78–79; Sanders 2004, 51–52.

*Fig. 9.1. KTU 1.40: 'The Ritual of National Unity'. From Bordreuil and Pardee (2009), supplementary CD, Text 9. Reprinted by kind permission of Projet PhoTEO, Mission de Ras Shamra.*

```
whether your 'beauty' be altered:
         (31') be it in your anger,
         be it in your im[pa]tience,
         be it in some turpitude that you should commit;
(32') whether your 'beauty' be altered:
         as concerns the sa[cr]ifices
         or as concerns the ṯʿ-sacrifice.
The sacrifice, it is sacrificed,
         the ṯʿ-sacrifice, it is offered,
         (33') the slaughtering is done,
May it be b[or]ne to the father of the sons of ʾIlu,
         may it be borne to the Circle-of-(34')the-Sons-of-ʾIlu,
         <to the Assembly-of-the-Sons-of-ʾIlu>,
         to Ṯukamuna-wa-Šunama;
here is the donkey.19
```
The invocation of social groups quoted above is repeated three times in the best-preserved version of the ritual,20 with minor variations each time. On one occasion the invocation of the 'son of Ugarit' is replaced with 'the daughter of Ugarit' and followed up with a reference to the 'well-being of the woman/wife'. This, along with the mentions of the 'oppressed ones' and 'impoverished ones' shows that the ritual's interest in identifying social groups goes beyond gentilics and also includes segments of the population identified by gender, financial circumstances or being subject to oppression.

The ritual has been studied a great deal,21 and most scholars have emphasised its integrative function. It highlights the various social groups seen to comprise Ugaritian society and unites them in a collective act of sacrifice in expiation of whatever misdeeds they have done against each other. Seth Sanders views this as particularly significant: a landmark in the subsumption of various sub-identities into one of the first true vernacular national identities – a sense that Ugarit is its people, united despite their diversity, a statement of ancient multiculturalism. We might not wish to go quite that far, although the importance of vernacular identities is something we'll return to later in this discussion. Nevertheless, on the face of it the idea that this is an integrative ritual of unity has much to recommend it.

But what do the gentilics actually mean here? In particular, attention has focused on the repeated phrase *ʾu l p* [GENTILIC], which the text's most recent editor, Pardee, translates as 'according to the statement of [GENTILIC]'.22 Most

<sup>19</sup> KTU 1.40 'Ritual of National Unity' 26–34. Translation from Pardee 2002, 79–83.

<sup>20</sup> Two or three other repetitions are probably lost in the fragmentary sections.

<sup>21</sup> de Moor and Sanders 1991; Pardee 2002, 78; Sanders 2004, 51.

<sup>22</sup> The Ugaritic is literally 'and in the [GENTILIC] mouth'. Given our interest in language it might be tempting to translate *p*, which can also mean 'voice', as 'language', but to the best of my knowledge there are no clear instances of it having that meaning in Ugaritic, nor would it be an expected meaning in Hebrew or Phoenician. Note that the ritual only uses adjectival forms such as *ḫty* – 'of Ḫatti' or

scholars, like him, see these as foreign ethnicities resident within the Kingdom of Ugarit, although Johannes de Moor and Paul Sanders (1991, 296–297) have an alternative view, considering these lines in fact to contrast with the 'foreigners within the walls'. They translate them as 'sin in the manner of [GENTILIC]' and interpret them as a list of 'arch sinners' to whom Ugaritian sin is compared.23 Pardee's translation seems the more probable one, but the idea that these are groups within the city doesn't seem entirely certain on internal grounds to this text. The Hurrians, Alašiyans, Qaṭiens and other gentilics are not directly associated with the reference to the 'foreigner within the walls', and while they need not be archetypes of sin as de Moor and Sanders suggest, it seems that they might just be legal or cultural standards according to which people might have sinned, without them necessarily being imagined to actually live in Ugarit. After all, while administrative texts refer to people identified as Alašiyans or Hittites resident in Ugarit, and Hurrians are plausible on linguistic and onomastic grounds, some of the other gentilics mentioned don't occur in other texts from Ugarit or can't be definitively pinned down: we can only guess at where the Qaṭiens and *Ddmy* are supposed to hail from.24 The unknown terms *Ǵbr* and *Qrzbl* may not even be gentilics at all. Although they follow the more certain gentilics, they also precede references to non-ethnic social groups – the oppressed and impoverished – so may represent some kind of other social group or identity.

I've sounded a note of caution here, but I still tend towards Pardee's interpretation as the most plausible for this text. It's telling, though, that even in a document that seems to tackle questions of identity and the social make-up of the polity directly, there remain considerable ambiguities even at the basic level of whether these people are considered to be part of Ugaritian society or not. As with the administrative texts, there is a strong sense that place-of-origin was an important dimension of how people were categorised, taxonomised and identified in Ugarit, at least by the elite, just as in other Levantine societies. But again we are left with little idea of the exact criteria by which these labels were applied, and in the light of 'Egyptian' Ṯipṭibaʿlu, we must assume that it was to a degree subjective and based on practice. It is also evident that ethnicity/place-of-origin was not the only social identity that was considered pertinent, since gender, wealth and level of oppression are all cited alongside it.

*ʾalṯy* – 'of Alašiya' rather than the 'sons of PLACE-NAME' formulation, which is usual when denoting citizens of a given location, cf. 'son of Ugarit' in line (26').

<sup>23</sup> de Moor and Sanders 1991, 296–297.

<sup>24</sup> Pardee (2002, 112, n. 114) suggests the latter refers to people from the region of Aleppo, while de Moor and Sanders (1991, 123) posit that *qṭy* denotes people from Gutium in the Zagros Mountains and mostly known from third-millennium Sumerian texts, and thus far removed from Ugarit both geographically and chronologically.

#### **Writing 'Hurrian' identity**

We must now directly address the question of Hurrian identity at Ugarit. Ever since the earliest days of Ugaritology, the vast majority of scholars have assumed that because of the presence of Hurrian language and Hurrian names, Ugarit must have had a significant 'Hurrian' minority. The form this was imagined to take has of course varied from one scholar to another, often shaped by the fashions and prejudices of the time. Schaeffer, for instance, envisaged a stratum of restive 'natives' ruled over by a more educated and internationalist Semitic colonial aristocracy. For him, Hurrian was the language of the streets and the bazaars, and incidents of destruction such as the mutilation of Egyptian statuary in the city were to be attributed to native mutinies against their overlords.25 In modified form, this colonialist model survives even in current scholarship, especially in German Ugaritology. For several decades now Manfred Dietrich, Oswald Loretz, Joaquín Sanmartín and Walter Mayer have promoted a view of the Ugaritian elite as intrusive, hailing from Arabia and superimposing a Semitic superstrate over an increasingly marginalised Hurrian culture.26 The idea here that Hurrian identity was formerly stronger but was declining in Ugarit's later years is very common, even among scholars who don't attribute it to invasion or political subjugation.27 In most cases, the argument is almost entirely linguistic, deriving from the highly restricted number and range of documents in which Hurrian was used in the city's final decades, contrasted with an evident heritage of Hurrian onomastics, socio-cultural elements and references to the Hurrian pantheon. Del Olmo Lete takes a different perspective, arguing that association between Hurrian language and ritual at Ugarit (see below, and also the presence of ritual texts in the 'House of the Hurrian Priest' [Chapter 6] and the use of the Hurrian-derived term *prln* – diviner) indicates that '[t]he Ugaritian cult seems then to have been in the hands of Hurrian personnel, which would be the equivalent of the Levite priestly class in the Hebrew Bible.'28

The question of the Hurrian language must be disconnected from Hurrian identity. We know that the term 'Hurrian' meant something at Ugarit, but it's not at all clear what that was, nor that language was necessarily a defining criterion. It was widely used across the ancient Near East as one of the many terms for inhabitants of the Mitanni Empire, and sometimes to the empire itself as a political entity: for example, the Mitanni king was often known as LUGAL ÉRINMEŠ Hurri – 'king of the

<sup>25</sup> Schaeffer 1939a.

<sup>26</sup> Dietrich and Loretz 1988; 1989; Mayer 1996; Dietrich and Mayer 1999; Dietrich *et al.* 2013. For why this suggestion must be discarded, see Boyes 2019b, n. 10.

<sup>27</sup> Some such as Mallet (2000) have even suggested that Ugarit itself was originally a 'Hurrian' foundation, at least in its Middle Bronze Age, urbanised form. See also the critical discussion by Buck (2018, 10–11), who thinksinstead that 'Hurrian' cultural influence spread in Syria during the seventeenth and sixteenth centuries because of the power vacuum in the region caused by Ḫattušili I's and Muršili's military campaigns. Veldhuis (2014, 279) thinks this cultural influence probably included the use of cuneiform in the northern Levant, believing that vestigial Hurrian elements in the region's cuneiform tradition point to its origins in Mitanni.

<sup>28</sup> del Olmo Lete 2018, 23.

Hurrian troops', while the Hittites referred to the Mitanni as 'the Hurrian enemy'.29 In Schwartz's view (2014, 268) thisissufficient to conclude that 'Hurrian' was understood as an ethnic, not just linguistic, term by the Hittites and others. It was occasionally used by people of Mitanni themselves, in the Hurrian language, such as in the lengthy Amarna Letter EA 24 from King Tušratta to his Egyptian counterpart.30 This letter was, of course, for foreign consumption, but the fact that it's in Hurrian rather than the usual Akkadian suggests that its author was not overly concerned with making allowances for foreign norms.

By the thirteenth century, however, the Mitanni Empire no longer existed, so at Ugarit 'Hurrian' cannot refer to a subject of that polity. Should we assume, then, that a sense of Hurrian identity outlived Mitanni and persisted among inhabitants of that region? It's certainly possible, but very hard to demonstrate.

There's little sign in the archaeology of the region that material culture was used in a distinctive way to articulate a single, coherent ethnic identity.31 However, as we might expect given the ethnolinguistic nature of the 'Hurrian' label, there has been relatively little discussion of material culture to accompany the linguistic and literary analysis. For example, a 420-page volume of *La Parola del Passato* on 'La civiltà dei Hurriti' published relatively recently in 2000 contained only sixteen pages on material culture, of which six were entirely given over to illustrations; going back further, Gernot Wilhelm's 1989 monograph on the Hurrians outsourced a relatively brief chapter on 'art and architecture' to Diana Stein in its English edition, having by Wilhelm's own admission barely covered material culture at all in the German edition of six years earlier.32

Instead, influencesfrom neighbouring cultures are blended differently in different regions to produce a material culture that is both internally diverse and has much in common with aspects of that of the Levant, Anatolia and Mesopotamia.33 Furthermore, there are signs that the Mitanni elite may have positioned itself as to some extent culturally distinct from other segments of society, namely through the conspicuous presence of a number of Indic cultural elements: as well as Indo-European loanwords, there are also references to deities such as Indra and Mitra. The reasons for these elements are extremely unclear, and while they have traditionally been put down to an intrusive elite superstrate, they could also potentially be explained through cultural contacts of other kinds.34 There is so little evidence for when, where and how these elements entered Mitanni/Hurrian culture that speculating would be unproductive, and anyway well beyond the scope of the current discussion. Some of these loanwords

<sup>29</sup> de Martino 2014, 63.

<sup>30</sup> Moran 1992, 63–71; de Martino 2014, 64.

<sup>31</sup> There has not, so far as I can tell, been a lot of work done on Hurrian identity, especially in the Late Bronze Age. For a discussion of the topic as regards third-millennium Urkesh in north-eastern Syria, see Buccellati 2010; 2013.

<sup>32</sup> Salvini 2000, see especially Wilhelm 1989; Pecorella 2000.

<sup>33</sup> Diana Stein in Wilhelm 1989, 80–90.

<sup>34</sup> von Dassow 2008, 77–90.

made it to Ugarit – such as *maryannu*, the elite charioteer class, thought to be cognate with Sanskrit *marya*, meaning a young warrior; however, none of the Indic deities are mentioned at Ugarit, although many other Hurrian gods are.

It seems less justified to talk about 'Hurrian' material culture, then, than to discuss the more specific assemblages and practices of Hurrian-speaking societies on their own terms, as the material culture of Mitanni, Alalaḫ, Nuzi and so on, and to think about the specific characteristics, innovations, hybridisations and patterns of each. This is not, of course, to imply that they did not overlap to a certain extent, but to argue that the existence and form of a pan-Hurrian identity expressed through material culture is something that would have to be demonstrated, and has not been. Even if we could delineate a Hurrian identity, expressed through material culture, it's unlikely we would be able to identify examples of it at Ugarit. We simply don't have the kind of fine-grained detail in either the types of objects recovered or their archaeological contexts necessary to delineate subtle differences in how assemblages are made up in different households or different parts of the city (let alone the Kingdom more broadly), by which we might identify different social groups.

At this point it's best to place the idea of 'Hurrians' at Ugarit to one side. We are currently in no position to know with any certainty what Ugaritian writers really meant by the term 'Hurrian', nor to assess how that correlated with the social reality or with how the people so labelled saw themselves. Since it seems unlikely to coincide with the linguistic definition used by modern scholarship, we should reserve the term 'Hurrian' specifically for the language and not talk about 'Hurrians'. We can, however, still talk about the choices being made regarding the Hurrian *language* at Ugarit, use of writing systems, and how these related to wider questions of identity and social change.

Texts from Ugarit written in the Hurrian language are almost all cultic in nature. As mentioned above, logosyllabic cuneiform outnumbers alphabetic for these almost three-to-one (counting Ugaritic/Hurrian bilinguals); however, scholarly attention has focused mainly on the alphabetic texts, with the notable exception of curiosities like the 'musical scores'.35 Alphabetic cuneiform texts in Hurrian fall into two main genres: sacrificialrituals and hymns. Likewise, in the Ugaritic/Hurrian bilinguals, the Hurrian sections seem to be hymns or prayers. This has led a number of scholars to conclude that by the end of the Late Bronze Age, Hurrian was probably not a widely spoken language at Ugarit but was used almost solely in the cult sphere, and even more precisely for 'l'aspect lyrique du culte ougaritain'.36

More recently, Vita has taken an opposing view and argued that Hurrian did indeed remain a living language in Ugarit even at the end of the Late Bronze Age.

<sup>35</sup> The multilingual word lists account for the vast majority of the logosyllabic Hurrian from Ugarit; the rest is made up of twenty musical texts, two letters (one Akkado-Hurrian; see below) and an Akkado-Hurrian wisdom text (Lam 2015, n. 3).

<sup>36</sup> 'The lyrical aspect of Ugaritian cult', Pardee 1996, 75. For the restrictedness of uses of Hurrian, see Mayer 1996; Dietrich and Mayer 1999, 74; Sanmartín 1999–2000; van Soldt 2003.

As he points out, it was not entirely absent from Ugarit outside the cultic sphere. There are a number of lexicographic texts, which are scholarly in nature and probably used in literate education, so don't necessarily indicate that Hurrian was actually spoken, but there are also two fragmentary letters written in logosyllabic cuneiform: RS 11.853 and RS 23.031, from the Royal Palace and Southern City respectively. The former is in Hurrian and refers to a messenger and the sending of men, perhaps soldiers; the latter is unpublished but said to be in a mixed language with elements of Akkadian and Hurrian. They probably originate outside Ugarit – RS 11.853 is likely from Karkemiš and the mixed language of RS 23.032 finds parallels at Qaṭna – and so belong to the genre of international correspondence. Vita (2009, 225–227) argues that rather than being rare exceptions that prove the rule of Hurrian's restrictedness, we should instead see these as examples of people choosing to communicate in Hurrian even in a genre in which the use of Akkadian was a very strong convention: even in the Amarna letters from the Hurrian-speaking kingdom of Mitanni, only one is in the Hurrian language. Vita also believes that the pattern of cross-influence between Hurrian and other languages at Ugarit is consistent with its being a living language. It has been widely observed that evidence of Hurrian influence on the Akkadian of Ugarit declines over time, while there is much more cross-pollination between Hurrian and Ugaritic. As Vita points out, this is consistent with Hurrian and Ugaritic both being living languages and so subject to everyday mutual influence, while, as predominantly a written language at Ugarit, Akkadian would have been more isolated from this interplay.37

Van Soldt has done very interesting work with Hurrian onomastics, teasing out some intriguing details of how Hurrian names were integrated into wider Ugaritian society and the contextual nature of their prestige.38 Both Semitic and Hurrian names occur within the same families, he points out, and the trend seems to be towards more children having Semitic names than their parents. This implies changes in onomastic fashion and is consistent with a decline in the prestige of Hurrian-language names. We could think of several possible reasons for this, which may connect to wider questions of identity and social integration; however, without any evidence for how language was used in articulating identity, it is impossible to choose between them. If there was a Hurrian ethnic identity, with language as a key boundary, then we might argue that this identity was becoming increasingly marginalised or stigmatised, perhaps subsumed into a new, more totalising concept of 'Ugaritianness'. On the other hand, it could be purely a linguistic process, and not significant in the construction of social identities at Ugarit. Or, on the third hand, not an indication of further-reaching linguistic change at all, but merely a phenomenon in itself – part of the everyday currents of fad and fashion that are a constant of human naming.

<sup>37</sup> Vita 2009.

<sup>38</sup> van Soldt 2003. See also earlier work, less in-depth but with a wider geographical scope, by Nadav Na'aman (1994).

The clearest insight into what's going on here comes with the elite. Just as the Hurrian language wasrestricted almost entirely to the religioussphere in officialtexts, and especially to material to do with music and chanting, the use of Hurrian-language names was also more appropriate in some circumstances than others. Broadly speaking, high-status Ugaritians show the same relatively free choice between Semitic- and Hurrian-derived names as does the general population, but with a few additional quirks. Unlike in most other coastal Syrian kingdoms, Ugarit's kings never have Hurrian names, although several princes do.39 It wasn't the case that Hurrian names were used only for children not expected to accede to the throne – ʿAmmiṯtamru II's crown prince had a Hurrian name – so it may instead be that Hurrian-named princes took Semitic throne names when they became king. Unfortunately, we can't tell for certain: we only hear of ʿAmmiṯtamru's son because in the scandal surrounding the divorcing of his mother he was given the choice whether to side with his father and inherit the throne or side with her and forfeit his inheritance. The next king has a Semitic name but we have no way of knowing whether this is the same person with a new name, or another prince who replaced him. While Hurrian names seem not to have been preferred at the highest tier of Ugaritian society, among senior officials like *sakinu*s (governors) and administrators, they were extremely popular: 69% of the former have Hurrian names and 75% of the latter (though sample sizes are small).40

Hurrian linguistic elements were thus used in very particular ways, with differing levels of prestige and desirability depending on the context. It is productive to compare the situation with that in the Ḫatti. The archives of Ḫattuša also contained a significant quantity of Hurrian written material and there too it was largely confined to the cult sphere, sometimes appearing in bilinguals with the primary local language. The factors motivating changes in fashion are a bit more transparent here than at Ugarit, generally coinciding with the political fortunes of Hurrian-speaking regions. In the thirteenth century, Hurrian elements became especially prominent during floruit of Queen Puduḫepa, who came from the region of Kizzuwatna, which had a sizeable Hurrian-speaking population.41 The use of Hurrian names among the Hittites shows both similarities and differences to that at Ugarit. Like at Ugarit, 'local' and Hurrian names both occur within the same families, and Hurrian names were particularly popular among the Hittite elite during the thirteenth century. There isn't the same reluctance for kings to bear Hurrian names, however. Many Hittite rulers had both Hittite and Hurrian names (such as Muršili III/Urḫi-Tešub), although there

<sup>39</sup> van Soldt 2003, 685.

<sup>40</sup> Van Soldt is less interested in speculating on whether or how this would relate to the existence of actual 'Hurrians' in Ugarit, although he does say he doesn't think Hurrian was widely spoken by the final decades of Ugarit's existence. Lipiński (2016), discussing Hurrian names in an article mainly focused on the southern Levant, is prepared to go further than van Soldt in using the onomastics as a basis for drawing demographic conclusions. His approach is somewhat unsophisticated, however, since he largely uses the presence of theophoric names mentioning Hurrian deities as a proxy for the existence of Hurrian-identifying people.

<sup>41</sup> Giorgieri 2013.

are no examples of people with Hittite names choosing Hurrian ones upon gaining the throne. Stefano de Martino has suggested that the prestige of Hurrian names at Ḫattuša was directly tied to the prevailing closeness of ties with Kizzuwatna and south-east Anatolia in general.42

Similar prestige for Hurrian onomastics within elite culture is apparent elsewhere in northern Syria during the Late Bronze Age, though as we would expect, it's not identical from site to site. At Karkemiš, for example – a traditionally Hurrian-speaking city within the heartland of Mitanni – Hurrian names were particularly preferred, even by the cadet branch of the Hittite royal family, which was parachuted in to rule the city after its conquest. In Amurru, Hurrian names saw an upsurge in popularity after its own incorporation into the Hittite sphere, de Martino suggests in imitation of Karkemiš.43 In Ugarit, the Hurrian language (along with north Syrian cultural elements such as the Hurrian-named deities and their associated mythology) seem rather deep-rooted in the local culture so are perhaps not down to its own brush with Hittite imperialism. A great mystery here is Ugarit's history before the mid-fourteenth century, and especially its unknown relationship with Mitanni (see Chapter 1).

Despite all the uncertainties, I think it makes sense to posit a regional Syro-Anatolian elite culture in which the Hurrian language, and perhaps also certain other cultural elements associated with it, carried certain shared connotations and associations. I would hesitate to call this a Hurrian elite identity; it might rather be based on standards of prestige that gained currency during the Mitanni ascendancy, or even before that. Ugarit's elites seem to have participated in, and positioned themselves in relation to, this shared set of linguistic and cultural elements, just as other regional elites did, but of course exactly how they adopted, manipulated and utilised these features varied according to local context and historically contingent events such as the Hittite Great King's marriage to a Kizzuwatnean woman. By the later thirteenth century, Ugarit shared the general sense that Hurrian was particularly suited to religious ritual, but while Hurrian names were *à la mode* in Ḫattuša and Amurru, the situation was more complex and ambivalent at Ugarit, which may hint at how people in the kingdom wanted to situate themselves with regard to their neighbours and these Syro-Anatolian ideas of prestige.44

In summary, the way Hurrian was written down at Ugarit should be seen more in terms of changing elite culture than as a proxy for the fortunes of a community within the city of people identifying themselves as 'Hurrians'. This doesn't preclude such groups having existed, of course, but they remain intangible to us. If they did,

<sup>42</sup> de Martino 2010, 18.

<sup>43</sup> de Martino 2010, 22.

<sup>44</sup> We could also note, however, that even within Mitanni, there were social factors we don't really understand governing elite naming conventions. At the highest social levels, Indo-European-derived names were preferred to Hurrian ones, which von Dassow (2008, 86–87) has strongly, and convincingly, argued is due to socio-cultural factors rather than an indication that the Mitanni elite was composed of 'Indo-Europeans', or even speakers of an Indo-European language.

then there may well have been connections between their circumstances and the prestige of Hurrian culture among the elite. If, as has been suggested,45 the split between logosyllabic and alphabetic cuneiform for writing Hurrian is largely chronological – that is, there was a gradual shift from logosyllabic to alphabetic cuneiform – then this would be another part of a wider move on the part of Ugarit's elites to use script as a means of emphasising their connection with the general population.46 This trend should be seen as an act of negotiation, a careful tightrope act in which Ugarit's elites played up their populism through acts such as the official adoption of the vernacular language and a distinctively Levantine alphabetic writing system, while also being unable to fully abandon the traditional systems, ideology and practices of an international model of Bronze Age eliteness in which they were fundamentally and inextricably entwined. By switching to alphabetic cuneiform for their Hurrian ritual writing, they were again attempting to have their cake and eat it, asserting their alignment with the local and the vernacular while also integrating traditional international elite 'vocabulary' into that new, populist form.

#### **Looking for Cyprus-town**

Beyond Hurrian, the main non-standard language and script represented in texts potentially written in the kingdom of Ugarit is Cypro-Minoan. Nine examples of Cypro-Minoan have been found at the city: one complete tablet, four fragments (two from the same tablet), two labels, a silver bowl and a pithos rim.47 Three fragments and the pithos rim were all from the House of Yabninu, a further fragment from the House of Rašapʾabu, the complete tablet from the House of Rapʾanu, the two labels from the House of ʾUrtenu and the bowl from a house on the acropolis, near the house of the High Priest.48 Most of the private individuals in whose archives this material has been found had some other connection with Cyprus: Yabninu's tablet store seems to have a particular interest in overseas trade and managing foreigners in Ugarit's territory, and Cypriot imports were found in his house; Rašapʾabu was the *akil kāri* – overseer of the quay – and Rapʾanu's tablets also included Akkadian diplomatic correspondence with Cyprus. The Cypro-Minoan from Ugarit is often seen as typologically distinct from that on the island, traditionally being dubbed 'CM3'. Some more recent work has questioned whether this distinction is really so cut and dried; in particular, Silvia Ferrara argued in favour of seeing the corpus as more integrated and the differences

<sup>45</sup> Vita (2009, 222) attributes a perceived gradual decline in the use of logosyllabic cuneiform for Hurrian to unspecified 'Hittite pressure'.

<sup>46</sup> See Boyes 2019a and Chapter 11 in this volume.

<sup>47</sup> Steele (2018, 204) considers the silver bowl inscription questionable as an example of Cypro-Minoan. A further Cypro-Minoan inscription was found on a cylinder seal at nearby Latakia (Steele 2018, 202).

<sup>48</sup> Ferrara 2012; 2013.

as palaeographic rather than typological.49 Nevertheless, the complete tablet from the House of Rapʾanu, RS 20.25 (##215 in the Cypro-Minoan corpus) (Fig. 9.2) is in form far more similar to cuneiform tablets from Ugarit than are any Cypro-Minoan texts from Cyprus itself.50 For this reason, a number of Cypro-Minoan specialists have come to believe that at least some of the Cypro-Minoan found at Ugarit was actually produced there, and influenced by local writing practices, rather than being imported.<sup>51</sup>

Although Cypro-Minoan is mostly undeciphered, certain aspects of RS 20.25/##215 have encouraged commentators to draw conclusions about its content. From the layout of the text on the tablet, it resembles a list, and one repeated string of signs has been interpreted by Miguel Valério as rendering a version of the Ugaritic word for son, *bn*. 52 This would potentially make it an example of a list of personnel (a familiar type among Ugaritian administrative texts) written in Ugaritic language but Cypriot script.

Given the many uncertainties, we should refrain from drawing firm conclusions as to what language this tablet is written in. Even without this knowledge, it points to contact between Cypriot and Ugaritian traditions of writing, and someone familiar with one experimenting with aspects of another.53 This is by no means implausible or even surprising, given how close contacts were between Ugarit and Cyprus. We have already seen that the so-called Ritual of National Unity refers repeatedly to Alašiyans, but they are also one of the most commonly mentioned ethnic groups within the Ugaritic and Akkadian administrative tablets. Numerous individuals within these texts are described as Alašiyan, pointing to the likely presence of people originating on the island in the Kingdom of Ugarit.

What is potentially more significant, however, are indications that Alašiyan presence in Ugarit went beyond merely individual travellers and merchants. The personal name Bn-ʾAlṯn (Cypriotson) seems to point toward second-generation (at least) immigrants,54 and a possible hint at something more substantial comes in KTU 4.102, a bilingual administrative tablet listing people by household. The main text is in Ugaritic and Alphabetic Cuneiform, but on the left edge is an Akkadian label reading URU.*A-la-ši-ia*, that is, Alašiya preceded by the logogram for town or city rather than the expected KUR – land.55 It is not unknown for some alternation

<sup>49</sup> Ferrara 2012; 2013.

<sup>50</sup> Ferrara 2012; Steele 2018, 107–108.

<sup>51</sup> Ferrara seems to have changed her views: in her corpus she judges that the Cypro-Minoan examples from Ugarit are 'not unlikely' to have been produced there (Ferrara 2012, 171ff.), but more recently she has argued that the tablet fragments from Yabninu's archive are probably imported (Ferrara 2016, 235). Steele considers this unlikely (pers. comm.) and thinks the Ugarit inscriptions were most likely produced by Cypriots resident in the city (Steele 2018, 203–204).

<sup>52</sup> Masson 1973; 1974, 30–35; Valério 2016, 346–396, and esp. 364–367.

<sup>53</sup> Steele 2018, 109. For a comparison of writing practices and the material characteristics of tablets produced in Cyprus and Ugarit, see Steele and Boyes forthcoming.

<sup>54</sup> Astour 1970, 122.

<sup>55</sup> This preceding logogram can either be read as a determinative not intended for pronunciation or as standing for the equivalent Akkadian words *ālu* – town and *mātu* – land.

*Fig. 9.2. Cypro-Minoan tablet RS 20.25/##215, from the House of Rapʾanu. Photo by Silvia Ferrara; used with permission.*

between these two to occur, especially where the same names are used for territories and their capitals. Although we are now confident in placing it on Cyprus, we do not know exactly what kind of geopolitical entity Alašiya was. Some estimates place it in the vicinity of the modern town of Alassa, in which case we may well have a case of continuity of name and the possibility that Cypriot Alašiya could be both a land and a city. However, given that this document is a census of families (wives and children as well as just men) presumed to be in the territory of Ugarit, Monroe reads the label instead as 'Cyprus-town', denoting some kind of district or village within either the city or kingdom of Ugarit.56 This is a wonderfully evocative and enticing notion, conjuring up a Cypriot community clustered in their own particular corner

of town. Some of the people listed have obviously West Semitic names, so are we to imagine a mixed community, one where locals live too, or perhaps where second- or third-generation immigrants are starting to adapt to the local onomastic practices? It is far from certain that this is the correct interpretation of these words, and we should want something a little more substantial before we start imagining Ugarit having a defined Cyprus-town district, complete with Cypriot restaurants and other cultural elements, but it's an intriguing prospect.

If correct, it would represent a fascinating contrast with Ugarit's apparent reluctance for Anatolian visitors to convert their presence into a more permanent arrangement: diplomatic letter RS 17.130+ concerns merchants from the Anatolian city of Ura and an agreement with the Hittites whereby their presence in Ugarit was limited and their ability to own property in the Kingdom curtailed.57 We potentially

<sup>56</sup> Monroe 2009, 220.

<sup>57</sup> Monroe 2009, 178–179.

have a situation where an Alašiyan immigrant community existed, possibly in a separate ghetto but nevertheless accepted and administered by the Ugaritian authorities in the same way as everyone else, whereas Anatolian merchants were socially distinct, their administration a matter of international diplomacy and Ugarit's elites apparently concerned by their growing social and economic importance. Of course, the reconstruction is highly speculative, but it would make sense within the wider social and political context within which Ugarit found itself. As a proud and independently minded kingdom nevertheless under the thumb of Hittite imperialism, it's easy to see how Ugarit would be nervous of increasing Hittite presence and economic power. Alašiya was powerful – its king was a 'Great King', treated as an equal of the Pharaoh and Hittite ruler – but this seems to be derived largely from its economic and strategic significance as chief supplier of copper; there are no indications that it attempted to exert dominance over its neighbours either politically or economically. Ugarit's elites could probably afford to be more relaxed about resident communities of people tracing their origins to its close neighbour and long-term trading partner.

Despite all this, there is little in the way of conclusive material culture evidence to confirm the presence of Cypriots in the kingdom of Ugarit. Unlike with the 'Hurrians', Cyprus has a relatively well-defined, well-understood and distinctive material culture repertoire, although it should be stressed that it is not entirely self-contained, including several overlaps and mutual influences with both Levantine and Mediterranean objects.58 As we've mentioned, Cypriot material culture – especially pottery – is known from Ugarit, but – contrary to a general assumption that is still taken for granted too often in Near Eastern archaeology – pots don't necessarily equal people. It's not enough to show that material culture associated with a particular place is present; we must instead focus on the practices that surround that material culture: the production, consumption, trade and ideology to show that people were living in Ugarit in distinctive ways that we associate with people from Cyprus.

By and large, this isn't what we see with Cypriot material culture in Ugarit. Among the largest class of published objects – pottery – we're dealing mainly with decorated finewares, especially jugs and bowls; that is, relatively prestigious items that were widely traded in the east Mediterranean. Provenance analysis tends to support the excavators' assumptions that the majority was imported.59 However, the 'corpora' of Ugarit's pottery presented in *Ugaritica* II and VII60 can't be considered reliable guides to the full range of ceramic material recovered from the site. The material

<sup>58</sup> Voskos and Knapp 2008.

<sup>59</sup> Renson *et al.* (2013), for example, presents lead isotope analysis results of White Slip II pottery from Minet el-Beida in comparison to samples from Hala Sultan Tekke and the pottery production centre at Sanidha-Moutti tou Ayiou Serkou in the south-east Troödos. Several of the sherds from Minet el-Beida seemed to match clays from Sanidha, but others were from a different, unknown source. It should be noted that the authors do leave open the possibility that this might have been in the vicinity of Minet el-Beida itself (233).

<sup>60</sup> Schaeffer and Chenet 1949; Courtois and Courtois 1978.

they present mostly hails from tombs and has thus undergone at least two selection processes prior to recovery: firstly by the people choosing suitable grave goods and secondly by the ancient looters who left this material behind. Additional rounds of selection occurred in collection and publication. The cumulative upshot of these various winnowings is a disproportionate presentation of decorated material and especially Mediterranean (predominantly Aegean) imports. As presented in the corpora, Aegean imports greatly outnumber either Cypriot or locally produced pottery. This certainly does not reflect reality. More recent commentators have emphasised that the proportion of imported pottery at Ugarit is negligible – less than 1% of the total61 – and it would be surprising if even this small amount was mainly comprised of Aegean material. At the Levantine site with the next-highest amount of Mycenaean pottery after Ugarit, Tell Abu Hawam, Cypriot material outnumbered Aegean wares by around forty to one.62 Jean-Yves Monchambert confirms that in the 1975 and 1976 excavations, Cypriot pottery was by far the most common class of imported ceramics.63 There can be little doubt that the prominence of Mycenaean imports in the original published reports from Ugarit is more representative of Schaeffer's preconceptions and biases than anything else. It's doubtful whether the kind of Cypriot pottery that would be useful for demonstrating the presence of actual Cypriots – low-prestige coarsewares, kitchen-kit and so on – would have made it into these publications if it were present. More usefully, Monchambert (1983, 28) does discuss locally produced Cypriot-style ceramics from areas excavated in the 1970s, but these are finewares – flasks, kraters and so-called milk-bowls.

Ifspecific archaeological proof of the existence of a community or communities of resident Cypriots at Ugarit is lacking, we can at least say that the Cypro-Minoan-inscribed objects from the site fit well within a general context of hybrid objects from Cyprus and the Levant that blend features or traditions from both places, as well as influences from elsewhere, such as the Aegean or Anatolia. Perhaps the most prominent example of this around the end of the Late Bronze Age is the so-called LH IIIC (or White Painted Wheelmade III) pottery, which develops out of the indigenous Cypriot tradition, but incorporates Aegean and Levantine features, and was widely produced and distributed around the east Mediterranean at this time.64 Often assumed to be straightforwardly 'Mycenaean' and taken as evidence for Aegean migrants, it in fact shows the popularity of hybrid material culture forms. There are other examples too: Rude Style bell kraters incorporated an Aegean shape into the Cypriot repertoire, and

<sup>61</sup> Monchambert 1983, 26. This imbalance is also apparent even in more recent excavation work. In his publication of the ceramics collected from residential buildings in 1975 and 1976, Monchambert (2008, 11) calculates that around 70% of imported material was kept and published, compared to only around 2.5% of locally-produced pottery.

<sup>62</sup> Åström 1993, 312.

<sup>63</sup> Monchambert 2008, 245.

<sup>64</sup> Sherratt 2003, 45. Sherratt notes that the Levantine production of 'Cypriot' pottery began in LB II with stirrup jars (as well as their contents, she assumes), including 'possibly' at Ugarit, before full-scale 'LH IIIC' production took off around the end of the thirteenth century.

were decorated with motifs whose origins lay in the Aegean, Cyprus and the Levant.65 Likewise, Cypriot bronzework draws heavily on overseas influences, especially in the anthropomorphic figurines, which recall the well-known metal gods of the Levant, but also blend in influences from the Aegean, Anatolia and Cyprus itself.66 Architecturally, parallels have been seen between the 'Bâtiment à la Colonne' in Enkomi and the 'Temple with Rhytons' in Ugarit's residential area, and the intramural ashlar chamber tombs of Enkomi may draw inspiration from those of Ugarit.67

We could go on indefinitely; there can be no doubt that contacts between Cyprus and the Levant were extremely close, and that this combined with the island's status as a trading nexus and middle-man to result in extremely hybrid material culture and practices that incorporated elements from far and wide. For these artistic and artisanal skills to be transferred we have to assume not just merchants visiting, exchanging their wares and leaving, but craftspeople and others moving back and forth and probably spending extended periods of time overseas. At the level of palace personnel, this movement of skilled workers is well attested in Late Bronze Age correspondence, and we have direct testimony of this occurring between Ugarit and Alašiya in the scribal field: RS 94.2177+, an Akkadian diplomatic letter from the king of Alašiya, found in the house of ʾUrtenu in Ugarit, includes a personal letter on its reverse from the Ugaritian scribe, asking for furniture to be sent to him in Cyprus.68


15' all your good words, my lord, I pronounced before my lord since I am your good servant in post here. And, my lord, those things which I have asked for, 20' have them brought to me: 1 table of very good quality 5 chairs. In the hand of your messenger when he comes, have them brought.69

The status of palace society in Cyprus during the Late Bronze Age remains extremely uncertain: although the diplomatic texts from Alašiya seem to point towards a

<sup>65</sup> Voskos and Knapp 2008, 669.

<sup>66</sup> Negbi 1976, 29–39; Voskos and Knapp 2008, 670ff. For the 'smiting god' pose, cf. also the well-known 'Baʿlu stele' from Ugarit, now in the Louvre.

<sup>67</sup> Negbi 2005, 11–12.

<sup>68</sup> Ferrara 2016, 239.

<sup>69</sup> Lackenbacher and Malbran-Labat 2016, 39; translation by the author, after the editors' French.

centralised, monarchical polity similar to those seen elsewhere in the east Mediterranean and Near East, this is not matched by the archaeology, which has not been able to clearly locate a 'capital' or palatial architecture. Economically too, it is often thought the island was characterised by decentralisation and 'private enterprise' (piracy?), rather than a more orthodox 'palace system'. It's worth asking exactly what kind of organisation this Ugaritian scribe and other travelling specialists might have found themselves attached to: a royal court bureaucracy similar to the ones they had left, or a Jabba's Palace of gangsters, freebooters, scum and villainy? More to the point, in this sort of decentralised context it would probably be a mistake to imagine that such movement of people was confined to palace personnel and elite transactions. For this reason, even though we can't directly prove there were Cypriots living and working in Ugarit long-term, it would be extremely surprising if there weren't.

### **Hittite(s) and Egyptian(s) at Ugarit**

Alongside the scripts and languages we've already discussed, there are also a number of Hittite, Hieroglyphic Luwian and Egyptian inscriptions from Ugarit. It has long been recognised that these mostly originated outside the kingdom and so had a fundamentally different relationship with any Anatolian or Egyptian immigrants and their identities than the cases discussed above.

It's difficult to get exact numbers on Egyptian hieroglyphic inscriptions found at Ugarit. The corpus has received little direct attention and to my knowledge has not been collated in a single place.70 Bordreuil and Pardee (1989) record 97 inscribed objects, but more are likely to have been found in the three decades since that collection was published (Table 9.1).

It's immediately clear that this is a dramatically different kind of corpus to the purpose-made clay tablets we've mainly been discussing so far. All these items are, in one way or another, prestige goods and the vast majority are probably imports. The largest class of inscribed object is, unsurprisingly, scarabs. These were widely distributed in the east Mediterranean, and while they were also produced outside of Egypt by others keen to cash in on demand, they are certainly objects of trade and prestige. The second most common object type is even more typical of international exchange and of elite trappings. Alabaster vessels were a common Egyptian diplomatic gift to the Levant and occur with particular frequency in Ugarit and Byblos, cities known for their close political and cultural connections with Egypt (although it has been noted that it can be difficult to distinguish between genuine Egyptian products and Levantine versions).71 Many of the inscriptions themselves are pharaonic cartouches, adding to the Egyptianising cachet.

<sup>70</sup> Important discussions of individual or multiple objects can be found in Schaeffer *et al.* 1956; Giveon 1981; Vita and Galán 1997; Yon 2006.

<sup>71</sup> Feldman 2002a; Sparks 2003.


*Table 9.1. Objects bearing Egyptian hieroglyphic inscriptions from Ugarit.72* 

The find-spots further emphasise that we're dealing with elite products from the world of Ugarit's upper echelons. The highest concentrations of inscribed Aegyptiaca come from the Royal Palace and the area of the Acropolis around the two principal temples and the house of the High Priest, although it seems most areas of the tell produced some examples. While many are personal prestige items, there is also a significant group of stelae, statues and other items associated with temple contexts, most likely cultic dedications. It's not uncommon for Egyptian dedications to appear in cult sanctuaries in Levantine cities with which they had close relations: the presence of such Aegyptiaca on Ugarit's Acropolis is comparable to the Egyptian dedications to Baʿalat Gebal/Hathor in Byblos. We can't say for certain whether such dedications were made in Egypt and brought to Ugarit in a finished state or whether they were the work of Egyptian craftsmen sent to the Levant to produce them, but we've already mentioned the existence of a letter alluding to a request

by the king of Ugarit for an Egyptian sculptor to be sent to create an image of the pharaoh to stand in the temple of Baʿlu (RS 88.2158).73 Either way, most of these were probably not the work of long-term Egyptian immigrants integrated into Ugaritian society.

A couple of items are less clear-cut, however. Probably the most famous Egyptian-style object from Ugarit is the so-called marriage-vase of Niqmaddu II (RS 15.239) found in Room 31 of the Royal Palace, near the Central Archive. The two adjoining fragments from an alabaster vessel are illustrated with an image of a Syrian-style man and an Egyptian-style woman. The inscription clearly labels the man as Niqmaddu, king of Ugarit. There's been a great deal of discussion as to whether this represents a wedding and, if so, whether the woman is indeed an Egyptian. We'll return to these questions in a later chapter. For now, what concerns us is where the vase was created. Stone vessels such as these, both decorated and not, were often used as diplomatic gifts. However,

<sup>72</sup>Bordreuil and Pardee 1989.

<sup>73</sup> Lackenbacher 1995.

this example is unique among the Egyptian-style stone vessels at Ugarit both in its decoration and its reference to an Ugaritian king. It also seems to alter certain aspects of Egyptian iconography, including details of the woman's headdress and the replacement of an Egyptian-style uraeus-frieze with Syrian ibex-heads, which has led commentators, ever since its first publication, to see it as probably originating outside Egypt.74 Marian Feldman (2002a) argues persuasively that the result is a deliberately hybrid object, most likely produced on the instructions of Niqmaddu himself and making careful use of ambiguity to associate him with Egypt's international prestige. The craftsperson responsible was evidently highly familiar with, and competent in, Egyptian techniques and artistic conventions, as well as the hieroglyphic writing system. An Egyptian artist serving at Niqmaddu's court or resident in the city would seem an obvious candidate.

The second item to discuss is a bronze sword found as part of a cache with other bronze weapons in the 'House of the Armourer' in the well-to-do residential east of the Royal Palace and dating from the late thirteenth century. The sword is a long, straight, cut-and-thrust blade of non-Egyptian type. Lucia Vagnetti (2000, 317) sees typological similarities with central European and northern Italian weapons but believes this to be of Near Eastern manufacture, possibly inspired by a foreign model. We can't be sure it was made in Ugarit itself, or whether the cartouche was always part of the sword or engraved at some point after manufacture; however, as we've already mentioned, textual evidence from the House of ʾUrtenu points to surprisingly close relations between Ugarit and the court of Merneptah (despite the fact that Ugarit was still theoretically a Hittite vassal) so it wouldn't be at all surprising if the inscription at least was added there. Schaeffer (1956, 169) notes that the hieroglyphs were perfectly executed, but quotes Vandier's opinion that the engraver was nevertheless not an Egyptian, although he possessed 'une certaine experience'.

Finally, it's worth mentioning again the hieroglyphic seal of Ṯipṭibaʿlu. We don't know whether this was created in Ugarit or was something he picked up or was given during trade activities with Egyptians, but again, it's not inconceivable that it may be an example of the writing of hieroglyphs in Ugarit.

There's a great deal in common across these possible uses of Egyptian hieroglyphs from Ugarit. None of them are certain, all are hybridising, prestige objects with international associations, all are produced to a high standard. It seems very clear that what we're dealing with here is not hieroglyphs as a means of self-expression or identity-negotiation on the part of an Egyptian community at Ugarit, but an attempt by the highest echelons of Ugaritian society to tap into Egypt and Egyptian culture's international cachet. If the way Hurrian was used at Ugarit was a careful attempt to juggle populism and a traditional, regional elite culture, then the use of Egyptian hieroglyphs is just the opposite: a fairly straightforward and unambiguous deployment of the traditional

<sup>74</sup> Desroches-Noblecourt 1956; Bryan 1996, 60; Feldman 2002a.

cultural vocabulary of the Bronze Age international jet-set.75 These objects were made for elite use and their message was wholly for elite consumption. Egyptians may well have been involved in their creation – we have already mentioned that texts record the occasional Egyptian in the city – but these are not Egyptian objects; they are products firstly of the Ugaritian elite and secondly of Late Bronze Age globalisation.

We turn, finally, to the Hittites. There's a better case to be made for Hittite communities in the kingdom of Ugarit than there is for Egyptians, although they may not have been resident full-time. As we've already discussed, the Ugaritian authorities were apparently at pains to make sure that Hittite merchants remained transitory visitors, not full-time settlers who could own property. It's quite possible that this temporary merchant community is what's being referred to in KTU 4.149 with its record of a measure of wine *l ḫty . maḫdh* – 'to the Hittite(s) in Maʾḫadu (Minet el-Beida)'.76 The vast majority of examples of languages or scripts from Ugarit come in the form of seal-impressions on Akkadian lettersfrom elsewhere in the Hittite sphere of influence. There are over fifty of these, sometimes with multiple seal-impressions appearing on a single tablet. Both Hittite cuneiform and Luwian hieroglyphs appear; many of the seals are bilingual/biscriptal. Several in Luwian alone also exist, particularly from Hittite vassals such as Karkemiš and Amurru.77 There are also two actual seals – a copy of the great seal of Muršili II, found in the Royal Palace, and a much less elaborate steatite seal in hieroglyphic Luwian.78

Evidence for the actual reading and writing of Anatolian scripts at Ugarit is considerably more limited. There is a single letter in Hittite cuneiform, RS 17.109, sent from an unknown foreign location, and a column of Hittite in the trilingual (Sumerian, Akkadian and Hittite) magical text RS 25.421+.79 The latter is the only instance of Anatolian writing from Ugarit for which any argument can be made for its production within the city itself.

The near-total absence of Hittite cuneiform and Luwian hieroglyphic usage at Ugarit is surprising both because of the kingdom's political incorporation into the Hittite Empire (other Hittite vassals used Luwian especially) and because of the presumed significantseasonal Anatolian merchant population we've already discussed. The two Hittite seals found in Ugarit can best be explained as the property of Hittite agents or functionaries either visiting the city or semi-permanently stationed there, but overall their small number is consistent with a very small Hittite official presence in keeping with the hands-off style of control generally thought to have existed. The lack of material associated with merchants is more perplexing: did these people

<sup>75</sup> Compare the highly Egyptianising material culture of royal burials at Byblos in the Middle Bronze Age. See chapter 2 of Boyes 2013, with further references. On the hybrid elite artistic culture of Late Bronze Age Mediterranean globalisation, see Feldman 2002b; 2006.

<sup>76</sup> Astour 1970, 121.

<sup>77</sup> Schaeffer *et al.* 1956.

<sup>78</sup> Schaeffer *et al.* 1956.

<sup>79</sup> Schaeffer *et al.* 1956; Laroche 1968; Bordreuil and Pardee 1989.

not keep records, send letters, seal documents? Perhaps the numbers were smaller than the texts imply, perhaps they weren't literate, perhaps they took their writing materials home with them when they left? Certainly, there is nothing whatsoever to suggest Anatolian writing systems were being used by Anatolians at Ugarit as a means of defining and delineating an ethnic or social identity. Rather, when it comes to questions of social identity, it's the absence of Hittite writing that is significant. Compared with the relatively enthusiastic take-up of Anatolian scripts by other Syrian vassals – especially Luwian, which would persist in the region long after the Hittite Empire itself had fragmented – the rejection of these writing systems by Ugarit's elite – particularly in emblematic contexts such as personal and offical seals – can only be interpreted as a political statement of self-identity, asserting their distinctiveness and independence from the Hittite world.

### **Minority writing systems and defining identities at Ugarit**

This rather long survey of the 'minority' writing-systems of Ugarit and their associated languages hasshown that while their patterns of use and archaeological contexts differ considerably, they have in common that they belong mainly to the world of Ugarit's elite. It's possible, to varying degrees, to identify either individuals or groups within the kingdom who may have belonged to the linguistic or ethnic communities associated with these practices, but it's almost impossible to trace any connection between the writing practices and how these people or groups thought about themselves or articulated their own identities. Instead, the way these scripts and languages were used and not used tell us a great deal about how Ugarit's leaders and high-status inhabitants positioned and defined themselves in relation to others, and how they thought about and presented the polity they headed. When thinking about the ethnic make-up of the Ugaritian state, previous discussions have often focused on the various gentilics of the administrative texts, and on the 'Ritual for National Unity' (KTU 1.40 and 1.84) as evidence for a sort of proto-multiculturalism, a sense that the people of Ugarit recognised and celebrated their ethnic diversity (not to mention gender and other categories) and utilised public ritual – sometimes performed bilingually – to integrate these various identities within a single nation.80

I don't disagree with this, but I do hope that this discussion has shown how this was a rather situational agenda. Diversity and multiculturalism came much more to the fore in certain spheres of public life than others, and some groups were accepted and integrated within Ugaritian society much more than others, most likely for reasons connected to wider political imperatives: contrast the apparent treatment of Alašiyans with that of Hittite merchants. These agendas were doubtless enacted and reflected in many different ways, one of which was script and language. Perhaps the subtlest and most ambiguous of these concerns Hurrian. I have argued here that we shouldn't try to use the language as a proxy for the presence or absence of

<sup>80</sup> Sanders 2004.

'Hurrians' or a 'Hurrian' culture in Ugarit: it is far too unclear what those terms would really mean, even if they could be directly indexed by language. But in how Ugarit's authorities deployed the Hurrian language, we can begin to shed light on how they engaged with regional structures of high-status behaviour and especially the heritage of Mitanni. The transition from Hurrian being written mainly logosyllabically to the use of alphabetic cuneiform seems to represent an adaptation and conscious appropriation of this regional tradition to a specifically Ugaritian form, which fits within the wider linguistic and cultural patterns that we return to again and again in this book.

The lyrical, ritual use of Hurrian places this ideological programme within Ugarit's own public sphere. When spoken and performed, its use was outward-facing, directed towards and engaging with the wider public. The writing practices used to codify such rituals don't come across in their performance, of course, so if the change to Alphabetic Cuneiform was a 'Levantisation' of the pan-regional Mitanni tradition, then it was hardly one that would have been readily interpretable by the vast majority of the population. So we must ask who this message was intended for, if such it was. The only answer can be the elites themselves, and the bureaucracy that enacted their will. This careful, ambivalent recasting of a regional tradition and Ugarit's relationship to it can only be a story the elites were telling themselves about who they were and who they wanted to be.

But this wasn't the only story they were telling, as the Egyptian hieroglyphic material tells us. As we've seen, the use of Aegyptiaca at Ugarit is wholly within the standard tradition of Late Bronze Age globalised elite culture. Individual objects such as Niqmaddu's 'marriage-vase' or especially the Merneptah sword might be hybrid objects reflecting an internationalism less focused on the great powers of the Amarna Age, but there seems to be little ambivalence about the prestige of Egypt. With the possible exception of the temple dedications, the Aegyptiaca from Ugarit belong almost entirely to a very restricted elite sphere: they were intended to be seen principally by other members of the elite, as well as their guests and visitors from other kingdoms. It's hard to know who would have had access to the Temple of Baʿlu, whether there was potential for a general public to see these items or whether the grand temples of the Acropolis would also have been a rather restricted elite space. It's also worth stressing that many of the Egyptian objects – including some of those in the temple – would have been antiques of several centuries old by the time of Ugarit's destruction, and we might question whether their main resonance would have been as icons of contemporary elite globalism or as venerable relics of Ugarit's own history.

Elite identity in Ugarit, as expressed through use of and engagement with scripts and languages other than the city's two main ones, was not a single, coherent thing. Like all identities it was context-dependent, but even when talking primarily to themselves, there seems to have been a degree of ambivalence and contradiction in their attitude towards international standards of prestige and elite culture. It seems that Ugarit's elites weren't just walking a careful line between independence and the political exigencies of life as part of the Late Bronze Age political networks, they were also unsure themselves of how much they wanted to see themselves in relation to these norms. Independent-minded, yes, but not so independent-minded that they didn't want the trappings of Egyptian prestige.

### Chapter 10

### Social change in Late Bronze Age Ugarit

Because of the focus on texts largely belonging to the final century or so of the city's existence, and archaeologists' primary interest in uncovering the city of this period, the image we have of Ugarit has tended to be synchronic, frozen at the time of its destruction, a kind of social and archaeological death mask. Inasmuch as scholars have considered Ugarit's physical nature and social practices at all, as opposed to merely its textual corpus and political history, it is this late thirteenth-/early twelfth-century image of the city which exertsthe greatest influence over ourthinking. But cities don't resemble their death masks throughout their whole lives any more than people do, and in this book where we're concerned with society and social practices, such a static and unchanging image is not sufficient. If we're to understand the social contexts within which Ugarit's people practised and interpreted writing, and the impacts these had on society itself, then it's essential that we pay attention to wider social changes. Because of the synchronic focus and often problematic quality of the excavations – especially in their earlier years – much of this discussion will necessarily contain an element of speculation; but this doesn't negate the usefulness of the exercise.

The idea of change is fundamental to the model of practice and agency we laid out in Chapter 2. There we discussed the ideas of Bourdieu and Giddens, whose notions of habitus and structuration position human practice as bound up in a dialectical cycle with social structure: people act in relation to the social context they find themselves in – including institutions, ideas, social norms, ideologies, power relationships and so on. This context doesn't predetermine practice in a mechanistic way, but it does enable, constrain, prevent, favour etc. certain courses of action. People may be conscious of the structures within which they have been socialised or not, they may choose to act with them or against them, to reproduce them or alter them; but practice is always in relation to this inherited social context. And it is in this process of reproducing, altering or discarding social structures through practice that these models become not just theories of social practice but of social change. This was not something Bourdieu was much interested in – indeed, he somewhat rejected the idea that his theory had relevance for understanding social change – but Giddens paid rather more attention to the subject, articulating a model whereby this ongoing background churn of constant, low-level change could be punctuated by occasional episodes of more dramatic transformation.1

This model of social change is useful both because it directly links it to social practice and agency, and because it avoids some of the value-laden and judgemental approaches to social change that have often dominated discussions, especially of the end of the Late Bronze Age. There is a long history of studying changes in societies through notions of 'progress' or 'evolution', often associated with distinct stages such as tribal bands, chiefdoms, states and so on. This is homogenising – forcing all societies into a single, unilinear ladder of development – and inevitably chauvinistic, either implicitly or explicitly projecting modern Western society as an end-point and any other social form as a step on the way or deviation from the path.2 While most associated with nineteenth- and early twentieth-century perspectives, such (neo-) evolutionist approaches have never entirely died out, especially in the United States. Perhaps the clearest demonstration of this is in the continuing fascination with progress's dark twin, 'collapse'. Students of the end of the Bronze Age can hardly have avoided this notion as applied to the upheavals that marked the end of the thirteenth century and beginning of the twelfth, in which Ugarit is a key case study.3 While thinking in terms of 'collapse' doesn't necessarily require subscribing to an evolutionist ladder of social change, it does carry with it a certain value judgement (cf. also parallel terms such as 'crisis') that implies complexity is good, that societies should tend to grow larger, more complex, more literate and so on and that any move to a different form is a step back, a fall or a failure.

The idea of collapse looms large over Ugarit – partly because when you know a place from its death mask, it's hard not to fixate on its death, but also because the dramatic events of the end of the Late Bronze Age bring the city into a much larger and more enigmatic discussion spanning much of the east Mediterranean and Near East, well-equipped with crowd-pleasing elements like wars, invasions, cryptic and portentous textual sources and ill-defined immigrants/refugees who can be blamed for it all. It's not my intention here to get into the wider debate about what happened across the wider region at the beginning of the twelfth century, but inevitably it will colour our discussions in this chapter, since a great deal of what previous work there has been on changes in Ugaritian society has been focused to a large extent on explaining the background to the city's destruction and why it wasn't subsequently reoccupied.

<sup>1</sup> Giddens 1984, 244–256; Joas and Knöbl 2009, 305–307.

<sup>2</sup> The literature on this is vast and I can't explore it in detail here, but good discussions of these approaches can be found in Tainter 1988 and especially Yoffee 2005. See also chapter 1 of Boyes 2013.

<sup>3</sup> See, recently, Cline 2014 and Schwartz and Nichols 2010 or, in a rather less academic vein, Diamond 2005, as well as older examples such as Tainter 1988 and Yoffee and Cowgill 1988. Again, a full literature review and discussion can be found in Boyes 2013.

Based on the evidence available, there are four main areas of social change that can be productively discussed for Ugarit in the Late Bronze Age, although they all interrelate to some degree:


#### **The social impact of natural phenomena**

In his early excavations at Ugarit, Schaeffer believed that all over the site he saw evidence of a major destruction some time before the city's final downfall at the end of the Late Bronze Age. He doesn't describe these signs in detail, but he does mention traces of fire and cracked and collapsed walls. Most ofthese, he says, were subsequently repaired or rebuilt.4 In Schaeffer's view,this destruction marked the boundary between the second and third levels of the Late Bronze Age and could be identified textually as the fire in the royal palace referred to by the king of Tyre in Amarna Letter EA 151: 'Fire destroyed the palace at Ugarit; (rather), it destroyed half of it and so half of it has disappeared.'5

Although the letter mentions only a fire, Schaeffer believed this event was an earthquake, and from the textual reference could be dated to the mid-fourteenth century BC. This kick-started a discussion of the role of earthquakes in shaping the urban fabric of Ugarit and its development during the Late Bronze Age which persists to the present day, and which relates to a number of other east Mediterranean sites. An early suggestion was that this was the same earthquake that destroyed the walls of Jericho, as recounted in Joshua 6.20,6 and although the nature of the theories has been modified considerably since then, an echo of this multi-site, epochal approach to seismic activity can be seen in Amos Nur and Eric Cline's suggestion that an 'earthquake storm' befell sites from Ugarit and Troy to the Aegean in the run-up to the end of the Bronze Age.7

Schaeffer's fourteenth-century earthquake is no longer widely believed in. Later excavators – in particular Olivier Callot – have considered it unlikely that physical evidence of such an event would still be visible in the city of the early twelfth century, given the amount of constant change and refurbishment taking place. They have also pointed out that EA 151 makes no mention of an earthquake causing widespread destruction in the city (or any other city, come to that); merely a fire in half the palace. For this reason, more recent work has proposed that the evidence

<sup>4</sup> Schaeffer 1948, 9.

<sup>5</sup> EA 151. Translation from Moran 1992, 238.

<sup>6</sup> Chandler 1964; North 1973, 128, 130–131; Ambraseys 2006.

<sup>7</sup> Nur and Cline 2000.

of damage and rebuilding Schaeffer observed, and the earthquake responsible for them, should be placed in the mid-thirteenth century; notably for our purposes, around the same time as the official adoption of alphabetic cuneiform.<sup>8</sup> Others have gone further and attempted to tie this earthquake activity into the larger question of Ugarit's destruction at the end of the Bronze Age and its failure to be reoccupied. Nur and Cline's 'earthquake storm' hypothesis falls into this category,9 as does Horst Klengel's suggestion that seismic activity may have caused changes in the structure of Ugarit's harbour, diminishing its potential as a useful port.10

There are two main question that concern us here. The first is whether this horizon of destruction and rebuilding was indeed an earthquake; the second is what its social implications may have been.

Earthquakes are a ready explanation for ancient destructions when there's no obvious evidence to support alternative hypotheses such as violent activity. Particularly in a seismically active region such as the east Mediterranean, they're neither unlikely nor uncommon. They are, however, rather difficult to prove archaeologically and even more challenging to date precisely when written records are lacking. Earthquakes produce a number of distinctive physical effects on man-made structures including cracking, folding, tilting and collapses. Many of these are visible archaeologically, if a specialist takes the time to look for them.11 In several cases, however, including that of Ugarit, the archaeologists diagnosing the earthquake damage are not experts in the effects of seismic activity and don't attempt to demonstrate that the patterns of damage they observe correspond to those produced by earthquakesspecifically,rather than other forms of destruction. Since the damage seen at Ugarit is only described in loose, general terms, we're not in a position to assess the accuracy of this diagnosis. Dating can also be challenging since seismic activity could cause such effects at any point between the construction of the buildings and the present day. Fortunately in the case of Ugarit, we are told that excavators observed plentiful signs of repairs, which does support the ancient dating of the destruction event.

I'm not arguing here that there was no earthquake in Ugarit around 1250 BC – in fact, it seems perfectly plausible – but I do want to sound a note of caution about the bases of argumentation in the decades-long discussion of the city's seismic history. Schaeffer'sintroduction of the idea of an earthquake in the city seems a little arbitrary from the evidence he presents, and was subject to both redating and disconnection from the textually mentioned event he used to supportit – the fire in the palace, which has since been attributed to alternative causes, such as hostile military activity.12 The

<sup>8</sup> Callot 1986, 748; 1994, 204–205; 2011; Singer 1999, 631; Yon 2006, 36.

<sup>9</sup> For a cautious appraisal of Nur and Cline 2000, see Knapp and Manning 2016.

<sup>10</sup> Klengel 1992, 151.

<sup>11</sup> Ambraseys 2006; Rodríguez-Pascua *et al.* 2011. Kreimerman (2017b), however, findsthese archaeological correlates of earthquakes problematic since in his view they are not unique to seismic events, cannot easily be identified solely from published data and often cannot be shown to have directly caused a given destruction horizon.

<sup>12</sup> See, for instance, Rainey 1965, 110 or Singer 1999, 630–631.

*Fig. 10.1. Phases of construction in the South City Trench. Drawn by the author after Callot (1994, fig. 310).*

earthquake hypothesis has thus been recast and resituated to better fit the evidence, but it has still not been conclusively shown that this *was* an earthquake as opposed to some other kind of destruction event – which surely should be a prerequisite before we start building grander theories on this foundation.

What does seem fairly clear is that there was some sort of event in the mid-thirteenth century that caused widespread damage across the city. In many cases this was not irreparable; in others, buildings had to be demolished, which allowed for either rebuilding or clearance of the site for something else. Many of the residential buildings in Ugarit show signs of this. In the 'South City Trench' (Fig. 10.1), we can observe how residents balanced the desire to rebuild in new ways with the fact that this was not a complete discontinuity. As with earlier Late Bronze Age building, the existing street plan was largely preserved, necessitating some unusual architectural shapes, but in places buildings were also extended into public space. In one case, a road in Insula XIII was almost completely built over. A balance also needed to be struck between new uses of space and the residual dictates of cult and ritual practice. For example, in Insula 6 a house was demolished but its tomb remained. When a new building,

House C, was constructed partially over the site of the older residence, a small alleyway was left, affording access to the old tomb, which now was essentially outdoors rather than internal as was usual.

More dramatic changes are apparent with the site's major elite structures. As we mentioned in Chapter 2, the after-effects of fire are likely to have been significant in the Royal Palace. If this was the fourteenth-century fire mentioned in EA 151, then for perhaps nearly 200 years of its existence, parts of the palace's ground floor may have been essentially abandoned and in places the masonry partially melted by the effects of the heat, such that it couldn't be easily demolished or repaired.13 Although the archaeology of the palace has not been fully published, Yon has suggested that further modifications continued throughoutthe thirteenth century, probably including repairs and restructuring following damage in the possible earthquake of ca. 1250.14 There were also changes to the fortified postern gate guarding the palace's western entrance.

Most notable, however, are the temples of Baʿlu and Dagan on the Acropolis. As we discussed in Chapter 7, the temples, originally constructed in the Middle Bronze Age, were destroyed in the mid-thirteenth century, most likely as part of the same possible earthquake horizon.15 The Temple of Baʿlu was rebuilt, but the Temple of Dagan seemingly was not – either because Dagan no longer held the same significance in thirteenth-century Ugarit as he once had, or else because they simply didn't get round to it before the destruction of the city early in the twelfth century.

Whatever happened in the mid-thirteenth century – let's assume for the sake of argument that it was indeed an earthquake – thus had a significant effect on the urban fabric of the city, and this in turn must have had important repercussions for its social life. Was the destruction of the ancient, sacred temples of the Acropolis seen as a judgement of the gods? Of course, we cannot know, but we do know that it affected religious practice. Choices were made about which temples should be prioritised for rebuilding and cult activities were adapted on the site of the temple that was not rebuilt. Elsewhere too, from the palace to the private residences of the South City, rebuilding and repairs likely highlighted tensions between continuity and change, between those who wanted to take advantage of the opportunity to claim a slightly bigger share of what was on offer – whether by extending their houses or building on what had once been public land – and those who found themselves marginalised and edged out, forced to access their onetime family tomb through an alleyway behind someone else's house, which now stood where their own home used to be.

We're probably not overreaching the evidence to suggest that this may have caused social friction, particularly in the early years after the destructions. Rapid change often does, especially when some stand to gain and others to lose out. We should also remember the possibility of loss of life from the earthquake (or whatever it was) itself and the opportunities for work and profit that widespread building work might offer. Disaster capitalism isn't unique to the modern world.

<sup>13</sup> Margueron 1995b, 191–192.

<sup>14</sup> Yon 2006, 36.

<sup>15</sup> Callot 2011.

More speculatively, we might also wonder whether this event may have prompted a greater degree of introspection and reflection among Ugarit's people about how they wanted to present themselves and what they saw their place in the world as being. This is, after all, inherent in choices about what to rebuild and how, whether to continue to practise religious cult on the site of a temple that isn't there anymore and so on. Reflecting on social practice and choices would be even more pressing in a society where disasters such as earthquakes weren't seen as random acts of geology but the actions of divinities and potential judgements on people's actions. This isn't to say that the destructions around 1250 are directly linked to changes in writing practices and the official adoption of alphabetic cuneiform around the same time – apart from anything else, our chronology is nowhere near precise enough to determine which came first – but these events and their aftermath may have fed into wider questions that were being addressed at the time regarding local identity and Ugarit's place in regional 'globalised' networks of politics, trade and culture.

The impact of nature on Ugarit in the thirteenth century is not limited to earthquakes. Anotherfactor, and potentially much more significant overthe long term, has increasingly come to light in recent years, namely climatic changes. Scientific analysis of pollen samples from Tell Tweini/Gibala at the southern border of the Kingdom of Ugarit and from Hala Sultan Tekke in south-eastern Cyprus indicates a significant climatic shift in the later thirteenth century. From then until the ninth century, the region's climate was characterised by instability and drought.16 Similar palynological results have also been obtained for the southern Levant, showing the largest reduction in tree pollen in the whole Bronze Age – specifically in olive trees, a staple subsistence product. Since there was no concomitant increase in the pollen of cultivated crops, researchers concluded that this deforestation was due to climatic factors rather than human action.17 Evidence of climatic instability has even been detected as far west as the Aegean,18 so it seems clear that this phenomenon was regional, if not global, in scope and we should not imagine that Ugarit was immune, despite the absence of scientific climate data from the city itself. On the contrary, this regional data accords well with the long-recognised textual references to famine in Ḫatti in texts found at

<sup>16</sup> Kaniewski *et al.* 2010; Drake 2012; Kaniewski and Van Campo 2017.

<sup>17</sup> Langgut *et al.* 2013. It's notable here that the motif of the burning sun, associated with drought and death, features prominently in Ugaritic mythological texts of the thirteenth century, notably – but not only – in the Baʿlu epic. Drought was, and is, a constant feature of Levantine life and is also referenced in later texts, such as the Hebrew Bible, but as Wikander (2014) has shown, the burning 'drought sun' motif is much less prominent in first-millennium Israelite writing. How much of this (if any) is due to climatic changes as opposed to religious is naturally an open question, and well beyond our current scope. We shouldn't rush to see the Ugaritic literary topos as a direct reflection of climate change when we have so little understanding of Ugaritian religion and literary culture outside this period – especially since, as Wikander points out (255) the 'drought sun' is much less apparent in ritual texts, which might be seen as an attempt to improve the climatic situation. Nevertheless, these literary features are of relevance in understanding Ugaritian responses to, and understandings of, their potentially changing climate.

<sup>18</sup> Drake 2012; Weiberg and Finné 2018.

Ugarit and elsewhere from the very end of the Late Bronze Age.19 There are mentions of famine in Ugarit itself in RS 18.038 (PRU V, 60) and RS 94.2002+.<sup>20</sup>

Such climatic changes are unlikely to have brought about the fragmentation of east Mediterranean societies on their own, but probably representated long-term sources of increased stress, which exacerbated other problems. In particular, resultant shortages of food and water would have contributed to social processes occurring in the Kingdom of Ugarit and other parts of the region during the thirteenth and early twelfth centuries BC. As well as the political tensions with the Hittites, which we see in textual sources such as those mentioned above, climate change is likely to have been implicated in the changes of demography and settlement pattern, which are our next major topic to discuss.

#### **Changing rural ways of life**

Much previous research into changes in Ugaritian society during the thirteenth and twelfth centuries BC has focused on the question of demography, settlement patterns and in particular the balance between sedentary and pastoralist or nomadic ways of life. In the late 70s and 80s, Mario Liverani proposed an influential model whereby Ugarit's decline and fall was attributed to 'desertion' of rural populations to join nomadic pastoralist groups. Combined with the increasing excusal of elite groups such as merchants and *maryannu*-charioteers from tax or service, he argued, this undermined the state's resources, leaving it unable to weather the Sea People crisis when it came along.21 Certainly, there is evidence pointing towards shifts in settlement patterns both in the Kingdom of Ugarit and the wider Levant, but as we'll see, there are a number of problems with Liverani's analysis which call for an alternative interpretation.

Assessing settlement patterns and ways of life in the Kingdom of Ugarit directly is challenging for reasons we've mentioned repeatedly throughout this book. There has been a certain amount of archaeological investigation of larger towns such as Ras Ibn Hani, Ras el-Bassit and Tell Tweini/Gibala, but by and large our understanding of Ugarit's hinterland has been drawn from administrative texts from Ras Shamra itself, which tell us little about rural culture, identities and ways of life. What information we do have points to important differences between Ugarit and secondary centres: Tell Tweini, Ras Ibn Hani and Ras el-Bassit all remained occupied through the end of the Late Bronze Age and into the Iron Age (or, at least, they were swiftly reoccupied), and continuity in village names from Ugaritic records to the present day is indicative of continuity of occupation in smaller settlements. Within these excavated secondary centres, locally produced Early Iron Age ceramics show marked continuity with Late

<sup>19</sup> Summarised by Divon 2008; see also Singer 1999, 715–719.

<sup>20</sup> Climate change as a cause of the Late Bronze Age collapse was first seriously proposed by Carpenter (1966), and the idea of famine in Ḫatti has subsequently been taken up by too many scholars to list, but on its relationship with Ugarit see especially Astour 1965; Singer 1999, 715–719.

<sup>21</sup> Liverani 1987.

Bronze Age forms. Foreign-derived forms, however, point to rather different markets and patterns of consumption from place to place, with Lione du Piêd noting both imported and locally produced Aegean- and Cypriot-style fine wares at Ras Ibn Hani, but not at Ras el-Bassit.22 It seems, then, that there was a degree of variation in practice, material culture preferences and ways of life in different parts of Ugarit's realm, though at present we're not able to go much beyond thisinto specifics. It's also fair to conclude that whatever caused the capital to be destroyed and abandoned in the early twelfth century didn't affect other parts of the kingdom in the same way. Of course, this doesn't necessarily equate to an urban/rural dichotomy or a 'desertion' of rural populations in the vein Liverani suggests.

In fact, our information on sedentism and nomadism in rural Ugarit is virtually non-existent, so, to make headway with this question, it's necessary to examine comparative evidence from elsewhere in Syria and the Levant. In particular, we need to look at the emergence of the peoples known in Assyrian sources as the Aramaeans, who are associated with nomadic ways of life and whose territory in the early first millennium BC probably included the former kingdom of Ugarit.

It used to be thought that the Aramaeans were an intrusive nomadic group that entered Syria around the twelfth century before settling and forming their own states in time for their appearance in Assyrian records in the tenth century. More recent work has preferred to see their origins as internal to the region.23 Current ethnographic perspectives on nomadic and pastoralist groups emphasise that they aren't rigidly demarcated from sedentary societies. On the contrary, research into modern nomadic and pastoralist people stresses that they are integrated into and part of wider society, even if they preserve their own customs and ways of life. Far from nomads and pastoralists being separate from – and in some way inimical to – sedentary 'civilisation', the relationship is both porous and symbiotic. People can and do regularly shift between sedentary and non-sedentary ways of life in response to circumstances, and this need not necessarily entail a shift in their other customs, identities, economic activities or other ways of life. In place of the traditional mutually antagonistic categories of barbarous-but-noble nomads and civilised town-dwellers, we now imagine broad spectra encompassing a near-infinite array of configurations of activities and ways of life.24

Non-sedentary populations had always been part of Near Eastern society. As well as the disparate groups who would later be known as Aramaeans, others are attested, such as the Shasu and especially the ʿapiru, notable for their prominent appearances in the Amarna Letters and the longstanding desire of some scholars to link the name with the Hebrews.25 The ʿapiru are well documented at Ugarit,such asin the Akkadian letter RS 27.238 (PRU 4, 107–8).

<sup>22</sup> du Piêd 2006–2007.

<sup>23</sup> Younger 2007.

<sup>24</sup> Porter 2012; Meijer 2014; Sader 2014; Younger 2014; Honeychurch and Makarewicz 2016.

<sup>25</sup> While this identification cannot be entirely discounted, nor can it be proven, and is probably unlikely (Fleming 2012).

There was indeed a general expansion in non-sedentary lifestyles around the end of the Late Bronze Age. In both the northern and southern Levant, where survey work has been done it tends to support the idea of a move away from large, nucleated coastal centres in favour of smaller and more dispersed settlements in the hinterlands. Many mostly abandoned sites, including Ugarit, preserve evidence of small-scale inhabitation by pastoralist groups at the end of the second millennium. These weren't merely 'squatters',26 but undertook some limited repairs and restoration of structures.27 As the Early Iron Age matured and the initial crisis of the end of the Late Bronze Age faded, there then seems to have been a trend towards more sedentism again, prompting the emergence of new social identities and political structures, such as those of the Aramaean states in the north and arguably Israel in the south. It's not so simple as seeing this as a swift reversal of trends towards nomadism and pastoralism, however, since the same period also offered new opportunities for the expansion of nomadic and pastoralist lifestyles, as the domestication of the dromedary opened up trade routes across the desert towards Mesopotamia.

The topographic and environmental differences between various parts of the Levant and Syria mean that we shouldn't assume that nomadic and pastoral ways of life were the same across the region.28 The hills around Ugarit are rugged and were heavily forested in antiquity, unlike the more open spaces of the Syrian interior or the south-eastern Levant. Nevertheless, the textual references and limited evidence of occupation of the ruins of the capital after its destruction do point towards the existence of such groups in the Kingdom, even if at a smaller scale or with different patterns of subsistence and behaviour than elsewhere.

On the other hand, we shouldn't overstate the move towards non-sedentary and non-urban ways of life in the last phase of the Late Bronze Age. If anything, the archaeological evidence from the capital points to increased population, which the excavators have attributed to an influx ofrural populations into the metropolis. They speculate this might have been motivated by benefiting from royal favours, which were seen to be freely distributed there.29 This manifests in increased subdivision of residential units to accommodate more people and the reduction of public space such as squares and streets to allow older houses to be extended and new ones to be built.

Liverani was probably right to highlight the possibility of Ugarit's rural population becoming uprooted and the knock-on impact this may have had for the urban state which relied on the hinterland's labour. But his model is predicated on an outdated Marxist understanding of an antagonistic two-sector model of Ugaritic society, with an exploitative palatial elite controlling its dependents, and a free rural peasantry.30

<sup>26</sup> *e.g.* Caubet 1992, 123.

<sup>27</sup> Callot 2008.

<sup>28</sup> The relevance of geography and topography to nomadism is shown, for example, by Buccelatti's (1990) discussion of rural lifestyles in the Khabur region.

<sup>29</sup> Yon 1987, 118; 1992, 114–115; 2006, 68; Callot 1994, 205.

<sup>30</sup> Heltzer (1988) notably offered a similar approach around the same time.

His understanding of the relationship between sedentary and non-sedentary population groups is the dichotomous, mutually inimical one we have already discussed and rejected. Even small but important details in his argument cannot be sustained, such as the idea that maritime trade was circumscribed by a Mycenaean thalassocracy and as such unavailable to inject new resources into the system.31 Instead, we should see Ugarit's rural population as adopting a range of strategies for coping with stresses to subsistence and security probably caused by famine and drought resulting from climate change. Some may well have adopted less sedentary – or partially non-sedentary – ways of life, at least for a while, while others sought the security and opportunity of the metropolis. Both responses are likely to have brought with them an array of new social stresses and tensions – from overcrowding and potential hostility between long-time city-dwellers and newcomers to further diminution and unpredictability of rural agriculture, to the potential for new forms of social identity and practice as people adopted new lifestyles.

#### **Political change and the Hittite Empire**

The changes we've discussed so far would have affected all members of Ugarit'ssociety, but perhaps the ordinary population most of all. Elites are likely to have been insulated to some extent from the effects of denser urban occupation or scarce resources, their property most likely the first to be repaired orrebuilt after crises such as earthquakes. We now turn to an issue that would have affected these elites most markedly while probably having a much less pronounced effect on the common people of the kingdom – the impact and social effects of Ugarit's incorporation into the Hittite sphere of influence in the fourteenth century, during the reign of Niqmaddu II, with the subsequent need for the elite to come to terms with what it meant to be both Ugaritian and a vassal within a larger (and increasingly fraught) political network. To do this, it's helpful to briefly re-examine Ugarit's political history in the fourteenth and thirteenth centuries.32

Ostensibly, Ugarit was incorporated into the Hittite realm after appealing for aid in the face of hostile action by neighbouring Syrian polities. More long-term, both the internecine strife and Ugarit's response were consequences of growing Hittite expansion into north Syria, both directly and through political influence at a distance. Niqmaddu's decision to invite the Hittites in is generally seen as bowing to the inevitable, the Ugaritian king having the good sense to jump before he was pushed.33

It seems to have counted for something with its new Hittite overlords that, nominally at least, Ugarit came under Hittite control peacefully and of its own

<sup>31</sup> Even decades later, we still lack solid evidence of particular Mycenaean interest in seafaring, let alone of their ability or desire to control Mediterranean sea-routes to the exclusion of others (Monroe 2009, 294).

<sup>32</sup> What follows is in part adapted from Boyes 2019a.

<sup>33</sup> For more detailed discussion of these events, see Singer 1999; Freu 2006; Altman 2008; Devecchi 2013.

volition. No doubt Ugarit's status as a prosperous trading centre also helped: the Hittites didn't want to jeopardise that juicy revenue by excessive meddling.34 Ugarit was initially granted generous territorial gains from its neighbours,35 and control remained relatively 'light touch' so long as Ugarit's elites behaved. The Hittites did not impose their own system of weights and measures, for example,36 and although political directives were occasionally sent from Ḫattuša (or, more commonly, from the Great King's viceroy at Karkemiš) and Hittite imperial agents such as the 'Sons of the King' (DUMU.LUGAL)37 did intervene in Ugarit from time to time, there was less direct imperial bureaucracy in place at Ugarit than at some other Syrian vassal polities. In the archives of Emar, for example, we read of the 'Overseer of the Land' (UGULA. KALAM.MA), a high-ranking and apparently peripatetic imperial official with fingers in many pies, from administrative and judicial oversight to intelligence gathering and even cult practice, possibly comparable in role to the 'Lord of the Watchtower' (*bēl madgalti*) who acted as a district governor for directly administered territory in Anatolia proper.38 There seems to have been no equivalent post in Ugarit, and these responsibilities remained within the purview of the Ugaritian king. When a higher authority was needed, this was usually the king in Karkemiš rather the Great King at Ḫattuša himself, although some important matters with diplomatic implications, such as the scandal surrounding ʿAmmiṯtamru II's divorce (see Chapter 11 below), did make it all the way to the court in Ḫatti.

But the Hittites didn't afford Ugarit's rulers an entirely free hand. Niqmaddu II's son ʾArḫalba seemingly did something that crossed a line, and he was apparently deposed and replaced by his brother Niqmepaʿ. Many scholarsspeculate that he joined or supported a rebellion by the neighbouring state of Nuhašše. Singer concludes that this traditional reconstruction is plausible but that the evidence for it is little more than circumstantial, relying mainly on the opening of RS 17.349B+, a treaty between Muršili and ʾArḫalba's successor Niqmepaʿ, which talks of placing him on his father's throne and restoring lands to him. The treaty largely restates the terms of Ugarit's incorporation into the Hittite sphere, with a few punishing tweaks – most notably the removal of Ugarit's vassal Siyannu-Ušnatu from its territory and its transfer to the direct suzerainty of Karkemiš, reducing Ugarit's territory by around a third.39 The most constant burdens, however, and – if surviving correspondence is any indication – the ones that chafed the Ugaritian kings the most, were the routine tribute

<sup>34</sup> Altman (2003) sees a fundamental dichotomy in Hittite imperialism between voluntary and conquered vassals, with differing legal treatment resulting from the terms of a state's incorporation into the Hittite sphere of influence, though he notes that the demands imposed on a vassal do not necessarily correlate with these categories.

<sup>35</sup> Singer 1999.

<sup>36</sup> Monroe 2009, 51ff.

<sup>37</sup> Imparati (1975) showed that these were officials rather than literal royal offspring. Indeed, many of them had Syrian names.

<sup>38</sup> Beckman 1992, 47; 1995, 28.

<sup>39</sup> Singer 1999, 637–638.

demands and the requirement that vassals present themselves regularly at the court of their overlord. Ḫatti's treaties with Ugarit stipulated an extremely high level of tribute even by the standards of Syrian vassals, and lend credence to the notion that their relatively light touch was motivated by a desire to reap the benefits of Ugaritian trading operations. Dissatisfaction with Ugarit's fulfilment of these obligations is a recurring feature of correspondence from Karkemiš and Ḫattuša in Ugarit's final years.

The internal instability of the Hittite Empire itself added an additional complication to this political landscape, one that became increasingly pronounced as the thirteenth century gave way to the twelfth. In-fighting, intrigue and factionalism within the Hittite royal family arguably began its critical phase with Muwatalli II's temporary relocation of the capital to Tarḫuntašša early in the thirteenth century and the subsequent establishment there of a breakaway dynasty styling themselves as Great Kings, in flagrant challenge to the kings at Ḫattuša. When Ḫattušili III usurped his nephew (Muwatalli's son) Urḫi-Teššub to seize the throne, it cemented the monarchy's descent into preoccupation with internal concerns. As we've seen, these dynastic squabbles came at a bad time. As well as famine, there was the growing threat of Assyria in the east, which culminated in the Hittite defeat at the battle of Nihriya ca. 1237 – events which were reported to Ugarit by the Assyrians themselves, in what seems to be a barely disguised attempt to invite closer relations between them and Ugarit's rulers.

Recent work has significantly revised our understanding of the end of the Hittite monarchy at Ḫattuša and the fragmentation of the empire. The site's excavators now believe the site was at least partially abandoned before the final destruction came, the higher-status segments of its population having packed their bags and left some time before.40 We don't know where they went. The most plausible destination would be one of the Syrian vassal states, but as yet no ruling dynasty there can be shown to descend from the last king at Ḫattuša, Šuppiluliuma II.41 If the king survived and his line continued, its traces have yet to be identified. It's uncertain whether Ḫattuša's abandonment was completed shortly before the destruction of Ugarit, around the same time, or just afterwards. Unpublished letters T 93-12 and T 96-1 from Tell Sabi-Abyad both seem to allude to the Hittite capital's end, placing it around 1190.42 Our current estimate for the date of the destruction of Ugarit is around 1195–1185, which would allow any sequence, although most scholars seem to tend towards the later end of thisrange. Up to around five years of post-Hittite life istherefore possible at Ugarit, though cannot be assumed with any confidence.43 If the Hittite Empire *had*

<sup>40</sup> Seeher 2001.

<sup>41</sup> Although several Hittite vassal dynasties were cadet branches of the imperial royal house from earlier generations.

<sup>42</sup> Cohen and d'Alfonso 2008, 15 and n. 54.

<sup>43</sup> Wiener (2017, 53) recently suggested that the end of Ḫattuša could have been completed 'by c.1200', but without explaining his basis for this date. Seeher (2001) doesn't offer a precise date but most other scholars writing since his reinterpretation of the city's end as one of gradual abandonment

effectively ceased to exist before the final end of Ugarit, then the likeliest scenario is that political suzerainty over Ugarit and other north Syrian vassals would have shifted fully to Karkemiš. That city already exercised most day-to-day direct control over the Hittite vassals in the region by the late thirteenth century, with the authorities in Ḫattuša mainly getting involved on issues regarding tribute, security or particularly high-level diplomatic concerns. There are no known instances of the last Hittite ruler, Šuppiluliuma II involving himself directly in Syrian affairs, and Bryce thinks it likely that even while he was still on his throne, the viceroy in Karkemiš 'exercised an almost independent role in Syria'.44 If this is correct, then the political transition from nominal Hittite vassalage effectively managed by Karkemišto the fullsuzerainty of Karkemiš itself may not have been particularly hard-felt. Whether there would have been ideological or emotional effects of the downfall of the Hittite Empire for the people of Ugarit – and what they might have been – is hard even to guess at.

The question that concerns us now is what effects all this had on social practice within Ugarit. By far the most important aspect of this is the matter of identities – how elites saw themselves, articulated their status and prestige, and their positions in relation to both local popular culture and globalised political and cultural networks. This is something we'll address in detail in the next chapter.45 Otherwise, there were a few sporadic social issues which crop up – such as a wrangle between Niqmepaʿ and Ḫattušili III over the economic problems caused for Ugarit by an influx of wealthy semi-permanently resident merchantsfrom the Anatolian port of Ura (RS 17.130+),46 or Ugarit being required to contribute militarily to imperial conflictssuch asthe Battle of Qadeš – but what's most striking about Ugarit's incorporation into the Hittite Empire is how little socio-cultural impact it seems to have had. From language to material culture, there is little sign of an 'Anatolianisation' of Ugaritian practice – there is no sign that Hittite or Luwian were written at the site, and we can guess that they were spoken only by visitors from Anatolia or other north Syrian polities.47 This contrasts with sites like Tell Afis to the east, where excavators have observed a standardisation of pottery shapes in the period after its conquest and the incorporation of Anatolian shapes into the local repertoire. Letters written in Hittite found at that site attest to

<sup>45</sup> See also Boyes 2019a.

rather than sudden destruction have favoured a date in the first decade or so of the twelfth century, which is in accord with the Sabi-Abyad letters. Bryce (2005, 328) goes with 'very early twelfth century', but also thinks Ugarit may have destroyed before this point. Beckman (2007, 111) opts for a slightly looser timeframe, proposing that the abandonment of Ḫattuša took place 'gradually over the early decades of the twelfth century'. The date of Ugarit's destruction is placed archaeologically within a vague 'first quarter of the twelfth century' window, but RS 86.2230 allows it to be slightly narrowed down since this letter was sent from Egypt by the notorious vizier Bay, who was at the height of his political powers 1197–1192. Assuming the city was destroyed shortly after this, Yon (1992, 119–120) suggests a date between 1195 and 1185. She favours its end slightly after that of Ḫattuša, but was writing a decade before Seeher's article.

<sup>44</sup> Bryce 2005, 328.

<sup>46</sup> Monroe 2009, 178–179.

<sup>47</sup> On the question of Hittite-speaking communities at Ugarit, see Chapter 9 above.

the presence there of people who could read, speak and, presumably, write Hittite.48 Nor is there much sign of Hittite influence on legal procedures at Ugarit, in contrast to the situation at, for example, Emar, where in the thirteenth century we see Hittite officials acting as judges and as witnesses for legal documents, not just in 'political' cases of interest to the imperial authorities, but even in 'private' local matters such as family law.49

While the apparently weak transformative power of Hittite domination on Ugarit's society and practice can in part be explained through imperial reticence to spook the goose that laid the golden eggs, the lack of almost any kind of emulation or adaptation of Hittite practices can only be interpreted as a deliberate choice by the people of Ugarit and as a conscious rejection of Hittite imperial culture. At a surface level, the social impact of Hittite political control seems to have been rather slight, but there is a world of difference between not doing something unreflexively and not doing that same thing once it has become highlighted, charged and called attention to; in Bourdieu's terms, between 'doxic' and 'orthodox/heterodox' practice.

#### **The end of the Bronze Age and the fall of Ugarit**

The theme that unites all the aspects of social change we've discussed so far in this chapter is their incorporation into wider narratives regarding the upheavals of the early twelfth century and the transition from the Late Bronze to Early Iron Age. Because of its ample textual resources and the fact that some of this written material seems to directly address major preoccupations of many working on this period of instability – namely possible violent invasion by seaborne attackers of mysterious origin – Ugarit has become an important case study and source of evidence for broader attempts to unpick what exactly was going on at this time, and what combination of factors combined to bring about the downfall of established elites and the globalised culture within which they were enmeshed, from the Aegean to the Levant.

That Ugarit was finally destroyed by violent action is now treated as fairly certain. The Late Bronze Age levels end suddenly in a major destruction horizon and arrowheads are reported to have been found across much of the site. The blame for the destruction – like so many others across the Near East – is laid at the feet of the perpetually enigmatic 'Sea Peoples'. Traditionally, these have been seen by Near Eastern scholarship as a confederation of violently destructive refugees, probably coming from the Aegean, and with designs on settling in the Levant and Egypt.50 The past decade has seen certain revisions and improvements to this picture – particularly when it comes to methodologies for determining whether migration has taken place, based

<sup>48</sup> Archi and Venturi 2013.

<sup>49</sup> Pruzsinsky 2007; Van Exel 2010. At Ugarit, the most extensive legal impact of the Hittites is apparently the implementation of a presumption that everyone should do *ilku* (corvée labour) unless otherwise stated (Márquez Rowe 1999, 171–178; 2006, 235ff.; van Soldt 2010, 98).

<sup>50</sup> I summarise and argue against this model in Boyes 2013, esp. 91–111.

on more than simply the presence of foreign-looking pottery.51 Mass migration is surely the majority opinion (although many scholars see this as occurring alongside a plethora of other factors constituting a 'systems collapse' or, as Cline (2014, chapter 5) puts it, a 'perfect storm'), but another strand of research has argued for a more economic vision of the Sea People phenomenon. This is particularly associated with Susan Sherratt, who published influential work in the late 90s and early 2000s which positioned the Sea Peoples not as marauding migrants but practitioners of a kind of 'aggressively open economy which was highly subversive to the centrally controlled, formal élite exchange systems which constituted an important part of the political basis of established powers'52 and based in places like Cyprus and the Levant. This kind of freebooting, unregulated drive for profit and commercial advantage – and the political position it could bring – has, throughout history, often been only thinly and impermanently separated from out-and-out piracy, racketeering and other forms of violent rapaciousness. One person's merchant or legitimate businessman is another's pirate, Viking or 'legitimate businessman'. Sherratt suggests that the violent destructions attributed to the Sea Peoples are likely to be symptoms of an environment in which armed aggression, privateering and coastal raiding were probably endemic.53

Sherratt's suggestion has been criticised in part on the grounds that it proposes an overly sharp distinction between palatial and privatised trade for the Late Bronze Age, and an overly antagonistic relationship between the two,54 but we needn't accept a purely economic paradigm shift model as the cause for the upheavals at the end of the Late Bronze Age for this notion of the Sea Peoples as deracinated, economically active merchants, traders and raiders of the east Mediterranean to be useful; at any event, it seems to me greatly preferable to the 'mass migrants of unknown origins' model. Indeed, a similar (though distinct) idea of the Sea Peoples as pirates has been developed recently by Louise Hitchcock and Aren Maeir (2014; 2016; 2018).

This whole messy debate is of great importance for how we understand the wider processes of the end of the Late Bronze Age, but in itself, the nature of the Sea Peoples isn't our concern here.55 What matters in particular is the question of time-span: was Ugarit's destruction a sudden event that came essentially out of the blue, while things were otherwise business as normal,56 or was it the culmination of a longer process of economic change, political instability and fragmentation of traditional Bronze Age ideologies of power and prestige? The answer seems to be a bit of both.

<sup>51</sup> A major contribution to this improvement was Yasur-Landau 2010, and see also Fischer and Bürge 2017 for a recent cross-section of approaches to the topic. While these remain broadly in favour of a modified version of the traditional mass-migration hypothesis, some scholars continue to sound valuable notes of caution, such as Millek 2017 on the difficulty of attributing Levantine destruction layers of this period to violent invasion.

<sup>52</sup> Sherratt 1998, 301 and see also Artzy 1998; Sherratt 2003.

<sup>53</sup> Sherratt 1998, 305–306.

<sup>54</sup> Bell 2006, 112; Cline 2014, 153; Yasur-Landau 2017.

<sup>55</sup> I outline my views on this in more detail in Boyes 2013, esp. 91–111.

<sup>56</sup> In Astour's view (1965, 254), '[n]o decay whatsoever, either material or spiritual, can be observed at Ugarit on the eve of its destruction. The city fell at the height of its vitality, suddenly, the result of a terrible catastrophe.'

From textual sources, it certainly appears that Ugarit's trade routes remained open till the end; tablets from the House of ʾUrtenu, for example, show that international diplomacy and commerce continued during the city's final years. On the other hand, other tablets indicate a degree of anxiety about raiders and a certain amount of wrangling between Ugarit and its neighbours about whose responsibility the insecurity was and whether they could help each other fight them off.57 An exchange between the king of Ugarit and Ešuwara, the governor of Alašiya, is worth citing here (RS 20.18/ *Ugaritica* 5.22). The Cypriot denies that his people are responsible for seaborne raids on Ugarit's territory, instead claiming Ugaritians were themselves to blame:

As for the matter concerning those enemies: (it was) the people from your country (and) your own ships (who) did this! And (it was) the people from your country (who) committed these transgression(s)… I am writing to inform you and protect you. Be aware!58

The picture is thus one of confusion, anxiety and a certain breakdown of international relations, which is consistent with the idea of the Late Bronze Age globalised system unravelling towards the end. Add this to likely long-term factors such as climate change and resulting famine and the disintegration of the Hittite Empire and we can hardly think in terms of 'business as usual' at Ugarit during its last few decades, even if life inevitably went on and the final destruction, when it did eventually come, is not likely to have been predictable in its exact timing or form.

The most plausible model for the end of Ugarit is thus a proximate cause that was relatively sudden – a violent attack – but a much longer period of instability, insecurity and fraying of the international diplomatic and economic networks that underpinned traditional Bronze Age authority, most likely coupled with climate change and famine. What effect would these have had on the social practices of Ugarit? Very probably, the effects would have been felt differently among differentsectors of the population. Matters of trade, diplomacy and elite prestige would have most directly affected the political elites and Ugarit's successful international merchants (and these groups of course overlapped to a very significant degree); effects on the wider population are likely to have been more indirect – changes in the availability of foreign goods in the markets, a sense of lessened security from attack, changes in how their rulers sought to legitimise and articulate their prestige (see Chapter 11). Famine, on the other hand, probably struck the poorest hardest, and would have had a much greater impact in the rural hinterland than the other factors we've discussed.

#### **Social change and the practice of writing**

In this chapter we've been able to highlight a number of possible social changes affecting the Kingdom of Ugarit, especially during its last 50–75 years. This is exactly the period in which alphabetic cuneiform was adopted and utilised in the city. How, then, did the changes we've identified interact with writing practices?

<sup>57</sup> For a detailed history of this period and discussion of the sources, see Singer 1999, 719ff.

<sup>58</sup> Translation from Cline 2014, 151, after Bryce 2005, 334.

Naturally, there is very little in the way of evidence to support direct causes and effects, but we can reasonably hypothesise some effects. The first concerns access to, and visibility of, writing, which we discussed in Chapters 7 and 8. To date, all our evidence for writing in the Kingdom of Ugarit comes from the capital (or nearby, closely-related sites such as Minet el-Beida and Ras Ibn Hani). We can plausibly argue that writing was not entirely absent from the hinterland, but we have only the most rudimentary idea of what form it took, how permanent or transitory a presence it was or what routes there were for rural people to become literate and to access the educated, literate professions centred in the metropolis (as, for example, ʾIlimilku seems to have done). Even within the city of Ugarit itself, writing is unlikely to have been widely visible, even though collections of tablets were quite widely distributed across the site – there wasn't a culture of monumental or public writing, and tablet collections seem to have been stored securely in the less-accessible upper storeys of private residences. The only places where inscribed objects seem to have been displayed in anything approaching a 'public' way were the main sanctuaries of the Acropolis, and even then the chances are that these votive objects would have been mostly out of sight in ritually circumscribed sacred spaces – it's very unlikely that just anyone could wander in and look at inscribed votive objects at their leisure.

The social changes of the later thirteenth century wouldn't have overturned this situation, but they may well have nuanced it in some ways. Increased population density in the capital is likely to have brought more people into closer proximity to writing – firstly through an influx of people from the hinterland to city, where routes into literate education are likely to have been (slightly) more numerous, and secondly through the more densely-packed living conditions, which would generally have raised the probability of people brushing up against writing practices or inscribed objects – especially if writing was being done in courtyards or on rooftop terraces.

The earthquake may also have brought writing out into the open, at least temporarily. The destruction of upper storeys would have scattered tablets stored there both over lower floors but also into the streets, where they would have been visible to passers-by, those clearing the damage and builders repairing or rebuilding. They might also have ended up being incorporated into the fabric or fill of other structures, as seems perhaps to have occurred with a tablet from a lexical text found on the threshold of House A of the City Centre.59 The situation is even clearer in the Acropolis temples, where, as we have discussed, inscribed votive objects continued to be deposited on the site for the Temple of Dagan even after its destruction. The two stelae inscribed in alphabetic cuneiform, while small, are likely to have been considerably more visible to passers-by than earlier inscribed votives would have been. Whether this has anything to do with the unique choice of the vernacular script and language for something approaching monumental inscription is intriguing to speculate on, but a conclusive answer is not optainable from the present evidence.

<sup>59</sup> Yon *et al.* 1987, 33.

These longer-term factors, and in particular climatic changes, may also have affected technical aspects relating to the production of writing materials, though this is impossible to demonstrate conclusively or to quantify given how little we know about the *chaîne opératoire* of various forms of writing at Ugarit. For instance, increased temperatures and drought may have made suitable clay harder (literally!) to gather and scarcity of water may have reduced the amount available for writing. The availability of beeswax for writing boards might also have been affected by hotter, drier climates, since these can affect the numbers and behaviour of bees, both directly and through effects on habitats and flowering plants. Higher temperatures would also affect practical considerations such as working time for both clay and wax (though in different ways), or the viability of working in exposed areas such as rooftop terraces during particular seasons or at certain times of day. In general, it seems reasonable to suggest that a hotter, drier climate would have been inconvenient for writing on clay, as it probably was in many aspects of human practice, though evidently not to the degree that it caused people to stop doing it. Most likely it was just another small way in which life was getting harder during the later thirteenth and early twelfth centuries.

Much more important for affecting writing practices in Ugarit are likely to have been the geopolitical changes we have discussed – first Ugarit's incorporation into the Hittite Empire, and then the disintegration of that empire along with the gradual fraying of numerous other elements of the fabric of Late Bronze Age globalised culture. These changes didn't affect Ugarit's writing practices directly – there is no evidence for anything as crude as Hittite-mandated, or at least -inspired changes in script or writing system, such as we may see in Emar with its 'Hittite' and 'Syro-Hittite' scripts. But they did have profound effects on how the people of Ugarit – and especially the elites – saw themselves and their place in wider east Mediterranean/Near Eastern society, and how they used writing and other forms of material culture practice to articulate and negotiate this. These questions of writing practice and identity will be the focus of the next chapter.

### Chapter 11

### Writing practices and elite identity: imperialism, resistance and vernacularisation

The last two chapters have shown, in differing ways, the important relationship between writing practices and identity. This chapter aims to complete our exploration of this theme by considering probably the most important area in which writing was used to shape identities at Ugarit – or at least the most visible in the evidence that survives to us: the negotiation of elite identities. It very much appears that writing practices – most notably the adoption of alphabetic cuneiform and the vernacular Ugaritic language – were tightly bound up with Ugarit's elites' attemptsto define their position relative to the globalised networks of elite politics, culture and trade and their responses to the unravelling of these networks towards the end of the thirteenth century. To understand this, it's first necessary to discuss Ugaritian elite responses to Hittite overlordship and their engagement with international prestige culture.

As we discussed in the last chapter, Ugarit's relationship with Hittite imperial overlordship was a carefully judged tightrope of resistance and compliance in which local elites had to balance their desires to exercise power and assert their own political and social identities with the necessity of not overstepping conforming to, and indeed utilising, accepted international norms.1 To an extent, this was surely true of any Syrian vassal, many of which had initially resisted Hittite overlordship in a way Ugarit had not.2 But Ugarit's resistance seems to have gone a little further; its incorporation of Anatolian culture somewhat less enthusiastic.

There was little adoption in Ugarit of Hittite-style practice or material culture.3 Pottery styles did not change and the Hittite and Luwian languages were not used. What Hittite influence can be discerned in Ugarit's material culture can't be chronologically tied to the period of imperial domination.4 At elite levels, material culture

<sup>1</sup> Boyes 2019a.

<sup>2</sup> Mukiš, Niya, Nuhašše and even Karkemiš had all had to be forcibly incorporated into the Hittite sphere of influence, and yet they or their territories became home to 'Neo-Hittite' states in the Iron Age.

<sup>3</sup> Boyes 2019a.

<sup>4</sup> Glatz 2013.

reflects a guarded, partial engagement with Anatolian prestige culture. For example, it has been argued that the development of the signet ring in north Syria stemmed from a desire to reconcile the Levantine and Mesopotamian tradition of cylinder-seals with the Hittite preference for stamp-seals, since the application of the rings – half-rolled, half-pressed – can be seen as a compromise between the two. The rings were widely adopted as personal seals by Ugarit's monarchs and other high officials in the later fourteenth and thirteenth centuries, but traditional cylinder-seals also continued to be used, often by the same individuals, perhaps when acting in a different capacity. The continued usage of the so-called dynastic seal of King Yaqaru by Ugarit's Late Bronze Age kings is a prime example of this.5 While the form of Ugaritian signet rings might be seen as an adaptation or hybridisation due to Anatolian influence, and Hittite features can sometimes be discerned in their iconography, the designs are often distinctly Syrian in character, including incorporating alphabetic cuneiform inscriptions, as in the seals of ʿAmmiṯtamru II and Niqmaddu III. This contrasts with seals from, for example, Amurru, where hieroglyphic Luwian was used in royal glyptic, or the widespread fondness in northern Syria for seal-rings depicting the Hittite king or sun-god.6 It's possible that outside the Ugaritian capital, attitudes differed slightly. The only inscribed items from Gibala/Tell Tweini are a bronze seal-ring and a biconvex seal, both inscribed in hieroglyphic Luwian.7

Even limited engagement with Anatolian material culture such as this was unusual, however. Instead, Ugarit's elites looked much more towards Egypt as a source of exotica and a model for the visual language of prestige. We've already mentioned several examples of Egyptian high-status objects in this book, especially in the religious sphere. Perhaps most notable of these is the statue of Merneptah, which the Ugaritian king wanted to erect in the Temple of Baʿlu in the late thirteenth century – the ingratiating tone of this gesture stands in stark contrast to the often frosty and uncooperative tenor of diplomatic contacts with the Hittite world at this time (see below). The 'Merneptah sword' discussed in Chapter 9 also indicates the valorisation of this pharaoh in late thirteenth-century Ugarit, and the desire to associate his name even with non-Egyptian objects. Even more telling is the so-called Niqmaddu Marriage Vase. As we saw in Chapter 9, this seems to be a locally made object in highly Egyptianising style, which also seems designed to imply that Niqmaddu married an Egyptian princess, something that appears unlikely from what we know about Egyptian royal marriage policy.8

An important object for understanding the strength of Ugarit's elite fascination with Egypt is on first appearance one of the less remarkable – an alabaster vase

<sup>5</sup> Herbordt 2005, 43–44; Kabatiarova 2006, 39ff.

<sup>6</sup> There is some debate about whether the figure in these rings is the Hittite king dressed as the sun-god (as in his title as overlord, My Sun), or merely the sun itself. See Güterbock 1993; Beckman 2002; Venturi 2010.

<sup>7</sup> Lebrun and Tavernier 2012; Bretschneider *et al.* 2019, 6; Bretschneider and Jans 2019b, 158–159.

<sup>8</sup> Feldman 2002a.

fragment with the cartouche of Horemheb.9 This object has apparently never been photographed or drawn, and has perhaps not even been seen since its cursory mention in preliminary publications. We lack even dimensions for it. Alabaster vessels were common Egyptian diplomatic gifts, and those with royal cartouches are particularly associated with royal contexts in sites such as Ugarit and Byblos. What makes this one especially interesting is that Horemheb reigned at the turn of the thirteenth century at a time when Ugarit's incorporation into the Hittite Empire was still relatively recent and when Hittite-Egyptian enmity was strong. Even then, it seems, Ugarit's royal family was accepting and displaying diplomatic gifts from the Egyptian court. This apparent insubordination may be connected to ʾArḫalba's midemeanour and removal – as well as anti-Hittite, the Nuhašše rebellion was pro-Egypt.

Alongside these diplomatic and material links with Egypt, we can see increasing signs of Ugarit's rulers chafing against Hittite overlordship as the thirteenth century progressed. Beginning especially with the reign of ʿAmmiṯtamru II, contacts between Ugarit and its overlords were increasingly characterised by instances of the former pushing its luck in the pursuit of its own interests in ways that aroused displeasure and reprimands at Karkemiš and Ḫattuša, but seemingly little in the way of actual repercussions.10 Ugarit's rulers were apparently adept at walking the line between autonomy and outright rebellion, asserting a distinct local identity and independent goals without ever becoming such a problem or embarrassment that the Hittites were compelled to look away from their internal political intrigues or wider regional conflicts to put them back in their box.

The first major example of this, if we exclude ʾArḫalba's possible failed rebellion early on, is the lengthy diplomatic scandal surrounding ʿAmmiṯtamru II's divorce. The Ugaritian king had married the daughter of King Bentešina of Amurru, a princess (whose name is not preserved) who was also niece of the Hittite great king. This marriage into the Hittite royal family probably represented something of an honour for a vassal ruler. Unfortunately, the princess did not please ʿAmmiṯtamru. After claiming that she 'sought trouble' for him, but without specifying the nature of this misdemeanour (most scholars assume either political plotting or adultery), he divorced her and exiled her back to Amurru. His anger not sated, he then thought better of this and demanded that she be sent back to Ugarit so he could punish her further. The ensuing legal and diplomatic wrangling takes up at least fifteen documents from Ugarit and allows us to trace the matter as it was passed up the imperial hierarchy until it reached the highest level – the Great King's court at Ḫattuša itself. The affair outlasted the reign of Bentešina and continued to be negotiated with his son – the woman's brother –Šaušgamuwa. In the end, ʿAmmiṯtamru paid a huge sum of blood money – 1400 shekels of gold – to the royal family of Amurru to secure his former wife's return and his right to do whatever he wanted to her without fear of

<sup>9</sup> Schaeffer 1954, 41; Caubet 1991, 233. Sparks (2003, 36) refers to it as a lid, but earlier publications merely say a vase or a fragment.

<sup>10</sup> Bordreuil and Pardee (2009, 13–15) term it 'passive resistance'.

reprisal.11 This torrid business may also have cost him his crown prince: at one point ʿAmmiṯtamru's son Utri-Šarruma was given the option of siding with his father and retaining his right to the throne, or of siding with his mother and forfeiting his chance to be king. It's not clear what choice he made. The next king of Ugarit was ʾIbiranu, but we cannot exclude the possibility that this was a throne-name used by the same person. What is abundantly clear from all this is that ʿAmmiṯtamru was far from a meek vassal; on the contrary, he was a strong-willed and assertive ruler whose rage and hatred still blaze brightly from the ancient texts and who was prepared to go to extreme lengths to satisfy his emotional and political needs, even if that meant potentially offending his imperial overlord and opening up diplomatic rifts with his neighbours in Amurru and the viceregal authorities in Karkemiš.

Perhaps surprisingly, this wasn't the last marital connection between the royal houses of Ugarit and Ḫattuša. It's less clear what happened with the second, but that doesn't seem to have gone particularly well either. RS 17.226 and RS 17.355 date from the reign of ʿAmmurapi and deal with the disposition of property belonging to a former queen who was leaving Ugarit. The complicating factor is that she was also the daughter of the Hittite ruler. This has often been interpreted as the aftermath of a second royal divorce during ʿAmmurapi's reign, but Singer has cautioned that we're not explicitly told either that there was a divorce or that the woman was ʿAmmurapi's wife. It is also possible, he suggests, that she was the widow of the former king, Niqmaddu III.12 Even so, for her to leave Ugarit after his death rather than continuing to live as queen with all the rank and property she had formerly enjoyed is unusual and points to more going on in this affair than we are currently able to reconstruct.

A number of other letters from Karkemiš and Ḫattuša record a litany of further transgressions, slights and examples of Ugarit being uncooperative. RS 17.247 comes from a Hittite 'son of the king' Piḫawalwi (probably a rank rather than a literal indication of familialrelationship to the monarch). Bluntly addressing ʾIbiranu without his royal title or formal courtesies, he angrily accuses him of not paying tribute and of failing to make the trip to the imperial court to pay homage to the Great King at any point since his accession:13

Thus speaks Piḫawalwi, Son of the King. To Ibiranu, my son, say: Here, for My Sun, all goes well.

Since you assumed the kingship of Ugarit, why have you never come to My Sun? And why have you not sent messengers regularly? Now, in this matter My Sun is very irritated. So send messengers to My Sun with all haste and have them bring here the King's gifts and my gifts.14

<sup>11</sup> Singer 1999, 680–681 and now Thomas 2019, who carefully and clearly delineates the ways in which ʿAmmiṯtamru's former wife was systematically extricated from the networks of relationships which afforded her status and protection, and ultimately was redefined as property to be granted from one king to another in the same manner as land or other royal grants.

<sup>12</sup> Singer 1999, 702–703.

<sup>13</sup> Singer 1999, 684.

<sup>14</sup> Translation based in part on the French of Nougayrol 1956, 191.

In KTU 2.36+, the Hittite queen Puduḫepa writes to Niqmaddu III, again complaining about a lack of tribute and possibly a failure to appear before the queen on a visit to Ḫatti, while in RS 34.136 it is this time the king of Karkemiš complaining about inadequate tribute for Hittite officials.15 Numerous other letters feature similar complaints, but reluctance to provide tribute was not the limit of Ugarit's avoidance of its duties as a vassal. As the Late Bronze Age reached its final phase and the letters back and forth increasingly tell of blossoming crises across the region, it is clear that Ugarit's rulers were putting their own interests ahead of those of their embattled overlords. In KTU 2.39 and RS 13.007B ʿAmmurapi was reprimanded by the kings of Ḫatti and Karkemiš respectively for the familiar failure to appear at court, but also for not sending grain supplies to alleviate the Hittite famine. RS 20.212 also complains about Ugarit's failure to ship grain to Ḫatti and details the dispatch of two agents to supervise a future shipment, noting that it is 'a matter of life and death'. In RS 34.143, the king of Karkemiš accuses the king of Ugarit of lying about the location of his army and the condition of his chariotry as a pretext for not sending military assistance as ordered. The soldiers he does send, allegedly, are worthless, with the best warriors being kept back in Ugarit itself. In another letter from Ugarit's final years, RS 20.238, the king of Ugarit claims his army is in Ḫatti and his navy in Lukka, but Singer is doubtful as to how far this is to be believed: 'Even if the king of Ugarit reluctantly fulfilled some of his military obligations, the constant reprimands from Karkemiš and from Ḫatti leave little doubt that he kept the best part of his army within the borders of his kingdom, as indeed any sensible ruler would do in a similar situation.'16

Ugarit's elite resistance of Hittite control and elite culture is thrown into even sharper relief by exploring developments in the region during the Early Iron Age. In particular, we're concerned with the polities generally – though somewhat misleadingly – known as the Neo- or Syro-Hittite states. These include a number of Ugarit's most prominent neighbours, including Alalaḫ and Karkemiš – cities that had been under Hittite overlordship in the Late Bronze Age, and in the Early Iron Age continued to exhibit markedly Anatolian-derived cultural elements such as the use of the Luwian language and hieroglyphic writing system, Hittite titulary and iconographical elements. In Assyrian and biblical texts of the first millennium, they're simply referred to as Hittite. It was for a while believed that large-scale migrations from Anatolia must have been responsible for these cultural features, but this now appears unlikely. Archaeology has produced no clear evidence of mass Anatolian immigration and, as Bryce notes, Syria would hardly have been an attractive destination for refugees given that it was suffering from much the same political instability and food shortage as Anatolia itself, as well as being considerably closer to the threat of Assyrian expansion.17

<sup>15</sup> Singer 1999, 693–694.

<sup>16</sup> Singer 1999, 707–708, 716, 723–724.

<sup>17</sup> Bryce 2012.

The traditional label as Hittite successorstatesrisks overshadowing the significant material and cultural influences of non-Anatolian societies on these polities. Mesopotamian influence unsurprisingly continued to be strong, and Aegeaninfluenced material culture was popular at many centres, to the extent that several scholars believe some of them were settled by immigrant Aegean 'Sea Peoples', an idea that has only gained momentum following Hawkins' reading of the Luwian name of one of these states as Palastin.18 The 'Aegeanness' of this material culture is itself not clear-cut. The LH IIIC pottery found at Tell Taʿyinat and several other sites in the Orontes basin is locally produced rather than imported; that at Tell Afis is likely to have originated in the Amuq or at the furthest in Cyprus.19 Indeed, what is generally termed LH IIIC in the Levant finds its direct inspiration in Cyprus rather than the Aegean, although certainly the Cypriot White Wheelmade III is Aegeanising.20

If there was migration into these north Syrian polities, it was far more probably small-scale and gradual than sudden and *en masse*, whether it originated in Anatolia, the Aegean, Cyprus or anywhere else. What's clear is that these societies blended cultural elements from diverse sources to create a mélange distinctive to each of them; neither straightforwardly 'Hittite' nor 'Syrian' nor 'Aegean' nor even homogeneous between themselves, but each an individual and complex realisation of cultural hybridisation and social identities. These responded to the Late Bronze Age prestige of the Hittite Empire, as they did, to lesser degrees, to the rising star of Assyria or the commercial expansion of Cyprus, but they developed these in new ways, merging and transforming them to create a set of distinctly new, Iron Age sociocultural forms.

The problem for understanding social changes in these polities in the Late Bronze Age and immediately thereafter is that much of the archaeology remains extremely patchy and texts are not available for a period of several centuries. At Karkemiš, for example – ostensibly the Neo-Hittite centre *par excellence* and likely one of the key centres for the development and dissemination of the kinds of hybridising culture we're interested in21 – only very little, fragmentary, material has been found that can be reliably dated to the period of Hittite domination.22 Since Šuppiluliuma I mentions leaving the temples of the acropolis intact when he captured the city, it has been speculated that this means the city's main Late Bronze Age administrative centre was not built on top of the tell;23 but if so, it has yet to be found. Many of the changes that were likely taking place in social identity, language, writing, ideology, demographics and so forth consequently remain opaque to us.24 What evidence we have, at Karkemiš

<sup>22</sup> Marchetti 2012; 2016.

<sup>18</sup> See, for example, Singer 2017, with further references; Hawkins 2009.

<sup>19</sup> Venturi 2010.

<sup>20</sup> Sherratt 2003.

<sup>21</sup> Mora 2014, 94.

<sup>23</sup> *e.g.* Aro 2013, 250–251.

<sup>24</sup> Recent work by the site's present Turco-Italian excavators has produced some evidence that Karkemiš shared elements of Late Bronze Age elite display – namely, orthostats bearing lion iconography – with other sites in the region, such as Aleppo, Alalaḫ and Emar (Marchetti 2016), but as yet this is a rather

and elsewhere, is clearest concerning the elites. Nevertheless, the broad strokes at least are clear: to varying degrees and in diverse ways, several north Syrian polities embraced and appropriated aspects of Hittite culture. Especially in fields such as kingship and elite display, traditional models of prestige and legitimacy continued to be followed, emphasising the importance of descent and international recognition.25 These were often expressed in ways borrowed from Hittite imperialism, in service of local dynasties and identities. Thus we have rulers styling themselves as 'Great Kings' or 'Country Lords', and utilising the iconography of Hittite power, including the monumental Luwian hieroglyphic writing system.

Ugarit presents a very different picture. Since it was destroyed at the end of the Late Bronze Age and not subsequently rebuilt, it of course didn't have the opportunity either to develop into a state of this kind or to interact with the polities that did. It's unclear whether any of the former Kingdom fell within the territory of one of the 'Syro-Hittite' states; if so, the likeliest candidate is Palastin/Pattin with its centre at Taʿyinat just across the mountains to the north of Ugarit, which seems to have encompassed much of the territory of Ugarit's long-time rivals Mukiš, Niya, Nuhašše. Although the mountains present something of a boundary, geographically the Kingdom of Ugarit is not far removed from the developments we have been discussing. However, there is good evidence that its people's attitudes towards the models of power and prestige represented by the Hittite Empire were developing in a different direction in the decades leading up to its destruction. We've discussed at length how Ugarit's elites resisted rather than appropriated elements of Hittite rule, but differences are apparent even in the use of Aegean-derived objects, which were popular at Ugarit. Unlike in the 'Syro-Hittite' polities, Agean pottery at Ugarit was mainly imported, and only small amounts of LH IIIC have been found in the city's final phase. The small numbers might be explained as an expected consequence of its destruction except for the fact that Tell Tweini, a town at the Kingdom's southern border, which appears to have continued without interruption into the Early Iron Age, exhibits a similar pattern.

#### **Resistance, localism and vernacularisation**

Ugarit's relationship with its Hittite overlords has two main phases. The first, encompassing the reigns of Niqmaddu IIto Niqmepaʿ in the mid-fourteenth to mid-thirteenth

scant basis upon which to build any hypotheses. There are a few small hints that despite the prevalence of Luwian, a degree of Akkadian literacy remained in the Syro-Hittite polities in the Early Iron Age: at Hama, an Akkadian letter from Anah was found, and at Karkemiš the eighth-century regent Yariri claims proficiency in numerous languages, including Akkadian (Aro 2013, 260). This suggests both that Akkadian was not entirely dropped for diplomatic purposes with the fall of the Late Bronze Age regional systems and that by the eighth century it was sufficiently uncommon to be worth boasting about. However, the balance between Luwian and other languages and writing systems, and how exactly this changed with time, presently remain a mystery to us.

<sup>25</sup> Ponchia 2011.

centuries, is characterised by Hittite strength. Relatively few examples of Ugaritian resistance are documented and potential instances,such asʾArḫalba's possible rebellion, overplayed their hand and were met with firm reprisals. The second phase, beginning with the mid-thirteenth-century reign of ʿAmmiṯtamru II and running until the end of the city not long into the twelfth century, sees Ugarit much more confident and sure-footed, its elites resisting their imperial Hittite entanglement in numerous relatively minor ways that allowed them to achieve their own objectives – whether those be vengeance on an ex-wife, not paying tribute or withholding supplies or military aid for the benefit of Ugarit itself – without overstepping the mark and eliciting a serious crackdown by the authoritiesin Karkemiš orḪattuša. Chronologically, this second period coincides perfectly with the adoption and use of the alphabetic cuneiform script and the vernacular Ugaritic language in the city's internal writing practices. This is no coincidence; the 'vernacularisation' of Ugarit's writing practices is fundamentally bound up in the same elite concerns about identity, culture and global political networks.

If Ugarit is indeed an example of 'top-down' vernacularisation, as we discussed in Chapter 4, then parallels can be sought in post-colonial states such as Rwanda or South Africa,26 and this invites us to consider the process in terms of Ugarit's relationship with – and resistance to – its imperial overlords and the globalised networks they represent. In pursuing such comparisons, due caution is obviously necessary since the natures of ancient and modern imperialisms are not the same, and modern capitalist globalisation – while having many things in common with the globalisation we see in the Late Bronze Age east Mediterranean and Near East (see Chapter 2) – is nevertheless quite distinct in some very important ways. Thus, although the Hittite Empire dominated Ugarit and served as its imperial overlord, there was, as we've seen, apparently little attempt on either side to 'Anatolianise' Ugarit, and certainly no indication of any compulsion or even encouragement for the people of Ugarit to use the Hittite language. The 'hegemonic' language and associated writing practices in Ugarit were not Hittite at all, but Akkadian; in fact, Hittite writing practices were just as dominated by these as Ugarit's were.

In this sense, Ugarit was engaged against two different fronts of hegemony – politically, with the Hittite Empire, and culturally, with dominant 'global culture' of the Late Bronze Age, which had its origins in Mesopotamia but by this point was to some degree deterritorialised, and a characteristic of the network itself (cf. Chapters 2 and 5). These two fronts were obviously and inextricably interconnected: to engage with the Hittite Empire, Ugarit was also required to utilise Mesopotamian-derived writing practices such as the use of logosyllabic cuneiform and the Akkadian language – to borrow the terms used by Grewal (2008), they had network power that had long-since passed the threshold of being essentially mandatory for those wishing to participate

<sup>26</sup> Kamwangamalu 2013.

in the international system.27 As in any example of resistance to the power of globalisation, the elites of Ugarit inevitably found that in negotiating their position in relation to one, either the political or the cultural, they inevitably also had to adopt a stance regarding the other.

Indeed, there are some signs that Ugaritian elite resistance towards their position at the end of the thirteenth century wasn't just focused on the Hittites specifically, but on the whole 'great game' of Late Bronze Age international politics. In a recent study of Ugarit's most important literary work, the Baʿlu epic, Aaron Tugendhaft has argued that, far from asserting that the status quo reflects an eternal and solid divinely ordained order of the kind seen in the roughly contemporary Babylonian creation-myth *Enu*̄*ma Eliš*, the poem is centred on the contingency and arbitrariness of rule, the emptiness and contradictions of the kinship metaphors traditionally used to justify and organise international relations.28 It is a poem written in the language of international diplomacy; no coincidence, Tugendhaft asserts, that its named writer ʾIlimilku was a senior courtier who seems to have been heavily involved in diplomatic relations, perhaps even serving as an envoy to Ḫattuša at one point.<sup>29</sup>

If Tugendhaft isright, then the epic of Baʿlu is a window on to the political mindset of Ugarit's elite in the final quarter of the thirteenth century. In his view theirs was a pragmatic, somewhat cynical philosophy in which power is merely one's ability to

<sup>27</sup> Grewal's model is more closely concerned with human agency than this simple summary may imply. In his view, 'standards' can take two forms, *mediating standards* – which allow people to interact and co-ordinate their practice within networks by their inherent qualities, language being a good example – and *membership standards* – where the achievement or performance of some agreed-upon marker is treated as necessary to participate in interactions. The latter include things such as regulatory standards or upholding certain social norms. Initially, people might adopt a standard and incorporate themselves into a given network of users for one of a number of reasons – Grewal highlights rational cost-benefit analyses, compulsion or simple chance (such as, for example, the use of a particular language or writing system based upon the location or circumstances of birth). As numbers of users grow, however, so does the 'network power', such that further people might adopt the 'standard' not for any intrinsic benefit it offers or even as a matter of free choice, but simply because they want or need to interact with others and that is the only viable way to do so. Thus agency exists, but in line with the structuration models of society outlined in Chapter 2, it isn't unconstrained: human ability to make decisions and act accordingly is constantly exercised in relation to, and contextualised by, the social structures and path-dependent courses of action shaped by previous agency and practice.

<sup>28</sup> Tugendhaft 2012a; 2012b; 2018. Previous interpretations of the political agenda of the poem have not seen it as anywhere near as subversive. Quite the opposite: Wyatt (2002; 2005) suggests that it may in fact be a *celebration* of links with the Hittite Empire, on the occasion of Niqmaddu III's marriage to Tudḫaliya IV's daughter, and that it was intended to flatter the Great King: 'It served the double purpose of celebrating the king's wedding and flattering the bride's father, the Hittite emperor Tudhaliya IV, with a treatment of the old martial ideology (KTU 1.1–2), a transparent reference in the palace-building episode to the establishment of a royal house (KTU 1.3–4), which itself might allow a pun on the play between a material house (palace, temple) and a dynastic house (offspring) such as lay behind 2 Sam. 7 […] The conflict with Mot (KTU 1.5–6) also allowed construal as the triumph of fecundity over death, appropriate to the celebration of a marriage' (Wyatt 2005, 252).

<sup>29</sup> RS 19.070, 8–11. For an alternative perspective on the life and experience of ʾIlimilku, see Wyatt 2015.

impose it and the traditional Late Bronze Age norms of kinship and diplomacy are little more than hollow words, impossible to implement consistently or fairly.

As a servant of [Niqmaddu], Ilimilku wrote a poem that betrays a profound comprehension of the workings of Bronze Age politics. He exposes the underlying premises that ground the politics of his world with remarkable discernment. It is reasonable to conclude that this acuity was a product of the exceptional age to which he was witness.

Was Niqmaddu privy to his scribe's wisdom? And if so, what did the king hope to gain from its dissemination? Perhaps Niqmaddu wished to enlighten his fellows about the basis of Hittite sovereignty over Ugarit. What remains remarkable, though, is the hardheadedness of the Ugaritic response. Where others might have shouted louder about the power of their own god to counteract the theologico-political claims of an imperial overlord, the strategy of the Baal Cycle is to undercut all claims to politics' divine foundation.30

This brings us back to the question of identity. The interconnection of vernacularisation and identity runs deep, from the proto-nationalism of Sanders (2004) to the nation-building of African stateslooking to define themselvesin the wake of colonialism, and far more besides. This intertwining has occasionally been questioned, such as by Pollock, who warns that the association of language with identity is a very European way of thinking and doesn't necessarily apply cross-culturally. In his own case study region, for example, he argues that, '[l]anguage was never the "indispensable pole of identification" in South Asia before modernity made it such'.31 For similar reasons, he also criticises the connection often made between use of the vernacular and elite legitimation. In his view, this too is a product of western modernity and its specific ideological hangups (in this case, he suggests, a concern with the need to reconcile traditional monarchies with the emerging capitalism of the nineteenth century).32 'Nothing compels us to believe,' he notes, 'that legitimation, or its higher-order form, ideology – two key components in the social analysis of capitalist modernity – have anything like the salience in noncapitalist nonmodernity that scholars have attributed to them.'33 For Pollock, the premodern vernacularisation seen in South East Asia was not a social change resulting from concerns about identity or elite legitimacy, but from politics and aesthetics.

These points are well taken, but the examples given above of how Ugarit's elites negotiated their position relative to the Hittite Empire and wider international networks through political actions, choices about art, iconography and other material culture and rituals such as the 'Ritual of National Unity' do, I think, point to an interest in defining a distinct local (elite) identity. Against this background, the

<sup>30</sup> Tugendhaft 2012a, 246.

<sup>31</sup> Pollock 2006, 511.

<sup>32</sup> Pollock 2006, 518–520.

<sup>33</sup> Pollock 2006, 517. Pollock uses 'ideology' here in its narrower sense of 'false consciousness, misrecognition, discourse of false necessity, and the like', not just 'any idea-system' (519). On problems with the ideas of ancient legitimacy and legitimation see also Lendon 2006.

linguistic and scriptal choices involved in vernacularisation fit in well; it's not a case of an identity or nationalism centred around the vernacular language, but of language and writing practices being employed as one among many tools in the enacting of this characteristically Ugaritian elite identity. An important thing to note, however, is that this is a pragmatic or constrained choice: Ugarit's rulers and other elites were not free to express themselves however they pleased. Despite their willingness to push the boundaries of Hittite tolerance, those boundaries still existed and Ugaritian elite agency had to work within those parameters; engagement with Egyptian material culture waxes and wanes in time with the prevailing attitude between Egypt and Ḫattuša, elite culture and prestige display still had to function within the expected norms of Late Bronze Age internationalism, and the vernacular was only used for a limited range of mostly internal functions.

Similar processes are apparent in neighbouring regions, including Ḫatti itself. There, of course, the political situation and associated power dynamics were very different to Ugarit, but nevertheless, trends towards increasing vernacularisation are apparent, most obviously in the apparent spread of the Luwian language and the hieroglyphic writing system used for it. Luwian Hieroglyphics gained in popularity from the fourteenth century onwards, and were the only writing system used for monumental inscriptions. It has been suggested that this may indicate that a growing proportion, or even the majority, of the population may have spoken Luwian rather than Hittite; signs of Luwian linguistic interference in thirteenth-century Hittite-language royal inscriptions may indicate that by this time even the elite were effectively bilingual. The ultimate conclusion of this process is apparent in the 'Syro-Hittite' states of the Early Iron Age, where Luwian, not Hittite, wasthe language used for official purposes, and hieroglyphs rather than cuneiform the script.34

An even closer parallel is to be found in Phoenicia to the south. There, like at Ugarit, diplomatic correspondence of the Late Bronze Age points to a strategic and careful engagement with the dominant political power (in this case Egypt, though unlike the Ugaritian situation, there was no formal imperial control). Elements of art and iconography, as well as certain elite burial practices, seem markedly more Syrian-looking towards the end of the Bronze Age and entering the transition to the Iron Age. The overall impression is that great international powers such as Egypt have lost their lustre for Phoenician elite display, and there is a growing interest in emphasising their Levantine-ness. The rise of Phoenician as a prestige, written language – perhaps first in the experimental form of the Byblos syllabary, then, to a limited extent in linear form on inscribed arrowheads and then, definitively, in the Byblian royal inscriptions of the tenth century – demonstrate again the rise of the vernacular at the expense of 'global' languages such as Akkadian and Egyptian. As with the 'Syro-Hittite' states, by the Early Iron Age this process seems to have successfully run its course, and asfar as we can tell from surviving evidence, all official

<sup>34</sup> Payne 2010.

Phoenician business of the early first millennium was conducted in the vernacular.<sup>35</sup> We don't have any surviving literary texts from this period, but there is little reason to doubt these too would have been written in Phoenician.

The changing tide of attitudes towards international languages and scripts is well illustrated by the Egyptian tale known as the 'Report of Wenamun'. This purports to be the factual account of an Egyptian envoy sent to Byblos to procure cedar for the royal barque of Amon, around 1075 BC. In fact, there's good reason to believe it was written somewhat later and is a fictional, perhaps satirical tale;36 nevertheless, it can give us a sense of how the Levantine coast at the end of the second millennium was perceived in Egypt. Two noteworthy scenes attest to the status of Egyptian. In the first, Zakarbaʿlu, the king of Byblos, reacts with sarcastic scorn at the suggestion that he might want to erect a hieroglyphic stele commemorating his trade with Egypt:

Now then, you should rejoice and order that a stele be raised for you, saying on it: 'Amon-Re, Sovereign of the gods, sent me Amon of the Road, his messenger – life, prosperity, health! – and Wen-Amon, his human messenger, in search of timber for the great and august ship of Amon-Re, Sovereign of the gods. I felled it. I transported it. I provided it with my ships and my crews. I ensured that they will arrive in Egypt, so as to beg Amon for fifty years of life for myself, over and above my destiny.' And it may happen that, in the fulness of time, a messenger may arrive from the land of Egypt, who knows writing and will read your name on the stele. And you will receive water [in] the West, like the gods who are there!

And he said to me: 'What you have said to me is a great testimony of words!'37

In the second scene, having left Byblos and been shipwrecked on Cyprus, the unfortunate Wenamun finds him beset by murderous locals, battles his way to the home of the town's female ruler and frustratedly exclaims to her entourage, 'Isn't there any one of you who understands Egyptian?!' It turns out one person does, but the overall impression is hardly of a language widely understood and highly respected in the eastern Mediterranean.38

Ugarit's vernacularisation shouldn't be understood in isolation, then, but as part of a trend encompassing much of the coastal eastern Mediterranean and Anatolia. This included linguistic vernacularisation, often hand-in-hand with the emergence of new scripts, but extended beyond it into areas of politics, material culture and other

<sup>35</sup> Boyes 2013.

<sup>36</sup> Boyes 2013, 30–31.

<sup>37</sup> Translation John A. Wilson in Pritchard 1969, 28. Wilson notes in footnote 37 that he cannot be sure whether the irony in Zakarbaʿlu's reply is deliberate or not, but it surely is, given that Wenamun goes on to promise that if he returns home successfully he will be able to obtain something for him, and given the overall tone of the story, which wastes few opportunities to demonstrate Wenamun's ineffectiveness, frosty welcome in the Levant and the lack of respect shown him.

<sup>38</sup> It's hard not to think of the scene in *Indiana Jones and the Last Crusade* where hapless English academic Marcus Brody tries to find a speaker of English – or, at a push, ancient Greek – in the bustling streets of İskenderun.

forms of elite practice. In each region the particular political, social and cultural context shaped the local dynamics, but the common denominator is an increased focus on emphasising local culture and local identity at the expense of globalised elite culture. This can be seen in terms of the idea of 'glocalisation', where local identities can become increasingly strongly expressed in reaction to the perceived homogenising and marginalising forces of global networks.39 This was an idea I first explored in my PhD dissertation in 2013, and its power has since been vividly and lamentably demonstrated by the resurgent success of far-right populist nationalism in Europe and America and the associated Brexit farce in the United Kingdom.

In the modern context, it is this populist element which forms the link between local identities, the elite and legitimation: elites have both fed and exploited growing localist sentiments within large sectors of the wider population in order to legitimise and bolster their position. This is nowhere more apparent than in the British Conservative party, whose attempts to dig itself out of, first, internal division, then ever-deepening political crisis have seen it redefine itself ever more as a party of hard-right radical nationalism and isolation. At the same time, these same elites are, of course, fundamentally implicated in, and reliant upon, the globalist networks that underpin both national economies and their private wealth and lifestyles.

In the ancient world, as Pollock pointed out, we should not assume the same need forrulersto legitimise their authority or appeal to a wider population. In Ugarit, Ḫatti and Phoenicia there was not even a veneer of democracy. And yet, rulers didn't rule absolutely, irrevocably and without need to maintain consensus. Authority must be enacted, power must be expressed in a way that is recognised, or in a sense it doesn't really exist at all. Status isn't a quality a particular person or group has, so much as a significance attached to their practice and trappings by those around them. This is why conspicuous consumption, ritual and elite excess are all important elements of ancient elite behaviour. Elites have to be seen to be elite in order to maintain and reproduce this position. But the exact form this display takes, the messages encoded in it, are subject to shaping dependent on the goals and requirements of those involved, what they want to achieve and whom they want to 'speak to'. So who, then, were these new displays of localism and vernacularisation aimed at in the Late Bronze Age east Mediterranean? Was this 'populism', in the sense that it was aimed at appealing to the masses, playing on some sense of local pride and suspicion toward global networks, in order to strengthen and advance the interests of elites? Elites who at the same time were also interacting with each other through those selfsame networks?

Some scholars have thought so. Both Annick Payne and Ilya Yakubovich, for example, see the rise of Luwian in Anatolia in terms of an elite attempt to speak to, and derive support from, the general population.40 Both point out the visual distinctiveness of Luwian hieroglyphs compared to cuneiform and highight the possibility

<sup>39</sup> Boyes 2013; 2019a.

<sup>40</sup> Yakubovich 2008; Payne 2010.

that this may have allowed inscriptions to communicate more effectively even to a largely illiterate population. For Payne (2010, 121), the pictographic aspect of the script may have enabled even those who couldn't, strictly speaking, read it to discern something of the meaning of royal inscriptions. For Yakubovich, it is more a matter of appealing to their sense of local identity: even if most of the population could not decipher the meaning of the inscriptions, they might still be able to recognise that they were written in the local script associated with their own, Luwian language, and from this get a sense that their rulers stood in solidarity with them, aligned with the general Anatolian population, not the foreign, international elite culture represented by cuneiform. This is, of course, up for debate. We can question how many ordinary people would have seen these inscriptions, whether they would have interpreted them in this way or whether the Hittite rulers would even have cared what the general populace thought. Ultimately, that is a matter for elsewhere; what matters to us is whether the same could be true in Ugarit.

This seems unlikely. As we've discussed several times throughout this volume, and especially in Chapter 7, writing was not highly visible in the kingdom of Ugarit. Writing *practices* may well have been part of everyday life and something even illiterate people would have seen going on, but the writing itself – the actual signs on the tablets or other objects – is very unlikely to have been something most people saw close up. Even if they did, the visual difference between alphabetic and logographic cuneiform – while readily noticeable to someone familiar with the systems – is nowhere near as pronounced as that between Luwian hieroglyphs and Hittite cuneiform. To an ordinary, non-literate member of Ugaritian society, one set of tiny wedge-shaped impressions in clay probably looked very much like another.

Although it is, as Sanders says, something of a landmark in world scriptal history and a major upheaval in terms of the kingdom's own writing culture, the large-scale vernacularisation of Ugarit's bureaucratic and literary culture and the adoption of a new, alphabetic script for writing it, was very likely completely invisible to most members of Ugaritian society. Even if they had been informed of it, I doubt whether it would have had the same significance to them as to the literate officials and intellectuals who actually practised writing, or to those of us who study it. Likewise, the exotica, diplomatic gifts and lavish temple dedications with which Ugarit's elites signalled their status, aspirations and overseas connections would hardly have been visible to the great majority of their subjects. The new localism we have discerned in east Mediterranean society at the end of the Late Bronze Age was not populism, or at least surviving evidence gives little indication that it was aimed at the general populus.

The only possible audience for this cultural signalling is the elite itself. Perhaps we might include certain foreign guests – likely similarly elevated in status – but in general, there seems to be a concern to confine the most potent markers of local autonomy and Ugaritian identity, such as the language and script, to internal rather than external contexts. This 'glocalisation' was not about Ugarit's elite legitimating itself in the eyes of the general public but about talking to itself, reimagining its own ethos. If there was a popular sentiment against the globalised networks and great powers of Late Bronze Age elite culture, then these actions probably represent Ugarit's rulers trying to convince themselves they are somehow different; it seems unlikely they would have done much to engage the wider population. Unlike many of the other societies we've discussed in this chapter, Ugarit failed to survive the crises of the end of the Late Bronze Age and the polity was not re-established afterwards. We might be forgiven for wondering, then, whether this solipsistic elite strategy of talking to themselves and crafting a new sense of local Ugaritian elite identity solely for their own consumption had anything to do with that. This is one of the questions we will be addressing in the next chapter.

### Chapter 12

### The impact and legacy of alphabetic cuneiform

In this chapter, we will consider the end of writing practices at Ugarit and what effect – if any – they had on the expansion of alphabetic writing in the Early Iron Age.

The traditional assumption is that the Ugaritic language and the alphabetic cuneiform script met sudden ends with the destruction of the city in violent action by seaborne raiders or invaders around 15 years into the twelfth century BC, which we've already discussed. All this is very probably correct, but I'd like to spend a little time thinking about why we believe this and the processes by which these linguistic and scriptal practices ceased. It is, after all, clear that the destruction of a town or city doesn't automatically mean that that settlement is permanently abandoned, or that its socio-cultural traditions come to an end. The end of the Late Bronze Age – and history as a whole – are filled with examples of cities being swiftly reoccupied and rebuilt after destructions and of language, script and other practices continuing even where reoccupation did not take place.

We see this, for example, in the sites of Tell Kazel and Tell Keisan in southern Syria and northern Israel respectively. At both towns, Late Bronze Age destruction layers have been identified but are followed by immediate Iron Age reoccupation. The same is true of several sites within the Kingdom of Ugarit itself, such as Gibala/Tell Tweini, Ras el-Bassit and Tell Sukas. While they haven't been investigated archaeologically, numerous smaller villages display onomastic continuity from the Ugaritic records right up to the present day, which points to a certain degree of continuity at least in social memory, if not in actual occupation.1 Continuity is also clearly apparent elsewhere in northern Syria. The Late Bronze Age levels at Karkemiš are not well known archaeologically, but the city is mentioned as a victim of the 'Sea Peoples' in the Year 8 inscription on the mortuary temple of Ramesses III at Medinet Habu in Egypt; in the Early Iron Age it was a major power among the 'Syro-Hittite' polities of

<sup>1</sup> Caubet 1992; Badre and Gubel 1999–2000; Niemeyer 2000, 92; Bretschneider and Van Lerberghe 2008, 43; Vansteenhuyse 2010.

northern Syria. In the latter case, as with other north Syrian states of the period, we can observe the potential for language and script to survive destructions – even the dissolution of the polity associated with them – and adapt themselves to new contexts. In this case the Luwian language, which was widely spoken in Hittite Anatolia and was presumably introduced to these polities during Hittite overlordship, became the vernacular during the Early Iron Age. The Luwian hieroglyphic script, which had formerly been used almost exclusively for Hittite monumental inscriptions, was now used for all known writing at these sites.

In the Aegean too we witness the differing fates of script, language and other cultural practices through periods of dramatic social upheaval. By the end of the Bronze Age, writing there was strongly associated with, and mostly limited to, the Mycenaean palace administrations. The Linear B script and its associated writing pratices fell out of use with the end of those institutions, but the Greek language it encoded of course continued. In wider culture, the transition into the Iron Age in the Aegean was marked by a combination of both continuity and rupture.

The inverse situation can also be true: we can see examples of dramatic changes in writing practices where otherwise continuity rather than rupture was the order of the day. This is the case in the transition to the Iron Age in Phoenicia. The major centres there have produced little sign of violent destruction at the end of the Late Bronze Age and their cultures continue uninterrupted into the first millennium; however, major changes in writing practice are apparent, with the apparent end of the use of logosyllabic cuneiform and Akkadian, and the rise of the linear alphabet used for writing Phoenician.2

There is thus no one-to-one correlation between site destruction and changes in linguistic or writing practices. Destroyed sites can be rebuilt and their practices can continue. Sites that were not destroyed can nevertheless radically change the way they wrote and writing's place in society. Clearly we must look more closely at the processes by which writing practices cease to be carried out.

#### **The end of writing practices**

There has been relatively little well-developed research into how scripts and writing practices fall out of use. The most important contribution remains that of Stephen Houston, John Baines and Jerrold Cooper (2003). An edited volume in 2008 responding to and broadening the scope of that article (Baines *et al.* 2008) represents the only major attempt at developing its work further, but beyond the editors' own chapters, the focus was more on individual case studies rather than the development of general theory or methodological frameworks for approaching script disappearance.

Despite the fact that we are evidently only beginning to think in detail about these matters, these two works make a number of useful advances over what little

<sup>2</sup> Boyes 2013, esp. chapter 2.

work had been done before. The first, in line with the discussion above, is to decouple script from language, both in terms of how it is approached and how the loss of one might lead to the loss of the other. They also seek to nuance the way both connect to concepts of polity, culture or 'civilisation'. For Mark Geller (1997), script, language and civilisation were so closely interconnected that the end of use of the cuneiform script could be considered the end of 'Mesopotamian civilisation'.3 Houston, Baines and Cooper (2003, 432–433) point out the significant differences between script and language with regard to their positions in human social existence: the latter is an innate biological faculty and no human community has ever ceased linguistic communication, even if the languages used to exercise that faculty can come, go and exist alongside each other. Writing, on the other hand, is not a hardwired part of human behaviour. The complete end of writing practices has been attested in numerous societies and historical contexts. Partly because of this lack of a biological aspect to writing, these authors advocate against the rather common metaphor of 'script death' to describe such disappearances. At a more practical level, they also argue that the number of users a script needs to survive can be significantly smaller than for a language, provided that there is a social investment in its use.4 This means that the social and cultural dynamics of script extinction function rather differently. Baines (2008, 351) concludes that the loss of script is frequently associated with that of one or more of 'languages, polities, and cultures or civilizations', but not in a predetermined way; these features might be lost in any combination – or independently – dependent on the specific socio-cultural circumstances and the agency of the actors involved. The loss of one might precipitate the loss of others, or it might not.

They also rightly highlight the importance of distinguishing between different kinds of writing practices when talking about the end of a script. They focus in particular on the difference between reading and writing.<sup>5</sup> This allows for considerable nuance as script use varies along both axes ('response' and 'production' respectively, in Houston's (2008) terms). We can even expand this if we also embrace 'ancillary' practices such as the manufacture of particular writing materials or particular ways in which writing or inscribed objects were used. In this sense, then, the cessation of particular writing practices should be understood in similar terms to the cessation of any other human practice, or indeed, to any social change at all. It's appropriate, then,

<sup>3</sup> 'When the script of a language is no longer intelligible nor the language spoken, a remarkable event results – the death of a civilization, a clearly definable historical boundary which can almost not be determined by any other objective means' (Geller 1997, 44).

<sup>4</sup> Of course, there is no absolute number of users necessary for the survival of either script or language; it depends very much on the specific socio-cultural circumstances, with elements such as degree of isolation or exposure to alternatives, the range and context of uses of the script/language in question, and its socio-political prestige all being important contributors (Crystal 2000, 11–19; Houston *et al.* 2003, 433).

<sup>5</sup> Houston *et al.* 2003; Houston 2008, 232 and *passim*.

that these authors also stress the importance of so-called script communities – groups 'that use [...] and disseminate [...] writing across generations through apprenticeship and other modes of training, and which might be united by a common identity centred around the use of a particular script, even if not all members are equally competent in its use'.6 This idea owes much to the 'textual communities' developed by Brian Stock (1983; 1996) to understand medieval heresies, a key aspect of which was that the community could form around a particular text even if most members were not literate – only a single literate interpreter was necessary, around whom a group could organise. This emphasis on the diversity of social groups engaged in writing practices further nuances the idea of script obsolescence, allowing for a situation such as the numerous 'small deaths' envisaged for the Terminal Classic Mayan script by Houston (2008), whereby obsolescence doesn't occur all at once or evenly, but by degrees, as different groups cease to engage in and pass on their practices, potentially at different times and for differing reasons.

#### **The fall of Ugarit and the fate of its people**

To understand the end of alphabetic cuneiform and the Ugaritic language we must first understand how the destruction of the city took place, and secondly, the fate of Ugarit's population afterwards, which links into a third question: why was Ugarit not reoccupied? I should say at the outset that, as with many of the questions posed in this book, there aren't necessarily good answers to all of these at the present time; nevertheless, we can make explicit exactly what it is we need to know.

So, the first question: how did Ugarit's destruction take place? The great majority of what we know comes from extrapolating from the written documentation of its final phase. As we saw in Chapter 11, numerous texts refer to seaborne raiders and a sense of threat hanging over the city. Given the site's destruction not long afterwards, it is perfectly reasonable to assume that one led to the other, although no text directly describes the events of Ugarit's fall. This historical reconstruction is seemingly supported by archaeological data, although this has not been published in great detail. Yon (1992, 117) refers to collapsed walls and signs of burning throughout the city, as well as large numbers of arrowheads scattered across the site, which she interprets as signs of 'violent fighting'. She mentions an ongoing typological study of these weapons, which might have illuminated somewhat who they might have belonged to, but the results seem not to have seen the light of day. While there is little reason to doubt that Ugarit met its end in violent action, then, our information on the nature of this end must be acknowledged to be rather vague.

Question two: what happened to Ugarit's population? The expectation in ancient Near Eastern urban warfare seems to have been that defeated defenders would be treated extremely harshly. Even where negotiation resulted in a city surrendering without conflict,this'peaceful'solution could be expected to involve the forced deportation of the population, and possibly their enslavement, as biblical references attest:

<sup>6</sup> Houston *et al.* 2003, 431, n. 2; Houston 2008, 232 and *passim*.

When you approach a town to attack it, you shall offer it terms of peace. If it responds peaceably and lets you in, all the people present there shall serve you as forced labour.7

Make your peace with me and come out to me, so that you may all eat from your vines and your fig trees and drink water from your cisterns, until I come (at the end of the stage of combat of that same campaign) and take you away to a land like your own, a land of grain and new wine, or bread and vineyards, of olive oil and honey.8

When a city did not surrender, Deuteronomy 20, 12–14 implies that its entire male population could expect to be slaughtered, the women and children to be carried off as booty alongside their livestock and possessions. Archaeological evidence from Lachish supports the idea that a violent sack would involve extensive loss of life. Four mass graves were found at the site, associated with a destruction horizon interpreted as the sack of the city. This has usually been identified as the destruction by Sennacherib in 701 BC, though the tombs themselves appear to be Late Bronze Age and datable material culture from the mass graves is sparse so there's a possibility that they might relate to a destruction occurring at the end of the Bronze Age, around the same time as the fall of Ugarit. Regardless of whether this was a destruction contemporary with that of Ugarit or five centuries later, the graves show how brutal such events could be. They contained skeletal remains from more than 1500 individuals, almost half women and children. Israel Eph'al (2009, 33–34) estimates that this must have represented a large proportion of the total population of a town the size of Lachish. It is worth noting, however, that few of the bones showed injuries from weapons, so their deaths may have occurred from disease or starvation associated with a siege, rather than in a deliberate massacre during or after the conquest of the city. Nevertheless, the end result was the same: a large proportion of the town's population did not survive its destruction.

No such charnel deposits have come to light at Ugarit. Human remains are, on the contrary, notably lacking from the city, even in the (mostly plundered) tombs from before its destruction. This doesn't necessarily indicate that no such massacre occurred: Yon (1992, 117) argues that conditions at the site are such that little might be expected to survive. Moreover, skeletal remains from destruction horizons are rare in the Near East. Igor Kreimerman has suggested this is most probably due to the fact that the looting of a conquered city could take days or even longer, and corpses would be likely to be cleared out to facilitate this pillage prior to a final deliberate burning-down of the settlement, which would occur at the end of an invader's interest in a site. With largely mud-brick and stone buildings such as those of Ugarit, wholesale destructions are unlikely to have occurred incidentally in the course of fighting, but would require planning and systematic effort.<sup>9</sup> However, we should also note that most of our information on what what was involved in sacking an ancient Near Eastern city relates to military action by soldiers fighting for states such as Assyria or Israel, rather than the ragtag, potentially stateless, marauders often envisaged for the Sea

<sup>7</sup> Deuteronomy 20, 10–11, adapted from Eph'al 2009, 46.

<sup>8</sup> 2 Kings 18, 31–32, cited by Eph'al 2009, 47.

<sup>9</sup> Kreimerman 2017a, esp. 20–22. On the destruction of cities, see also Kreimerman 2017b.

Peoples. Since we know so little about who Ugarit's destroyers were or what they hoped to gain from the act, it's impossible to say how likely they would be to have acted in accordance with the later norms of Near Eastern urban conquest.10

If a general massacre did happen, the chances are it wasn't a total extermination. Even at Lachish, where large numbers of human remains were found, this is likely to have been the case. If the dead there do relate to Sennacherib's conquest, Eph'al argues, then we must still assume that some of the population were taken captive and sent into exile, as is suggested by relief depictions of the battle, which show the Assyrian ruler surveying the prisoners and booty taken from the city.11 In Greece, the fourth-century BC destruction of Olynthos by Philip II is well-documented by text and archaeology – Florence Gaignerot-Driessen calls it 'the most paradigmatic case of a destroyed city' (2013, 285). It was razed, plundered and its population enslaved or decimated. And yet Diodorus (19.52, 19.61) mentions numerous survivors and the name of Olynthos survived on inscriptions. 'It is thus clear on the one hand that the site was partly reoccupied and on the other hand that a small community of Olynthians survived in the immediate area. So we can conclude that despite the destruction of the urban center, the loss of part of the population and the end of military power, the city of Olynthus continued to exist.'12 In the Kingdom of Ugarit, Yon sees signs of evacuation first at Ras Ibn Hani and then at Ugarit itself. The former was, perhaps, more orderly than the latter: more possessions seem to have been taken from Ras Ibn Hani, while at Ugarit Yon discerns evidence of surprise and haste: overturned tripods and equipment, even abandoned washing-up.13 In her view, the likeliest immediate refuge for any potential refugees would have been the villages a little way inland on the plateau and in the mountains.14 In the longer term, we might imagine some joining the resettlement of sites like Ras Ibn Hani or Gibala, or else dispersing into neighbouring regions or joining nonsedentary communities.

This brings us to the third and final question: why wasn't Ugarit reoccupied to any great extent, when other sites in the Kingdom were? The first thing to make clear is that Ugarit was not wholly abandoned after its destruction. There are signs of occupation in the Early Iron Age, some of which involve quite substantial modifications to the surviving structures. For instance, the moving of large stone water troughs or the construction of ovens. Marguerite Yon and Annie Caubet are both clear that this is not permanent, organised reoccupation, however, and it is usually attributed to transient populations of nomads or pastoralists.15 By contrast, Ras Ibn

<sup>10</sup> Differences are certainly evident between cultures in the treatment of conquered cities. For example, in contrast with the Assyrian predilection for razing conquered cities, Egyptian documents and imagery, for example, consistently point to a preference for leaving them intact after their plunder, probably to allow ongoing economic exploitation (Hasel 2016, 211).

<sup>11</sup> Eph'al 2009, 34.

<sup>12</sup> Gaignerot-Driessen 2013, 285.

<sup>13</sup> Yon 1992, 117; Bounni *et al.* 1998, 53.

<sup>14</sup> Yon 1992, 119.

<sup>15</sup> Caubet 1992, 123; Yon 1992, 119.

Hani's reoccupation involved the erection of new structures, at least on the crest of the tell, and the continuation of traditional local ceramic traditions, as well as the presence of Aegean-inspired 'LH IIIC' pottery.16

We have no direct evidence for why second-order centres in the Kingdom were subject to new permanent settlements while the capital was not. It may be a longterm effect of actions taken during Ugarit's destruction – sources on the siege and conquest of cities in the ancient Near East often refer to the deliberate destruction of agricultural resources, such as the burning of fruit trees or the wrecking of fields.<sup>17</sup> This destruction of food resources is implied by KTU 2.61, which seems to be a letter from another settlement in the kingdom to the capital and reports:

When your messenger arrived, the army was humiliated and the city was sacked. Our food in the threshing floors was burnt and the vineyards were also destroyed. Our city is sacked. May you know it! May you know it!18

If such wrecking of agriculture occurred at Ras Shamra, it may well have made Ugarit a less attractive site for reoccupation, especially by refugees who were perhaps few in number and somewhat in disarray.

There would also have been psychological and ideological effects. If, as Yon thinks, people initially evacuated outlying towns like Ras Ibn Hani to seek refuge in the capital, and the Kingdom's final stand took place there, then it's likely that it would be at Ras Shamra that fighting would have been fiercest and the destruction the most intense. This would mean not only that there would be less remaining that was in a fit state to reoccupy, but also that the ruined capital may have been associated with the bad memories, trauma and ghosts of its violent defeat. It might also have been symbolic of a polity that had failed: of kings and urban intitutions that, in the final reckoning, had been shown to be unable to protect their people or themselves. Maybe they were even seen in some way to have precipitated this destruction, whether through direct political acts, inaction or through inviting the displeasure of the gods. Again, we can see parallels for this kind of phenomenon elsewhere in the ancient world. At Dreros in Greece, there was no reoccupation after its destruction in the late third/early second century BC. Gaignerot-Driessen sees this as the result

<sup>16</sup> Caubet 1992, 124–127.

<sup>17</sup> For example, 2 Kings 3, 19: 'You shall conquer every fortified town and every splendid city, you shall fell every good tree and stop up every spring, and every fertile field you shall ruin with stones'. Cf. Deuteronomy 20, 19–20, which bans the destruction of fruit trees during a siege, or Tiglath-Pileser III's boast that he cut down all the gardens and orchards of Damascus when he besieged it in 733 BC (Tadmor 1994, 78, cited by Eph'al 2009, 52, see also 14–15). Similar actions are recorded by the Egyptians and Hittites (Hasel 2016, 214–215). However, Cole (1997) cautions against taking the documentary accounts over-literally. In his view, the economic and subsistence importance of orchards makes it unlikely they would have been destroyed peremptorily. Instead, he suggests the accounts mix a hyperbolic literary trope with elements of psychological warfare whereby trees would have been destroyed gradually to weaken defenders' resolve and encourage surrender.

<sup>18</sup> Translation from Singer 1999, 726.

of the invaders deliberately 'killing' the site – not just destroying its structures but enacting a kind of ritualised annulling of the polis through destruction and forced relocation. Afterwards, she suggests, it was 'as if the place had become forbidden or a taboo placed upon it'.19

#### **The end of Ugaritian writing practices**

The background we've established over the preceding few pages sets us up to clarify, as best we can given the evidence currently available, how Ugarit's distinctive writing practices ceased. The first thing that can swiftly be affirmed is that the most distinctively Ugaritian practice – the use of alphabetic cuneiform – did indeed stop with the destruction of the city. To the best of my knowledge, only one person has seriously proposed otherwise – Benjamin Sass, who suggested that the 'short alphabet' inscriptions found outside the city, in Lebanon, northern Israel and the Mediterranean, need not be Late Bronze Age in date.He justifiably highlightsthe poor archaeological context information for most of these items and moves from this to wondering whether 'the writing employed in parts of Syria (south of Ugarit) and Phoenicia in the twelfth–tenth centuries was alphabetic cuneiform which had outlived the end of Ugarit by two centuries and more before being replaced by the linear alphabet?'20

This is not impossible, but there is little reason to suppose it was the case. While Sass is right to draw attention to the lack of good archaeological data on the origins of most of these objects, those that do have decent stratigraphies hail from the end of the Late Bronze Age: thus, the Phoenician-inscribed handle from Sarepta21 or the ivory rod from Tiryns in Greece.22 None are from secure Early Iron Age strata. Ultimately, Sass's suggestion is in service of his attempt to lower the date for the widespread adoption of the linear alphabet, which we have already discussed and rejected (see Chapter 3). Outside of this, there's no good reason to think it was the case.

I've attempted to show in this book how late thirteenth-century writing practices at Ugarit were the result ofthe kingdom'sspecific socio-cultural context and especially of contradictory but intertwined impulses towards localism and the need and desire to engage with international culture. These and other factors resulted in a writing

<sup>19</sup> Gaignerot-Driessen 2013, 292–294. The Greek word used for such city-killings, *kataskaphe*, is rich in associations. It is also used to describe a punishment for certain severe crimes such as murder and treason, whereby a person's house is razed to the ground. This could be accompanied by other measures intended to signal the annulment of a household or kinship line, such as the seizure of property, the exhumation of ancestral remains or lack of burial for the criminal and even their descendants. It's easy to see, then, how when applied to a city, such a destruction was not 'just' a physical destruction (Connor 1985). On the often-overlooked trauma and other emotional impacts of living through civic destructions and other such events, see Hitchcock 2013.

<sup>20</sup> Sass 2005, 53.

<sup>21</sup> Khalifeh 1988, 113.

<sup>22</sup> Tropper and Vita 2010, 693. To be fair to Sass, although this item was discovered in 2002/3, it was not published until 2010, five years after he published the suggestion we are discussing.

culture characterised by the principal use of both alphabetic and logosyllabic cuneiform, and the Ugaritic and Akkadian languages respectively, as well as a smattering of otherscripts and languagesin certain specific contexts. As hasfrequently been noted, the general pattern is that alphabetic cuneiform was used for internal purposes and those associated with civic life and identity – administration, ritual, mythology and so on – while logosyllabic cuneiform predominated in genres that pertained to globalised political and economic culture: diplomacy, treaties and so on. This distribution does not represent an absolute division between the local and the global, however: alphabetic cuneiform practices were fundamentally based on the requirements and internalised prestige of the logosyllabic system – the choice of a cuneiformised version of the linear alphabet, the use of clay tablets of similar type and layout to those familiar from the logosyllabic system. These were not parallel systems, but a single writing culture whose different facets reflected the complex cultural realities of Ugarit and its people's positioning relative to each other and to their neighbours.

In terms of who was using Ugarit's various scripts – the script communities – we have identified a number, with varying degrees of certainty. The largest and most obvious is the urban elite literati – the politicians, businessmen, administrators, priests, magicians and poets (often many of those roles being embodied by the same person) who emerged out of the traditional 'scribal' education system. These seem to have been the driving force behind much of the kingdom's writing culture – certainly inasmuch as it survives today. As well as alphabetic cuneiform/Ugaritic, they probably account for most, if not all, of the logosyllabic/Akkadian practices. the likeliest scenario for the emergence of alphabetic cuneiform – at least as a standardised script written in cuneiform style on clay tablets – is the top-down initiative of some member of the elite, probably a king or crown prince. Most likely, its creation was bound up with elite self-presentation – namely an attempt to negotiate an elite identity that walked the line between vernacular localism and pragmatic acceptance of the requirements of the globalised networks of trade, politics and culture within which Ugarit's higher-status citizens were thoroughly implicated. For this reason, the alphabetic cuneiform script and the writing practices that went along with it were very probably strongly associated with the city and the urban elite in the minds of those who used it. I've speculated that there might have been a prior, linear version of this script, from which the surviving letter-forms were 'cuneiformised' for writing on clay. Perhaps this continued in parallel alongside the clay-based system, on perishable materials. Since this is, at best, highly speculative, it would be pointless to guess at whether such a system would have shared the fate of its cuneiform counterpart.

We also suggested the possibility of sub-elite writing, less rigidly centred on the capital – craft producers, provincial notaries and such-like. While I think these can reasonably be inferred, the evidence for them is certainly much sketchier than for the previous category. This writing was probably mostly in alphabetic cuneiform, and perhaps also a related linear version of the script on perishable materials, though this is even less certain than the existence of these writers in the first place. These writers may also account for some of the minority scripts and languages found in the kingdom, especially where those appear on crafted objects rather than standard clay tablets – items such as the Merneptah sword. We can't definitively prove their existence, much less delineate the natures and boundaries of their script communities. What is clear, however, is that if sub-elite and non-urban writers existed in the Kingdom of Ugarit, these writers would still not have been entirely separate from the fortunes of the metropolis or its leading citizens. Literate craftspeople would have been integrated into economic networks centred on the city; village notaries and functionaries are most likely to have existed to facilitate administration of the hinterland, even if they engaged in other writing practices on the side.

The destruction of the city of Ugarit ended its existence as a polity. We don't know what became ofits kings,queens andhighofficials.Perhapstheirrole asleaderswouldhavemade them priority targets for the attackers; perhaps it would have put them to the front of the queue when it came to escape. Either way, once the city fell, they had little left to be kings, queens or high officials of. If they did escape, they left their archives and libraries behind, whether through haste or an understanding that they would have no need of them now. Without the state there was no need of alphabetic cuneiform administration, no need of ideologically charged epics valorising the civic deity or the city's mythological patrimony. Without the temples, there was no need to inscribe votive objects. Without the economic and political networks centred on the capital, the need to write letters – and the ability to send them safely – would have been greatly diminished. Even those writers outside the city – if they existed – would have had less reason to practise their skills in the absence of the networks of exchange, consumption and administration that supported them.

At the same time, the end of Hittite imperial control and the disintegration of wider regional networks of trade, politics and elite culture meant that there was no longer a need to maintain a bureaucracy able to do business in the globalised communication standards of Akkadian and logosyllabic cuneiform. Even in parts of the Levant where the Late Bronze Age polities were not destroyed, such as Phoenicia, this resulted in the end of these writing practices and a move away from the clay tablet-based format that they entailed. In Ugarit, it's no surprise at all that this spelt the end of Akkadian and logosyllabic writing; but the end of clay-based, cuneiform as a communication standard would also have impacted the local alphabetic script, which, for all its localist aspirations, was fundamentally set up so as to integrate with the pre-existing logosyllabic writing culture.

Perhaps even more significantly for the end of alphabetic cuneiform, the script's close association with the urban elite, its practices, ideology and identity would very probably have made it a tainted brand after the city's fall. It's likely there were survivors. It's likely some of those survivors may have been literate and may have incorporated themselves into towns and villages where alphabetic and logosyllabic cuneiform had been at least known, and perhaps quite well established – Ras Ibn Hani being the most obvious candidate. Even at this former palatial centre, however, there seems to have been no attempt to rebuild a literate bureaucracy or to re-establish the use of Ugarit's most characteristic scripts for any other purpose. Perhaps they wrote on perishable materials in a related script, perhaps they didn't write at all; either way, the inhabitants of these Early Iron Age settlements had little need or desire to resurrect the practices of the Bronze Age capital, so fundamentally grounded were they in a social, cultural and political context that no longer obtained. They were no longer invested in being Ugaritian; no longer beholden to a globalised network that required Akkadian and cuneiform.23

This is unlikely to be a surprising model for Ugarit specialists. It is, I think, what lies unspoken behind the usual brief statement that Ugaritic and alphabetic cuneiform ended with the city. Nevertheless, it is worth laying out in full, not only for its own sake, but also to highlight the gaps in our knowledge and the fact that the end of these practices was not a one-off event – the destruction of a city – but a series of choices and actions following on from that. When was alphabetic cuneiform gone? In a single day or two, when the city fell and the political structure that created and sustained its writing culture collapsed? Days or weeks later when the city was burned down? When a literate survivor gave up trying to persuade anyone in his new community that the script was worth using? When someone impressed a wedge for the last time? When the last person decided not to pass the script on to their children? When the last person who could read or write it died?

These are not questions we can answer, and I don't ask them as a basis for speculation, but to highlight that even in a case of violent destruction, the end of a city or polity's writing practices is likely to have been a succession of 'small deaths' founded in agency and people's responses to actions outside of their control, and that potentially, these could have been drawn out over quite some time.24

#### **The legacy of alphabetic cuneiform**

The ancient impact of alphabetic cuneiform beyond the Kingdom of Ugarit and in the years after its destruction is, at first glance, negligible. With the dissolution of Ugarit as a polity and the dispersal and eventual death of the last people familiar with the script and its related writing practices, it can seem as if it passed entirely out of historical memory. There are no confirmed examples of alphabetic cuneiform after Ugarit's

<sup>23</sup> We should also not underestimate the psychological trauma of living through the destruction of one's city. There has been very little work on the archaeology of emotion, for the obvious reason of the difficulty of reconstructing it from a partial material record. Nevertheless, more recent and contemporary comparanda suggest the effects of survivors of the fall of Ugarit are likely to have been profound. On this, and for material evidence that can be used to begin to approach the question of emotional effects of destruction, see Hitchcock 2013. In the case of Ugarit, unfortunately, the extreme lack of any archaeological data from the twelfth century and from settlements outside the capital means that little can, at present, be done.

<sup>24</sup> For a thorough discussion of a more recent example of the protracted decline and extinction by degrees of a set of writing practices, see Houston 2008 on the fate of Maya after the Spanish conquest.

destruction, no elements of linear script or palaeography that have been suggested to derive from an Ugaritian source. To all intents and purposes, as generally understood, alphabetic cuneiform was a dead end and passed on nothing to the writing practices of succeeding ages.

However, the model of writing practice I've put forward in this book is not one of rigidly demarcated and separable systems, but of a network of practices that relate to each other to greater or lesser degrees.25 Alphabetic cuneiform writing practices, those using linear forms of the alphabet and even logosyllabic cuneiform were shaped by ongoing connections. Ugarit's writing practices didn't arise, live and die in a vacuum, but were entangled and intertwined with those of other polities around it. The question isn't whether Ugarit's writing practices had an impact or effect on writing elsewhere in the region, it's the extent to which the subtle flows of influence, reaction, response and innovation are now recoverable to us, especially given that evidence for the twelfth and eleventh centuries in the region is even sketchier and more difficult to work with than that for the Late Bronze Age.

In the last chapter I touched on the rise of Phoenician as a parallel example of vernacularisation to that seen at Ugarit. The adoption by Phoenician polities – apparently first Byblos and then others – of distinctively Levantine alphabetic script for writing the local language at a time when these polities were keen to assert their autonomy and define their position in the region has much in common with what I've suggested was going on at Ugarit. Where the Ugaritian impulse to vernacularisation was curtailed by destruction and political dissolution, that of the kingdoms of the coastal central Levant was allowed to run its course and eventually resulted in the complete replacement of logosyllabic cuneiform and the Akkadian language by the linear alphabet and Phoenician for official writing. The same is true further south still, where cuneiform and Egyptian scripts also gave way to linear alphabets and emerging Hebrew and Aramaic vernaculars. It's unlikely, I think, that these vernacularising movements themselves were directly prompted by Ugarit's example. Arguably the first sign of linguistic and scriptal vernacularisation in Phoenicia is the Byblos syllabary, which probably predates alphabetic cuneiform. More to the point, as we discussed last chapter, the Late Bronze Age eastern Mediterranean was in some senses a globalised network, in which ideas, ideologies and cultural trends could be decentred and disembedded, moving back and forth along the connections without clear-cut geographical points of origin or sequences of cause and effect. But if the importance of Ugarit's choices regarding script didn't ripple out to its neighbours in a linear way, they may still have mattered in shaping ideas of what did and didn't work in terms of vernacular writing and its appropriation for state purposes; not so much a progenitor as an example to be borne in mind by other Levantine writers, politicians and ideologues as they were making their own decisions.

To assess this possibility, we need to pin down with a bit more clarity what writing practices and writing culture are likely to have looked like in Phoenicia and the southern Levant in the Late Bronze/Early Iron Age transition. The earliest surviving

<sup>25</sup> See also Boyes 2019b.

example of writing in the Phoenician script is the early royal inscriptions of Byblos – disputed in date but most commonly assigned to the tenth century – and a number of inscribed arrowheads, almost all of which lack secure archaeological provenances but may be slightly earlier. It is all but inconceivable that this represents the full scope of writing practices in the coastal central Levant between the thirteenth and tenth centuries. There are a number of reasons to think that a much more extensive body of writing is likely to have existed on perishable materials, but simply does not survive. The simplest of these is that the Phoenician alphabet is, after all, a linear one, and linear scripts are best suited to writing with a pen or brush on flat surfaces, not to chiselling into limestone or engraving on hard metal. What's more, Reinhard Lehmann has argued that the palaeography of the Byblos inscriptions shows signs of developments in handwriting and letter-form – a 'calligraphic turn' as he calls it – that are most likely to have occurred in pen-and-ink.26

Numerous later sources – all dubious to some degree – have suggested the existence of Phoenician literary material on perishable materials in an unbroken tradition from the Bronze Age into the first millennium BC. The most commonly cited genre of this material is civic historical annals. The earliest and best of these sources is the Egyptian account of Wenamun, which purports to be an eleventh-century factual account but is probably fictional and written a century or more later. In it, the king of Byblos has the 'annals of his fathers' brought out for consultation. These seem to include economic records of the city's trade dealings with Egypt – this would be material relating to the Late Bronze Age. The script and language of these documents is not specified, but they're said to be 'unfurled', implying scrolls. The script was thus most likely envisaged to be either Phoenician or one of the Egyptian ones.

Greek historians of the Classical and Hellenistic periods also allude to Phoenician polities' long historical memories. Herodotos (2.44) claims to have learned Tyre's millennia-long history from the priests of the temple of Melqart. Several centuries later, Josephus claimed in his *Jewish Antiquities* to have consulted Tyrian annals stretching back to the Late Bronze/Early Iron Age transition, as translated into Greek by Menander of Ephesos. In his later, more defensive work *Against Apion*, Josephus presents Menander more as a historian than a direct translator. Several scholars have argued that Josephus was greatly overstating his access to Tyrian sources, and probably even Menander's. Giovanni Garbini sees the latter as a Hellenistic compiler of earlier historical material, and thinks any of Josephus's information that did originally come from Tyrian annals was probably third- or fourth-hand at best.27

Even more doubtful are the historical sources mentioned in the *Praeperatio Evangelica* of the third-to-fourth-century AD Christian bishop and historian Eusebius. In his discussion of Phoenician myth, history and religion he quotes fragments of two otherwise unattested earlier writers, Porphyry and Philo of Byblos. Both these historians refer to a Phoenician historian named Sanchuniathon who lived and worked around the time of the Trojan War – *i.e.*, ca. 1200 BC – and whose surviving works

<sup>26</sup> Lehmann 2008; 2019.

<sup>27</sup> Garbini 1980. See also Van Seters 1983, 195–199.

transmitted ancient knowledge and history from before this time to their own period. According to Porphyry, Sanchuniathon drew on Bronze Age king-lists and historical records, which he received from a priest and which were associated with the king of Beirut. These accounts are confused, however, and incorporate obviously legendary material. Porphyry, for example (*Praep. Ev*. 1.9.21), claims that Sanchuniathon was a contemporary of the legendary Assyrian queen Semiramis, whose historical inspiration, Šammurammat, in fact lived in the late ninth century, not the thirteenth. Philo offers an even more tenuous and unlikely tale, claiming that Sanchuniathon discovered ancient historical records in temples of Amun.28

The legends of Sanchuniathon are routinely, and rightly, discounted as concoctions of Hellenistic and later history. While 'Sanchuniathon' as a name has a good Semitic root, there is little reason to believe he was a real figure, or if he was, that he lived in the Late Bronze Age. It seems much more likely that 'before the Trojan War' was merely a rhetorical device to imbue his supposed testimony with the authority of great age. As Albert Baumgarten (1981, 77–82) and others have pointed out, Philo was a Euhemerist and his Phoenician history is thoroughly permeated by the beliefs, approaches and language of this tradition. His reference to Sanchuniathon finding ancient hidden knowledge in Egyptian temples bears more than a passing resemblance to Euhemeros' own account of his discovery of ancient hidden knowledge inscribed on a golden pillar in a temple on a distant, invented isle. These are clearly the tropes of early esotericism and fringe scholarship rather than a reliable guide to what writing and literary culture did or did not exist in the central Levantine coast at the end of the second millennium BC.

A final reference to Phoenician historical literature from the Late Bronze/Early Iron Age transition is so unlikely as to scarcely be worth mentioning: the fourthcentury AD Latin story *Ephemeris belli troiani* claims to be a Latin translation based on an eyewitness account of the Trojan War found during the Neronian period inscribed on tablets in the tomb of 'Dictys of Crete' and written in an unknown script that

<sup>28</sup> The Greek is: τοῖς ἀπὸ τῶν ἀδύτων εὑρεθεῖσιν ἀποκρύφοις Ἀμμουνέων γράμμασι συγκειμένοις ἅ δὴ οὐκ ἦν πᾶσι γνώριμα – *to the hidden texts found in the adyta of the temples of Amun, composed in letters which were not known to everyone*. In the first flush of excitement after the decipherment of Ugaritic, Otto Eissfeldt, a colleague at Halle of one of the decipherers, Hans Bauer, argued that Philo may in fact have been referring to alphabetic cuneiform inscriptions here (Eissfeldt 1939, 8–12, esp. 10–12). In his view, Philo's Phoenician-centric worldview made the usual Egyptian interpretation impossible, with the alternative being that these texts were written in a 'Phoenician' script, but not the one that was widely known. Alphabetic cuneiform fitted the bill, Eissfeldt believed, and in this case Ἀμμουνέων should be interpreted not as a reference to the Egyptian god Amun, but to the region around Mt Amanus – *i.e.*, the Kingdom of Ugarit.

Later on, Philo refers to a figure Eisirios – possibly a corruption of Osiris or perhaps of εἴς Σύριος (Baumgarten 1981, 216, n. 7, 232) – as the 'discoverer of the three letters'. This remains an enigmatic phrase, but Eissfeldt enthusiastically connects it to the three additional signs, I,i,, u and 1 added to the original 27-letter cuneiform alphabet some time after its creation (Eissfeldt 1939, 58–62). Eissfeldt's Ugaritian interpretations of Philo have not been widely taken up. Baumgarten (1981, 233–235) admits to not knowing what the 'three letters' are, but probably rightly deems the alphabetic cuneiform ones unlikely due to the great chronological separation involved.

Nero assumed to be Phoenician. This has rarely been seen as anything more than an elaborate literary frame-narrative, although Arthur Evans did wonder if it might preserve some memory of Linear B.29 Certainly there's no reason to treat it as a genuine source on the existence of Phoenician accounts of the Trojan War.

What does all this add up to? On the one hand, a number of tangential or late referencesto Phoenician historical writing, all of which are subjectto a significant degree of doubt. Individually, none can be taken as reliable. On the other hand, together they do point to at least a general belief in the existence of historical records in the Levant which stretched back well beyond those of the classical world. We could dismiss this as nothing more than a recurring literary topos, whereby historians appealed vaguely to the mysterious and ancient civilisations of the east to lend authority to their writings. But while they certainly contain distortions,30 they also preserve kernels of what seem like authentic historical knowledge. By and large Josephus's Phoenician history marries up with what we know from Assyrian and other sources. More persuasively, the Hebrew Bible both refers to and preserves signficant portions of comparable annalistic material from the earlyfirst millennium BC southern Levant, primarily in the books ofKings and Chronicles. Kings refers to various kinds of written material: 'the book of the chronicles of the kings of Israel' (1 Kings 14.19, 29 etc.), 'book of the chronicles of Solomon' (1 Kings 11.41), 'the book of the law' (2 Kings 22.8, 11) and diplomatic correspondence between Solomon and Hiram of Tyre (1 Kings 5.1–9). It also mentions writers working in the courts of Solomon and other kings.31

Of course, opinions differ on exactly how accurate the historical accounts of Kings and Chronicles are, what kinds of sources their authors had access to and when they date from. This is not the place for a detailed discussion, but most scholars agree that they are based on annalistic and other records stitched together and elaborated in places. Most see this writing tradition as beginning around the time of the United Monarchy in the eleventh to tenth century,32 although Alan Millard has argued for an unbroken literary tradition stretching back into the Bronze Age.33 While I have some doubts as to the character of the United Monarchy, the existence of such a writing culture seems to me perfectly plausible, at least in some parts of the region. It is certainly not stretching credulity to imagine a well-developed set of writing practices using the linear script and perishable materials in the southern Levant by around the tenth century, including historical records, economic and administrative materials and diplomatic correspondence. In other words, much the same kind of material we have from Ugarit in the thirteenth century. This argument obviously connects, then, with

<sup>29</sup> Evans 1909, 108–110, quoted and discussed by Bennett 1991, 566–567. See also Quinn 2018, 147.

<sup>30</sup> See Boyes 2012 for discussion of one aspect of distortion in Josephus.

<sup>31</sup> Millard 2010, 155.

<sup>32</sup> *e.g.* Na'aman 1999; Halpern and Lemaire 2010. Sanders (2008, 101–102) thinks there may have been 'annals of Solomon' but that they could have been written by 'Phoenician' rather than local scribes. It is worth noting that some historians and archaeologists have doubted the existence of the United Monarchy itself, notably Finkelstein (1996; 2010). See also Lemche and Thompson 1994.

<sup>33</sup> Millard 2010, 157.

the discussion in Chapter 4 about the nature and extent of linear alphabetic writing in the Late Bronze Age, and whether its sporadic and idiosyncratic survivals are representative of the bulk of ancient practice. My conclusions are essentially the same in both cases: despite the absence of conclusive evidence, it seems more likely than not that writing was being used extensively for a wide range of functions – mostly, but not exclusively, official or elite – broadly comparable to those we see at Ugarit. It's only the lack of monumental writing on durable surfaces such as stone which give the impression of an absence of writing. This might seem like arguing from silence, but we know for a fact that this is what occurred in northern Syria with Luwian. The Luwian script is wholly unattested for a long time following the fall of the Hittite Empire but re-emerges in the monuments of the Syro-Hittite politites of the Early Iron Age. The only possible explanation for this continuity across the lacuna is that writing continued on materials which do not survive.

If annals, diplomacy and other genres of writing using linear alphabetic script existed in Israel/Palestine around the eleventh–tenth centuries and possibly earlier, it seems very likely they also did in what is now Lebanon. In other words, we have to contend with a diverse and well-developed writing culture significantly before the Byblos royal inscriptions. There is relatively little evidence for violent upheaval in Phoenicia in the Late Bronze/Early Iron Age transition, so this is even more likely to be an unbroken development of what came before than is the case further south. The choices regarding writing practice made by writers and their patrons in Late Bronze Age polities in the southern and, especially, central coastal Levant thus shaped the writing culture of the Early Iron Age.

These choices didn't take place in a vacuum, but were shaped and informed by the socio-political climate of the day and by writers' thoughts and feelings regarding other writing practices with which they were in contact. Ugarit was part of the same cultural, political, economic and scriptal network. At least some of these Phoenician and southern Levantine writers would have been in contact with Ugarit and perfectly well aware of how writing was practised there. We can see this from the spread of alphabetic cuneiform to sites like Sarepta, Beth Shemesh or Kamid el-Loz, or from diplomatic letters from Phoenicia written in alphabetic cuneiform, such as KTU 2.40 and 2.44. Their decisions about how to carry out their own practice and what to pass on to their successors would have been shaped by their views on what they felt had worked or not in Ugarit, just as they would have been by their responses to other writing practices and scripts with which they were in contact, such as logosyllabic cuneiform, Egyptian hieratic, Cypro-Minoan and so on. This is, perhaps, a slight and indirect measure of influence and legacy for the writers of Ugarit, but it is a real one, and an important consequence of an interlinked network view of culture and writing practices across the region.

### Chapter 13

### Conclusion: the social context of writing practices at Ugarit

Over the last twelve chapters we've explored numerous aspects of the relationship between writing and its social and cultural contexts at Ugarit. Central to my methodology is that writing is a form of social practice and is thus deeply intertwined with other forms of practice, with material culture, with belief systems and ideology. These relationships are ever-changing, reproduced, negotiated and transformed by human agency operating within and against the social structures in which people find themselves. I argued in Chapter 2 that to really understand writing practices in a given society, it's not enough to just understand the 'writing system', the structure and rules of how the script operates. Rather, we need a holistic approach, an explicitly material one which I've called an 'archaeology of writing'.

Throughout this volume, however, this methodological ideal has found itself colliding with a landscape of data and publication which is often not conducive to answering the kinds of questions it asks. For example, I advocated a detailed and multi-level contextual analysis of inscribed objects that would enable us to understand not just their own materiality, but their relationships with other, potentially uninscribed, items and with the physical structures of Ugarit. As with the archaeology of other kinds of material culture, this would enable us to understand in more detail how they were used, for what purposes and with what other items; it would allow us to explore object biographies – not just production, which tends to dominate discussions of inscriptions, but distribution, exchange, use and disposal. This detailed 'immediate' context could then be combined with an understanding of the cultural situation and contingent historical circumstances to begin reconstructing the mesh of interrelationships between writing and the other practices and objects of Ugaritian culture; to begin to understand which choices people had made and why, not just in an abstract, disembedded way but in a manner that fundamentally recognises that this agency was exercised in the context of a specific time, place and set of circumstances.

The problem has been that all too often at Ugarit, this methodology struggles at the first stage – the detailed analysis of the immediate context is necessarily vague or even impossible. Often, we don't know exactly where inscribed items were found. Orif we do, the stratigraphic information islacking,rendering our understanding flat and atemporal. What limited amount we know about the inscribed objects dwarfs the extremely minimal – and very uneven – information we have about the rest of Ugarit's material culture assemblage, preventing us from establishing the crucial relationships between inscribed items and other objects, and thus from properly understanding the use of space in the city and the kinds of other activities that might have been associated with writing. What's especially frustrating is that often this information was gathered during excavation; it is the lack of priority accorded to uninscribed archaeological material in study and publication over several decades which has led to this situation. Ugarit is an excellent case study of the power of path-dependence in archaeology and epigraphy: a continuous focus on the most popular area of research across nearly a century – in this case the tablet corpus – has led to a situation where answering other kinds of questions, or approaching it in another way, becomes incredibly difficult.

As a result, this volume asks a lot of questions which more often than not it has only been able to answer in broad terms or with speculation, if at all. What was the demographic breakdown of writing at Ugarit? What other professions were tied into writing practices, broadly conceived? What did the social landscape of writing look like, both within and outside the city? How did those not directly implicated in Ugarit's writing culture think about it and encounter it? The speculative and openended responses to these questions may at times have been frustrating. I would be lying if I pretended it had not been so for me too on occasion. But we should not allow the lack of firm answers at this stage to overshadow either the value of the methodology and the questions themselves, or the conclusions that I think we can reasonably draw where the data is strongest. In particular, there is a good deal we can say with some confidence about the relationship between writing, elite identity and Ugarit's relationship with international political and economic networks in the thirteenth and twelfth centuries BC.

#### **The place of writing in Ugaritian society**

Histories of Ugarit have tended to focus on the elites. This is neither unusual nor unexpected given the kind of evidence that typically survives from the ancient world in general and from Ugarit in particular. Especially when research tends to concentrate on written material, quite a small segment of society and their activities can attain an undue prominence in our thinking and assumptions. This is not to say that Ugarit's political and mercantile elite were not important – they self-evidently were. And writing was very clearly an important tool they used in going about their business. But literacy may have been quite limited even within this group. Within the city as a whole, it was certainly very much a skill possessed by a small minority. Few would dispute this in factual terms, but Ugarit's 'scribal culture' and political history still account for the great majority of research into the kingdom. The lives and practices of a great majority of the population – especially those living outside the capital – are tremendously neglected.

Data is scarce and this is, ultimately, a book about the social context of writing practices, so there's only so much we can do here to redress that imbalance. But where possible I have attempted to clarify how much, and in what manner, writing was present in the lives of ordinary Ugaritians. The results are paradoxical, with writing practices both more and less pervasive than we might assume. On the one hand, I have explored the possibility – and I think the likelihood – that the practice of writing and inscribed objects themselves were not confined just to the capital and its ancillary sites of Ras Ibn Hani and Minet el-Beida, where tablets have been found. It seems very probable that writing was present in the villages and estates of the hinterland in the form of itinerant administrators, perhaps local notaries, and travelling literate merchants and soldiers. It is also quite likely that some people from the countryside came into contact with writing practices when they travelled to the metropolis, most plausibly in the form of administration taking place at gates, such as in the vicinity of the West Tablet Store. There is also the possibility of people in the hinterland being involved either in writing directly or in the wider networks of practice relating to it – as literate craftspeople or gatherers of materials for producing writing materials. Writing practices, we must recognise, include not just the writer themselves, but people who gather clay, beekeepers who supply wax, woodcutters and the craftspeople who put together writing boards. Our attempts to identify literate women offered no direct evidence, but the people involved in these wider 'supply chains' relating to the practice of writing have the potential to be much more diverse than the elitist clique of men often evisaged for the capital's scribes. Finally, we can also observe that some sort of route existed for social mobility from the hinterland to the inner circles of the capital's literate elite, as shown by ʾIlimilku, who achieved the high office of *ṯ*ʿ*y* and is today the most famous writer in Ugarit because of his mythological works but identified himself with reference to his origins in the village of Šubbanu.

On the other hand, as far as we can tell, writing was not an especially visible part of the fabric of the city or the hinterland. Although widely distributed across the tell, inscribed material was shut away in tablet stores on inaccessible upper storeys, or hidden away in dark and restricted temples. When inscribed objects were exhibited, as with dedicated objects orstelae commemorating sacrifices, they still fall well short of what we might call 'public' or monumental writing. It's not clear who the intended audience for these inscriptions was, but their character, size and placement don't suggest general reading was a high priority. These objects are solely restricted to religious dedicatory contexts; there are no signs at all of writing as part of a public political agenda, either in the capital or the wider kingdom. As we've discussed, this lack of interest in monumental and political inscription is in sharp contrast to many of Ugarit's neighbours, and exemplifies the different interests and choices made by its elites in their use of writing.

#### **Writing and elite ideology**

Writing was political at Ugarit. This much is very clear. From its earliest arrival in the city, writing had been associated with other societies and especially with the great international powers. We can perhaps suggest a difference between Egyptian writing – which seems to have been closely associated with Egypt specifically, and which maintained a prestige at Ugarit throughout the thirteenth and early twelfth centuries – and logosyllabic cuneiform, which may have betokened Late Bronze Age globalism in general, rather than Babylonia or Assyria more particularly. Ugarit's elites had a complex relationship with logosyllabic cuneiform, which likely informed their decision to create their own script and the form they chose for this vernacular writing.

It's very probable that cuneiform writing was practised in Ugarit well before the mid-fourteenth century when the first surviving documents hail from. It was a practical necessity of involvement in international networks and something that rulers used in markers of their personal identity and office, such as Yaqaru's royal seal. This didn't change with the incorporation of Ugarit into the Hittite Empire, although something seems to have occurred to 'clean out' earlier records from before this time. Rather, the key time of change seems to be the mid-thirteenth century, around the time of the rise to power of ʿAmmiṯtamru II. We can observe major upheavals in several aspects of Ugaritian life at this time. A destruction – presumed to be an earthquake – affected buildings all over the site, prompting extensive changes to the urban fabric. These included the construction of new houses and the expansion of some existing ones at the expense of public space. The palace seems to have been restructured and the construction of the royal residence at Ras Ibn Hani may also be a result of this. As we discussed in detail, there were also important implications for the religious sites of the city with the destruction of the temples of Baʿlu and Dagan, only one of which was rebuilt.

While this was going on, ʿAmmiṯtamru was apparently adopting a more assertive line in foreign policy. From the mid-thirteenth century onward, a succession of Ugarit's kings can be seen to be at loggerheads with their overlords over everything from tribute to military assistance. Initially at least, it's unclear how much of this change was down to a general sense of unhappiness with imperial control in Ugarit and how much down to ʿAmmiṯtamru's own combative and domineering personality, on clear display during his divorce proceedings. But ʿAmmiṯtamru seems to have set a tone that his successors were happy to continue, and if Tugendhaft is right in his interpretation of the themes of the Baʿlu Cycle, then a more general cynicism and scepticism towards international political networks does seem to have existed among the kingdom's elites. It isn't too much of a stretch to see this as resulting both from

a period of strife and introspection inherent in choices about how to rebuild the city and from the effects of famine and the general fraying on international relations as the end of the Bronze Age approached.

This is the historical and cultural context in which the choice to adopt alphabetic cuneiform has to be understood. As a marker of a distinctive local identity it makes a lot of sense in this climate. It seems to have been a top-down imposition, perhaps adopting and adapting a linear script already in use on perishable materials at Ugarit or nearby. As an alphabetic script it has clear resonances with Levantine culture as set against the global culture symbolised by logosyllabic cuneiform. But we should not misunderstand the nature of this nativist, vernacular statement. The new script was used almost entirely for internal purposes and is deeply pragmatic in its use of the cuneiform style and document formats. In doing so, it adheres to the deeply ingrained idea of the prestige of cuneiform writing and its appropriateness in official contexts, as well as being able to be used efficiently alongside continuing logosyllabic writing practices that had to be maintained for Ugarit to continue to play its role on the international stage. From the restricted visibility of writing that we've discussed, and the visual similarity between the two varieties of cuneiform to the untrained eye, it's clear that alphabetic cuneiform was not a populist appeal aimed at convincing the masses that their rulers were in touch; rather it was a message from those rulers to themselves, a solipsistic demonstration of who they consider themselves to be, hedged against the limitations under which they were forced to operate.

### Bibliography


Dunand, M. (1945) *Byblia Grammata: Documents et recherches sur le développement de l'écriture en Phénicie*, Beirut.

Eidem, J. (2002) 'The clay they wrote on – Old Babylonian letters as artefacts' in Al-Gailani Werr, L., Curtis, J.E. and Martin, H. (eds) *Of Pots and Plans: papers on the archaeology and history of Mesopotamia and Syria presented to David Oates in honour of his seventy-fifth birthday*, London, 74–81.

Eissfeldt, O. (1939) 'Ras Shamra und Sanchuniathon', *Beiträge zur Religionsgeschichte des Altertums* 4.


Grewal, D.S. (2008) *Network Power: the social dynamics of globalization*, New Haven.

Grivelet, S. (2001) 'Introduction', *International Journal of the Sociology of Language* 150, 1–10.

Gröndahl, F. (1967) *Die Personennamen der Texte aus Ugarit*, Rome.

Guillaume-Pey, C. (2018) 'Des rebelles armés de feuilles blanches aux déchiffreurs de pierres. Appropriation de l'écriture chez les Sora et d'autres groupes tribaux de l'Inde (Adivasi)', *Terrain* 70, 62–81.

Guillaume-Pey, C. (forthcoming) 'A script "good to drink". Invention of an alphabet and emergence of a religious movement among the Sora (central India)' in Boyes, P.J., Steele, P.M. and Elvira Astoreca, N. (eds) *The Social and Cultural Contexts of Historic Writing Practices*, Oxford.

Güterbock, H.G. (1993) 'Sungod or King?' in Mellink, M.J., Porada, E. and Özgüç, T. (eds) *Aspects of Art and Iconography: Anatolia and its neighbours. Studies in honor of Nimet Özgüç*, Ankara, 225–226.

Hall, J.M. (2002) *Hellenicity. Between ethnicity and culture,* Chicago.

Hallo, W.W. (1996) *Origins: the ancient Near Eastern background of some modern Western institutions*, Leiden.

Halpern, B. and Lemaire, A. (2010) 'The composition of Kings' in Lemaire, A., Halpern, B. and Adams, M.J. (eds) *The Books of Kings: sources, composition, historiography and reception*, Leiden, 123–153.

Halverson, J. (1992a) 'Goody and the implosion of the literacy thesis', *Man* 27, 301–317.

Halverson, J. (1992b) 'Havelock on Greek orality and literacy', *Journal of the History of Ideas* 53.1, 148–163.

Hamidović, D. (2014) 'Alphabetical inscriptions from the Sealand', *Studia Mesopotamica* 1, 137–155.

Hamilton, G.J. (2006) *The Origins of the West Semitic Alphabet in Egyptian Scripts*, Washington.

Hamilton, G.J. (2014) 'Reconceptualizing the periods of early alphabetic scripts' in Hackett, J.A. and Aufrecht, W.E. (eds) '*An Eye for Form'. Epigraphic essays in honor of Frank Moore Cross*, Winona Lake, 30–55.

Haring, B. (2019) 'Ancient Egypt and the earliest known stages of alphabetic writing' in Boyes, P.J. and Steele, P.M. (eds) *Understanding Relations Between Scripts II: early alphabets*, Oxford, 53–67.

Harmanşah, Ö. (2013) *Cities and the Shaping of Memory in the Ancient Near East*, Cambridge.

Harris, R. (1963) 'The organization and administration of the cloister in ancient Babylonia', *Journal of the Economic and Social History of the Orient* 6.2, 121–157.

Harris, W.V. (1989) *Ancient Literacy*, Cambridge, Massachusetts.

Hasel, M.G. (2016) 'The archaeology of destruction: methodological desiderata' in Ganor, S., Kreimerman, I., Streit, K. and Mumcuoglu, M. (eds) *From Sha'ar Hagolan to Shaarayim: essays in honor of Prof. Yosef Garinkel*, Jerusalem, 205–228.

Haugen, E. (1966) 'Language, dialect, nation', *American Anthropologist* 68, 922–935.

Havelock, E.A. (1963) *Preface to Plato*, Cambridge, Massachussetts.

Havelock, E.A. (1982) *The Literate Revolution in Greece and its Cultural Consequences*, Princeton.

Havelock, E.A. (1986) *The Muse Learns to Write. Reflections on orality and literacy from antiquity to the present*, New Haven.

Hawkins, J.D. (2009) 'Cilicia, the Amuq, and Aleppo: new light in a Dark Age', *Near Eastern Archaeology* 72.4, 164–173.

Hawley, R. (2008) 'On the alphabetic scribal curriculum at Ugarit' in Biggs, R.D., Myers, J. and Roth, M.T. (eds) *Proceedings of the 51e Rencontre Assyriologique Internationale*, Chicago, 57–67.

Hawley, R., Pardee, D. and Roche-Hawley, C. (2015) 'The scribal culture of Ugarit', *Journal of Ancient Near Eastern History* 2.2, 229–267.

Heier, S. (forthcoming) 'The visibility of runic writing and its relation to Viking Age society' in Boyes, P.J., Steele, P.M. and Elvira Astoreca, N. (eds) *The Social and Cultural Contexts of Historic Writing Practices*, Oxford.

Helck, W. (1976) 'Ägyptische Statuen im Ausland. Ein chronologisches Problem', *Ugarit-Forschungen* 8, 101–114.

Heltzer, M. (1976) *The Rural Community in Ancient Ugarit*, Wiesbaden.

#### *Bibliography* 293


Loretz, O. (1994) 'Mari, Ugarit und Byblos' in Acquaro, E. *Biblo. Una città e la sua cultura*, Rome, 113–124.


Skelton, C. (2010) 'Re-examining the Pylos Megaron Tablets', *Kadmos* 48.1–2, 107–123.

