Frauke Albersmeier **The Concept of Moral Progress**

## **Practical Philosophy**

Edited by Herlinde Pauer-Studer, Neil Roughley, Peter Schaber and Ralf Stoecker

# **Volume 24**

# Frauke Albersmeier **The Concept of Moral Progress**

ISBN 978-3-11-079883-8 e-ISBN (PDF) 978-3-11-079891-3 e-ISBN (EPUB) 978-3-11-079901-9 ISSN 2197-9243 DOI https://doi.org/10.1515/9783110798913

This work is licensed under the Creative Commons Attribution 4.0 International License. For details go to http://creativecommons.org/licenses/by/4.0/.

Creative Commons license terms for re-use do not apply to any content (such as graphs, figures, photos, excerpts, etc.) not original to the Open Access publication and further permission may be required from the rights holder. The obligation to research and clear permission lies solely with the party re-using the material.

#### **Library of Congress Control Number: 2022944083**

#### **Bibliographic information published by the Deutsche Nationalbibliothek**

The Deutsche Nationalbibliothek lists this publication in the Deutsche Nationalbibliografie; detailed bibliographic data are available on the Internet at http://dnb.dnb.de.

© 2022 with the author(s), published by Walter de Gruyter GmbH, Berlin/Boston. This book is published with open access at www.degruyter.com. Printing and binding: CPI books GmbH, Leck

www.degruyter.com

D61

## Acknowledgments

This book is a revised version of my dissertation of the same title, which I defended at the Heinrich Heine University Düsseldorf in January 2020.Iam indebted to Christoph Kann for his excellent supervision and to Frank Dietrich, who acted as the second thesis referee. For critical feedback on this work and many helpful discussions of related topics, I thank, first and foremost, Alexander Christian. I have also benefitted from conversations with David Hommen, Oscar Horta and Gottfried Vosgerau. I thank the German Research Foundation for funding the research project in which I have been employed during my doctoral studies and the University and State Library Düsseldorf for supporting the open access publication of this work. For the many ways in which each of them has supported and encouraged me, I am deeply grateful to my parents and my grandparents.

## Contents

#### Introduction 1

#### Methodological Preliminaries 7


#### Moral Progress: Conceptual Commitments, Pragmatic Expectations 25


#### Ethics and the Idea of Moral Progress 44


#### The Phenomenon of Moral Progress 98


#### Moral Progress and Moral Motivation: Improvement as a Fetish? 180


Conclusion 229

Bibliography 235

Index 245

## Introduction

It is part of the self-image of members of many cultures that they belong to progressed societies. Industrially developed liberal democracies are typically regarded (at least from within) to be the result of progress in many different realms technological, scientific, and political. Certain groups within these societies do not merely view themselves as progressed but also mark their political outlook as progressive, associating certain socio-political agendas with the perpetuation of a presumably progressive project. While such agendas typically revolve around distinctly moral considerations pertaining to the protection of the interests of individuals, the case for supposedly politically progressive objectives is rarely cast in the language of "moral progress."

Many of us welcome progress with respect to values such as equality, social inclusion, political participation. Some moral agents are actively involved in working toward these changes and in battling what they see as moral evils, such as the failure to aid people fleeing hunger and terrorism, the practice of animal experimentation, or the destruction of ecosystems. But we rarely address any of these causes as manifestations or instances of a need for moral progress. Talk of "moral progress" is notably absent from moral discourse. Until recently, this was true of discourse in moral philosophy, and it continues to be the case for public discourse. This absence is puzzling because as moral agents, we should be expected to be interested in determining how we are doing morally, how we are doing compared to those that have come before us and lived in less progressed circumstances, and whether our efforts contribute to a more comprehensive development toward the better or whether we are treading water. Yet, these questions are not discussed in terms of moral progress.

There are several possible reasons why appeals to "moral progress" are often eschewed. An aversion to "moralizing" debates might play a role, as well as the intention to avoid being presumptuous in issuing judgments of moral progress (or moral judgments more generally, for that matter), but also an uncertainty about whether specific achieved changes constitute moral progress or whether there has been moral progress, all things considered. Maybe moral agents in general have a bleaker outlook: maybe they doubt that further moral progress is even possible or feel that progress has become obsolete and our mission as moral agents is rather to form "the resistance to the perpetual danger of relapse" (Adorno 2003, 145). Rather than holding reservations against the idea of moral progress or applying the notion of moral progress in normative discourse, the agents participating in this discourse might simply not see any need to speak of moral progress. Those who are working to advance certain causes for moral reasons care about these causes rather than progress—or moral progress—as such. Furthermore, where there is need to mark a change for the morally better explicitly, it might be the case that the notion of progress itself—where 'progress' remains unspecified—is thought to suffice, because progress is already regarded as somewhat morally loaded. But the idea of moral progress is not superfluous or redundant.We miss something if we never ask whether change for the better is moral progress and if we never think of our moral goals in terms of moral progress.

The relative absence of the concept of moral progress¹ from our moral discourse is also puzzling because, supposedly, the idea of moral progress is highly important motivationally for moral agents. Michele Moody-Adams claims as much when she says that it is "a necessary presupposition of action for beings like us" (Moody-Adams 2017, 153). In fact, the present volume will expand on existing views of the motivational significance of belief in the possibility of moral progress. Even in its most abstract form, the idea of moral progress—moral improvement, wherever it may lead—might be a proper object of a moral agent's desire—something to strive for, despite its abstractness. This idea will be vetted in the final chapter, which will carve out an even more positive motivational role for the concept of moral progress than it has been assigned so far.

If we assume that moral agents do care about how they do morally, when we observe that they do fight for moral causes and contrast this with the lack of explicit appeals to the notion of moral progress, we find that the concept of moral progress plays a largely implicit—but not insignificant—role in our moral lives. The aim of the present investigation is to make the concept of moral progress explicit. Uncertainty about the existence or the possibility of moral progress can only be removed once uncertainty about the content of the concept has been removed. Judgments about the occurrence of moral progress serve our moral orientation by marking accomplishments and goals (§ 2.3.). They can be better articulated and justified based on a clear grasp of what they attribute to the evaluatee.

The little attention that has been given to moral progress thus far contrasts with the way the more general idea of progress has been treated. Progress in the general sense is a recurring theme in the history of philosophy, perhaps associated foremost with the era of Enlightenment—where it was proposed as a viable possibility and an objective (Condorcet 1963) as well as criticized (Rousseau

In the following, it will often be necessary to distinguish between the phenomenon of moral progress, the concept, and the term. I will mark the concept either by using small caps (moral progress) or by speaking of the concept of moral progress, and I will use single quotation marks for the term 'moral progress.'

1965). It has received attention as a historiographical idea in philosophy (Bury 1920) as well as criticism for serving as a self-applied accolade (e.g., Lasch 1991) and today, some find it necessary to come to its defense and express opposition to a diagnosed contemporary skepticism toward progress (Pinker 2018). Progress in general is, however, not the topic here. We will be concerned with moral progress, a concept that continues to be an underexplored topic in philosophy.

There are few book-length investigations into the topic of moral progress and the ones that exist focus on other aspects than the development of a detailed account of the concept moral progress. These works share an interest in reconciling the idea of moral progress with a naturalistic picture of (human) moral agents and the preconditions of morality. Peter Singer might be credited with reviving interest in the topic of moral progress with his book The Expanding Circle (2011), first published in 1981. Therein, he proposed the idea that moral progress consists foremost in the enlargement of the domain of moral concern, i.e., in recognizing more beings as worthy of moral consideration. He also dealt in length with the compatibility of the idea of moral progress with the fact that human moral capacities are products of evolution. This theme is picked up in recent additions to the still limited debate about moral progress. Philip Kitcher is also concerned with a naturalist account of moral progress in his book The Ethical Project (2011) and Allen Buchanan and Russell Powell even entitle their contribution The Evolution of Moral Progress (Buchanan and Powell 2018). Their perspective resonates with that of Singer in that they are crucially concerned with arguing for the flexibility and susceptibility of human morality for purposeful change for the better given its roots in our evolutionary prehistory. Kitcher, likewise, seeks to give a naturalistically minded "how possibly"-explanation of the occurrence of moral progress (Kitcher 2011, 12). They each take different approaches to characterizing the concept of moral progress. For Singer, it is an outgrowth of his fundamental conceptual commitment to seeing morality as characterized by disinterested justification, whereas Kitcher characterizes morality as a set of rules which serve social order and thus, moral progress as an increase in functionality. Buchanan and Powell choose a "bottom-up" approach and provide a list of paradigmatic types of moral progress (such as improvements in moral concepts (Buchanan and Powell 2018, 54), or "understanding of virtues" (Buchanan and Powell 2018, 55)).

In works on moral progress, the clarification of the very concept does not often take center stage. Rather, more specific questions are addressed, for instance, about the scope and criteria of or obstacles to moral progress, the application of the concept to particular phenomena, or the role of the idea of moral progress in specific historical authors, e.g., in Hume (Cohen 2000; A. C. Baier 1991), Kant (cf., e. g., Kleingeld 1999, Lindstedt 1999, Anderson-Gold 2001, Stroud 2005, and Moran 2012), or Whitehead (Bellantoni 2004). Furthermore, there are controversies regarding the relevance of overcoming culpable or inculpable ignorance as a mode of moral progress (Moody-Adams 1997, Moody-Adams 1999, and Wieland 2017), the continuity (Moody-Adams 1999) or potential for discontinuity (Stokes 2017) in moral discourse, the relevance of moral philosophy (Nussbaum 2007) and even metaethical insights (Elzanowski 2013) for furthering moral progress, and the capability of different metaethical perspectives to make sense of moral progress. It is sometimes argued that the phenomenon of moral progress is a challenge to relativism (Prinz 2007; Coliva and Moruzzi 2012), it is seen as a desideratum for constructivist views to develop accounts of moral progress (Arruda 2017), and moral realists sometimes almost casually mention moral progress as an aspect of arguments in their favor (Smith 1994; Enoch 2011), but its role is seldom spelled out in detail. One exception here is Michael Huemer's argument from progress toward liberal values to moral realism via an inference to the best explanation (Huemer 2016). Anti-realists typically accept that the appearance that there has been moral progress poses a challenge to their view but seek to explain the supposed implication of moral realism away, for instance by pointing to non-moral standards of evaluation (Prinz 2007) or to the example of anti-realism in philosophy of science and its responses to the no miracle-argument (Wilson 2011).

Bold claims either to the effect that moral progress does not exist (Posner 1998) or that an unprecedented level of moral progress has been achieved at the current state of civilization (Lachs 2001) typically invite criticism, provoking opponents to point out reasons for modest hope for progress (Nussbaum 2007) and the overlooked downsides of supposed progress, respectively (Pokrovsky 2001, Light 2001, Schmidt 2001, and Willett 2001). The phenomenon of disagreement about the current stage of progress is mirrored in a conceptual controversy over the scope of judgments concerning moral progress, where there is disagreement regarding whether moral progress can be attributed to circumscribed realms of morality when it entails moral regress in other realms (Moody-Adams 1999; Dixon 2005).

Approaches to characterizing moral progress as such often appear to take a preliminary or skeptical approach to the possibility of defining the concept. This is due, typically, to an awareness of the challenge to find an account that is agreeable from different normative (more so than metaethical) perspectives. For instance, Dale Jamieson characterizes as a "naïve conception" his definition in terms of improvements in states of affairs and increases in the prevalence of right acts (Jamieson 2002). Others decidedly substitute material normative conditions (Macklin 1977) or a "proxy property" as a criterion of moral progress (Evans 2017) for an explicit definition. Proposals of provisional definitions and conceptually uncommitted replacements for definitions bespeak the need to make moral progress the topic of a philosophical conversation, while at the same time showing doubts about the possibility of defining what that conversation is about to the satisfaction of all participants. The present investigation aims to move past this ambiguity. If moral progress is a moral phenomenon, we should strive to clarify the concept that applies to it. A clearly stated, plausible, argued conceptual account of moral progress that will be adaptable to diverse normative perspectives is what is sought here. Based on the proposal for such an account, the present investigation will then also inquire into the moral worth of a concern for moral progress.

Chapter 1 will lay out the methodological rationale for the subsequent investigation. It will problematize the apparent congeniality of typical ways of approaching the topic of moral progress to the traditional method of conceptual analysis and portray (broadly) Carnapian explication as a viable alternative for arriving at a definition of moral progress.

Chapter 2 will disclose the conceptual commitments regarding progress and morality presupposed in this work. It will offer a characterization of morality that revolves around the practice of making judgments, moral awareness, and moral considerability, and it will address the way the capacity for moral progress and moral agency are often thought to interrelate. Chapter 2 will also discuss the most important pragmatic considerations pertaining to the concept of moral progress as well as some of the reservations with which this concept is met.

In Chapter 3, the most important ways in which specific normative theories could inform the explication of moral progress will be reviewed. Following this, widening the perspective beyond normative ethics toward classical political philosophy will generate helpful insights regarding the potential inherent evaluative conflicts involved in moral progress-judgments. Chapter 3 will also address the supposed connection between moral progress and moral realism and discuss on what grounds, if any, an investigation into the concept of moral progress can set concerns about the metaphysical implications of moral progress aside.

The explication of moral progress will be given in Chapter 4. It will be developed along theoretical considerations centering on moral progress made by individuals and by decomposing the question of what moral progress is into questions about the theoretical and practical changes required to justify talk of moral progress. It will be argued that three concepts of moral progress must be distinguished: the concepts of actual moral progress, dispositional moral progress, and impactful moral progress cover several types of moral accomplishments, while also distinguishing moral progress from mere ethical progress as well as from non-agent-related improvements in states of affairs.

Chapter 5 is then concerned with a challenge to the motivational value—and thus, practical relevance—of the concept of moral progress. Drawing on a debate in metaethics, an objection to the idea that moral agents ought to desire moral progress will be developed: the objection that harboring such a desire would fetishize moral progress. The refutation of this objection will reveal the special moral worth that attaches to the motivation to bring about moral progress.

## 1 Methodological Preliminaries

When the concept of moral progress is under consideration, with what kind of subject matter are we concerned to begin with and how should it be approached? This first chapter² will lay out the methodological rationale of the intended investigation into the concept of moral progress. Philosophy claims a special competence with the conceptual work that needs to be done before an attempt at an answer to a question such as "Has there been moral progress?" can even be made. But what is the object we are dealing with in working on the concept of moral progress and what is a philosophically viable way to engage with it? Is our primary aim the clarification or rather the revision of the concept, and how exactly ought such clarification or revision be achieved? When investigating the concept of moral progress, are we dealing with a mental representation, an ability or disposition, a set of socially embedded rules or rather an abstract object of some kind? If we seek to propose a novel concept of moral progress, are we thereby inevitably changing the topic—or might, alternatively, our concepts turn out to be immune to our efforts to transform them?

The methodological approach laid out in this first chapter is not specific to explicating a moral concept. It does not take its starting point in moral semantics but, on the contrary, will focus on a method originally introduced to amend scientific terminology (i.e., the terminology of natural science, not ethics). For the most part, the methodological remarks made here will be concerned with philosophical work on concepts generally, and moral progress will be regarded as just one such concept.

## 1.1 Moral Progress, Intuitions, and the Limits of Analysis

In work on the concept of moral progress, the provision of examples of moral progress is often more than a way to illustrate the range of applicability of some working definition of moral progress. It is common initially to invite assent to classifications of particular historical or contemporary developments as instances of moral progress. The single most commonly cited instance, of course, is the abolition of chattel slavery in Britain and the U.S. in the 1800s (though the abolition of slavery in other countries often goes unmentioned). It often seems that assent to such classifications is sought in order to make the case that engaging with the topic is worthwhile at all. Motivating one's topic by pointing to spe-

Portions of this chapter are taken, with modifications, from Albersmeier (2020).

Open Access. © 2022 the author(s), published by De Gruyter. This work is licensed under the Creative Commons Attribution 4.0 International License. https://doi.org/10.1515/9783110798913-003

cific instantiations is not unusual, but when cases rather than theoretical considerations are the object of appeal, this might indicate that there is (an anticipation of) skepticism toward the topic at hand. The appeal of the very concept of moral progress seems to be intuitive rather than theoretical. Often, engagement with the concept is based on the compelling impression that some things have changed for the (morally) better—not on the concept's indispensability to theoretical ethical discourse. On the contrary, making the concept explicit in a way that is acceptable to different participants in the ethical discourse will prove difficult because of the complex theoretical ramifications of settling for any particular definition of moral progress. In principle, authors addressing the topic could also count on intuitive assent to a claim as general as "there has been moral progress." The fact that they do not might be interpreted as evidence that the viability of the concept itself—its non-emptiness or usefulness—is in question, so that such general claims have little significance for the issue at hand. It seems that a general skepticism must be overcome concerning the concept's applicability or coherence even for those who would agree that improving the world in some sense is a moral requirement. At least, the audience is assumed to be critical about the very concept of moral progress (cf. Buchanan and Powell 2018, 4–11). In such circumstances, what is sought are actual developments that apparently call for being classified as instances of moral progress by any reasonable observer.

Intuitions about particular instances of improvements in our collective moral lives may also be evoked in order to illustrate specific characteristics of progressive developments. For instance, the assertion that "I cannot own you, and you cannot own me, period. Only yesterday, it was otherwise" (Godlovitch 1998, 272) not only points out the relative recency of some moral accomplishments, but also conveys what Godlovitch takes to be the essence of moral progress: that it establishes new moral certainties (see § 3.4.).

The cases that are indicated by an appeal to intuition as well as typifications of such cases may also be relied on in a more extensive way, i.e., by using them as test cases for the definition of moral progress that is put forward. Processes like the abolition of slavery are often viewed as benchmark cases—instances that will have to be covered by an acceptable definition of moral progress. For instance, Buchanan and Powell recommend a "bottom-up" approach to characterizing moral progress that "begins by identifying paradigmatic instances of moral progress" (cf. Buchanan and Powell 2018, 45) and builds a theory on the typification of these specific instances.

When actual cases are relied on in this way, the approach to defining moral progress that is thus taken begins to resemble traditional conceptual analysis, a way of engaging with concepts that is now often seen as somewhat discredited. As it will be understood here, the term 'conceptual analysis' denotes both a process and its result. The desired result is a definition that decomposes a concept (the analysandum), e.g., moral progress, into concepts specifying singly necessary and jointly sufficient conditions for its application (the analysans) (Beaney 2018, § 6). The procedure by which this result is to be reached starts from a provisionally framed or hitherto accepted analysis (i.e., a definition), which is then confronted with hypothetical cases in order to adjust the proposed analysis so that it will exclude counterexamples. From a (some would say overly) simplified perspective, an analysis of moral progress would be completed when a (narrow) reflective equilibrium is reached between the general characterization of moral progress and the specifics of the particular cases we want to classify as morally progressive (cf. Daly 2010, 49). We could, for instance, start with an idea about moral progress like the one captured in the first disjunct of what Dale Jamieson calls the "naïve conception" of moral progress, the idea that "[m]oral progress occurs when a subsequent state of affairs is better than a preceding one, or when right acts become increasingly prevalent" (Jamieson 2002, 318). The first part of this characterization construes moral progress as an improvement in states of affairs. Upon pondering instances in which the state of affairs has improved, we might find this analysis over-inclusive. For instance, we might want to exclude changes that are not in any way related to changes in the beliefs and behavior of any moral agents (or so I shall argue in the following). On the other hand, we might equally grow uncomfortable with a rather Kantian analysis of moral progress that excluded any changes not brought about by the adoption of a certain behavior that is based on a grasp of moral duty. Some changes that seem to be clear instances of moral progress, such as the abolition of slavery, might have been possible because some stakeholders changed their behavior out of a recognition of only a legal duty, based on selfinterested economical reasoning, or, as has been argued, on a sense of "national honor" (Appiah 2010). In view of cases that are perceived as counterexamples to the proposed definition, the definition would have to be amended to fit the cases.

If the way conceptual analysis has just been described marks a paradigmatic way of understanding the method (cf. Margolis and Laurence 2011), it closely aligns the standard account of conceptual analysis to the use that has been made of it in the tradition of ordinary language philosophy. A prime example of an explicit account of the method thus understood can be found in Grice's description of his view of the nature of a conceptual analysis (cf. Nimtz 2012).

To be looking for a conceptual analysis of a given expression E is to be in a position to apply or withhold E in particular cases, but to be looking for a general characterization of the types of cases in which one would apply E rather than withhold it. And we may notice that in reaching one's conceptual analysis of E, one makes use of one's ability to apply and withhold E, for the characteristic procedure is to think up a possible general characterization of one's use of E and then to test it by trying to find or imagine a particular situation which fits the suggested characterization and yet would not be a situation in which one would apply E. If one fails, after careful consideration on these lines, to find any such situation, then one is more or less confident that the suggested characterization of the use of E is satisfactory. But one could not test a suggested characterization in this way, unless one relied on one's ability to apply or withhold E in particular cases. (Grice 1989, 174, emphasis in original)

What Grice describes here is the establishment of consistency between intuitions about particular instances and a general definition, i.e., a narrow reflective equilibrium. Grice even describes it as a trimming of the general characterization to judgments about particular cases, with the latter determining the former.

That conceptual analysis seeks a narrow reflective equilibrium is sometimes said to be an oversimplification. Defenders of the method typically describe it as a more comprehensive way of evaluating definitions, which has to take into account empirical knowledge, self-observation, uncontroversial background assumptions and more (Nimtz 2012, 219). Reliance on such additional information would make the equilibrium that is sought in a conceptual analysis wider. This reconstruction may render conceptual analysis a more comprehensive and thus potentially more widely applicable method, but it does so at the expense of the method's distinguishability from different approaches to engaging with concepts. Therefore, rather than supposedly increasing conceptual analysis's philosophical viability by construing it as a process that seeks a wide(r) reflective equilibrium, here, the method will be interpreted in the more restricted way outlined above that renders it a distinct process with a more limited objective (reaching a narrow reflective equilibrium). On this view, some approaches that purport to rely on conceptual analysis might have to be reclassified.³

In practice, engagement with the concept of moral progress is hardly ever restricted to the pursuit of such a narrow reflective equilibrium. Instead, more theoretical considerations enter into the evaluation of possible analyses of moral progress, such as metaethical parsimony, or a commitment to some kind of naturalism. This transgression of "pure" traditional conceptual analysis is unsurprising given the loss of confidence in the method's merits and reliabil-

For instance, when Frank Jackson, who has presented the most influential defense of "conceptual analysis" in recent years, admits of "sensible adjustments to folk concepts" (Jackson 1998, 47), the objects of analysis, based on theoretical considerations and empirical discoveries (Jackson 1998, 44), he is obviously no longer pursuing a narrow reflective equilibrium.

ity. Just as the idea that moral progress occurs in the world might strike some as naïvely optimistic, the method of conceptual analysis has come to strike many as relying on a both naïve and pretentious view of the merits of relying on one's own conceptual competence as it is reflected in intuitive responses to applicability questions. The fact that conceptual analysis is grounded exclusively in conceptual competence gives rise to the supposed paradox of analysis. This paradox concerns the question of how any given conceptual analysis can be both informative and correct. If an analysis of moral progress can be identified as correct solely based on conceptual competence, neither the provider of the analysis nor the evaluator should have learned anything. On the other hand, we surely constantly learn new things about the objects picked out by our concepts—and some new insights might change what used to be taken as a conceptual truth. The rejection of the very distinction between analytic and synthetic sentences entails that conceptual analyses cannot have any claim to a special kind of truth. If meanings and conceptual content depend on the content of our wider "web" of beliefs (Quine 1976, 132), there is no way of isolating an unchanging concept from changing beliefs about the subject matter picked out by the concept.

Viewed in this way, conceptual analysis's limited merit seems to lie in rendering explicit what would otherwise be an implicit and supposedly shared defeasible understanding. A problem for relying on conceptual analysis in this capacity is posed by the fact that understanding of conceptual content often does not seem to be so widely shared at all: there is considerable disagreement on analyses for philosophically interesting concepts. Given the prominent role of intuitive judgments about particular cases in conceptual analysis, this disagreement is often traced to disagreement in intuitions. Experimental philosophy's discoveries of disagreement in conceptual intuitions have been taken to undermine the central presupposition for expecting meaningful revelations through conceptual analysis, at least about "folk" (i.e., shared) concepts: the faith in the general agreement in intuitions about concept applicability (cf., e.g., Weinberg, Nichols and Stich 2001). This problem concerns moral progress as it pertains to morality itself.What, precisely, the concept morality indicates is itself a contested question. Consequently, the potential for disagreement is a manifest problem in the case of the concept of moral progress.When we approach the definition of moral progress as a matter of conceptual analysis, we face the threat of plenty of intuition stand-offs—for instance, between those who are convinced that "moral progress is an improvement in states of affairs" and those who intuit that "moral progress is the increase of the share of acts that are done out of a recognition of duty." The disagreement about the intension of the concept of moral progress is obvious and any overlap in its extension could—depending on the analyses of the concepts mentioned in the analysans—be purely contingent. If both parties feel confident that their definition captures perfectly what they want to be covered by the term 'moral progress,' we need more to work on than these intuitions in order to move forward. Conceptual analysis alone does not get us very far.

Further problems with the intuitions that go into analyses pertain to their ontology and evidential role. For one thing, it is initially unclear what intuitions are. When the conceptual analyst has the intuition that the establishment of women's suffrage constituted moral progress, is she experiencing an "intellectual seeming" (Bealer 1998), holding an ordinary belief or making an ordinary judgment (Williamson 2007), or possibly making a commitment to see things a certain way (Hommen and Albersmeier 2020)? There is broad agreement that intuitions are "psychological phenomena" (Cohnitz and Häggqvist 2010, 2). The question of precisely what kinds of mental states or episodes intuitions are, however, remains unsettled. This disturbing circumstance is one of the reasons why Herman Cappelen (2012) advises philosophers to abandon the bad habit of "intuition talk."⁴ A related problem concerns the evidential value of intuitions. Why is it that intuitions should have a say in the definitions of our concepts to begin with? Are these seemings, beliefs or commitments so related to the purported objects of analysis—concepts—that they can reliably inform us about them?

The question leads directly to a more fundamental one concerning the nature of concepts. What is the concept of moral progress? Is itamental representation, an abstract object such as a Platonic idea, or something else? If it were, for instance, an abstract object, how could it be that our intellectual seemings inform us about a concept (Goldman 2007)? The answer to the question about the ontology of concepts will be somewhat reverse-engineered in the following section. After noticing some further challenges to conceptual analysis, the next section will portray an alternative method for working on concepts and the final section of this chapter will present the view of the nature of concepts that aligns most smoothly with the ontological presuppositions of both methods, as they purport to apply to "concepts."

Another alleged flaw of conceptual analysis is its commitment to a descriptive theory of meaning or conceptual content. According to the conceptual analyst, the descriptions that people reflecting on their concepts can produce in their definition give the meaning or content of these concepts. If we defined moral progress as improvement in states of affairs, this description would give us the content of the concept moral progress or the meaning of the

The exposition of the problem concerning the ontology of concepts is adapted from Hommen and Albersmeier (2020, 72–73).

term 'moral progress.' But if we can, in the way the conceptual analyst claims, determine what the content of a concept is, different analyses would reveal different concepts. Accepting this runs contrary to the intuition that concept users share concepts despite holding differing, and even partly erroneous, associated beliefs. If one takes individuals citing different application conditions to be possessing different concepts, apparent disagreement about features of their subject matter is turned into mere verbal or conceptual disagreement (Laurence and Margolis 2003, 272, n. 27). The person who thinks that moral progress consists in improvement in states of affairs and the person who thinks that moral progress consists in change in the moral views of moral agents could not really disagree with each other about if and where moral progress exists in the world i.e., they would not make claims with "exclusionary content," claims that could not both be correct (cf. Khoo and Knobe 2018 for an account of how the two could still be thought to disagree)—but would merely be talking about different things, as they each would have a different concept of moral progress. This is so because the descriptive theory of meaning is internalist, i.e., it locates meaning "in the head" (Putnam 1975b, 144). Externalist views, motivated by thought experiments proposed by Kripke (Kripke 1980) and Putnam (1975b), focus on the idea that the meaning of terms—and content of concepts—depends in part on facts about the world which are not necessarily reflected in the descriptive accounts concept users can produce of their concepts based on intuitive understanding. In this way, externalists can account for substantive disagreement because they offer a view of how concepts are shared: concepts are shared by virtue of relating concept users to the same things in the world falling under the concept. If externalism is correct, the descriptive meaning we can state with an analysis may always get the nature of our concepts' referents wrong.

This is a problem given that conceptual analyses may sometimes be—and have at least traditionally been—advanced as real definitions of the referents of concepts (Beaney 2018, § 2). Note, for instance, how Grice passes from talking about linguistic intuitions to talking about the nature of the things to which he applies his term (this move is made within a clarification of the topics that distinguish philosophical from any everyday conceptual analyses):

One might (wanting a conceptual analysis) ask such a question as 'What is a battle?' 'What is a game?' […] But the nature of battles, games […] would not be regarded […] as falling within the subject matter of philosophy. So that to practice conceptual analysis is not necessarily to practice philosophy; some further condition or conditions must be satisfied for a piece of conceptual analysis to count as a piece of philosophy. (Grice 1989, 174–175)

The ontological leap from conceptual intuitions to the nature of things seems problematic in itself. Should we expect, for instance, our concept of moral progress to pick out a real moral kind in the world? If that is so, the concern about the evidential role of intuitions becomes even more pressing: why should we expect our intuitions to inform us correctly not only about our concept, but about the phenomenon that is moral progress?

All these issues are problems only as long as conceptual analysis is viewed as anything more than a clarificatory procedure. If conceptual analysis is understood as first and foremost an articulation of preexisting conceptual commitments—not the way to ultimately settle what a concept is supposed to pick out—it may be assigned an important but limited place within a more comprehensive examination and constructive engagement with a concept such as moral progress. In order to arrive at a theoretically interesting and useful concept of moral progress, one needs to move beyond analysis.

#### 1.2 Improving Concepts

Instead of viewing the prime task of an investigation of the concept of moral progress as that of providing an analysis of it, it will here be taken to be the task of developing an explication. Whereas an analysis is offered as an elucidation of one concept, an explication proposes an alternative concept to take the place of the original concept in specific contexts. Explication in this sense was characterized by Carnap as a process of replacing an inexact (pretheoretical) concept (or term) with a more exact one for the purposes of scientific theorybuilding:

explication consists in transforming a given more or less inexact concept into an exact one or, rather, in replacing the first by the second. We call the given concept (or the term used for it) the explicandum, and the exact concept proposed to take the place of the first (or the term proposed for it) the explicatum. (Carnap 1962, 3)

The explicandum may be taken from everyday language or a "previous stage in the development of scientific language" and the explicatum is to be "given by explicit rules for its use, for example, by a definition" (Carnap 1962, 3). The first step toward an explication is the selection of the concept that is to be replaced, i.e., the clarification of the explicandum. Here, Carnap envisions "[a]n indication of the meaning with the help of some examples for its intended use and other examples for uses not now intended," possibly supplemented by "[a]n informal explanation in general terms" (Carnap 1962, 4). Providing examples of cases in which a concept is supposed to apply along with a clarification in the form of a definition is precisely what conceptual analysis does—so the method seems to be a natural candidate for accomplishing the preparatory step toward an explication.

As far as the explication itself is concerned, the first among the requirements of adequacy for the explicatum it produces is similarity to the explicandum. "[H]owever, […] considerable differences are permitted" (Carnap 1962, 7)—a fact that makes for a stark contrast to conceptual analysis and has provoked a line of criticism of explication most powerfully articulated by Strawson: that explications amount to "chang[ing] the subject" (Strawson 1963, 506). This concern will be revisited in a moment.

The two most important requirements are exactness and fruitfulness. The former pertains to the characterization of the explicatum which is required to "introduce the explicatum into a well-connected system of scientific concepts" in an exact way, "for instance, in the form of a definition" (Carnap 1962, 7). Fruitfulness refers to the explicatum's usefulness "for the formulation of many universal statements" (Carnap 1962, 7). A final and clearly subordinate requirement is simplicity: the explicatum ought to be simple, but only to the extent to which the other requirements allow (Carnap 1962, 7). The relevant notion of simplicity here appears to be "syntactic simplicity" (Brun 2016, 1224) which, with respect to theories, "concerns the number and complexity of hypotheses" (Baker 2016, § 1) and in the given context must relate not to hypotheses but to the characterization of the explicatum—the number and complexity of the rules specifying its use.

The key structural difference between conceptual analysis and explication is that the former seeks to decompose one and the same concept, whereas the latter actually produces a second, new concept. The basic aspiration in proposing a conceptual analysis is to integrate all cases intuitively covered by the concept into the general account of the concept and not result in a multiplication of concepts.⁵ Conceptual analysis strives for simplicity on the level of the system of

The attempt to analyze the concept knowledge is the attempt to integrate all cases in which one would apply the term 'knowledge' into a general description of what it denotes, which does not cover any of the cases in which one would withhold the term. If there had been no commitment to cover all cases of knowledge and exclude all non-cases, Gettier's discovery of hitherto covered non-cases would not have caused much trouble. But the trouble it has caused persists. With respect to this situation, a recent suggestion precisely has been to abandon attempts at a conceptual analysis of knowledge in favor of different explications of the concept for different purposes (Olsson 2015). This suggestion is only comprehensible if a key difference between conceptual analysis and explication is the former's commitment to conceptual unity. This fact should not be overlooked due to the possibility that in the course of an analysis terms might well prove to be ambiguous. For instance, the analysis of a concept such as integrity might well involve a process of distinguishing distinct uses of the term 'integrity' (most obviously,

concepts, accepting complexity on the level of the individual concept, whereas the reverse is true of explication. Some have found conceptual analysis's commitment to conceptual unity simply unreasonable—a misguided "counsel of perfection" (Tillman 1967, 602). It certainly can serve at least as a distinguishing feature of analysis vis-à-vis explication, as the latter does not share the investment in conceptual unity, which is evident from the explicitly weakened similarity requirement (despite some controversy about just how weak the similarity requirement is; cf. Strawson 1963, Carnap 1963, and Maher 2007).

One could characterize the difference in the two methods as one of significance assigned to antecedent understanding: 'conceptual analysis' refers to a process which seeks to elicit the preexisting understanding an epistemic subject has of what the concept picks out. If the ultimate target of the investigation served or constituted by conceptual analysis is not the concept, but that which falls under the concept, conceptual analysis can be said to invest greater confidence in the existing concept's adequacy with respect to the object picked out. Explications, on the other hand, are proposed because the subject's preexisting understanding is taken to be in need of improvement—at least in certain contexts. A better match with empirical insight into the characteristics of the phenomena covered by a preexisting concept is but one pragmatic interest that can prompt an explication. In tailoring concepts to pragmatic (or specifically theoretical) purposes, an explication takes into account more considerations than merely the balancing of intuitive applicability judgments with a general characterization of the subject matter. In other words, explications establish wide rather than narrow reflective equilibria, with considerable weight given to pragmatic considerations.

Explications are proposed based on the acknowledgment that there are other legitimate purposes which require using terms differently than in the way established by the explication. A conceptual analysis, in contrast, is proposed as the one account of the concept—or rather, the account of what it indicates. The structural difference relates to a difference in purposes: the plurality of purposes presupposed by the explicator who takes one of them to set the explication task at hand contrasts with the one authoritative purpose an analysis seeks to fulfill, best understood as providing something close to a real definition (cf. Beaney 2018, § 2). Now, explications can aim at moving toward a real definition. But they do not necessarily purport to offer one. Whether or not the ob-

uses with regard to physical vs. moral integrity)—very much like in the course of the clarification of an explicandum. Still, once a distinct kind of integrity is identified as the analysans, the aim is to produce one comprehensive characterization of that concept—not to add another concept inspired by the chosen one to the conceptual scheme, as an explication would.

jects falling under an explicated concept form a real distinct kind of entities, can be left open by the explicator.

Explication has attracted a lot of interest lately (e. g., Carus 2007, Wagner 2012, Dutilh Novaes and Reck 2017, Pinder 2017, and Schupbach 2017) as a paradigmatic approach to "conceptual engineering"—the refinement of concepts, specifically for theoretical purposes. Explication has been repeatedly suggested as a viable way for revising normative, not merely scientific concepts (e.g., Shepherd and Justus 2015; Brun 2016). The only problem with this suggestion seems to be that it is less clear what makes a normative concept fruitful compared to what makes a scientific concept fruitful. As far as the concept of moral progress is concerned, we might foremost want it to fit into an exhaustive, consistent conceptual framework with little conceptual overlap—we want it to draw plausible distinctions to other morally desirable changes and to morally desirable events that lack the procedural character of progress. We should expect it to pick out a distinct type of moral change without leaving neighboring types of changes for the better unaccounted for.We might want it to be applicable in different normative contexts (such as contexts constituted by individual or social developments) and to bring out the similarities or underlying commonalities in these diverse contexts which warrant the judgment of moral progress, i.e., to give us an idea about how different progressive developments could be given a unified account. We want it to have the greatest possible applicability while being maximally distinctive in order for it to play an indispensable role in a comprehensive system of ethical concepts. Similarly to the case of scientific concepts, we want to be able to appeal to moral progress to make many generalizations—i.e., we want it to be fruitful in a sense akin to that which Carnap claimed for scientific concepts—and probably not simply based on ethical theories but on sociological, economic or political theories, too, i.e., we want our concept to be interdisciplinarily applicable. Eventually, a concept of moral progress might also be seen as fruitful when it allows us to compare the specific theories of progress advanced from diverse theoretical standpoints.

Like analyses, explications produce definitions, but they present them in a different mode. Explications do not clarify—possibly idiosyncratic—preexisting commitments or claim to capture shared concepts but rather propose ways of thinking with respect to further development, making a contribution to an ongoing, theoretically informed discussion. It is in this mode thatadefinition for moral progress will be proposed here.

Explication avoids several of the abovementioned problems with conceptual analysis. There is no paradox of explication (pace Dutilh Novaes and Reck 2017), as explications do not seek to inform us about an existing concept but rather to propose new ones. Problems with intuitions are mitigated, because intuitions are systematically devalued in explications in favor of further theoretical considerations and pragmatic interests. There is, of course, a persistent problem with intuitive disagreement with respect to the acceptability of the new concept, but in generating the concept, intuitions are not as decisive anymore. Also, an explication can be ontologically uncommitted in the sense that it may propose a concept for pragmatic purposes whatever its referent may turn out to be or not to be. Furthermore, explication does not rely on the analytic-synthetic distinction: explications do not seek to reveal conceptual truths, they seek to give definitions as mere "starting points" for further investigations (Quine 1980, 35).

However, like the account of conceptual analysis, the reconstruction of the basic elements of explication presented above does not do justice to the application of the method. Explicating a concept such as moral progress is not a linear succession of a presentation of the explicandum, followed by a single step of presenting its replacement. Rather, the ensuing investigation will display the search for an explication of moral progress. In the course of this search, there is no neat separation of the consideration of preexisting ideas of moral progress—the characterization of the explicandum—and the theorizing that will inform the choice of its successor—the explicatum. Chapter 2 will present some of the basic conceptual presuppositions made about constituents of the concept of moral progress—both the explicandum and the explicatum. It will present some of the constraints on a deviation from an intuitively plausible account of moral progress, i.e., it will elaborate on the similarity requirement for an explication of moral progress. Chapter 3 will supplement theoretical considerations that can inform the explication. The delineation of the explicatum proposed here and alternative concepts—circulated in the literature or pretheoretically held—will continue through Chapter 4.What our explicandum—our prior concept of moral progress—was, or rather what different possible explicanda have been, will only transpire as the selection and characterization of the explicatum moves along.

The investigation aims at producing an explication in that it seeks to give a proposal as to how alternatively defined concepts of moral progress could be replaced, it does so based on theoretical and further pragmatic considerations. It does not seek to capture everything that might intuitively fall under the concept of moral progress. The central idea is that the few existing accounts of moral progress are in need of revision to make the concept clearer and more fruitful both for ethical theory and nonprofessional moral theorizing.

#### 1.3 A Better Way of Thinking about Moral Change

Two of the problems associated with the method of conceptual analysis are not overcome simply by switching to presenting a definition in the mode of an explication. These concern the nature of concepts and the implications of a semantic theory for the shareability of concepts. Let us first consider the question of what we take to be our subject matter when we say we are analyzing or explicating concepts such as the concept of moral progress.

At its historical origins, in Platonic philosophy (Beaney 2018, §§ 1.1, 2, and 4), conceptual analysis was associated with objectivism about concepts, the view that concepts are abstract objects (cf., e. g., Kann 2011, 326–327). Concepts in terms of Platonic forms account for at least one of the abovementioned challenges to the method of conceptual analysis, the ontological leap from analyzing concepts to giving real definitions—because from a Platonic point of view, inquiring into the content of a concept precisely is to inquire into the essence of a thing. However, today, the view of concepts that is widely supported in philosophy is arguably subjectivism, the supposed "default position in cognitive science" (Margolis and Laurence 2011, § 1.1), which construes concepts as mental particulars. Proponents of conceptual analysis not only casually presuppose subjectivism, but this ontological position is also deliberately turned to as a source ofadefense of conceptual analysis. Subjectivism about concepts seems to solve is the problem of the evidential role of intuitions. As Alvin Goldman argues: "It's part of the nature of concepts (in the personal psychological sense) that possessing a concept tends to give rise to beliefs and intuitions that accord with the contents of the concept" (Goldman 2007, 15).

So, conceptual analysis has been combined with both objectivist and subjectivist accounts of the ontology of concepts.What, if any, are the ontological commitments of explication as far as its subject matter, concepts, are concerned? In his characterization of explication, Carnap speaks of the explicandum as "the given concept (or the term used for it)" and of the explicatum as "the exact concept proposed to take the place of the first (or the term proposed for it)." The question of whether the subject matter of an explication is, after all, a term or a concept is addressed by Georg Brun (2016). Presupposing an objectivist stance on the nature of concepts in Carnap, according to which they are the meanings of terms, he observes that the explicandum cannot easily be equated either with a term or its meaning (i.e., a concept):

Carnapian meanings are not ambiguous; terms are ambiguous if they are used with more than one meaning. Therefore, the first step of an explication, the clarification of the explicandum, must be understood as dealing with terms. But explicanda cannot simply be terms either since clarifying the explicandum calls for selecting one meaning of a given term and this implies that a concept is identified, not merely a term. These arguments show that "explicandum" in fact covers two different items. The potentially ambiguous starting point of an explication is a term, but the unambiguous result of the first step, which the explicatum replaces, is a concept. (Brun 2016, 1216)

The clarification of the explicandum thus singles out one of the different meanings associated with a term. The explication may associate a new meaning (concept) with the same term or introduce a new term for it. Brun suggests to "free the method of explication from its association with Carnap's theory of concepts" (according to which they are meanings) and instead to take "'concept' to refer to an elementary linguistic entity, a 'term,' together with rules for its use" (Brun 2016, 1216–1217).

Like explications, analyses are also concerned with ways to use terms. When Grice describes the method, he makes no mention whatsoever of a "concept":

To be looking for a conceptual analysis of a given expression E is to be in a position to apply or withhold E in particular cases, but to be looking for a general characterization of the types of cases in which one would apply E rather than withhold it. (Grice 1989, 174)

The starting point for a conceptual analysis is a linguistic capacity ("to be in a position to apply or withhold E in particular cases"). Of course, Grice casually alternates between aiming at a characterization of the linguistic practice that reflects this capacity ("a general characterization of one's use of E") and a characterization of the referents of the terms used in the course of this practice ("a general characterization of the types of cases in which one would apply E"). In any case, a concept—as something distinct from the term and the practice of using it —does not enter the description he offers of the process.

Both conceptual analysis and explication purportedly apply to concepts, but talk mostly of words and their use. If one thinks that these methods really are concerned with concepts, their focus on words and the way they are used (rather than on concepts as abstract objects or mental particulars) suggests characterizing concepts in accordance with a view that seeks to establish a middle ground between objectivism and subjectivism. This supposed intermediate ontological position is a "cognitivist" version of pragmatism—a broadly Wittgensteinian approach to concepts according to which they are ultimately "logical constructions out of the practices and abilities of concept-exercising creatures" (Glock 2011, 147). The circularity in this characterization is overcome when those practices and abilities are identified as "ways in which subjects do or could conceive of objects as having properties" (Glock 2011, 157). Cognitivism opposes viewing concepts as items on which the subject operates cognitively, as subjectivism and objectivism would have it: "[cognitivism] agrees with objectivism in denying that concepts are mental particulars, while at the same time maintaining, with subjectivism, that they have an ineliminable mental or cognitive dimension" (Glock 2010, 117). It claims that ways in which the subject operates come closest to what concepts are. From a cognitivist point of view, concept-talk is eventually an abstraction, but "indispensable […] for describing our linguistic practices and abilities" (Glock 2010, 158). Cognitivism therefore claims ontological parsimony as an advantage over both, subjectivism and objectivism. Also, its ability to render concepts shareable is proposed as an advantage over subjectivism (Glock 2009, 2010, and 2011).

Any truly indispensable construction at least suggests that there really is something of which it is a re-construction. So, a choice has to be made about whether to speak of concepts as "logical constructions" (Glock 2011, 147) or "ways in which subjects do or could conceive of objects" (Glock 2011, 157) and in the following, I will opt for the latter. Capacities for "ways of conceiving of things" are ultimately what concept-talk in the established methods that purportedly apply to concepts relates to—so this is what concepts will be taken to be here. The concept Moral progress is a way of thinking about a special type of change for the better, expressed in a certain way of applying the term 'moral progress.'

Cognitivism offers the vocabulary for the following redescription of what is done when an explication is given. On a cognitivist view, a concept is a way of conceiving of a certain subject matter, a lexicalized concept is a way of conceiving of things that is reflected in a way of applying linguistic expressions consistent with this way of thinking, and the term is that which is applied in the exercise of a lexicalized concept. The concept of moral progress then is a way of thinking about moral change, and since it is a lexicalized concept, it is expressed in the way the term 'moral progress' is used. An explication of moral progress proposes a certain way of thinking and using words. The explicandum and the explicatum each are a way of conceiving of things that is reflected in a way of applying linguistic expressions consistent with this way of thinking. The clarification of the explicandum—the selection of "one meaning of a given term" (Brun 2016, 1216)—would then be the selection of one way of conceiving of the subject matter at hand (moral change) which is reflected in one way of applying a term ('moral progress') to it. The explication would be the process of replacing the selected way of conceiving of moral change with another one which is governed by explicit rules that integrate the new way of thinking into an established way of theorizing about the larger context of the subject matter in question (morality). Characterizing the present project as an explication of moral progress thus amounts to stating that it is not one that purports to preserve and make explicit a presumably already shared way of thinking about (a type of) moral change—but rather one that develops a way of thinking about (types of) moral changes that may well deviate from those of other individuals or from the proponents' former ways of thinking.

One worry about this approach is that such a revisionist method risks changing the topic that is selected by the term 'moral progress.' This concern has been at the heart of a classical objection to the method of explication. Peter Strawson has criticized Carnap's methodological approach on the grounds that the replacement of one concept by another (more exact and supposedly fruitful one) that is accomplished in an explication amounts to "chang[ing] the subject" (Strawson 1963, 506). When the philosopher proposes a more exact and possibly even formalized concept to replace any concept of philosophical interest, she inevitably changes what the conversation is about. Since an explication is decidedly advanced as a way of replacing—rather than clarifying persisting—concepts, it is a way of changing the topic. An explication exchanges a concept—the way of thinking that is reflected in the use of a term—by changing the rules of that use as given by a definition, i.e., its descriptive meaning. When the extension of the expression 'moral progress' changes, its subject matter changes and we are not concerned with one and the same topic pre- and post-explication.

The concern about changing the subject through an explication—or any kind of "conceptual engineering"—can be turned on its head by proponents of externalist theories of meaning. If it is assumed that concept contents depend constitutively on features outside the mind of the concept user, explications (i.e., modifications of descriptive meanings) would be futile attempts to change contents, because when they depend on relations to the world, we cannot hope to change them by changing definitions (Cappelen 2018, 64).

So, it seems that we are facing a dilemma: either we can change concepts, but then the effect will be that we fail to keep talking about the same thing, i.e., we fail to share a concept diachronically, or we share concepts because our concepts relate us to the same things in the world, but then we cannot hope to replace our concepts by proposing new definitions. Either the concept of moral progress can be changed at will, but then we change the topic of the conversation instead of making progress on the continued topic "moral progress," or we cannot even hope to effect any change by proposing an explication of moral progress.

On the interpretation of the cognitivist approach to concepts favored here, viewing concepts as ways of thinking places one on the first rather than the second horn of this dilemma. Ways of thinking and abilities to apply words are not genuinely shared to begin with, as they are each individual's ways of thinking and abilities—and they can be changed at will by accepting a new definition of something as guiding one's interaction with the things falling under this definition. At the same time, the cognitivist approach cushions the blow of Strawson's objection: ways of thinking about things always only overlap more or less. The question of whether one is still talking about the same thing is to be decided in similar ways as the question of whether one has ever been talking about the same thing. Whether to treat a conversation as continuing a stable topic is to be decided by those engaged in the conversation.

On this level, the answer to the question of whether we are still talking about the same topic—in the way that it confronts us and is debated—becomes a normative one: we must decide whether we still acknowledge a certain way of talking about a subject matter as being about the same thing. It is very possible, though, that we take the question to be fundamentally about a descriptive matter: that we may think a reality that is independent of our linguistic and conceptual choices is the ultimate arbiter of topic continuity, correctness, or aptness even in the moral sphere. So, the way we think we can change—or not change —our concepts bears on our moral ontology. However, in arguing for certain ways of thinking and for acknowledging topic continuity or discontinuity, we cannot refer to such an independent reality if we are to make our case as convincing as possible to the widest possible audience (see § 3.5.4.). Rather, we need to identify overlaps and differences and present arguments for accepting that a new concept is a better way of thinking about a subject matter. This is what is attempted here.

On the view taken here—that in giving an explication, we propose a way of thinking about certain subject matter by proposing a way of talking about it both concepts and words are potentially subject to change. Intellectual progress—and ethical progress—is possible, if we can find that there is sufficient continuity or overlap in our ways of thinking to acknowledge the new way as a superior successor of the former—or, alternatively, if we can make the case that disruption is necessary because there was no coherent subject matter a former way of thinking picked out to begin with.

There are two possible success criteria for explications, viz., "acceptance" (Sawyer 2020, 564) or "uptake" (cf. Pinder 2017) by the linguistic community and "truth" (Sawyer 2020, 564) or, rather, adequacy. A straightforwardly realistic interpretation of adequacy would be that the explicatum picks out a distinct class of items, i.e., it succeeds in being a way of thinking about something. But this realist reading is not universally available. At the level on which we operate here, it is enough to note that we generally aim for acceptance, and that there are different ways of making sense of this aim: some will say that we aim for acceptance based on assumed adequacy and will have different accounts of what makes for adequacy. The puzzle of what it is—if anything—that makes it the case that our ways of thinking and talking are properly related to the things we assume they are about (if there are such things) cannot be solved here. The observation on which to end these methodological preliminaries is that any successful explication of the concept of moral progress may make it the case that we reclassify supposed instances of moral progress. The aim in proposing an explication is to produce a concept that will be seen as continuing those parts of the supposed conversation around "moral progress" which share the most plausible ways of thinking about a type of moral change as moral progress. The initial step toward proposing such an explication will be to anchor it in a clarification of what progress in general and the realm in which the sort under consideration here occurs—morality—will be taken to be.

## 2 Moral Progress: Conceptual Commitments, Pragmatic Expectations

The concept of moral progress is obviously an evaluative one. It might be thought that this fact is grounded in the concept's constitutive component progress, rather than the component moral (Macklin 1977, 371). The concept of progress itself seems to have not merelyadescriptive but also an evaluative component, which confers a "pro-attitude, favorable evaluation, or positive assessment expressed when any judgment is made that progress of some sort has occurred" (Macklin 1977, 373). Notably, the concept expressed by the term 'moral' does not have an evaluative or normative component on its own. We can supplement the term 'moral' with terms like 'ontology,' 'psychology' or 'discourse,' and while the phenomena captured by such kinds of terms are identifiable in part by their relations to evaluations made by moral agents, the term 'moral psychology,' e. g., carries a descriptive sense and names either a metaethical discipline or a specific part of psychological reality. If there is to be an evaluative element in moral progress and if the concept expressed by the term 'moral' does not contribute it, it must come from the concept progress.

The idea that it is progress rather than reference to morality that contributes the evaluative content to the compound moral progress can be taken one step further. It could be claimed that the evaluative standard on which judgments of progress in general rely simply is a moral one:

All progress, as distinguished from evolution or development, is in a sense moral progress, in other words, it is a movement towards a better state of affairs. (Ginsberg 1944, 11)

On this view, 'moral progress' is the default specification of 'progress' (making 'moral progress' a pleonasm). It is by virtue of denoting an approximation of "a better state of affairs" that the "notion of progress […] from the outset had an ethical connotation" (Ginsberg 1944, 10). If this is so, then the accusation that "earlier writers" have been "lumping" improvements in different domains of human activities "together under the heading of 'progress'" (Buchanan and Powell 2018, 18) would be beside the point, as all these writers would have been concerned with progress that was unified by being, in some sense, moral progress. Notably, if all progress was moral progress, this would make the current investigation relevant for contexts where progress as such is the concern.

That unspecified progress-judgments ought to be interpreted as judgments of moral progress is at least not obvious: entirely unspecified progress claims may also be summarizing judgments about a variety of specific types of progress (relativized to some field of action), where the concluding progress-judgment is based on standards of those fields, not moral ones. In contrast, authors like Ginsberg take unspecified progress to be moral progress by default. This idea may even carry over to judgments of progressive change that do involve an explicit qualification providing a reference to the non-moral standard (e. g., technology) that is of primary interest. Since the notion of progress appears to have an inherent "ethical connotation," one might think that even otherwise specified progress (e.g., technological progress) is—by virtue of being the product of the actions of moral agents—morally constrained, such that it may ever only be called progressive (either in an unspecified or a specified way) if it not only advances the specific field in which it occurs but is also conducive to moral aims. Stan Godlovitch makes an argument to this effect, claiming that it is accordance with the moral code that justifies judgments of progress in any non-moral domain. For something to be fully progressive, it must be progressive with respect to the demands of morality (Godlovitch 1998, 282, 3). If this idea were widely shared, it would explain why it would sound odd to many if someone spoke, for instance, of "progress in the field of torture methods" or any other field of activity that is widely regarded as inherently morally wrong.

#### 2.1 Progress

In contrast to what has been indicated thus far, 'progress' does not necessarily entail a connotation of improvement. Natural language allows us, for instance, to speak of a disease or of climate change as progressing, and neither is usually valued by the speaker. However, when we speak of progress as occurring in some area rather than specific processes as progressing, the positive evaluative connotation may well be triggered.

A plausible starting assumption for distinguishing between different notions of progress is that they will all share three features: progress will always be identified with a process, where this process is a type of change through which some changing entity persists, and this process will be directed in contrast to random change without a pattern or goal. In this section, I will briefly differentiate five concepts of progress. The list does not aim to be exhaustive, but shall illustrate the fact that the evaluative component of the concept of moral progress does not stem from the concept of progress as such, but from one of its versions.

First of all, we might speak of "genetic progress" simply to refer to the mere succession of all events (Rapp 1992, 26). When we use 'progress' in this sense, it is not clear that we are presupposing that there is something that persists through change. 'Progress' in this genetic sense does not even seem to speak of something (more specific than the world) as changing. If "event sequences" still constitute a "change of state" (Godlovitch 1998, 272), this seems to be only the state of the world that sets the stage for the succession of events. Also, in this minimal sense of progress, the only thing that is being presupposed as having a direction is time itself, which may not be a trivial conception (Markosian 2016) but is a widely shared one. Genetic progress, then, is simply the passing of time. Obviously, an evaluative component is missing from this concept. The genetic concept is certainly the least fruitful for thinking about moral progress.

More often, progress is understood in a teleological sense. This might even be its default reading. The notion of progress is typically appealed to in order to express the idea of change being oriented toward a goal, as constituting a step forward (Braun 1990, 177 fn 57). The goal toward which teleological progress is oriented is, importantly, thought to be a pre-determined goal. Where this concept is applied to natural changes, it "presupposes the existence of norms in nature," not the mere predictability of change, but its conformity to laws that pre-determine a specific goal (Godlovitch 1998, 273). Typical examples would include the unfolding of a game (Godlovitch 1998, 273) or development toward biological maturity (Godlovitch 1998, 285). The example of games illustrates that the succession of states in teleological progress, unlike its goal, need not be fixed (Godlovitch 1998, 273). The concept of teleological progress is obviously of interest for the explication of moral progress, because we could think of moral progress as development toward a goal that is fixed by the true moral theory. With regard to this idea, it is important to note that there is an epistemic constraint on judgments of teleological progress. We are probably not justified in assuming that a succession of states is teleological "unless we already know how sequences of that type generally unfold" (Godlovitch 1998, 274).

In contrast to teleological progress, there is progress that is not predestined to end in a specific state. Such open-ended progress does not aim at a substantively fixed goal, because the principles it follows underdetermine this result. Biological or cultural evolution would be the prime examples for such open-ended pattern-obeying change (Schurz 2011). They occur within varying circumstances which preclude the predetermination of a goal.While there is no materially specifiable end-state in open-ended progress, there is a formally describable goal for each sequence of change, given by the inner logic of principles governing the change and specifying its purpose—for instance, adaptation. In some cases, there may be a plurality of equally functional temporary goal-states that conform to the principles governing the change. Something like this is occasionally said to be the case for moral progress (Moody-Adams 1999).

A subtype of such open-ended change would be transformative progress in the normative realm, where normative goals are themselves being reconsidered (Buchanan 2012; Buchanan and Powell 2018). It might be viewed as a desideratum of moral progress that it is transformative, in that the moral ends moral agents actually accept are not treated as fixed, but subject to scrutiny. Transformative progress is first of all made with respect to morality in the descriptive sense, i.e., in the moral views and practices actually adopted by moral agents. It is a contested metaphysical question of whether thereby morality "itself" is being transformed. It would be possible to think of the actual transformative change being made as merely the epistemological side of moral progress and to assume that moral progress is fundamentally an approximation to a pre-determined state, i.e., teleological progress. This state would be specified by the true account of what the actually accepted morality should be like—even if we, as moral agents, do not yet understand (and may never understand) what this end-state would have to be like. However, deciding whether moral progress is at bottom teleological is not necessary to carry out an investigation into the concept of moral progress, as I will argue in § 3.5.We can speak in a meaningful way about moral progress without making a commitment to the idea that it must culminate in a predetermined end-state—and without ruling out this idea.

The final type of progress to address here is progress in the sense of "amelioration" (Godlovitch 1998) or improvement. Whereas the former distinctions of types of progress turned on structural differences, improvement is set apart from the former concepts in that the evaluative component that supposedly marks progress in general is clearly present. Traditionally, 'progress,' when used in an unspecified way, often referred to improvement of "the human condition" (Meek Lange 2011). Improvement of this kind might be thought to consist in changes in a number of different spheres, such as knowledge, economic welfare or happiness. Transformative progress in the normative realm could likewise be seen as consisting in one such improvement. Improvement or ameliorative progress seems to be of prime importance for the concept of moral progress. Moral progress might simply be understood as improvement in morality (see § 2.2.).

For now, we will simply classify moral progress as a kind of improvement. Qua progress, it is a process of change undergone by something that persists through this change and it is directed—whether this direction is given byapredetermined end state or the inner principles of moral progress can be left open. Given the availability of concepts of progress that do not involveacomponent conveying a "pro-attitude" or "favorable evaluation" (pace Macklin 1977, 371), we must assume that it is the component 'moral' that selects the improvement-reading of 'moral progress.' Moral progress must be a kind of process by which something that is morally evaluable, constitutive of morality or simply falls within the domain of morality becomes better.

If we characterize moral progress as a kind of improvement, we can see that it requires two different antonyms. 'Moral regress' could denote the relapse into a prior state, whereas 'moral decline' would refer to a deterioration into a new negatively evaluated state.

## 2.2 Morality

Once we have determined that progress, when it is specified as moral progress, will have to be ameliorative progress (i.e., will consist in improvements), we have to clarify what else the component moral might be thought to contribute to the concept under consideration, over and above selecting the ameliorative reading. To see how this might be done, let us look once more at Ginsberg's equation of progress and moral progress and, more importantly, the way in which he sets this wide reading of 'moral progress' aside:

All progress […] is in a sense moral progress, in other words, it is a movement towards a better state of affairs. We can, however, also speak of moral progress as meaning progress in morality. This seems to consist in the clarification of moral ideas, the removal of inconsistencies between moral rules,aclearer understanding of human needs and purposes, the enlargement of the area within which moral principles are applied and increasing approximation of behavior to the demands of the moral code. It may also involve the discovery of fresh moral principles and the emergence of new rights and duties with changing circumstances. (Ginsberg 1944, 11)

Ginsberg continues with an enumeration of items in the field of morality susceptible to improvement. He mentions that moral progress might involve the enlargement of moral norms' area of application, an approximation of behavior to moral demands or the emergence of new rights and duties. While this gives us an idea about how moral progress is more specific than progress in general, its characterization is still unsystematic and only provides us with an eclectic list of conditions with unclear interconnections. Also, there are axiological tensions within this list, for instance, the enlargement of the area of application of moral principles might entail that it is no longer human needs and purposes that are at the center of moral concern.

The least controversial idea we find in the above-quoted passage is that what the specification moral does to progress is to locate the relevant change in morality. Now, if moral progress is progress in morality, it seems we have to say something about the nature of morality—or at least what we take to be the nature of morality. This, however, is notoriously difficult:

The question of the definition of morality is the question of identifying the target of moral theorizing. Identifying this target enables us to see different moral theories as attempting to capture the very same thing. […] There does not seem to be much reason to think that a single definition of morality will be applicable to all moral discussions. […] What counts as definitional of morality […] is controversial. Moreover, the line between what is part of a definition, in the sense at issue, and what is part of a moral theory, is not entirely sharp. (Gert and Gert 2016)

This passage points us to the crucial challenge we will be facing in this investigation, which is to find a definition of the concept moral progress that will be acceptable from different theoretical perspectives and for many practical purposes. However, if it is the case that the prospect for finding such a definition in the case of morality is bleak, the same seems to be the case for a definition of moral progress.

The differentiation between definitions and theories might be thought to mirror the differentiation between concepts and conceptions of morality (Gert and Gert 2016; cf. Rawls 1999, 5). Gert and Gert caution against this parallelization on the grounds that the terminology of "conceptions"

suggests a psychological distinction, and also suggests that many people have [such] conceptions […]. But the definition/theory distinction is not psychological, and only moral theorists typically have moral theories. (Gert and Gert 2016)

Here, concepts and conceptions will simply be taken to be more circumscribed and more comprehensive ways of thinking about a subject matter, respectively (cf. Chapter 1). Also, as will become clearer shortly, the notion of theory in the context of morality will be treated in a more liberal way. However, it will be helpful to keep the basic distinction between concepts and conceptions (or definitions and theories) in mind. The challenge for the present investigation will be to narrow down the account of moral progress to what could justifiably be passed of as the concept—and not a wider conception—so that it will be compatible with many such conceptions.

One complication of the task of defining morality is the fact that 'morality' is used in a descriptive and a normative sense, i.e., for actual moralities adopted or practiced by moral agents, and for an ideal understanding of what such moral beliefs and practices ideally ought to be (Gert and Gert 2016). When we are concerned with moral progress, it seems that we will have to be concerned with accounts of how actual morality can be changed to accord with what morality ought to be. This makes the present task possibly even more difficult, as some supposedly inevitably controversial assumptions with respect to both the starting point and the goal of moral progress will have to be made.

Instead of attempting to give an overview of different ways in which the question of "What is morality?" could be answered, this section will address three approaches to the phenomenon of morality that will be most relevant for the topic of moral progress as approached in this investigation. So, this section seeks to clarify the crucial conceptual commitments regarding the first constituent of the concept of moral progress.

#### 2.2.1 Moral Judgments

Whether morality ought to be characterized as basically a set of rules, a set of behaviors, character traits or values is controversial. When we are concerned with morality in the descriptive sense, we often accept several of these items as prima facie parts of morality. One way to approach the phenomenon that is morality is to characterize it as a practice of making judgments and then, possibly, to attempt to qualify what differentiates those judgments from other kinds of judgments.

When it comes to moral judgments, a pertinent question is whether we might take uttered moral judgments at face value with respect to the kind of mental states they express. Moral judgments such as "You were wrong to do that" or "It is good to give to charity" appear to be assertions that express beliefs. They purport to describe how things are. When we take moral judgments for what their surface grammar presents them to be, we are taking a cognitivist or rather, a descriptivist⁶ position. The term 'cognitivism' is commonly used to denote a position about the semantics of moral judgments—that they are truly assertions—but this is unfortunate, because the related thesis about moral psychology (that moral judgments express beliefs) is better labelled 'cognitivist,'⁷ it is logically separable from the semantic thesis, and the two can come apart (cf., e. g., Horgan and Timmons 2008, 233–235). Here, I will refer to the view that moral judgments are genuine assertions by which moral agents make truthevaluable statements as 'descriptivism.' Noncognitivists hold that moral judgments do not express belief-states, but noncognitive states such as emotions

Descriptivism in this sense is not to be confused with Frank Jackson's analytic descriptivism (Jackson 1998) nor with the descriptive theory of meaning (cf. Chapter 1).

This cognitivism (in moral psychology) is unrelated to cognitivism about concepts (§ 1.3.)

or, more generally, positive or negative attitudes. The corresponding nondescriptivist thesis is that moral judgments, even if they appear to be assertions, are in fact imperatives or expressions of noncognitive responses to situations and, as such, not capable of being true or false (van Roojen 2013). What is uncontroversial is that our moral discourse appears to be descriptivist and cognitivist—in exchanging moral arguments, we appear to communicate moral beliefs. However, nondescriptivists and noncognitivists see that as a misleading appearance.

In this work, I will keep to convention and speak of moral judgments in a way that presupposes descriptivism. I will treat moral judgments as expressing beliefs and being capable of being true or false. In an investigation into the concept of moral progress, this will be helpful, because it allows us to say, literally, that moral views can be improved by being corrected or that there can be erroneous moral beliefs and to discuss the idea that moral progress consists in an approximation of moral truth (see § 3.5.). However, I think that this decision will not fundamentally affect the explication of moral progress that is ultimately given here.

One can approach the characterization of morality by listing the structural or substantive features of moral judgments—for instance, that they are categorical (i.e., they evaluate actions irrespective of prudential or other non-moral considerations), that they must be universalizable (i.e., they must be transferable into a form that does not include singular terms), that they seek to be intersubjectively acceptable or aim at impartiality (Birnbacher 2013, 20–43, 413–417). Unfortunately, this is a list of increasingly controversial features. In fact, the very attempt at establishing moral judgments (and with them, moral norms, rules or principles) as being distinct from other types of judgments (and norms, rules, and principles) can be found to be objectionable. Pragmatists, for instance, are perfectly ready to accept that moral judgments are simply judgments that refer to social rules (A. O. Rorty 1995).

In one sense, it might be slightly misleading to speak of the practice of making judgments, because the practical side of morality might well be seen as consisting in something different than merely judging, viz., in action—and more specifically, in acting in accordance with or against moral judgments. Judgments are part of the psychological side of morality. Together with abstract entities like theories and propositions they might be said to form the theoretical domain of morality. The other domain is the realm of the morally evaluable conduct of moral agents. The evaluation and regulation of this realm is what moral judgments are made for. Action-guidance is a crucial, some would say the crucial, desideratum for moral theories and judgments. Consequently, there is at least one feature of moral judgments that is relatively uncontroversial, viz., that they are related to action (Birnbacher 2013, 12–19) or agency (Driver 2012, 147). The point of moral deliberation is to find answers to practical problems, the point of moral judgments is to be practically relevant. Some moral judgments evaluate actions in deontic terms, i.e., in terms that express their requiredness or compliance or non-compliance with moral norms (their "to-be-pursuedness"; Mackie 1977, 40). Some evaluate moral agents in aretaic terms that relate to character traits or motives. And some evaluate actions in terms of their quality, for instance, with respect to their consequences. In all that, moral judgments are restricted to objects of evaluation that are in one way or another related to the capacities and conduct of moral agents.

The idea that relatedness to action is an essential feature of moral judgments will be presupposed from here on. This is not supposed to exclude the possibility that states of affairs which are not properly related to actual moral agents (in the sense indicated above) can be evaluated in moral terms. When we say that "Every sentient being ought to be happy," we are making what appears to be a moral judgment about a state of affairs that is morally desirable, but that no moral agent can bring about. Evaluations of this kind do not obey the principle that "ought implies can," i.e., only what could be done can be morally asked of anyone. That is because they do not seem to ask anything from a specific or identifiable moral agent at all. As Broad puts it, a judgment of the kind "So and so ought to exist" requires "only logical possibility," which implies that "any being who could bring it about ought to try to do so. But it does not imply that there actually is any such being" (cf. Broad 1967, 141–142). This gives us a perspective that allows for thinking about such "impossible ought-clauses" as properly related to moral action, after all, and thus, as moral judgments proper. Such "oughtto-be" (Broad 1952) or "state of affairs"-related judgments (Pereboom 2014, 139) are different from "ought-to-do" judgments that prescribe a certain action, in that they provide us with ideals moral progress can aim at:

One way to think about this distinction is that an "ought to be" is an "ought" of axiological evaluation, or sometimes of axiological ideality, which does not (at least directly) entail a "can" claim, while an "ought to do" expresses a demand of an agent in a particular circumstance, which does entail that the agent can perform the indicated action. (Pereboom 2014, 139)

Moral progress, though, can only be thought to aim at realizing states of "axiological ideality," i.e., states in which some morally relevant value is optimally realized, if our moral theory or belief system provides for the realization of value as a moral objective. The idea that all progress is akin to moral progress by virtue of being a "movement towards a better state of affairs" (Ginsberg 1944, 11) expresses precisely this idea. When moral progress is concerned, the idea that it is supposed to involve the increase of value is often treated as a default position. In the following, I will presuppose that the realization of some non-moral values (such as happiness) is plausibly regarded as a moral objective.

Even though we can identify the increase in value as a possible moral aim, and even though it is possible to regard judgments about the desirability of such an increase as moral judgments absent a direct relation to any actual moral agents, this does not force us to accept any increase in such value as a prima facie incident of moral progress. The values we accept as morally relevant can be affected by other forces than moral agency—nice weather can make sentient beings happy. Although such axiological change can be morally desirable, its realization need not be recognized as moral progress (I will argue in § 3.1. that it should indeed not be recognized as such). In the following, I will occasionally refer to non-agency-related improvements (increase in value) as "axiological progress."

Thinking of morality as a practice of making judgments provides us with many aspects in respect to which moral progress might be made (or at least spelled out). Some supposed features of actual moral judgments can be seen as susceptible to improvement. For instance, more and more judgments made in a moral community could conform to the doctrine of universalizability, for instance, by being secularized (i.e., having references to a deity removed; cf. Birnbacher 2013, 36). However, as some of the features listed above are dependent on specific normative commitments about morality, they do not provide obvious starting points for elaborating on a notion of moral progress. The division of morality into a practical and a theoretical domain will be a more helpful perspective for working toward an explication of moral progress (see § 4.2.). In view of the widely shared assumption that moral progress is intimately related to positive changes in states of affairs, the realm of axiological (i.e., non-moral, but morally relevant) value will have to be kept in mind as a possible arena of moral progress, but the sufficiency of changes in axiological value for moral progress will have to be scrutinized.

#### 2.2.2 Moral Awareness and Moral Consideration

Another way to approach the question of the nature of morality is to address what we take to be characteristic of moral agents. The capacity to make moral judgments in the form of truth-evaluable assertions or forming a belief in a proposition that could be so asserted is, to some, essential to moral agency and thus, to morality (cf. § 2.2.3.). But our psychology is usually considered to encompass more than moral beliefs—we also have moral intuitions, emotions, and desires. These could all be recognized as different ways in which moral agents could be responsive to morally relevant facts, i.e., different modes of moral awareness.

It is sometimes suggested that for a moral agent to be praiseworthy, she need to be "responsive" to what makes her action right:

Praiseworthiness as responsiveness to moral reasons: for an agent to be morally praiseworthy for doing the right thing is for her to have done the right thing for the relevant moral reasons, that is, the reasons making it right. (Arpaly 2002, 226)

This idea gives us a desideratum for moral agency—that it be exercised in the form of adequate responses to what actually is morally relevant. While a plausible desideratum, it would be overly demanding as a condition on moral agency. Moral agents must be able to fail in responding to morally relevant facts and still be recognized as moral agents. They must be able to do badly as moral agents. Still, we could say that moral agents are those beings who are in some way responsive to the right kinds of reasons. We could think of moral awareness as the state of actualized responsiveness. In the following, I will presuppose that there could be many, both cognitive and non-cognitive, ways of becoming and being morally aware, i.e., I will assume both epistemological and motivational pluralism. Moral agents might think, believe, understand, feel or sense that something is morally relevant. In the previous section, I committed to a cognitivist and descriptivist reading of moral judgments. At this point, given the just articulated non-restrictive stance on what might constitute moral awareness, it is necessary to note that non-cognitive responses will be treated here as genuinely moral responses and might be characterized as moral judgments in a wide sense in which they are not covered by the cognitivist reading of moral judgments in the narrower sense.

Obviously, a question arises concerning what those "right kinds of reasons" are that moral agents are said to be responsive to qua moral agents. The answer that I propose here is that these moral reasons are constituted by situations in which anyone or anything that is morally considerable is involved. What moral agents should respond to are morally considerable entities, i.e., entities that can be taken into account in moral action. Human beings, nonhuman animals, living beings, and ecosystems are paradigmatic candidates for being morally considerable things—they are such that we could find it somewhat reasonable if someone wanted to take them into account in deliberating choices of courses of action. Taking considerable things into account in a moral way means to consider being considerate of them. To recognize considerable things as considerable means to recognize them as being such that they could be taken into account in this way.

Moral considerability can be given either a realist or an anti-realist reading. On a realist reading, something is morally considerable if it is, as a matter of fact, worthy of being considered in itself. On an anti-realist reading, what is morally considerable is what is such that moral agents want or happen to or could consistently consider it. On this reading, the characterization of moral awareness and considerability becomes somewhat circular, but this is unavoidable when we think of the epistemological flipside to this: moral agents can only seriously consider what they take to be considerable. If we want to allow for moral agents to be able to misidentify what is considerable (without failing to be a moral agent), the realist can give a straightforward answer as to what makes this the case. The anti-realist will have to point to something other than "facts" about who or what is considerable, e. g., a problem with the coherence of the moral agent's belief system or the rationality of her desires. The circularity among moral awareness and considerability is not vicious, because we can reasonably enter into moral deliberation and debate about the criteria of considerability—we can go beyond saying that morally considerable things raise awareness in moral agents and exchange independent reasons for taking different types of entities to be considerable.

We identify moral agents by identifying a consideration-indicating pattern in their behavior or by recognizing them as the kind of beings that are typically capable of consideration. If an agent behaves in a way for which a rationality-maximizing explanation can be given by claiming that she takes some entities around her to be such that she ought to be considerate of them, we can recognize her as a fellow moral agent. This agent's considerability-judgments need not be correct for that to be possible, and neither do ours. If there is sufficient similarity in our consideration-indicating behaviors, we can recognize each other as moral agents. We can also identify completely inconsiderate moral agents (as moral agents), if they are the kinds of beings that we normally expect to have the capacity to become morally aware.

We could conceive of borderline cases in which agents have such fundamentally different ideas about moral considerability that we cannot identify them as moral agents anymore. For instance, many would agree with Peter Singer that stones are paradigmatic non-considerable entities:

It would be nonsense to say that it was not in the interests of a stone to be kicked along the road by a schoolboy. A stone does not have interests because it cannot suffer. Nothing that we can do to it could possibly make any difference to its welfare. (Singer 2009, 37)

If an agent focused all her moral attention on taking all and only stones into consideration, we might not be able to identify her as a moral agent, as our attempts to reconstruct her moral belief system might not even get off the ground. Conduct expressing moral consideration of stones might be so alien that we could not square it with familiar patterns of behavior associated with enacted moral consideration. On the view of morality as characterized by moral responsiveness, awareness and considerability that is endorsed here, it is nevertheless conceptually possible that such an agent would be a moral agent, even though we could never become aware of that fact.

The fact that the category of moral considerability provides a possible perspective on the nature of morality at all and that there are questions about what makes entities morally considerable can be seen as a result at least of progress in moral philosophy. It is the result of the fact that the default idea of moral considerability—that all and only humans were exclusively or primarily morally considerable—is being increasingly challenged.

Moral awareness and views about moral considerability provide us with additional possible respects in which moral progress might be made. However, the latter is more of a substantive normative issue. In explicating the concept—not a comprehensive conception or theory—of moral progress, our focus will be on the role of moral awareness for moral progress instead.

#### 2.2.3 Moral Agency and the Capacity for Moral Progress

Given that we have just tried to identify ways in which morality can be characterized and given that we have done so in the context of an investigation into the concept of moral progress, it is noteworthy that the capacity to make moral progress is often suggested to be a condition for morality:

My main contention is that we could not properly speak of a morality, as opposed to a system of conventions, customs, or laws, until the question of the correctness or incorrectness, truth or falsity, of the rules prevalent in a community is asked, until, in other words, the prevalent rules are subjected to certain tests. It is only when the current rules are no longer regarded sacrosanct, as incapable of alteration or improvement, only when the current rules are contrasted with other possible, improved, ideal rules, that a group can be said to have a morality as opposed to mere set of taboos. (Baier 1958, 174, emphasis in original)⁸

Note that in Baier's account, the decisive idea is not that rules began to be revised—as rules of law and custom could likewise be revised based on internal and external criteria—but that they began to be revised with an eye for truth or correctness. In substance, Baier's basic argument about what makes morality special (Baier 1958, 175) isakind of open-question argument (cf. Moore 2004): we can ask of the results of our moral revisions "But is it moral?" We cannot ask a respective question about revisions in custom or positive law.

According to Baier, moralities—now conceived of as sets of norms—emerged when deliberate social change occurred—in the sense that it was deliberated (Baier 1958, 180). Such change requires in moral agents a capacity for reflection which enables them to actively revise moral beliefs. Thus, we could say that this view makes the capacity for moral progress a condition for moral agency and, therefore, morality. On this view, the rules that were customary before deliberation occurred must have naturally appeared and been unreflectively accepted. A similar picture is painted in Philip Kitcher's contribution to the contemporary debate on moral progress. Kitcher likewise thinks that moral progress—to which he refers as "the ethical project"—began only when our ancestors began to deliberate on and to seek to improve the rules they accepted, everything before that is regarded as morality's pre-history (Kitcher 2011).

These expositions are to be distinguished from even more intellectually demanding views of the essence of morality, such as the Kantian idea, advanced by Christine Korsgaard, that morality requires a capacity for "normative self-government," i.e., the capacity to distance oneself from one's motives for action and ask: "'but should I be moved in that way? Wanting that end inclines me to do that act, but does it really give me a reason to do that act?'" (Korsgaard 2006, 133). The capacity for normative self-government can likewise be seen as a capacity for progress, as its function is to get the moral agent to act on better reasons.What all these views have in common is the fact that they attribute to a capacity for reflection prime importance for the existence of morality.

The view of morality presupposed here is nowhere near as intellectually demanding. It might be thought that this comes at the cost of waiving a crucial function of morality, which is to ascribe moral responsibility and hold (moral) agents accountable. The capacity for critical self-reflection may be rightly considered a precondition for this practice. Holding moral agents who could not help but act on their (moral) responses accountable would, on the contrary, be unfair. However, it must be noted that holding others accountable—or rather justifying the practice of holding accountable—is hardly our only interest in morality, and it is not clear that it should be this pragmatic interest which determines the content of the concepts of moral agency and morality. Also, accountability is a relational notion. Sometimes, if we are unable to hold others accountable, this might be due not to a lack of capacities underlying responsibility in the other agent but rather in our epistemic restrictions or a lack of authority on our part. Certainly, capacities such as the ability to engage in self-reflection could improve the exercise of moral agency. However, I will assume that it is possible to fully attribute moral agency toabeing before these advantageous features take effect.

Before we move on to review some of the possible pragmatic expectations for the concept of moral progress, a brief terminological remark is in order: the term 'ethics' is sometimes used to mark the systematic reflection on morality. Based on the conceptual commitments laid out in this section, it seems that moral agents are not necessarily able to make ethical progress, as they are not necessarily able to reflect on their moral views or responses. Nevertheless, even moral agents who lack this capacity can undergo a change in these responses, and such change is to be distinguished from making practical changes. So, I will say that even though not all moral agents are capable of making ethical progress in a sophisticated sense, i.e., by reasoning their way to a new moral view, they can make ethical progress in a wide sense, i.e., by undergoing changes in their moral views, widely construed (manifest responses or judgments in the wide sense). The relation between ethical and moral progress will be discussed in § 4.2.

## 2.3 The Concept of Moral Progress: Pragmatic Expectations and Concerns

A key pragmatic consideration in explicating moral progress will of course be to generate an account of moral progress that will not be overly dependent on one specific normative theory or even metaethical point of view.We have already mentioned some of the respects in which a concept of moral progress might be said to be fruitful (§ 1.2.). Here, we will review some of the more specific purposes that might be associated with this concept. What is its potential relevance for moral deliberation and theorizing? Why should we take an interest in the concept of moral progress at all? After discussing some of the potential pragmatic interests attaching to the concept of moral progress, we will also register what kinds of reservations talk of "moral progress" might evoke.

When it comes to the pragmatic purposes the concept of moral progress might serve, first of all (1), as mentioned in the introduction, an important way in which the idea of moral progress might be thought to "earn its keep" (Godlovitch 1998, 284) is by featuring in the motivation of moral agents:

The idea of moral progress is a necessary presupposition of action for beings like us. We must believe that moral progress is possible and that it might have been realized in human experience, if we are to be confident that continued human action can have any morally constructive point. (Moody-Adams 2017, 153)

Moody-Adams takes the necessity of the belief in moral progress to be a "truth for moral psychology," even though the role of this belief is something short of providing action-guidance. Godlovitch's somewhat more modest claim is that the idea of moral progress is not necessary for action, but the belief in the possibility of success is crucial for efforts for improvement (Godlovitch 1998, 278). Similarly, Thomas Nagel argues:

It is evident that we are at a primitive stage of moral development. […] The idea that the basic principles of morality are known, and that the problems all come in their interpretation and application, is one of the most fantastic conceits to which our conceited species has been drawn. […] Not all of our ignorance in these areas is ethical, but a lot of it is. And the idea of the possibility of progress is an essential condition of moral progress. None of it is inevitable. (Nagel 1986, 186)

However, the quoted passage and its further context suggest that what Nagel takes to be decisive is not foremost belief in the success of progressive efforts, but awareness of fallibility. He is concerned with the necessity of corrections to the current system of moral beliefs rather than with a vision of the ultimate endpoint of ethical inquiry. It is still noteworthy that he casts this insight in the language of moral progress. The relevance of the idea of moral progress for moral motivation will be discussed in detail in Chapter 5.

(2) Some (defeasible) support would be given to the belief in future progress if progress could be identified in the past. Irrespective of this potential relevance for an optimistic moral outlook, registering moral progress in past developments is in itself something of interest to moral agents. Moody-Adams suggests that "the idea of moral progress is a plausible, critically important and morally constructive principle of historical interpretation" (Moody-Adams 2017, 153), i.e., we can rely on it in writing moral historiographies. In this way, moral progress can have an important backward-looking purpose. It can serve to locate others and ourselves (cf. Rapp 1992, 1) within moral history. In this way, moral progress can be an orientation concept. Presumably, it will be combined with a claim to knowledge or at least justification in this function, i.e., it will typically convey more than mere approval.

(3) The apparent existence of moral progress may also be appealed to in order to counter skepticism regarding the objectivity of morality or the possibility of moral knowledge. For instance, Michael Smith appeals to moral progress to refute the argument from disagreement against moral realism, shifting the focus from persistent disagreement to achieved agreements:

disagreements were removed inter alia via a process of moral argument. I am thinking in particular of the historical, and in some places still current, debates over slavery, workers' rights, women's rights, democracy and the like.We must not forget that there has been considerable moral progress, and that what moral progress consists in is the removal of entrenched disagreements of just the kind we currently face. (Smith 1984, 188)

The relation between the concept of moral progress and moral realism will be discussed in § 3.5.

(4) Obviously, the concept of moral progress features in a special subtype of comparative judgments. Occasionally but rarely, moral progress-judgments are treated precisely as ordinary comparative judgments. For instance, Ruth Macklin argues that moral progress does not only "compare […] units in temporal sequence," but can be used to rate "two contemporaneous […] societies or countries or cultures" (Macklin 1977, 370). Being "more morally progressed" then comes down to satisfying to a higher degree certain material conditions of moral progress. To Macklin, these are given by the principles of "humaneness" ("sensitivity to […] the pain and suffering of human beings") (Macklin 1977, 371) and the "principle of humanity ("recognition of […] the intrinsic worth of human beings") (Macklin 1977, 372). On such an account, the claim that "x is more morally progressed than y" (Macklin 1977, 370) reduces to saying that x better exemplifies the chosen principles. However, absent any "significant relations […] between the states of affairs being compared," i.e., "close causal, cultural, and temporal connections" (Jamieson 2002, 332), such a judgment is a mere comparative judgment. But "[n]ot all betterness relations are progress relations" (Rønnow-Rasmussen 2017, 140). What must at least be the case for a genuine moral progress-judgment to be issued about two cultures (societies or countries), is that some change must have taken place within each culture (and that it is possible to compare them with respect to this change). On the view taken here, evaluating separate contemporaneous cultures (or individuals) in terms of moral progress amounts to comparing two progressive developments, not two states of affairs.

(5) In addition to recording moral progress in our moral historiographies, we can also use 'moral progress' to give credit for a moral achievement, i.e., to convey recognition. A special mode of conveying recognition is to praise moral agents for the progress they have made.Whereas recognition of a moral achievement can be expressed to a third party, for the special speech acts of praising and blaming (in distinction to giving recognition) to have a point, they must be directed at the agents that have progressed.

(6) A particular function of expressing recognition of moral progress can be to mark an accomplishment as worthy of protection, i.e., to highlight that something is an improvement over a prior state and that measures should be taken to guard against regress. Adverse effects of this function of applying the idea of moral progress will be discussed in § 3.4.1.

(7) The idea of moral progress is present, at least in substance, in certain debates about how to evaluate past moral failures. See, for instance, these two appeals to moral change compared by Patrick Stokes:

Cheshire Calhoun, writing of the ways in which we become aware that what had seemed like innocuous practices are deeply sexist, argues that "feminist consciousness is a consciousness of inhabiting a new reality, of seeing what one did not, could not, see before and what others still do not, cannot, see" (Calhoun 1989: 390). Rorty too notes how moral change involves displacing old descriptions with new ones, such that "descriptions of situations which used to seem crazy seem sane" (R. Rorty 1991: 233). Any exculpation that lies in the thought "it was a different time" turns on this sense of living after a moral gestalt shift, such that those living before the shift could not see what we can see now. (Stokes 2017, 1829)

These are examples of making reference to moral change in order to account for prior moral insensitivity.⁹ On the views cited here, moral progress sometimes overcomes inculpable moral failure. In contrast, Michele Moody-Adams insists that moral progress typically overcomes "affected" or culpable moral ignorance. Because moral discourse is continuous, the ideas of which we have a deepened grasp now were in principle available to moral agents at earlier stages as well (Moody-Adams 1994). The idea of moral progress can be invoked when the overcoming of past failure is discussed, irrespective of how this failure is being evaluated.

(8) Marking moral progress can also serve to guide ongoing moral discourse and to guide action. For instance, consistent change toward a new moral conviction might be cited in justifying legal decisions that to some degree reverse prior legal practice (Annas 2002, 1814). The idea then is that legal systems (or other systems of social rules) have to catch up to a development in shared moral beliefs. It might even be thought that if an explicit account of moral progress is available, new proposed changes could be morally evaluated based on it (Macklin 1977, 380). However, this prospective use of an account of moral progress seems to be a problematic idea, because such an evaluation would fail to scrutinize the hitherto developed idea of moral progress itself. Furthermore, appealing to moral progress does not even seem to be necessary in this case, as the rel-

What I mean by "moral insensitivity" is commonly also referred to as "moral blindness," but this is a problematic term due to its derogatory use of "blindness" (see the following usage of the term: "We take as our starting point the fact that men sometimes act horrendously, and thus consider that any plausible ethical theory should seek to describe and probe … [b]ehavior which calls from us disgust and reprobation, and occasionally the charges that those who engage in such behavior are moral defects, who like the optically blind are morally blind" Baumrin 1986, 205). Unfortunately, the expression "moral incapacity" is already in use for a state of mind at the opposite end of the scale of moral awareness, i.e., the inability to act against a certain moral belief (Williams 1993).

evant evaluation requires only an idea of what would be good or right, not what would be progressive.

Despite these various pragmatic interests in it, several concerns could be raised about the concept of moral progress from different ethical vantage points. It could be argued that the concept of moral progress is redundant, because morality is inherently progressive anyway. This objection might be based on the idea that moral agency is marked by the capacity to reevaluate oneself and the norms one follows, or the idea that the moral agent's task is to improve the world on any given occasion for action. Either way, successful moral action would (necessarily or typically) result in improvement that can be registered as moral progress. The underlying ideas of these two versions of the objection will be discussed in §§ 3.2. and 3.1. In any case, moral progress might always be relevant as a marker of moral success, although I will argue that its role is more special than that.

Furthermore, it might be objected that the concept of moral progress is useless at best, because any diagnosis of moral progress is always underdetermined, or it might be argued that moral progress-judgments are actually false, i.e., that the concept is empty. Whether the concept of moral progress is inapplicable or empty obviously depends on how it is defined, and the explication put forward here will aim to produce a concept that is suitable for being operationalized and, ideally, not empty.

Moral progress might also be said to establish a morally undesirable perspective on the present. Thinking about the status quo as a result of moral progress might make moral agents complacent and distract them from persisting or emerging problems. This worry can simply be read as a warning and refers to the psychological challenge of not merely applying the concept of moral progress in a backward-looking fashion but remaining alert to the need for further progress (see Nagel's remark on the necessity of an awareness of fallibility above). The flipside to the objection could be that the idea if moral progress is, unfortunately, not a suitable source of motivation for future improvements. This objection will be spelled out and rebutted in Chapter 5.

In any case, assessing the merits and limits of the concept of moral progress requires to give an explicit account of it first. The following chapter considers the hints one might take in this endeavor from theories in normative ethics and addresses the supposed metaethical commitments of any talk of moral progress.

## 3 Ethics and the Idea of Moral Progress

When we think about where the investigation into the concept of moral progress fits on the map of ethical disciplines, the enterprise might be viewed as a straightforwardly metaethical one, insofar as explicating this concept inevitably means to elaborate on an idea of the very nature of morality (see § 2.2). In order to be able to say what moral progress might be, we need to address the character of the domain in which this particular kind of progress supposedly occurs—and characterizing morality as such is the province of metaethics.

Moral philosophers who work on the concept of moral progress nevertheless often approach this task in other ways than by drawing on explicit metaethical theorizing. What many seem to find much more important than laying out their metaethical presuppositions is the responsiveness of their account of moral progress to what they see as uncontroversial instances of the phenomenon. The provision of examples in work on the concept of moral progress is often more than a way to illustrate the range of applicability of one's account (cf. Chapter 1). Rather, it sometimes serves as a starting point for building an account of moral progress on the identification of the common features of these cases (e. g., Buchanan and Powell 2018) or by providing a list of material criteria for the identification of moral progress independent of a specific normative theory (e. g., Jamieson 2002). Frequently cited examples of moral progress include the accomplishments of the civil rights movement in the U.S. in the 1960s (Jamieson 2002, 336), the prohibition of cruel and unusual punishment (Macklin 1977, 327), including public execution (Jamieson 2002, 335), the existence and accomplishments of the animal rights movement or green movements (Godlovitch 1998, 278), and advances with respect to women's rights (Kitcher 2011, 145–153; Wilson 2011, 107) and the rights of people with disabilities (Nussbaum 2007b, 940). Of course, the one universally acclaimed, most frequently mentioned example is the abolition of chattel slavery in Britain and the U.S. in the 19th century (cf., e. g., Macklin 1977, 370, Godlovitch 1998, 272, Oshatz 2008, Wilson 2011, 107, Anderson 2014, and Stokes 2017, 1827). As Philip Kitcher puts it: "If there is one example in which the attribution of progress is almost incontrovertible, it is the abolition of the 'peculiar' institution, chattel slavery" (Kitcher 2011, 154).

Even when historical examples are not used to tailor an account of moral progress to these examples, they often have significant importance for work on the concept of moral progress, because they appeal to the audience's intuition that there really must be such a thing as moral progress for which some agreeable account is to be found. In this way, Godlovitch appeals to the abolition of slavery as an instance of a newly established moral obviousness, achieved in a relatively short period in human history: "I cannot own you, and you cannot own me, period. Only yesterday, it was otherwise" (Godlovitch 1998, 272). Such reminders of historical changes that strike us simply as changes for the better serve a motivational purpose, since the identification of a theoretical account of moral progress proves to be difficult, which creates an initial difficulty for engaging with the topic at all. The problem with proceeding from the collection of paradigmatic examples to the provision of an explicit definition of moral progress is often traced to the potential for disagreement over any account that might be given (e. g., Evans 2017, 76). Allegedly, agreement about the extension of the concept moral progress is reached easier than agreement on its intension.

One recent objection to the very attempt to give a unifying account of moral progress turns on the apparent plurality of forms of moral progress to be identified in the paradigmatic examples as well as the idea that any account of moral progress itself ought to be subject to revision in the course of further moral progress (Buchanan and Powell 2018, 44; Chapter 2; 92–94). Still, even an account of moral progress that remained, in light of this consideration, "pluralistic" and "provisional" (Buchanan and Powell 2018, 93) would have to employ some working definition of moral progress. As we shall see, even finding such a working definition is difficult and any result will likely be subject to critique from some particular normative point of view. In this respect, the problem of finding a definition of moral progress resembles challenges in applied ethics more than paradigmatic metaethical theorizing. In applied ethics, problems recognized as problems from many different normative angles have to be solved in a way acceptable to these many perspectives. Of course, there are theory-based ways of doing applied ethics that approach these problems by relying on the resources of one particular normative theory. But often, the task of the ethicist working on problems in applied ethics (such as the question of the permissibility of physician assisted suicide, the safeguarding of scientific integrity, or our duties with regard to nonhuman animals' suffering in the wild) is conceived of as finding solutions to actual moral problems without relying too heavily on normative presuppositions that might not be shared by all concerned with the problem. One of the crucial concerns about reliance on a specific normative theory in these contexts centers on the "absence of consensus on which theory" one ought to build one's case, i.e., moral pluralism (Arras 2016, § 3.1, emphasis in original). It is precisely due to the plurality of normative perspectives involved that the approach presently dominating, for instance, bioethical debates, principlism, seeks to replace appeals to a particular normative theory by reliance on principles, which are supposed to be acceptable from various normative angles (Beauchamp and Childress 2001). The task of finding a definition of moral progress is similar to the challenge of resolving problems in applied ethics given moral pluralism to some extent: only that instead of a moral problem, it deals with a phenomenon that is recognized as such by people looking it at from many normative points of view, and instead ofasolution to the problem, it consists in the search for a description of the phenomenon that will be acceptable from all these different viewpoints.

It makes sense to view moral progress as a "moral phenomenon" similar to phenomena such as moral dilemmas, supererogation, or moral luck¹⁰—which pose challenges to different ethical theories—in the following sense: like these other phenomena, moral progress is a feature of our moral experience. Certain developments simply "look […] like progress" (Stokes 2017, 1824)—or so the use of examples like the ones cited above would cause us to think. Any normative theory could be expected to have something to say about it. The difference from some of these other phenomena is that disagreement over the very possibility of moral progress is in fact located on the metaethical rather than the normative level. Compare this to the case of, for instance, supererogation, i.e., actions that are good to do but not required: a utilitarian might deny that supererogatory acts could exist, because the course of action that the moral agent is morally required to take simply is the best course of action; there is nothing beyond the best course of action that is morally good. Such an objection to the very existence of supererogation would be given on the level of normative theorizing. In contrast, any normative theory entails an account of moral progress: improvement in terms of the theory's own proliferation and implementation. If the possibility of moral progress is doubted, the skepticism will turn on the more fundamental, metaethical question of whether what seems to us to be an improvement (for instance, as described by some particular normative theory) really is an improvement. In this sense, moral progress might be more adequately described as a meta-moral phenomenon.

Theories on the level of normative ethics are in one sense obvious starters for elaborating on the concept of moral progress, and in another sense, nonstarters. On the one hand, normative theories—theories dealing with what is good and right—entail accounts of moral progress. But on the other hand, different theories in normative ethics—proposing different axiologies, different deontic principles, different paradigmatic problems and paradigmatic solutions, i.e., diverging conceptions of morality—will obviously give different accounts of improvement in morality. The fact that these theories will give different conceptions of moral progress is trivial, but spelling out these different conceptions likely is

Cf. https://philpapers.org/browse/moral-phenomena, last accessed June 28, 2022.

not—unless it is restricted to saying that moral progress consisted in increasing conformity to the provisions given by some specific theory. When it is done in a more comprehensive way, demonstrating how a certain normative theory would have us conceive of moral progress is a stand-alone project (cf., e. g., Singer 2011; Kitcher 2011)—always specific to a certain variety of the type of ethical theory in question. Just as any normative theory might be enriched or completed by a virtue theory which gives an account of what are valuable dispositions of the moral agent according to that theory (Hursthouse and Pettigrove 2018), any normative theory might be complemented by a theory of moral progress that states what individual and/or collective improvement consists in according to the theory. Elaborating such a theory is, however,adifferent project than the one pursued here.

In this chapter, we will see how perspectives from different normative theories (ethical theories, widely construed) nevertheless could inform or constrain our explication of moral progress (which will be given in the following chapter). The purpose is not to try and extract any accounts of moral progress from the respective normative theories but rather to use their most important "talking points" to point us toward necessary basic decisions about our concept of moral progress and toward possible challenges in making these decisions.

§ 3.1. will begin by reviewing the supposedly close relation between consequentialism and the topic of moral progress: utilitarians are often said to have a special affinity to progressive causes due to their commitment to inclusivity. I portray an influential utilitarian proposal for understanding moral progress that focuses precisely on the feature of a moral code's inclusivity—Peter Singer's account of moral progress in terms of an "Expanding Circle" of moral concern (Singer 2011). Another reason for thinking of the consequentialist perspective as instructive on the topic of moral progress is its focus on consequences. For judgments of moral progress, observed changes in states of affairs are crucial and the idea that moral progress must involve such changes features prominently in the "naïve conception" of moral progress. As far as implications for an explication of moral progress are concerned, the assessment of the concept of moral progress emerging from "progressive consequentialism" will be most important. Progressive consequentialism holds that in order to act rightly, moral agents need to improve the world (rather than bring about best consequences) (Jamieson and Elliot 2009). I will scrutinize and reject the resulting picture of moral progress that presents it as a change in states of affairs (rather than in moral agents and their performance) that may result from any singular morally right act.

§ 3.2. will consider a concern about the prospect of moral progress which might be broadly labelled a "Kantian reservation," as it can be found in Kant's engagement with the problem of human improvement. Basically, it points us to the question of how comprehensive we should take judgments of moral progress to be. If moral progress is attributed to moral agents rather than changes in states of affairs, as will be argued in § 3.1., should a judgment of moral progress encode an evaluation of the rightness of moral agents' acts, their motives or both? In our explication of moral progress, we will have to address these options in detail. Kant is stereotypically associated with a focus on motives, and his remarks on the possibility of progress can be used to illustrate the resulting reservation with respect to the occurrence and detectability of moral progress.

If moral progress-judgments are taken to evaluate the moral agent rather than the effects of his actions, and especially if they are thought to evaluate moral agents in terms of their mindset, it seems that a virtue ethical perspective on moral progress is invited. In § 3.3. I will confront this idea and argue that even based on the preceding differentiation between moral progress and improvements in states of affairs, and givenacommitment to evaluate moral agents more comprehensively than in terms of the rightness of their acts, it is not clear that the concept of moral progress ought to be explicated in terms of a virtue ethics account—or even in terms of a virtue theoretical account.

§ 3.4. will widen the perspective beyond the realm of normative ethics to include an observation concerning the paradoxical dynamics of moral progress to be found in John Stuart Mill's political work. His remarks on the course of progress (notably not specified as moral progress) in liberal societies are of special interest to the present project because the adoption of liberalism itself is widely regarded as a morally desirable accomplishment—i.e., a result of past progress. When the concept of moral progress is applied to further progressive developments, this application will therefore often be situated in the context of circumstances characterized by liberal values. Mill's discussion of progress within liberalism provides useful insights into the interrelation of moral and ethical progress.

§ 3.5. will turn to the one metaethical question that is, after all, frequently addressed in the debate about moral progress: whether talk of moral progress commits us to accepting some form of moral realism. The suspicion is that the concept of moral progress or the ways in which it is usually employed presuppose that there are moral facts which are independent of our evaluations of or responses to moral problems (or, more generally speaking, situations of moral significance) and which make our moral judgments true or false, so that moral progress would have to be thought of as the approximation of objective moral truth (and the realization of behavior that is prescribed by objectively true moral judgments). I will address the problem of this supposed progress-realism tie-up and briefly portray an attempt to characterize moral progress in antirealist terms of increased moral functioning—the pragmatist account elaborated by Philip Kitcher in The Ethical Project (2011). I discuss the reasons for which and the extent to which an explication of the concept of moral progress may remain ontologically uncommitted.

## 3.1 Consequentialist Concerns: Inclusivity and Better States of Affairs

Utilitarians are sometimes regarded as the "paradigmatic" progressives in the political sense of the term: utilitarians have historically been supportive of causes that are classified as "progressive" (women's suffrage, the decriminalization of homosexuality, the reduction of class privilege, etc.). This political tendency is grounded in a more basic feature of utilitarianism as a moral doctrine: its commitment to inclusivity (cf. Brink 2013, ix), which stems from its axiology. Utilitarianism is a type of consequentialist ethical theory, i.e., it determines what is right in terms of consequences or outcomes. Classical utilitarians were concerned with the rightness of acts and held that the right act is that which maximizes the good, where the only values relevant for the good are the value of pleasure and the disvalue of pain. The commitment to inclusivity, in the case of classical utilitarianism, is based on the axiological decision to select pleasure as the value to maximize and a commitment to impartiality or "neutral value," i.e., the maximization of value irrespective of whose value it is (Driver 2012, 26). As Jamieson puts it: "classical utilitarians are paradigm moral progressives because of their insistence on taking seriously the interests of all who are affected by an action" (Jamieson 2002, 329).¹¹ It is this theoretical commitment to comprehensive moral concern for all which predisposes classical utilitarians to endorse causes which are recognized today as successes of earlier efforts for moral progress or increasingly championed as urgent moral challenges of the present: causes that center on greater inclusivity for existing schemes of justice. An iconic example in this regard, frequently cited in animal ethics literature, is Jeremy Bentham's anticipa-

When other values and disvalues than pleasure and pain are inserted into the consequentialist axiology, the resulting moral theory is not necessarily so radically inclusive anymore, at least not in the sense of yielding, in theory, the same egalitarian moral community. Mill's attempt to separate higher from lower pleasures (Mill 2008, 164) might at least work toward the disadvantage of some moral considerees, and even interests (as mentioned by Jamieson) could be characterized in such a way that the class of bearers of interests is not coextensive with the class of those capable of experiencing pleasure and pain, such that the resulting morality is no longer as inclusive.

tion of the animal rights movement, in which the inclusivist spirit of his utilitarianism is clearly articulated:

The day may come when the rest of animal creation may acquire those rights which never could have been withholden from them but by the hand of tyranny. The French have already discovered that the blackness of the skin is no reason why a human being should be abandoned without redress to the caprice of a tormentor. It may one day come to be recognized that the number of legs, the villosity of the skin, or the termination of the os sacrum are reasons equally insufficient for abandoning a sensitive being to the same fate. […] The question is not, Can they reason? Nor Can they talk? But, Can they suffer? (Bentham 2007, 311, n. 1, emphasis in original)

Here, the move toward inclusion of groups of individuals who have so far not been covered by legal protection of their morally justified claims is presented as a moral requirement and realistic prospect. In contemporary work on the concept of moral progress, increasing inclusivity is widely regarded as a hallmark of moral progress.

According to Peter Singer, development toward greater inclusivity simply is the course that moral progress has taken in the past and that it must continue to take in the future. In The Expanding Circle (Singer 2011), he argues that the widening of the domain of moral concern—the extension of moral consideration and institutional protection toward former out-groups—can be observed over the course of human history (temporary setbacks notwithstanding) and should continue in the future because it is, importantly, what morality demands. He arrives at this normative conclusion based on a metamoral premise that is not exclusive but is nevertheless central to his utilitarian ethics: the idea that in order to engage in the practice of giving justifications, individuals must provide "disinterested reasons" for their judgments, i.e., reasons that can be accepted by all members of a group of "reasoning beings" (Singer 2011, 93). Singer argues that the idea of impartiality in terms of disinterestedness of reasons can be taken further

by taking the element of disinterestedness inherent in the idea of justifying one's conduct to society as a whole and extending this into the principle that to be ethical, a decision must give equal weight to the interests of all affected by it. (Singer 2011, 100)

By a process of elimination, he arrives at the "principle of impartial consideration of interests" as the basis for moral reasoning (Singer 2011, 110). The argument here is that other ethical standpoints that formally satisfy the condition of disinterestedness, i.e., the requirement not to privilege anyone's interests (on egoism, see Singer 2011, 103, and on absolute deontological principles, see Singer 2011, 107–108), cannot be recommended for acceptance based on disinterested reasons, at least not without "import[ing] the puzzling idea of an objective moral reality" (Singer 2011, 109). Other than by pointing to such an independent reality, Singer argues, the choice of a normative principle could only be guided either by subjective preferences or reference to their consequences for all affected, and only the latter option withstands the test of impartiality itself (Singer 2011, 103–110). The "principle of impartial consideration of interests" therefore emerges as the "ultimate basis for morality" (Singer 2011, 110)—and it implies that moralities that do not yet take the interests of all affected into account must expand their domain of moral concern. We will return to this line of reasoning, which connects metaethical and normative considerations, in § 3.5., where we will address how Singer later came to reconsider the metaethical grounds of his argument.

It is worthwhile to emphasize at this point that in calling utilitarians "paradigmatic progressives" because of their commitment to an inclusive morality we identify a specific preexisting flaw in the actual moralities that undergo moral progress: their exclusivity. This fact may usually simply be taken for granted, but it is noteworthy that the idea of progress in terms of greater inclusivity presupposes that the "moral point of departure" (see § 4.1.4.), the state of morality in need of improvement, is characterized by undue exclusivity. While this seems to be a widely shared assumption, it is a substantive normative and empirical assumption nonetheless. In principle, it is possible that moralities are in need of improvement in terms of an increase of exclusivity: there may be moralities acknowledging too many rather than too few types of objects of moral consideration. In actuality, the casual mode in which change toward greater inclusivity is often mentioned as a mode of moral progress might be seen as evidence of at least some already accomplished ethical progress, i.e., growing agreement in the identification of exclusivity as a prevalent flaw in existing moralities.

Aside from the commitment to inclusivity, another reason why utilitarians' perspective might be seen as especially relevant to the concept of moral progress is the focus on consequences. Classical utilitarians are act consequentialists;¹² they assess acts by the consequences they produce. Consequentialists more gen-

Utilitarians are consequentialists, but not vice versa; as Walter Sinnott-Armstrong observes, the term "consequentialism" (which was coined by Anscombe in an attack on classical utilitarians (Anscombe 1958)) is often used as a "family resemblance term to refer to any descendant of classic utilitarianism that remains close enough to its ancestor in the important respects" (Sinnott-Armstrong 2019, § 2). In addition, consequentialism comprises theories with entirely different deontic principles—egoism and egalitarianism are concerned with consequences as well. Still, "consequentialism" and "utilitarianism" are sometimes used interchangeably. Here, "consequentialism" will be used in a wider sense.

erally not only apply this reasoning directly to acts but can assess acts in virtue of the consequences of different kinds of items, such as the motives (Adams 1976) or rules (Hooker 2000) underlying the act. They might also assess motives or rules directly (instead of acts) in terms of their consequences (Sinnott-Armstrong 2019, § 5). Some consequentialists will not restrict their account to the application to any of these items, but will accept any kind of 'evaluand': "global consequentialists" or "global utilitarians" (Kagan 2000; Pettit and Smith 2000) view anything as having morally evaluable outcomes. They do not merely cover consequentialists' standard evaluanda—acts, rules, and motives—by also evaluating norms, character traits, and institutions directly in terms of their consequences (Ord MS, 3; Driver 2012, 146); they also "seek to take the [consequentialist] approach to its natural conclusion and assess everything directly in terms of its consequences" (Ord MS, 3). This natural conclusion is articulated by Parfit in the following way:

Consequentialism covers, not just acts and outcomes, but also desires, dispositions, beliefs, emotions, the colour of our eyes, the climate, and everything else. More exactly, C covers everything that could make outcomes better or worse. According to C, the best possible climate is the one that would make outcomes best. (Parfit 1984, 24).

Talk of "outcomes" with respect to properties such as eye colors or, more generally, with respect to objects rather than with respect to events or actions may seem odd in itself, but the more crucial break, not merely with other forms of consequentialism but with any ethical theory, comes sooner, namely, when global consequentialism leaves out any restriction to agency-related evaluands: when it is not merely global in the sense of ranging over all conceivable items in the domain of moral agency but also in the sense of ranging over any kind of item whatsoever. Because of this wide scope, the evaluation given by globally utilitarian judgments does not necessarily exhibit the "relatedness to action" anymore, that was accepted as a requirement for moral judgments earlier (see § 2.2.1). In view of this peculiarity, Julia Driver proposes to limit the range of evaluanda for global consequentialism to "feature[s] of agency or relevant to agency," pointing to "the practical nature of ethics" (Driver 2012, 147). The fact that evaluative terms are used in their "distinctively moral sense" means that they are used in a context that revolves around the choices of agents with respect to courses of action or entire ways of life (Driver 2012, 147–148). On this view, there is a sensible way in which consequentialism might be global, viz., when it accepts all kinds of evaluanda that are related to agency, but it ought not to be global in the sense of accepting just any evaluand in moral evaluation even eye colors or the climate.

In this latter—more restricted and thus less global—version, global consequentialism would be a normative theory compatible with the view of morality we endorsed earlier by including the requirement of relatedness to action for moral judgments. With that in mind, we should revisit the naïve conception of moral progress. According to Jamieson's proposal, a simple idea of moral progress can be stated as follows: "Moral progress occurs when a subsequent state of affairs is better than a preceding one, or when right acts become increasingly prevalent" (Jamieson 2002, 318). Supposedly, the first disjunct of the naïve conception is especially apt to accommodate the consequentialist view of moral progress. Before he argues that the naïve conception is acceptable from a wide range of normative perspectives, Jamieson notes that

it still might be thought that [the naïve conception] is more congenial to some normative views than to others. Consequentialists, for example, will be comfortable with the idea that states of affairs can be ordered according to their goodness (Jamieson 2002, 318–319).

When he qualifies the naïve conception, he says that "significant relations [must] obtain between the states of affairs being compared," i.e., "close causal, cultural, and temporal connections" (Jamieson 2002, 332). The mention of "cultural connections" makes it clear that this account of progress is, after all, supposed to capture changes that are related to the practices of agents—but the naïve conception never says so. Insofar as it speaks of improvements in states of affairs, it does not make it clear whether these should be restricted to include only improvements that are consequences of action. Without this restriction, the naïve conception allows for any increase in the value recognized by a consequentialist theory as the one to be increased to count as moral progress. A relevant improvement in states of affairs could then come about by a rainstorm ending a drought and the suffering caused by it or a genetic mutation that renders individuals immune to a painful disease—i.e., by processes that would be recognized as evaluands only by fully global consequentialists. This would entail that those who argue that not all "improvements from a moral point of view" are "instances of moral progress strictly speaking" (Buchanan and Powell 2018, 48) are wrong and that the world cannot simply become better in the sense of undergoing morally desirable changes, but that there is a literal way in which we could say that "the world" makes moral progress (cf. Rønnow-Rasmussen 2017, 147, who, however, argues that the world can only make moral progress if agents undergo certain positive changes; see § 4.2.2.).

Our earlier choice to require the property "relatedness to action" in any moral judgment, including the judgment that moral progress has occurred, forces us to reject such a wide conception of moral progress as is entailed by the first conjunct of the naïve conception in its unqualified form. The second conjunct on its own, though, accommodates only the evaluation of actions in terms of their rightness. While this would actually cover the conception of moral progress entailed by classical utilitarianism—which determines that acts are right when they maximize the good—it would not be easily adaptable to some non-consequentialist ethical perspectives, and it would not cover other forms of consequentialism either, which, as we just saw, might take other items than acts as their evaluands.

In view of the problem of accounting for even the consequentialist perspective (let alone the perspectives of virtue ethics or Kantian ethics) with the naïve conception in its original form, we might try to amend the first conjunct to at least cover all (except for fully global) consequentialist theories. The idea of moral progress consisting in an improvement in states of affairs is, after all, an attractive idea, so much so that it gives rise to the view that "[a]ll progress is in a sense moral progress" because "it is a movement towards a better state of affairs" (Ginsberg 1944, 11). In order to modify the naïve conception accordingly, we might want to speak of something like "improvements in states of affairs brought about by moral agents" instead of mere improvements in states of affairs.

Actually, the amendment just considered results in the kind of concept of moral progress that underlies a different proposal by Jamieson and Robert Elliot: a version of consequentialism they call "progressive consequentialism" (Jamieson and Elliot 2009). The background for the proposal is a certain criticism often leveled against consequentialist theories—the objection that they are overly demanding. Classical utilitarianism, for instance, with its requirement to maximize the good in any act that may count as right seems to require enormous efforts of moral agents in determining the right course of action as well as in terms of cutting back on the realization of their own ends. The demandingness objection to consequentialism is that it requires so much of agents that it renders the theory implausible (cf., e. g., Williams 1973). Jamieson and Elliot invoke the notion of progress in a solution to the demandingness problem which is based on the rejection of the requirement to maximize value and which they argue is superior to another consequentialist approach that rejects maximization, i.e., satisficing consequentialism (Slote 1984). Satisficing consequentialism caps the requirements faced by the moral agent by only demanding her to act so that consequences will be "good enough" (Slote 1984, 141). It relieves the moral agent from the task of finding and taking the best course of action, requiring her only to find an option that will do, morally. Jamieson and Elliot argue that while satisficing may be a good heuristic for decision-making—just as what it has been introduced (Simon 1955)—it is an objectionably arbitrary way to limit moral requirements, leaving the resulting moral provisions lacking in justification (Jamieson and Elliot 2009, 244). To remedy this failure of satisficing consequentialism while also establishing a way to limit moral requirements, they propose a progressive consequentialist approach, which "requires of agents […] that they act in such a way as to increase value in the world" (244), where "the baseline is the value of the world at time T2 [the time when the moral agent's impact on the world occurs] on the counterfactual assumption that the agent does not exist" (247). They describe the moral agent's task as set by progressive consequentialism and the resulting picture of overall improvement in the following terms:

Our mission as moral agents is to leave the world better than we found it. This struggle for improvement should be constant. The more we accomplish, the more that is demanded. Ourselves and others are held to ever higher standards as the world improves. (Jamieson and Elliot 2009, 245)

Now, it is not entirely clear whether they take the progress that is brought about by an action that is right from the perspective of progressive consequentialism to be moral progress, but this much is at least suggested not merely in the passage cited above but also when they indicate that their proposal aligns with the classical utilitarian perspective. In this context, they argue that Mill, in "recognizing the importance of moral progress" incidentally "sounds more like a progressive consequentialist than a textbook utilitarian" (Jamieson and Elliot 2009, 245, emphasis added). The alleged connection of their project to Mill's appreciation of moral progress suggests that they take themselves to be talking about the same thing as Mill: moral progress. The changes progressive consequentialism requires the moral agent to bring about supposedly qualify as instances of moral progress.

While bringing about moral progress initially might seem like something that exceeds what can be asked of the moral agent, in the form of progressive consequentialism it becomes something that reduces the strain on the moral agent: what is expected of her is mere progress. By substituting the initiation of progress for the more demanding task of maximizing utility, Jamieson and Elliot portray inducing moral progress as a moderate task. The worry that inducing moral progress might be a too moderate aim for the moral agent will be discussed in § 5.3.3. Here, we shall focus on the connection between right action and moral progress as established by progressive consequentialism.

That moral agents who succeed at improving the world (in morally relevant ways) should be recognized as bringing about moral progress is at least strongly suggested in the portrayal of progressive consequentialism and its connection to the work of Mill as well as by the naïve conception. But this way of using the term 'moral progress' is problematic. It locates moral progress in the states of affairs the moral agent manages to improve.

As a reaction to fully global consequentialism, we have just considered an amendment of the naïve conception in order to account for the fact that valuable changes must be related to agency and, thus, morality, in order to count as moral progress. In this version, the naïve conception would say that "moral progress occurs when improvements in states of affairs are brought about by moral agents, or when right acts become increasingly prevalent." This still allows for what progressive consequentialism suggests: that the moral agent may perform one singular act and manage to bring about moral progress as she manages to improve the states of affairs. This view is theoretically costly. It takes away a conceptual distinction between right action and the initiation of moral progress, between changes in states of affairs that are effected by singular acts of a moral agent and changes over time within the moral agent's acts themselves. This view makes the states of affairs the prior undergoer of moral progress. It is in states of affairs that durable and directed change occurs, not the moral agent's practices. Excluding this option was a welcome consequence of the earlier decision to require moral judgments to be related to action (§ 2.2.1.), and a motivation for accepting that the view of moral progress to be laid out in the following chapter should not accommodate the perspective of fully global utilitarianism. Undoubtedly, singular acts can have exceptionally good consequences that have a persisting impact. The saving of a life might be achieved in an instance and have positive consequences for years to come. But should we therefore say that the moral agent who manages to perform such a highly value-increasing act is bringing about moral progress? Using the term 'moral progress' in this way means to waive the alternative, which is to use it for processes that pertain to the agent and her performance itself. It does not do justice to the procedural nature of progress (see § 2.1.).

Of course, nothing strictly prohibits calling any valuable changes effected by a moral agent 'moral progress.' But the fact that we can simply call them 'improvements' or 'progress' or 'axiological progress' (in order to emphasize that they involve or are constituted by an increase in some value) and reserve the term 'moral progress' for improvements in the realm of what we might address as "moral performance" creates a strong pragmatic presumption in favor of the distinction. Improvements in moral agents' actions can be determined based on consequentialist criteria, i.e., in terms of their increasingly good outcomes. But this improvement in how moral agents do can only be addressed distinctly as moral progress if this term does not already apply to the outcome of each individual act on its own.

While progressive consequentialism may be an attractive alternative to maximizing and satisficing utilitarianism, equating what the successful progressive consequentialist moral agent achieves with moral progress commits one to a view of morality that renders talk of moral progress redundant: anytime the moral agent does not fail morally (because she changes the state of the world for the better) she brings about moral progress. The failure to distinguish between right action and the initiation of moral progress seems like a clear loss in fruitfulness of the concept of moral progress. To avoid it, one would have to reserve the term 'moral progress' to changes undergone by moral agents, not states of affairs. This will be the view of moral progress to be investigated here in more detail. It does not rule out an otherwise consequentialist approach to moral progress: the moral evaluation of the relevant changes could still be done in strictly consequentialist terms, i.e., exclusively via reference to the change in quality of the outcomes of the moral agent's actions over time.

Not only have classical utilitarians been especially friendly to paradigmatic progressive causes due to their commitment to inclusivity—a feature of morality with respect to which corrections in moral communities are regarded as important types of moral progress. Consequentialism in general appears to be especially relevant to the topic at hand because it can provide a straightforward and compelling answer to the question: what is moral progress? The answer, in short, seems to be: better outcomes. We have seen that it is, after all, a controversial question to ask how this seemingly simple answer can and should be understood. The consideration of (fully) global consequentialism and its deviation from our prior commitment to view moral judgments as intimately linked to agency have given rise to some queries concerning the naïve conception and its capacity to accommodate consequentialist conceptions of moral progress. The brief discussion of progressive consequentialism has pointed us to a worthwhile distinction between better states of affairs as the outcomes of right acts on the one hand and developments that are progressive in the sense of yielding increasingly good outcomes on the other hand. It seemed that only the latter ought to be viewed as candidates for manifestations of moral progress. The discussion has resulted in the decision to view moral progress as restricted not only to changes effected by moral agents but also to changes in which the moral agent continues to be intimately involved—that are, in some sense, her changes. The conclusion from these reflections is that if moral progress is partly constituted by improvements in states of affairs, these improvements cannot come about without the continued involvement of moral agents. Moral progress does not bypass the moral agent.

#### 3.2 A Kantian Reservation

The previous decision to draw a distinction between making or bringing about moral progress on the one hand and acting rightly on the other hand amounts to viewing moral progress as "located" or at least firmly "anchored" in moral agents themselves. This is already a fairly substantial conceptual commitment. On this view, the project of "making the world a better place" is to be distinguished from the project of bringing about moral progress, because the latter, unlike the former, requires continued agential involvement in the relevant improvements. A moral agent might succeed at improving the world through a singular act. By contrast, for such an act to be effecting moral progress, the relevant value increase would have to be grounded in lasting changes in (other) moral agents' practices. This delineation of a working account of moral progress was reached simply by rejecting the idea that states of affairs might be the undergoers of moral progress, which was motivated by the commitment to seeing moral judgments as being closely related to action.

This section considers a concern about the attributability of moral progress which will be relevant to our explication—a concern which I will broadly label a "Kantian reservation," as it is articulated or at least can be extracted from Kant's own engagement with the problem of human improvement. I will not delve into the details of Kant's moral philosophy or his philosophy of history and the question of their compatibility here. The question of whether Kant's philosophy can make room for moral progress and what exactly his conception of it is (cf., e. g., Kleingeld 1999, Lindstedt 1999, Anderson-Gold 2001, Stroud 2005, and Moran 2012) shall not concern us here. Instead, certain remarks in his engagement with the question of improvement in mankind will be used to articulate potential constraints on an account of moral progress. The considerations discussed here are not exclusive to Kant, but they are readily associated with him, and his work on the question of progress is a pertinent source for the matter at hand.

In The Contest of the Faculties, Kant addresses the "old question": "is the human race constantly progressing?" He is concerned with the question of whether progress has occurred in the past and, more importantly, whether it will necessarily continue in the future. He initially identifies a principled reason to reject the optimistic conjecture that mankind will continue to progress morally:

It may always be conceded that the proportion of good and evil elements inherent in our predisposition remains constant and can be neither augmented nor diminished in the same individual; how then should the quantity of good in our predisposition increase? For that would happen only through the freedom of the subject, for which purpose the subject would in turn require a greater reservoir of good than it now possesses. The effects cannot surpass the power of the efficient cause; thus the quantity of good mixed in man with the evil cannot exceed a certain measure beyond which it would be able to work its way up and thus ever proceed toward the better. (Kant 1979, 141)

The idea that is articulated here—that human nature is not amendable by the individual—must be read in the context of Kant's argumentative aim: he is delimiting possible answers to the question of the grounds on which we might affirm that mankind is continually changing for the better. What he establishes here is that the hope for improvement cannot be based on the expectation that the individual will transform and expand her naturally given moral capacities. Individual moral development is limited by the inherent "proportion of good" in a moral agent—there is no possibility of bootstrapping and extending moral progress beyond the activation of fixed individual predispositions.

This is not an objection to a concept of moral progress that centers on change in moral agents, but to a certain interpretation of what that agential change might entail. It is akin to a contemporary type of view that Buchanan and Powell call "evoconservative" and which combines the insight that moral capacities are evolved features with a certain idea of what the limits of these capacities are (Buchanan and Powell 2018, § 4). Ideas about evolutionary determined limits of moral progress are precisely what takes center stage in the most comprehensive contributions to the present debate on moral progress. A core concern in The Expanding Circle is to show that there is no biologically hard-wired psychological barrier to the expansion—that the content of moral norms adoptable by human beings is not biologically determined simply because the cognitive capacities underlying moral reasoning and the strategies for resolving conflicts among social creatures are products of evolution. For Singer, the emergence of reasoning capacities makes human moralities independent of their evolutionary heritage. As he puts it: "The blind forces of evolution have thrown up creatures with eyes. Being able to see, if they dislike the direction in which these same blind forces are taking them, they can change course." (Singer 2011, 169) The principle of disinterestedness in particular has the potential to transform customary and evolved strategies of conflict management into more inclusive moralities:

The idea of a disinterested defense of one's conduct emerges because of the social nature of human beings and the requirements of group living, but in the thought of reasoning beings, it takes on a logic of its own which leads to its extension beyond the bounds of the group. (Singer 2011, 114)

This transformative potential and the fact that it has manifested in human history is paramount to the argument that moral progress beyond the limits of systems of norms which enhance biological fitness is possible. In The Evolution of Moral Progress, Allen Buchanan and Russell Powell crucially rely on the observation that the "circle" of moral concern has, as a matter of fact, expanded as Singer describes and demands it: the extension of moral consideration to nonhuman animals and their inclusion in some schemes of legal protection is the first among several "inclusivist anomalies" they cite as evidence opposing the idea that morality is biologically determined (Buchanan and Powell 2018, 153). At some point, inclusivist tendencies become "anomalies" from the point of view of "ecoconservatism," i.e., the view that there are "strong evolutionary moral constraints" (Buchanan and Powell 2018, 152) in that morality can only develop within the limits set by its evolutionary history which favored tribal and strategic rather than inclusivist moral thinking (Buchanan and Powell 2018, 138–145). While they disagree with Singer's explanatory focus on the role of reason for the emergence of "inclusivist anomalies" (Buchanan and Powell 2018, 148) and his insinuation that moral progress is fully characterized by reference to the "expanding circle," they agree that the existence of inclusivist anomalies proves transformative moral progress to be possible despite morality's evolutionary basis, and they also hold that change toward greater inclusivity is, normatively, one of the most important respects in which morality ought to progress further.

Any specific suggestion regarding supposed limitations of human moral progress can always be criticized as a premature empirical conjecture (cf. Buchanan and Powell 2018, Chapter 4). Precisely because we do not know what boundaries our evolved moral capacities set to moral progress, we can hope for further progress and should work for any kind of worthwhile moral cause. If there were ultimate biological limits to moral progress, and if moral progress requires agential change, it could still come about where agents managed to exhaust their moral capacities to a greater extent, not expand them.

Incidentally, Kant would agree that improvement is possible, but his remarks on the type of improvement that we may expect point us to reasons for pessimism when it comes to moral progress—and to a potential conceptual constraint. Kant identifies a "historical sign" (Kant 1979, 151) for the "capacity of the human race to be the cause of its own advance toward the better" (Kant 1979, 151). He identifies it in the response of contemporaries to the French Revolution and the similar responses of people in other periods to political upheaval:

It is simply the mode of thinking of the spectators which reveals itself publicly in this game of great revolutions, and manifests such a universal yet disinterested sympathy for the players on one side against those on the other, even at the risk that this partiality could become very disadvantageous for them if discovered. […] owing to its disinterestedness, [it demonstrates] a moral character of humanity, at least in its predisposition (Kant 1979, 153)

Because the display of sympathy to one of the struggling parties is always risky, Kant takes it as a sign of genuine concern for the moral causes that are at stake and, as such, a force for further improvement, because its accomplishments will not be forgotten even if they may be rolled back temporarily (Kant 1979, 159– 161). So, Kant ultimately affirms that there is, has been and will be progress in the history of mankind. However, he makes it clear that the improvement he insists will certainly occur will not yield improvements in morality, strictly speaking:

What Profit Will Progress Toward the Better Yield Humanity? Not an ever-growing quantity of morality with regard to intention, but an increase of the products of legality in dutiful actions whatever their motives. That is, the profit (result) of man's striving toward the better can be assumed to reside alone in the good deeds. Of men, which will become better and better and more and more numerous; it resides alone in phenomena constituting the moral state of the human race. For we have only empirical data (experiences) upon which we are founding this prediction, namely, the physical cause of our actions as these actually occur as phenomena; and not the moral cause-the only one which can be established purely a priori—which contains the concept of duty with respect to what ought to happen. (Kant 1979, 165)

Not only do the responses of contemporaries to social upheaval, which serve as a "historical sign" for the capacity for improvement, reveal mere sympathy, i.e., a passionate response, rather than the good will, the results of the improvement that is to be expected are likewise not moral improvements in the ultimate sense—they are mere increases in dutiful action, not increases in action from duty. There is also an epistemic limitation involved in this result: we only face the phenomena of human practices and they do not allow inferences to the good will.

Though the sympathetic response of spectators "borders closely on enthusiasm" (Kant 1979, 153), and though "genuine enthusiasm always moves only toward what is ideal and, indeed, to what is purely moral, […] and it cannot be grafted onto self-interest," sympathy remains a passionate response and "as such deserves censure" (Kant 1979, 155). Kant can neither acclaim the forces nor the results of improvement as genuinely morally worthy. The improvement that is to be expected, which is not an increase in morality, is therefore, in one sense—maybe in its core sense—not moral progress.

Since we are not in the business of building a concept of moral progress on Kant's moral philosophy, but only want to use his thoughts as pointers to possible modifications on a hopefully fruitful concept of moral progress, we need not subscribe to his ethical theory in our engagement with moral progress. Instead, we can consider his reservation against processes that lack an increase in moral worth as a hint at a potential condition on the concept of moral progress—in the form of a requirement for the relevant change to be driven by the right motives. If we are in principle prepared to accept changes in moral agents' practices that yield better outcomes at least as candidates for moral progress, we should address a Kantian reservation against this idea of moral progress. It is a reservation against accepting any such improvement in morally relevant practices as moral progress—irrespective of the motives that drive moral agents.

It might be argued that by referring to Kant as a source for a potential requirement of right motives we have fallen prey to the false "classroom cliché" that "Kantians are concerned with the motives for our actions, while utilitarians only care about the results" (Arpaly 2002, 224). Nomy Arpaly notes that this conception stems from a failure to appreciate the distinction between an "action's moral desirability," i.e., its deontic status and its place in a ranking of alternative courses of action in terms of their quality, and "moral worth," i.e., the praiseworthiness or blameworthiness of the agent in virtue of the action. As far as the action's deontic status is concerned, Kant acknowledges that an individual acting from prudential reasons can do "the right thing" and, in turn, utilitarians can draw on more than the action's result to determine the agent's praiseworthiness. They, too, can care about motives. While this may be so, the discussion in this section supports labelling the reservations presented here "Kantian." To use the terminology employed by Arpaly: the question that emerges here is whether the concept moral progress applies in virtue of a) an increase in morally desirable action (where desirability might be evaluated in terms of consequences), which would be a natural answer for consequentialists to give, or b) in virtue of an increase in morally worthy action. The caveat Kant enters when he affirms the existence of progress—his insistence that it is not an increase in morality that is to be expected—seems to delineate a notion of progress in terms of option b), although this move is combined with the denial that this notion is ever applicable.

The choice between a) and b) concerns the comprehensiveness of the evaluation that underlies a moral progress-judgment. Should the evaluation of moral progress include more than rightness of actions in terms of their "legality" should it include (alongside, possibly, an evaluation of outcomes) a more comprehensive assessment of the moral agent that covers her motives—and if so, in what sense?

The Kantian reservation against accepting a mere increase in right actions as moral progress must be taken into account—as an option—when explicating moral progress. It points to a potential further distinction and an even narrower concept of moral progress than the one based on the distinction between morally desirable changes and morally desirable changes in moral agents' practices. We might end up requiring, in a Kantian spirit, that improvements in moral agents' practices involve moral worth-conveying motives or an improvement with respect to such motives in order to count as moral progress. For now, such a requirement is merely an option—to be revisited in the explication of moral progress in Chapter 4. Including such a requirement would not commit us to a more comprehensively Kantian view of moral progress: in a more optimistic spirit than Kant's, we could think that improvements of the relevant kind are possible and detectable. In addition, we would not need to give a Kantian specification of what makes right motives right. There is a wide range of potential requirements for "right motives"—we could be satisfied with agents being only minimally aware of the moral relevance of their action or we could eventually require them to pursue their own improvement actively in order for them to count as making moral progress. These questions, pertaining to the comprehensiveness and rigidity of the moral evaluation that is relevant to judgments of moral progress, will concern us in § 4.2.

#### 3.3 Moral Progress as the Cultivation of Virtues?

If we say that moral progress is located in the moral agent rather than in the worldly conditions that she might influence (§ 3.1.), and especially if we were to decide that an increase in conformity between agents' acts to the prescriptions of morality did not yet justify a judgment of moral progress (§ 3.2.), it might seem as if we were moving toward a virtue ethical approach to the phenomenon of moral progress. Arguing that it is the agent who must come to act better and not the states of affairs that must improve in order for moral progress to occur seems to amount to the claim that for moral progress to occur, moral agents must become better people. The major approaches to normative ethics are typically characterized and differentiated from one another by pointing out that consequentialism is crucially concerned with outcomes, deontology with acts and their conformity to duty, and virtue ethics is concerned with the moral agent and her character. Therefore, if the site of moral progress is the moral agent, it seems as if virtue ethics is inescapably the approach on which we must rely to characterize moral progress properly. In fact, a virtue ethical view of moral progress may force itself onto us as soon as we decide that it is moral agents who have to be the undergoers of moral progress. Consider the following brief summary of Stan Godlovitch's view of moral progress. It mentions changes that are readily associated with a consequentialist and, respectively, deontologist view of moral progress: a widening domain of moral concern (as described by Singer) and the growing stringency of moral requirements. However, the way in which these two aspects are connected in an account of the progressing moral agent is reminiscent of a virtue ethical perspective:

we progress morally by virtue of becoming ever more the kinds of beings for whom an everwidening moral scope and ever-encroaching moral stringency both spontaneously and reflectively influence our personal and collective choices. (Godlovitch 1998, 281)

The change in "the kinds of beings" they are that is undergone by progressing moral agents is crucially characterized in terms of the influence of moral considerations on their spontaneous and reflective choices. This sounds like progress in terms of character improvement, which is precisely the focus of virtue ethicists' concern.

Virtue ethics focuses on the moral agent and recommends the cultivation of virtues as a path to the good life, where the good life for the agent is likewise the morally good life.Virtues are complex dispositions to act on the relevant reasons that the agent discerns by applying practical wisdom or phronesis rather than abstract principles (Hursthouse and Pettigrove 2018, § 1). Virtue ethicists typically have reservations regarding the merits of abstract moral theory and doubt the possibility of codifying morality. In addition, they see the supersession of virtue ethics by abstract ethical theories as an ethical decline (Anscombe 1958; MacIntyre 1984).

Virtue ethics may be seen as especially conducive to developing a concept of moral progress. Its focus on the moral agent's character (instead of the evaluation of singular acts) presents us with a persisting subject of change and requires a wider perspective on the moral agent's performance for moral evaluation. Virtue ethicists caution against judging moral agents based on isolated acts, which are usually insufficient evidence for the moral agent's character and her motivating reasons (Sreenivasan 2002). As Moody-Adams puts it: "Statements concerning character traits are fundamentally elaborate inductive hypotheses rooted in the observation of patterns of behavior, and the circumstances in which such behavior is displayed" (Moody-Adams 1991, 117). What makes it so relevant for the topic at hand is that virtue ethics is uniquely concerned with moral improvement, i.e., the cultivation of virtues. Individual moral development for the better is an integral concern for virtue ethics, so that any virtue ethical theory might be plausibly understood as an account of (individual) moral progress. However, several considerations count against presupposing virtue ethics as the supreme approach to the very concept of moral progress.

First, Moral progress does not necessitate adopting a virtue ethical (or even virtue theoretical; see below) perspective, as it is not a virtue concept like courage, truthfulness, or modesty. This is so not only because it lacks material content, but because it does not obviously pick out a disposition in the first place. It indicates a process.

Second, moral progress need not even be construed as a meta-virtue concept in the strict sense that would relate to the kinds of improvements in moral agents on which virtue ethics focuses. That the way in which moral agents progress must be reconstructed in terms of their cultivation of certain dispositions as described by virtue ethics would be a substantive theoretical commitment. Several concerns about virtue ethics oppose relying heavily on it not only to explicate moral progress, but also to develop a more comprehensive theory of moral progress. i) Virtue ethics is focused on the moral agent herself in a way that is sometimes found to be problematic. Traditionally, it is eudemonistic, i.e., it is concerned with the moral agent's own good life as the ultimate value. This agent-centeredness has provoked the objection that virtue ethics does not provide the moral agent with proper motives for moral action (Hurka 2001, 246–247), i.e., charges of self-effacingness (§ 5.2.). The concern is that by making the moral agent strive for a good life, virtue ethics gives her the wrong kind of reasons for moral action. ii) Due to the rejection of abstract theoretical reasoning, virtue ethics is sometimes said to lack critical potential or to neglect intellectual resources for achieving moral progress. Samuel Scheffler, for instance, argues that aversion to abstract theorizing shows disregard for the commitment to certain values it exemplifies and the instrumental value it can have:

the justificatory enterprise in which much contemporary moral philosophy has been engaged […] is closely linked with larger social practices of moral inquiry and debate. These practices […] give expression to fundamental social values of truthfulness, accountability, openmindedness, and rational equality […] [and therefore are] important instrument[s] of social change. (Scheffler 1992)

iii) When virtue ethicists claim that the cultivation of virtues is dependent on familiarity with preexisting practices and tradition (MacIntyre 1984), they seem to import a tendency toward conservatism and relativity into morality which might be seen as antithetical to progress. I will assume that there is enough to these concerns to have initial reservations against a virtue ethical approach to moral progress.

Third, it is anything but obvious that we even have to think of moral progress as a meta-virtue concept in a much weaker sense, which does not establish an inherent connection to distinctly virtue ethical reasoning, but retains the focus on the moral agent's character. Certainly, "any plausible normative ethical theory will have something to say about" virtues (Hursthouse and Pettigrove 2018). Any normative theory might even have enough to say about morally desirable dispositions to furnish a virtue theory. Nonetheless, it is not clear that moral progress should be understood in this virtue-theoretical sense, either. The more weight we give to changes "on the ground" (Jamieson 2002, 318) in applying or withholding the concept of moral progress, where we take these to be changes in morally relevant states of affairs effected by moral agents, the less apt it would seem to characterize moral progress as a concept uniquely connected to (mere) dispositions, which might, after all, remain unactualized. A plausible weighting of requirements for moral progress in terms of changes in moral views and dispositions, actual practices, and outer states of affairs will occupy us throughout Chapter 4. It is not clear at the outset what the proper weighting of these factors is and whether it justifies a characterization of moral progress as a meta-virtue concept, even loosely speaking.

Virtue ethics seems to provide a useful perspective on moral progress in that it is crucially concerned with the moral agent's disposition to act well. This is what it seems we are likewise zeroing in on, if we decide that moral progress is a process undergone by moral agents (and not occurring in states of affairs) and that it requires a specific mindset on the part of the morally progressing agent. However, the congeniality of the virtue ethical perspective to the concept of moral progress may be superficial: dispositions for moral action—virtues—are of interest to and can be normatively characterized by any normative theory. Furthermore, it may turn out that moral progress requires changes in dispositions, but not only in dispositions.

#### 3.4 Liberty, Progress and Inevitable Ethical Regress

In this section, we shall widen our perspective to look beyond the field of normative ethics and take into account considerations pertaining to the political conditions under which moral progress may take place. This widened perspective will bring into view evaluative conflicts that may await if we require improvements both in moral agents' behavior and their mindset in order to attribute to them moral progress.

The proliferation of liberal values is commonly taken to be a manifestation of moral progress in and of itself. Liberal democracies are themselves regarded as products of past progress, in the course of which autocracies have given way to political systems that do justice to the moral relevance of individual persons, safeguard basic human rights as well as civil and political rights, limiting the power of the state. Michael Huemer takes the idea of liberal democracies being progressive accomplishments to a metaethical conclusion. He sees worldwide trends toward greater realization of liberal values (Huemer 2016, § 3) broadly summarized as the recognition of the "moral equality of persons," "respect for the dignity of the individual," and opposition to "gratuitous coercion and violence" (Huemer 2016, 1987)—as evidence of a striking convergence in moral views and according practices, in other words, evidence of moral progress. This convergence, according to Huemer, licenses belief in the objective truth of liberal values, i.e., moral realism. He diagnoses "a dramatic shift in human values over history," away from the glorification of warfare (Huemer 2016, 1988) and from acceptance of torture, death penalty, slavery (Huemer 2016, 1990), exhibited also in the abandonment of racist, sexist policies (Huemer 2016, 1992), the spread of democracies (Huemer 2016, 1992–1993), and decolonization (Huemer 2016, 1993). He concludes that "there has been enormous moral progress over human history," not merely changes in practices but changes in "moral beliefs" (Huemer 2016, 1993). As far as alternative ways of accounting for the observable changes are concerned, Huemer focuses on two evolutionary debunking explanations, arguing that the change that has occurred has been "too rapid to be explained by biological evolution" (Huemer 2016, 1995) and too uniform to be explained by cultural evolution (Huemer 2016, 1999). Instead, the objective correctness of liberal values—independent of our valuing them—and the efficacy of efforts toward a greater appreciation of correct values supposedly best explain why societies worldwide tend to reform institutions in accordance with liberal values (Huemer 2016, § 5).

We need not endorse the abductive inference to moral realism to appreciate the fact that liberal values set the stage for pending moral progress in many places around the world and that this situation is widely regarded as a moral accomplishment in itself. The practical challenges for further improvement that are encountered in this kind of setting are discussed by one of the most important proponents of liberalism: in the exposition of his views on the scope of individual liberty, John Stuart Mill addresses the at times antithetical relation between progress and liberty prominently and extensively (Mill 2008a). His remarks will be of interest to the present topic because they point us to the typical obstacles for achieving morally desirable change within a liberal democracy. They also point us to the theoretical tension between, on the one hand, the belief in the possibility of such change and, on the other hand, to the recognition of human fallibility, which is essential for the justification of key elements of the liberal agenda. In this respect, Mill's discussion of the benefits of freedom of expression is relevant to the present project because we can extract from it a perspective on moral progress according to which the proliferation of better moral views comes with a psychologically inevitable downside, a certain kind of ethical regress. According to Mill, the understanding of beliefs becomes shallower as their acceptances grows. With respect to moral beliefs, this suggests an inherent downside to the in principle desirable spread of improved moral views in the moral community: a decrease in the intensity of the moral agent's ethical involvement in morally progressive practices. This prospect is important to bear in mind when proposing an explication of moral progress—in which a requirement of ethical involvement may or may not be inscribed.

#### 3.4.1 The Practical Problem of Liberal Progress

In the first paragraph of On Liberty, Mill addresses the need to challenge the endpoints of prior progressive developments: the need for progress to transgress its own results (i.e., to be transformative progress, cf. § 2.1.). The case with which he is concerned is the transformation of democracies to liberal democracies. He states that the topic of his essay, "the nature and limits of the power which can be legitimately exercised by society over the individual," concerns a question that "in the stage of progress into which the more civilized portions of the species have now entered, […] requires a different and more fundamental treatment" than ever before (Mill 1859, 3). The relevant stage of "progress of human affairs" is marked by the transition of autocracy to democracy (Mill 1859, 5). At this stage, the democratic achievement of "self-government"—though occurring through the realization of what was once only a utopian ideal (Mill 1859, 6)—comes to be seen as flawed itself, because the term glosses over the fact that individuals are not ruled by themselves but by all (others), or rather by "the most numerous or the most active part of the people" (Mill 1859, 7, emphasis in original). Once political power has been transferred to the people, the question arises what limits the power of the people over their co-citizens (Mill 1859, 6). Mill mentions that the problem that this question concerns is recognized by many, as evidenced by the phrase "tyranny of the majority" (Mill 1859, 7). This particular mode of tyranny is much more pervasive than other forms of tyranny, residing not only in the branches of government, but also in public opinion (Mill 1859, 7). For this reason, protection "against the magistrate" as well as "against the prevailing opinion and feeling" is of prime importance (Mill 1859, 7, 8).

The problem of determining and securing the boundaries of individual liberty exemplifies in a special way a more general challenge. Once realized, a former formulation of an ideal can come under scrutiny itself and a deficit with respect to the hitherto developed understanding of the ideal comes into view. Ongoing progress comes with the challenge to reconsider and reevaluate earlier progressive achievements. For the ideal of liberty, the challenge is particularly profound, because the realization of liberty is itself instrumental to progress.

Mill sets out to reformulate the ideal of liberty in terms of individual liberty. The limit of individual freedom set by this ideal is grounded in others' right to "self-protection" (Mill 1859, 13, 14), so that the only justifying reason for the exercise of power over the individual is the prevention of "harm to others" (Mill 1859, 14). The harm principle excludes the protection of the individual's own good as a potential reason for interfering with her freedom. Thus clarified, the ideal of individual liberty sets a boundary for the pursuit of further morally desirable change. Further moral progress that would result in restrictions of hitherto granted individual liberty must be demonstrably covered by the harm principle: further moral improvement, if it is to be enforced on a societal level rather than individually pursued, must be shown to be a means to ensure the "prevention of harm to others." The restriction on the pursuit of moral agendas on a societal and especially on the state level is echoed in modern liberal work in the doctrine of liberal neutrality on the good (Dworkin 1978) and Rawls' exclusion of "comprehensive doctrines"—moral, religious or otherwise—from public reasoning (Rawls 2005, 441).

The establishment of this threshold for the legitimacy of efforts for moral progress gives rise to a paradox for moral agents who are morally invested in any value besides liberty. Respect for individual liberty can require them to tolerate—for moral reasons—what, according to their own additional moral commitments, seems morally wrong. This is the "paradox of moral tolerance" (Forst 2012, § 1), which Mill discusses in the context of freedom of religion (Mill 1859, 11–12). A commitment to liberty commands a paradoxical attitude toward one's other moral convictions and places a restriction on the pursuit of moral progress under the auspices of liberty, which can require the individual to tolerate preservation of what she herself would like to see improved. The paradox of toleration is merely a psychological challenge to the moral agent as long as she agrees that her additional moral agenda is precisely that—and additional agenda, and not a contribution to the further specification and better implementation of the ideal of individual liberty itself (i.e., an agenda that is not covered by the harm principle). But many controversies over supposedly progressive causes can be seen as struggles over the recognition of some agenda as a due correction to an already accepted ideal. Mill addresses this problem as well. In the context of his defense of individual liberty, it presents itself as a practical challenge for the implementation of further improvements covered by the harm principle.

The primacy of the ideal of freedom from coercion can createapowerful barrier for corrections to the accepted limits of liberty: a typical problem for progress in a liberal society is presented by the obstacle of misinterpreted individual freedom and overlooked obligations. Mill illustrates this using the example of disregard for parental duties and what we might refer to as misinterpreted procreative freedom:

liberty is often granted where it should be withheld, as well as withheld where it should be granted; and one of the cases in which, in the modern European world, the sentiment of liberty is the strongest, is a case where, in my view, it is altogether misplaced. […] It is in the case of children, that misapplied notions of liberty are a real obstacle to the fulfilment by the State of its duties. One would almost think that a man's children were supposed to be literally, and not metaphorically, a part of himself, so jealous is opinion of the smallest interference of law with his absolute and exclusive control over them; […] Hardly any one indeed will deny that it is one of the most sacred duties of the parents (or, as law and usage now stand, the father), after summoning a human being into the world, to give to that being an education fitting him to perform his part well in life towards others and towards himself. But while this is unanimously declared to be the father's duty, scarcely anybody, in this country, will bear to hear of obliging him to perform it. […] It still remains unrecognised, that to bring a child into existence without a fair prospect of being able, not only to provide food for its body, but instruction and training for its mind, is a moral crime, both against the unfortunate offspring and against society. (Mill 1859, 135– 136)

Mill does not explicitly argue here that the accomplishment of the hitherto achieved respect for individual liberty has amplified the problem of neglect of parental duties or the lack of governmental intervention—the problem might well be rooted in ideas about parental authority predating the proliferation of liberal values. At the very least, though, a strong but misplaced "sentiment of liberty" has been used as a new warrant for maintaining the status quo. The fact that the status quo appears to be covered by a newly esteemed moral principle contributes to a kind of moral inertia with respect to further improvements that are entailed by the same principle.

The paradoxical effect of progress with respect to liberty is even more pronounced where the relevant practices have not predated but have been formed by some progressive development themselves. The more a practice is valued as an accomplishment of past progress, the more difficult it becomes to establish further moral improvements that require a transformation in those earlier results of progress. Especially if the change for the better is thought of in categorical terms as a correction of a prior wrong (rather than in comparative terms as an improvement) it can lead to the need for further transformation being overlooked. For example, anthropocentrism (the moral doctrine that all humans matter morally and that they matter more than nonhuman beings) has come to be viewed as an advancement over more exclusivist doctrines (tribal, nationalist, racist, sexist doctrines). The fact that a certain change is regarded as an improvement provides a strategic reason to frame a further progressive cause as a continuation of that already welcomed change. So, proponents of the equal consideration of the interests of human and nonhuman animals (Singer 2009) or animal rights (Regan 2004; Donaldson and Kymlicka 2011) have often made an effort to present their case as a continuation of arguments and struggles for basic liberties of disadvantaged groups of humans—the women's rights, civil rights or disability movement. However, their arguments and practical conclusions have been met with conspicuous reservations by those who are supportive of "paradigmatically progressive causes" in the tradition of these movements—the political "Left" (Kymlicka and Donaldson 2014, 116). Arguments for the moral consideration of animals entails the restriction of (misconceived) human liberties (in the interaction with nonhumans). In other words, they seek to revise what is seen as a result of prior moral progress. The new cause is perceived as a transformation—and not merely a continuation—of the results of prior progress. In such a case, the new cause is likely to be met with reservation—to the extent that the prior result has been endorsed as an improvement. Something like this might be behind the "Left's" hesitancy to overcoming anthropocentrism. Once something is being accredited as an accomplishment, it becomes an obstacle for endorsing what is proposed as a due transformation of that accomplishment (for a related argument with a different focus, cf. Kymlicka and Donaldson 2014). The furthering of inclusivist causes within the limits set by anthropocentrism is a case in point.

Mill takes explicit note of this kind of tension between progress and liberty, but suggests that it will ultimately be of only temporary duration:

The spirit of improvement is not always a spirit of liberty, for it may aim at forcing improvements on an unwilling people; and the spirit of liberty, in so far as it resists such attempts, may ally itself locally and temporarily with the opponents of improvement […]. (Mill 1859, 90)

The real conflict is not necessarily to be found between liberty and progress, but between progress and liberty on the one hand and custom on the other—custom being the normative system that demands compliance without proper grounds for this demand:

The despotism of custom is everywhere the standing hindrance to human advancement, being in unceasing antagonism to that disposition to aim at something better than customary, which is called, according to circumstances, the spirit of liberty, or that of progress or improvement. […] the only unfailing and permanent source of improvement is liberty, since by it there are as many possible independent centres of improvement as there are individuals. The progressive principle, however, in either shape, whether as the love of liberty or of improvement, is antagonistic to the sway of Custom, involving at least emancipation from that yoke; and the contest between the two constitutes the chief interest of the history of mankind. (Mill 1859, 90)

So, it seems that the challenge the promoter of further improvement faces is the challenge to demonstrate that the recognition of past progress involves a misconception rendering what was once an accomplishment a mere custom that resists further improvement.

The practical problem for progress in liberty is an indicator of a deeper theoretical tension between the rationale for liberal values and a certain conception of moral progress. What underlies the practical challenge of promoting transformative change with respect to earlier results of moral progress is a paradoxical relation between the recognition of progress as progress and the reasoning behind a key element of the liberal agenda: freedom of expression. The next subsection will bring out the way this theoretical tension can inform our explication of moral progress.

#### 3.4.2 The Theoretical Paradoxes of Fallible Progress

Early in his essay, Mill notes that "man" is by nature "a progressive being" (Mill 1859, 15) but also that improvements may be forced upon individuals prior to a certain stage of maturity—after that, liberty is the enabling condition of further progress.

The early difficulties in the way of spontaneous progress are so great, that there is seldom any choice of means for overcoming them; and a ruler full of the spirit of improvement is warranted in the use of any expedients that will attain an end, perhaps otherwise unattainable. […] Liberty, as a principle, has no application to any state of things anterior to the time when mankind have become capable of being improved by free and equal discussion. (Mill 1859, 14–15)

Once humans mature—both individually and collectively—liberty becomes a precondition for further intellectual progress. Two of its central tenets, the "liberty of conscience" and freedom of expression, bestow critical epistemic benefits on the liberal society: they secure opportunities for "exchanging error for truth" and for "the clearer perception and livelier impression of truth, produced by its collision with error," respectively (Mill 1859, 23). These are two different benefits to be reaped by different parties to a disagreement. The first is an opportunity for belief revision—a desirable effect of a difference in opinions reflecting the "quality of the human mind […] that his errors are corrigible" (Mill 1859, 26). It is with respect to this epistemic benefit that Mill asserts that "[a]ll silencing of discussion is an assumption of infallibility" (Mill 1859, 23). It is because human beings are fallible but also able to correct their errors that a free exchange of opinions is of prime importance for intellectual progress, as it creates opportunities for corrections. But there is a second benefit to be had, viz., the deepening of understanding. As soon as a true belief "is not fully, frequently, and fearlessly discussed" anymore, it "will be held as a dead dogma, not a living truth" (Mill 1859, 45). The grasp of the truth, then, will be shallow and as a result, we might not be able to ascribe knowledge to the subject anymore:

assuming that the true opinion abides in the mind, but abides as a prejudice, a belief independent of, and proof against, argument—this is not the way in which truth ought to be held by a rational being. This is not knowing the truth. Truth, thus held, is but one superstition the more, accidentally clinging to the words which enunciate a truth. (Mill 1859, 45–46)

Here, Mill goes so far as to claim that the disappearance of dissenters causes adherents of an opinion to actually become immune to counterarguments, observing that this situation usually arises because when a "doctrine has taken its place, […] those who hold it have generally inherited, not adopted it" (Mill 1859, 51). The proliferation of true beliefs, therefore, is accompanied by a necessary theoretical regress: the more hold a true opinion takes in the epistemic community, the more the grasp of its truth is loosened and the less can we take it to be known rather than held as a "dead dogma." And we might add, the more an opinion is turned into a dogma, the less awareness of the first reason for valuing freedom of expression—human fallibility—is there with respect to the opinion in question. This prospect of an inevitable downside to the proliferation even of true beliefs opens up a problem for the expectation of a certain form or moral progress.

Mill's view of the inevitable ambivalence of intellectual progress clashes with the idea that moral progress involves the proliferation of true moral beliefs and their elevation to the status of indubitable moral axioms. This idea of moral progress is exemplarily articulated by Stan Godlovitch. He argues that moral progress amounts to

a ratcheting-up shift in what becomes simple moral obviousness or banality. This 'thoughtless' aspect represents, oddly, an advance upon the preparatory phases of moral reflectiveness which emerge first in isolated and often eccentric movements. The philosophical tradition dwells principally upon the reflective and deliberative aspect of the moral awareness individuals must cultivate as its fully mature stage. […] As individuals in a morally progressing culture, however, we have no more to revisit and fight through earlier cultural stages of moral awareness than we have, as individuals in a scientifically progressing culture, to fight through and overcome our culturally earlier stages of pre-Copernican, pre-Darwinian, or pre-Einsteinian science. (Godlovitch 1998, 280, emphasis in original)

To pretend otherwise would be to nourish what Charles Sanders Peirce called "paper-doubts" (Peirce 1966a, 208)—the insincere denial of what is actually (and rightly) taken for granted (Godlovitch 1998, 280).

Philosophers of very diverse stripes propose that philosophy shall take its start from one or another state of mind in which no man, least of all a beginner in philosophy, actually is. One proposes that you shall begin by doubting everything, and says that there is only one thing that you cannot doubt, as if doubting were "as easy as lying." Another proposes that we should begin by observing "the first impressions of sense," forgetting that our very percepts are the results of cognitive elaboration. But in truth, there is but one state of mind from which you can "set out," namely, the very state of mind in which you actually find yourself at the time you do "set out"—a state in which you are laden with an immense mass of cognition already formed, of which you cannot divest yourself if you would; and who knows whether, if you could, you would not have made all knowledge impossible to yourself? Do you call it doubting to write down on a piece of paper that you doubt? If so, doubt has nothing to do with any serious business. But do not make believe; if pedantry has not eaten all the reality out of you, recognize, as you must, that there is much that you do not doubt, in the least. Now that which you do not at all doubt, you must and do regard as infallible, absolute truth. (Peirce 1966b, 188, emphasis in original)

In this longish passage, Peirce emphatically advises against engaging in Cartesian doubt as a way of approaching philosophical problems. Godlovitch relates this line of thought specifically to the question of the possibility and implications of moral progress, arguing that progress in our moral views must involve the setting aside of contested questions of the past and "our setting out from there" (Godlovitch 1998, 281). When "a new obviousness" is created, it makes "possible new moral inquiry at its edge" (Godlovitch 1998, 284). Not only is it possible for us to adopt moral insights from the past in an attitude of trust, this seems to be a moral requirement in itself and a conceptual necessity for progress to occur. Advancement is only possible when earlier results can be assumed to be results—when they can be taken to be steppingstones for further moral inquiry. The instruction here is to refrain from employing methodic doubt when it comes to morality—to refrain from systematically suspending all adopted beliefs in order to build up one's own moral outlook anew—for the sake of contributing to the collective effort of making moral progress as a moral community.

Notably, Descartes himself would have been able to agree with the recommendation to trust in common moral wisdom at least to some extent: it is precisely in the realm of morality that his aspiration to presuppositionless knowledge is restricted by a recommendation to adopt the most widely shared moral beliefs as a working moral code, because they are the most likely to be correct (Kann 2001, 448–449; cf. Descartes 2003, 17–18). Precisely when it comes to morality, Descartes's sweeping methodic doubt is reined in by a concession to the need for practical guidance. These provisional morals, however, ultimately await supersession by a belief system rebuilt on rationalist grounds.

It is precisely the attempt to rebuild one's moral code on a more secure foundation than "inherited" moral beliefs provide that Godlovitch sees as the antithesis to achieving moral progress. To him, it is the settling of disputes over core moral ideas and the removing of certain concepts from the realm of serious questioning that constitutes the basis for further progress: "our present starting points [for further moral progress] need no longer revisit the stages already passed through," "opportunities for progress rest precisely on progress already made" (Godlovitch 1998, 278). Liberal values are among those that have, by past progress, been removed from the domain of serious dubitability. Values such as personal liberty and self-determination are "no longer a serious matter of opinion or controversy in many countries," but "take[n] comfortably and intuitively for granted" (Godlovitch 1998, 276). At the same time, the kind of pluralism of moral views that liberalism fosters is recognized as a "prudential Insurance Principle":

that society is best able to deal with future contingency which keeps in reserve, through toleration, the greatest number and variety of mutant (novel, nonconformist) notions which arise within it as a simple expression of random variation. (Godlovitch 1998, 277)

Here, keeping nonconformist ideas in store is regarded as instrumentally valuable for coping with changing circumstances. This is a different instrumental role for deviant moral thought than would be envisioned based on an appeal to Mill's rationale for freedom of expression. It is not the contribution to the vividness with which true moral beliefs are held,¹³ but the currently inapplicable beliefs' potential activation by different circumstances to which Godlovitch appeals, emphasizing the potential of "novel" moral notions. But what about existing "mu-

Appropriating Mill's ideas about the epistemic benefits of conflicts of opinions for the debate about moral progress and speaking of "true moral beliefs" might be a transgression in that Mill himself is commonly taken to have been a noncognitivist who would have had no place for talk of moral truth. Whether this is correct or whether Mill might have been a cognitivist, after all, (Macleod 2013) shall not concern us here. Either way, once we have allowed ourselves talk of moral truth, we can utilize the ideas on the epistemic benefits of dissent of a supposed noncognitivist and apply them to the realm of moral conflict.

tant" ideas that have been discounted as immoral? The upshot of the creation of moral obviousnesses would be that these disproven ideas have to be put to rest. This view conflicts with the rationale behind the key liberal tenet of the freedom of expression.

Paradoxically, while liberal values secure the preconditions of correcting error, their justification does not license confidence in "moral obviousnesses," because the recognition of fallibility commends a readiness to reevaluate any given moral belief. On the other hand, to the extent that handed down moral beliefs are adopted as accomplishments of past progress, i.e., as moral obviousnesses, they cease to be genuinely known, becoming "dead dogmas" instead at least according to Mill. It appears that the paradox of moral toleration is mirrored in two paradoxes of moral progress within the context of liberal reasoning: (1) just as it is paradoxical to tolerate what one judges to be wrong, it is paradoxical to axiomatize (qua adoption as a moral obviousness) moral convictions if one takes them to be fallible; (2) axiomatizing moral convictions that are believed to be the results of moral progress turns them into customarily adopted "dead dogmas"—it turns morality into custom and precludes genuine moral knowledge. The first paradox can also be expressed as a fundamental tension between two ways of thinking about moral progress: between recognition of prior progress as progress in terms of a correction of past error and the belief in the need for further progress.

Either way, securing past progress by elevating its results to the status of moral obviousnesses seems to be closed off from a theoretical point of view like the view underlying traditional liberal arguments for key civil rights (freedom of expression). But how, then, is advancement possible at all and especially on a large, intergenerational scale?

At this point, of course, it must be noted that Mill does not present a skeptical argument. He decidedly resolves the tension we have presented here as a paradox in favor of the accumulation of true beliefs. He considers the question of whether his remarks on the inevitable downsides of the proliferation of truth commit him to denying the possibility of intellectual progress:

Is it necessary that some part of mankind should persist in error, to enable any to realise the truth? […] As soon as mankind have unanimously accepted a truth, does the truth perish within them? […] Do the fruits of conquest perish by the very completeness of the victory? (Mill 1859, 55–56)

In his answer, he affirms, at the very least, the moral desirability of progress in terms of a proliferation of obviousnesses:

I affirm no such thing. As mankind improve, the number of doctrines which are no longer disputed or doubted will be constantly on the increase: and the well-being of mankind may almost be measured by the number and gravity of the truths which have reached the point of being uncontested. The cessation, on one question after another, of serious controversy, is one of the necessary incidents of the consolidation of opinion; a consolidation as salutary in the case of true opinions, as it is dangerous and noxious when the opinions are erroneous. (Mill 1859, 56)

The first part of his answer suggests that he takes the overall salutary consolidation of opinions to be a reasonable prospect. Still, the threat of the second paradox remains: Mill's characterization of the inevitable intellectual drawbacks from such a consolidation suggests that there is at least an ineliminable downside to moral progress that involves the consolidation of moral views. As debate over an opinion is settled,

not only the grounds of the opinion are forgotten in the absence of discussion, but too often the meaning of the opinion itself. The words which convey it, cease to suggest ideas, or suggest only a small portion of those they were originally employed to communicate. (Mill 1859, 50)

Applied to the issue of moral progress, this suggests that the resolution of moral conflicts—prima facie a potential instance of moral progress—inevitably leads to a loss in moral awareness, which prima facie counts toward ethical regress. Mill calls our attention to a tradeoff in intellectual progress that is relevant to our investigation into the concept of moral progress. He highlights the fact that there is a tension between achieving the proliferation of beliefs and preserving the "vividness" with which they are believed—the extent to which they are intellectually permeated.

The loss in the vividness of moral beliefs is a possible downside to moral progress. Whether it should lead us to qualify our moral progress-judgment depends on the requirements of ethical involvement we ultimately include in our concept of moral progress. If we choose to requireamoral agent to advance ethically as she improves her moral practice in order to count as progressing morally, if such a demand for ethical progress would include the requirement of a deepened grasp of the meaning and scope of accepted moral beliefs (cf. Moody-Adams 1999), and if Mill's psychological hypothesis about the link between the acceptance of opinions and the loosened grasp of their meaning was correct, then any moral progress that consisted in the proliferation of better moral views would have an inevitable ethical downside. We would be faced with the task of weighing the ethical progress that consisted in the adoption of the better moral belief and the improvement in the moral agent's practices against the ethical regress due to the loss of vividness in moral beliefs. If we prioritized the latter, it could be the case that the only progress that could be made is ethical progress in the modest philosophical sense of accumulating and/or refining ethical questions instead of providing answers. Building on the answers in the future by treating them as new starting points would be precluded. This conflicts with Godlovitch's view, as he attributes "irreversibility to moral progress of a sort that ratchets us up, and thereby closes off certain past options by creating ever-new and newly obvious moral points of departure" (Godlovitch 1998, 284).

The establishment of new moral obviousnesses and the deepening of understanding of moral concepts and beliefs seem to be antithetical expectations in moral progress. However, a slight shift in emphasis in the description of the possible results and the procedure of moral progress might work toward reconciling the two views:

[Moral progress] consists […] in the building up of a tradition which makes it possible for each generation to re-acquire, perhaps more easily, what has been achieved by former generations, and in turn, to prepare the ground for coming generations. (Ginsberg 1944, 11)

Rather than taking our starting points in others' obviousnesses, moral progress might require us to re-acquire the beliefs in which earlier progress resulted and not to regard them as "infallible, absolute truth" (Peirce 1966b, 188) but rather as temporary, reversible truth.

Mill's discussion of the tension between freedom, custom and progress is instructive for an inquiry into moral progress on several levels. On a practical level, Mill's remarks can serve to warn us of a common source of threat to progress: if the proliferation of moral beliefs comes with a loss in ethical understanding, it likewise creates a possible reason for subsequent regress, because badly understood moral ideas are deficient guides for moral action. On the other hand, the recognition of prior progress as progress (the marking of the results of progress as worthy of protection; see § 2.3.) can lead to a misclassification of past achievements as ultimate corrections, not mere improvements in need of further advancement. Thus, it can become a force that resists progress.

On the theoretical level, we have to take the paradoxes of progress into account in our explication of moral progress. Our account of moral agency, as laid out in § 2.2., seems to enable us to circumvent the second theoretical paradox: for morality to be turned into mere custom by a growth in acceptance of its provision, this growth would have to come with a complete loss of moral awareness, not merely a diminishing of vividness of moral beliefs. But the first paradox and its theoretical background must be kept in mind when explicating moral progress. The more ethical involvement our concept of moral progress demands from the moral agent, the more likely it is that we inscribe the conceptual grounds of an inevitable ethical downside into our account of moral progress. Before we can begin to work out this account, the supposed metaethical baggage of the very idea of moral progress must be addressed.

## 3.5 Moral Progress and Moral Objectivity

Both in works on the concept of moral progress and in debates in moral ontology, it is often suggested that an appeal to the notion of moral progress brings moral realism to the scene: affirming that there has been moral progress allegedly presupposes that there is an objective standard against which a development is being assessed—a moral standard existing independently of the responses of moral agents to the development in question. For those who view moral realism as an implausible stance on the nature of morality, this presents a problem with the very concept of moral progress. In this section, we shall see what the supposed affinity of progress and realism is grounded in and to what extent an investigation into the concept of moral progress can distance itself from questions about moral ontology.

#### 3.5.1 A Progress-Realism Tie-Up?

Moral progress is sometimes referred to in order to illustrate an intuitive attraction of moral realism. In his introduction to metaethics, Andrew Fisher includes the leap from progress to realism in a list of apparent instances of evidence for objective moral properties which composes "[a] presumptive argument in favour of realism" (Fisher 2011, 56–59):

[T]he fact that there is moral progress is arguably best explained in terms of moral realism. It seems undeniable that there has been some moral progress: we no longer send people up chimneys, or force children to mend looms, or keep slaves and so on. However, if there is progress this seems to imply that we are somehow moving closer to the truth of how the world actually ought to be. If moral realism is false then it seems that there could be no standard or benchmark, and it is hard to see why we would think moral progress was possible at all. (Fisher 2011, 59)

Here, Fisher portrays the inference to realism as an abductive one. In an argument for "[m]oral progress without moral realism," Catherine Wilson cites the "widespread assumption" that moral progress underlies a "transcendental argument" for moral realism (Wilson 2011, 97), which is, however, not spelled out in detail. As a transcendental argument, it would have to amount to something like the claim that (1) we think that moral progress exists and we think that it consists in approximating moral truth where this is not merely a matter of growing agreement, (2) this fact is not only best explained by the truth of moral realism but only possible if moral realism is true, i.e., if there are objective facts that render our progressed moral beliefs true or closer to the truth, and that, therefore, (3) moral realism is true (cf. § 3.5.3.).

Unlike Wilson and like Fisher, Dale Jamieson merely hints at a potential abductive inference to realism:

Moral realism, for example, would seem to provide a natural explanation for how states of affairs can be ranked according to their goodness. On this view, the point of moral language is to correspond to the moral order, and the role of moral action is to exemplify or conform to it. (Jamieson 2002, 320)

However, he treats the prospect of moral progress being tied to moral realism not as a welcome argument for realism, but rather as a challenge to provide an alternative explanation of our persistent impression that there is progress.

When realists talk about moral progress, they sometimes already interpret the concept in a realist fashion (more in line with the premise in a transcendental argument), taking the question to be whether there is evidence of progress, not whether presumed progress is evidence for moral realism:

Most people no longer think that slavery, racial discrimination, rape, or child abuse is acceptable. […] the sort of realism I have been defending is entitled to appeal to this kind of moral convergence as (defeasible) evidence of moral progress. (Brink 1989, 208–209)

Anti-realists seem to view the connection between progress and realism in similar ways, as they sometimes address moral progress as a "challenge" to their view. For instance, Jesse Prinz, speaks of "the challenge of progress" (Prinz 2007, 288) faced by a view like his, according to which the truth of moral judgments is a matter of their relation to the emotional responses of moral agents, not objective moral facts. Prinz combines sentimentalism and descriptivism into the idea that moral judgments are truth-apt, but the truth-makers are dispositions to feel in certain ways toward situations:

I think that moral claims can be true. Moral concepts are indexicals that refer to responsedependent properties. The good is that which is an object of approbation in a moral observer, and the bad is an object of disapprobation. (Prinz 2007, 288).

Moral judgments are, therefore, true only relative to moral agents' emotional dispositions, and not in an absolute way. Making sense of the appearance of moral progress becomes a challenge on this view:

As we move from slave states to slave-free states, as we struggle for gender equity, as we renounce torture and imperialism, are we not doing better than we did before? Our current values seem to be an improvement over the values of the past, and moral humility leads us to hope for continued improvement. (Prinz 2007, 289)

The threat of moral relativity is that it makes such change a mere "lateral repositioning in a space of equally acceptable options" (Prinz 2007, 289) and that it renders hope for future improvement irrational: if moral evaluation is a matter of our emotional response, on what grounds do we hope for these responses to change—and not only to change, but to improve? Notably, Prinz is concerned here with the challenge the apparent existence of moral progress poses to relativism. Usually, relativism is treated as a thesis not about moral ontology but rather as the thesis that "[t]he truth or falsity of moral judgments, or their justification, is not absolute or universal, but is relative to the traditions, convictions, or practices of a group of persons" (Gowans 2015, § 2), i.e., a thesis about the scope of validity of moral judgments. Relativists are in opposition to realists by virtue of opposing the thesis that true moral judgments are universally true. Whether this is all there is to objecting to moral realism depends on how realism is defined—if realism is thought to include an ontological thesis about the objective grounds of universally true moral judgments, then relativism is not synonymous with anti-realism. However, insofar as Prinz identifies the challenge posed by the appearance of moral progress to relativism as a challenge from universal standards and insofar as realists posit such universal standards, his acknowledgement of the "challenge of progress" can also be read as bearing witness to a supposed progress-realism tie-up.

The progress challenge is the challenge of accounting for the phenomenon that is moral progress. In normative ethics, to account for moral progress as a phenomenon of our moral lives is to spell out what a normative theory would say about the normative grounds of progressive change—to make sense of moral improvement in the terms of the respective theory (e. g., in consequentialist terms). In metaethics, the challenge is to account for the way the phenomenon of moral progress presents us with the objectivity of morality and the possibility of progress, i.e., to render permissible and warranted those judgments of moral progress that seem to be coming to us quite naturally. The idea that moral realism is the prime candidate for meeting the challenge is more often than not seen as a problem. Upon recognizing the supposed congeniality of realism to judgments of moral progress,Wilson turns to scientific anti-realism to find a model for accounting for progress in anti-realist terms (Wilson 2011). Jamieson complains that "[m]oral realism may respect our intuitions about moral progress, but it violates our metaphysical sensibilities" and explores the conduciveness of a view of "morality as a human construction" to judgments of moral progress (Jamieson 2002, 321). Sometimes epistemological and ontological worries about realist views of moral progress are combined to make claims to the effect that "any view of moral progress as the attainment of increasingly accurate knowledge of an unchanging moral reality is untenable" (Stokes 2017, 1825). As we have already seen, Peter Singer explicitly draws on the rejection of "the puzzling idea of an objective moral reality" in the argument for his account of moral progress (Singer 2011, 109; see § 3.1.). At the same time, he claims universal validity for his own chosen ethical principle of equal consideration of interests, arguing that

there is something in ethics which is eternal and universal, […] The process of reasoning we have been discussing is eternal and universal. That one's own interests are one among many sets of interests, no more important than the similar interests of others, is a conclusion that, in principle, any rational being can come to see. (Singer 2011, 105–106)

Observing the tension between this claim and Singer's rejection of objective moral truth, Michael Huemer argues that the idea of a "logical or quasi-logical obstacle to embracing concern for all members [of] one's own tribe while being indifferent or hostile to outsiders" is "dubious" (Huemer 2016, 1996) and since the claim that membership in one's tribe is morally insignificant is a "substantive, evaluative proposition," the proponent of this idea—i.e., Singer—"has joined the ranks of the ethical rationalists" (Huemer 2016, 1997). Huemer diagnoses Singer with an unadmitted ethical intuitionism that presupposes the existence of objective truths for the moral agent to grasp. Meanwhile, Singer has endorsed a realist basis for his normative views of moral progress himself. Thirty years after the original publication of The Expanding Circle, Singer addresses his former struggle with moral realism—and revises his position. In the afterword to the 2011 edition, he notes: "[i]n rereading my earlier text I can see how ambivalent I was about the idea of ethics being objectively true and rationally based" (Singer 2011, 198). He explains that his foremost concern was the motivational relevance of an objective morality and refers to how John Leslie Mackie expressed the grounds of this concern:

An objective good would be sought by anyone who was acquainted with it, not because of any contingent fact that this person, or every person is so constituted that he desires this end, but just because the end has to-be-pursuedness somehow built into it. (Mackie 1977, 40)

To Singer, the obstacle that must be overcome in order to be able to endorse the existence of objective normative truths is the conflict between the conviction that the grasp of such truths (i.e., true moral beliefs) would have to necessarily motivate the moral agent (i.e., motivational internalism about moral judgments) and the view that beliefs could not possibly be so motivating independent of contingent desires (as the Humean theory of motivation holds). Singer basically struggles with the metaethical trilemma at this point, which we will address in detail in Chapter 5. He resolves the conflict by rejecting the idea that our moral judgments must necessarily motivate us (Singer 2011, 202–203), thereby endorsing cognitivism and realism, affirming that our best moral judgments can be beliefs in objective moral truths (Singer 2011, 204).

As Singer's struggle with accepting a realist underpinning for his theory as well as others' firm rejection of realism in the context of discussions about moral progress illustrate, its supposed prima facie tie to realism is more often considered to be a burden than a benefit to the concept of moral progress. While it has been argued that "moral realism can fairly claim to have common sense and initial appearances on its side" (Sayre-McCord 2017), this fact can be debated when it comes to "folk" common sense (Pölzler 2016), and it is certainly not true of philosophical common sense, not even among those who are interested in and optimistic about the idea of moral progress.

#### 3.5.2 A Success Concept and a Success Theory

In what way can moral realism claim to have "common sense on its side?" What does moral realism amount to in the first place? The supposedly commonsensical core realist commitment is the descriptivist commitment that moral judgments can be true or false, combined with the conviction that some of them are true (Sayre-McCord 2017). The straightforward way of accounting for this supposed fact is to posit that there are objective moral facts, which are the truthmakers of our moral judgments. One problem with this view, however, is that it seems to introduce items of an implausibly peculiar nature, given another commonsensical idea about moral judgments. Moral judgments are not only true or false, but prescriptive—they are "imperatival," at least "this is what we all naturally think" (Baier 1958, 174). To account for the prescriptive or evaluative nature of moral judgments, their truth-makers would have to be, as Mackie put it, "queer":

If there were objective values, then they would be entities or qualities or relations of a very strange sort, utterly different from anything else in the universe. Correspondingly, if we were aware of them, it would have to be by some special faculty of moral perception or intuition, utterly different from our ordinary ways of knowing everything else. (Mackie 1977, 38)

Whether realists are to be distinguished from anti-realists by positing such peculiar (or any less peculiar, viz. natural) objective values or moral facts is contested. As was just noted, some would characterize realists simply as subscribing to descriptivism and to the thesis that there are some moral truths (Sayre-McCord 2017 and Smith 2013), arguing that moral realism does not come with "a distinctive metaphysical commitment over and above the commitment that comes with thinking moral claims can be true or false and some are true" (Sayre-McCord 2017). Others hold that realism should include just such a commitment (Enoch 2011). Since there is "no consensus as to what, precisely, it would take to be a moral realist" (Smith 2013, 17; cf. Sayre-McCord 1988), rejections of "moral realism" as the basis for moral progress-judgments remain somewhat unclear.

The way in which realism can be characterized without attributing to it a metaphysical commitment to the existence of objective values is useful to bring out the natural connection to the idea of moral progress. When realism is equated with the acceptance of descriptivism (or cognitivism) and the conviction that some moral claims are true, this makes for a tripartite distinction between realism, error theories (which combine descriptivism with the view that all moral claims are systematically false, because the states of affairs which would render them true do not obtain; Mackie 1977) and nondescriptivism (or noncognitivism, the view that moral judgments are not truth-evaluable in the first place). On this view, realist theories can be characterized as "success theories," in contrast to error theories, with which they share their view on moral semantics (Sayre-McCord 1988, 10). They are success theories because they allow for some moral claims to be true—they allow for the chance of being successful in our practice of making moral judgments. The concept of moral progress, for its part, might be characterized as a success concept, since we appeal to it to mark out accomplishments, corrections, and improvement. The "success-theoretical" analysis of such positive changes would be that in making moral progress, we approximate moral truth or accumulate more and more true moral beliefs (cf. § 3.4.2.).

However, many would include further requirements to qualify what moral realism amounts to. For instance, Geoffrey Sayre McCord argues that realism has a built-in requirement for an account of truth and meaning that is not peculiar to moral realism, but connects moral realism "seamless[ly]" to realism in other areas (Sayre-McCord 1988, 6). David Copp claims that realist theories are not only success theories in allowing for moral judgments to be true, but they are also "Success theories" "(with a capital 'S')" by virtue of being "compatible with a robust non-deflationary theory of truth" (Copp 2013, 110). Another requirement can be that the truth conditions of moral judgments do not mention "anyone's subjective states or […] capacities, conventions, or practices of any group of people" (Sayre-McCord 1988, 19). This requirement can be further specified to cover hypothetical perspectives of moral agents:

Realists believe that there are moral truths that obtain independently of any preferred perspective, in the sense that the moral standards that fix the moral facts are not made true by virtue of their ratification from within any given actual or hypothetical perspective. (Shafer-Landau 2003, 15)

That it is unclear whether realism requires acceptance of any of the above-mentioned—or possibly further—claims is unfortunate, because it renders unclear that to which moral progress appears to be tied, and which of the accounts of moral progress that have been given thus far can properly be called anti-realist.

If the aim of this section were to deliver a thorough evaluation of the progress-realism tie-up, we would face the problem that it was unclear what the target was. However, giving a comprehensive assessment of the relation between moral progress-judgments and a commitment to some version of moral realism is not the objective here. Instead, it is to establish the grounds on which this metaethical question may be safely set aside in the context of an explication of moral progress. One key motivating thought is that moral progress-judgments may not be so radically different from other moral judgments in their dependency on some specific metaethical interpretation (cf. Buchanan and Powell 2018, 394) that an explication of moral progress must await the resolution of the debate about the content and credibility of realist theories in moral ontology.

That said, in the following, I will reserve the term "realism" for positions that go beyond the minimal commitment to the truth of some moral judgments and include a metaphysical claim about response-independent moral facts that serve as truth-makers. At least prima facie, these seem to be the sorts of positions that provoke the reservations that are articulated by reference to "metaphysical sensibilities" (Jamieson 2002, 321) and the "puzzling idea of an objective moral reality" (Singer 2011, 109) with which proponents of moral progress struggle.

#### 3.5.3 From Progress to Moral Facts: Appearance and Explanation

One might think that the problem of accounting for their moral-ontological commitments is not a unique problem that pertains in any special way to judgments of moral progress at all and that it might therefore be negligible in a context like the present one. This is the view Buchanan and Powell adopt when they address questions about the metaethical implications of moral progress in the conclusion of their book on the compatibility of the idea of moral progress with an evolutionary underpinning of morality. They state that

the judgments that the abolition of slavery was moral progress or that the cessation of honor killings would be moral progress are not different in kind, not inherently more problematic as moral judgments, than the judgments that slavery and honor killings are morally wrong. (Buchanan and Powell 2018, 394)

Therefore, they feel no need to make any metaethical assumptions beyond accepting that moral judgments can at the very least have a justification (Buchanan and Powell 2018, 394).

Supposedly, if moral progress-judgments appear to imply objective standards against which the change in question is judged to be a change for the better, this situation is not so different from the situation with respect to other moral judgments. Moral claims in general are commonly assumed to have an objective appeal and yet much theorizing, conceptual analyzing and explicating in ethics proceeds without alluding to questions of moral ontology. This seems to be correct. Furthermore, the more the appearance of moral discourse in general is not merely taken to be descriptivist but also, in some sense, realist, the more "the plausibility of anti-realism […] depends […] on preserving our normal ways of speaking even while challenging the natural […] realistic interpretation of what is being said" (Sayre-McCord 1988, 2). That is to say, if we view moral judgments as carrying an implication not simply of being truth-evaluable but of being true in some objective way (or in a way that is consistent with the manner in which other claims are thought to be true, see § 5.2.2.), then we might simply proceed with the impression of objectivity and view it as the anti-realist's task to offer a plausible way of accounting for this impression that licenses holding on to it on many levels of ethical discourse. Moral progress might not be any more demanding of such an account of its supposed objectivity than other moral concepts. Just as nondescriptivists typically seek to license our practice of taking moral judgments to be truth-evaluable (e.g., Blackburn 1993), anti-realists might have to seek to license our treating moral progress and all other moral concepts as appealing to objective standards, i.e., to "capture what we mean when we make moral claims" (Sayre-McCord 1988, 22) in cases where what we mean is something objective. To the extent that there is a shared impression that "we don't construct values" but instead "bump into them" (Chappell 2016, 84), this holds for encountering values in general, not merely increasing value as in the case of moral progress.

This line of thought might provide the motivation to leave the metaethical debate aside at this point and move on with an explication of moral progress —whatever its ontological commitments. Either these commitments are vindicated by a realist account of morality or an anti-realist construal will reveal that objective moral facts exist merely in the universe of discourse, and offer us a plausible alternative way of thinking about "what we mean" by making moral progress-judgments that does not require us to revise our concept of moral progress all too thoroughly. Either way, the fate of the concept of moral progress ought not to be different from the fate of our concepts of rightness or wrongness, good and bad.

However, it is often presupposed that judgments of moral progress are peculiar in this regard, after all, and that the fate of moral progress does depend more heavily on the vindication of moral realism. To some, our impression that a certain change "just looks like progress" (cf. Stokes 2017, 1824) appears to be more intimately linked to the supposition of an objective moral reality than the impression that something "just looks like the right thing do to." Recall the suggestion that "[t]he notion of moral progress is central to moral epistemology. There is a widespread assumption that its very possibility furnishes the basis of a transcendental argument in favour of moral realism." (Wilson 2011).

In general, transcendental arguments are arguments that establish that "X is a necessary condition for the possibility of Y," and since it is taken as a given that Y is the case, "X must be the case too." Transcendental arguments are traditionally directed at skeptics about the respective X—skeptics who accept, however, that the respective Y is the case (Stern 2019). For the argument from moral progress to be an argument for moral realism of this type, it must take a special form. As we have already noted, for the possibility of moral progress (the Y) to be the basis of an anti-skeptical transcendental argument for realism (the X) it must be included in a premise that refers to what we accept as a given. Furthermore, the way in which moral progress must be taken as a given has to be a way that includes the appearance of objectivity. What the skeptic about moral realism will concede is that moral progress appears to occur.We can only hope to persuade a skeptic about moral realism by appeal to moral progress if we assume that we share an experience of the phenomenon of moral progress in which it appears to us that moral progress is change for the objectively better. If the skeptic does not share this impression, there is no common ground on which to base any transcendental reasoning to begin with. Given the very limited common ground about moral progress, a transcendental argument from moral progress (relying on a provisional, ad hoc characterization of moral progress) could only start from a premise such as: (P1) It appears to us that moral progress, i.e., the improvement of moral practices and/or views in accordance with an objective standard, is possible. Accordingly, there would have to be a premise or set of premises that would take us from the appearance of the occurrence of moral progress to moral realism. For the transcendental argument to be able to convert the skeptic about moral realism, what the transcendental premise would have to establish is not that the possibility of moral progress of the kind mentioned in (P1) depends decisively on the truth of moral realism (this is what the anti-skeptical transcendental argument as a whole is supposed to show), but that the fact that it appears to us that moral progress is possible depends decisively on the truth of moral realism—because the appearance that there is moral progress is the only claim with which the skeptic will agree. But the idea that moral realism is necessary for it to appear to us that moral progress is possible is rather peculiar. I will not try to explicate the potential transcendental premises here.What is important to see is that the very idea that moral progress furnishesatranscendental argument for moral realism relies on the view that an appeal to objectivity is part of the very concept of moral progress. A question arises as to why one would consider the phenomenon of moral progress but not the phenomenon of morally right or wrong action to be a basis for a transcendental argument for moral realism.

One reason could be the diachronic perspective expressed in moral progressjudgments. Maybe we think that a judgment that moral progress has occurred bespeaks a special commitment to objective standards, because it not only deems an action right or a practice good here and now, but because it articulates the view that the present practice is better than the one it has superseded, a view that is based on holding both practices to the same standard. This is especially salient in the case of moral progress-judgments applied to the moral agent's own practices or the practices in her own moral community. In judging that she herself has made moral progress by revising certain moral beliefs and changing her practice accordingly, a moral agent acknowledges not only that her former actions were wrong but also that her own former standards by which those actions seemed to be right was wrong. That is, her moral progress-judgment seems to imply acknowledgement of her own fallibility of her standards. Fallibility is precisely that of which realists claim to have a uniquely apt account. Sturgeon argues:

only a moral realist can be a genuine moral fallibilist. By fallibilism about an area of thought I understand the view that any of one's views about that area might in principle be mistaken. This is not skepticism: the concession is intended to be consistent with the view that the area is one in which there is genuine knowledge and warranted belief. But fallibilism has seemed to many philosophers an important and attractive form of epistemological modesty. (Sturgeon 1986, 127–128)

Are self-directed retrospective moral progress-judgments expressions of such modesty? If they generally implied an acknowledgement of fallibility and if the relevant kind of fallibility could only be accounted for based on the acceptance of moral realism, we would have identified a unique realist commitment in the concept of moral progress. Prinz opposes both parts of this claim. He argues that not all moral progress-judgments do involve an acknowledgment of fallibility and that those that do can be accounted for in antirealist terms.

Ordinarily, when we assess alternative values using the values that we already possess, […] [a]lternative values are perceived as defective, corrupt, or otherwise worse than our own. In these cases, progress is possible only in an empty backward-looking sense. From the present, we always seem to have better values than we had in the past for the trivial reason that we embrace our present values and no longer embrace our past values. (Prinz 2007, 289)

Prinz does not see the inability of his relativism to account for the condemnation of past wrongs as a failure. Instead, he sees attempts to condemn past wrongs (or present wrongs in other moral systems) as potential "self-deception" (Prinz 2007, 301–302), in case these supposed wrongs are not wrong by the supposed perpetrators' own lights (Prinz 2007, 301). To Prinz, the more interesting way of thinking about moral progress is in a forward-looking way:

The interesting cases are the ones in which we consider the possibility that an alternative value is better than a value we currently possess. This kind of comparison is progressive, or forward-looking. We can see that the moral future might be better than the moral present. (Prinz 2007, 289)

Those accepting that there might be universal moral truths can make sense of this prospective way of applying the concept of moral progress as an anticipation of the approximation or grasp of those truths. The only way Prinz can account for the forward-looking way of thinking about moral progress—and thus, for hope for moral progress—is by appeal to non-moral ("extra-moral") qualities of moral belief systems, such as "consistency, coherence with facts, stability" and the like (Prinz 2007, 292). Disagreement with this reconstruction could be evidence of an at least tacit commitment to the idea that orientation toward objective moral standards is inherent to the concept of moral progress. We may think that our backward-looking moral progress-judgments are not empty and that our hope for moral progress is hope for more than increased coherence and improvements with regard to other such non-moral factors. Progress-judgments applied to oneself bring this commitment to light in a special way. Still, other moral judgments may involve a similar commitment: an agent saying that an action is wrong may well by saying that it is actually wrong, not simply wrong by her own lights. In any case of moral disagreement, we are aware of the fallibility of moral standards—at the very least, the standards employed by the other party to the conflict. The appearance of moral objectivity—if and insofar as it is a part of our shared moral phenomenology—is not entirely peculiar to moral progress-judgments.

However, maybe in using the term 'moral progress,' we anticipate a special explanatory relation between (even if not a transcendental argument revolving around) progress and realism and maybe this anticipation informs our notion of moral progress. We have already seen Michael Huemer's take on an abductive argument from uniform social change to moral realism which concluded that the observable change shows that "societies progress towards the truth over time" (Huemer 2016, 2000). It is important to note that the argument from convergence over time is not reducible to a mere argument from agreement. The fact that it concerns a process of coming to moral agreement sets the argument revolving around apparent progress apart. This fact is not always emphasized even by realists. For example, when considering different ways anti-realists could undermine realists' arguments, David Enoch presents the following argument from agreement:


Enoch presents this as an argument open to the realist and open to attack for the anti-realist. He decidedly picks the argument from agreement rather than an argument from progress toward agreement, even though he immediately considers that this might be objectionable:

You may have doubts about this argument. For one thing, the argument might be stronger with a slightly different explanandum—perhaps, for instance, the phenomenon the explanation of which lends support to realism is not mere agreement, but […] the progress towards more and more agreement, the gradual elimination of disagreement (Enoch 2011, 196)

However, he goes on to explore the argument from mere agreement. The fact that he mentions convergence over time in the second premise may be taken as a sign of recognition of the relative force of the progress argument compared to the agreement argument, but ultimately, both are beside Enoch's main point, as he defends a nonnaturalist kind of realism based foremost on the argument that moral facts are not explanatorily, but "deliberatively indispensable" (Enoch 2011, 197–198; cf. § 3.5.5). He still acknowledges that having the argument from agreement available would have conferred further plausibility on realism (Enoch 2011, 198). The argument from progress toward agreement should be all the more interesting in this regard.

A convergence of views seems to be a more suggestive datum to support a realist view than the existence of agreement, which could, after all, be turned against the realist—for being too uniform to be reasonably interpreted as tracking a response-independent reality (Sauer 2019). Strikingly, realists who do not think that they have the argument from mere agreement at their disposal sometimes turn to the argument from achieved agreement. Michael Smith, for instance, points out: "We must not forget that there has been considerable moral progress, and that what moral progress consists in is the removal of entrenched disagreements of just the kind we currently face" (Smith 1994, 188). Changes in moral views lend themselves to interpretations in terms of moral learning. Overcoming disagreement is only one relevant phenomenon. The other type of apparent learning appears where views that had once been common sense are being abandoned in favor of agreement over new views. Viewing these types of developments rather than the state of agreement as the explanandum makes for a distinct argument for moral realism not from mere agreement, but from achieved agreement. In this argument, moral progress is the explanandum, moral realism the explanans. Arguments for the explanatory value of moral realism of this kind may be seen as corresponding to the no miracle argument for scientific realism according to which realism "is the only philosophy that doesn't make the success of science a miracle" (Putnam 1975a, 73; Psillos 1999, 70–75).

But of course, as Wilson notes (Wilson 2011, 101) the no miracle argument leaves constructive empiricists, among others, unimpressed (cf. van Fraassen 1980).Wilson goes along with their diagnosis that the matching relation between theory and world is "difficult to see" (Wilson 2011, 101) and rejects the idea that moral realism is required to make sense of the phenomenon we aptly characterize as moral progress. To see how the supposed progress-realism tie-up may be relaxed, it will be useful to consider an example of a deliberately anti-realist account of moral progress.

#### 3.5.4 No Miracle at All: Moral Progress in Antirealist Terms

One detailed articulation of the idea that moral progress is to be characterized without reference to independent moral truth has been put forward by Philip Kitcher. In The Ethical Project (2011) Kitcher argues for what he calls "pragmatic naturalism" and spells out an account of moral progress in terms of increased functionality. When Kitcher explains the sense in which he takes his account to be naturalist, a sharp opposition to moral realism is evident:

The naturalism consists in refusing to introduce mysterious entities—"spooks"—to explain the origin, evolution, and progress of ethical practice. Naturalists intend that no more things be dreamt of in their philosophies than there are in heaven and earth. They start from the inventory of the world allowed by the totality of bodies of well-grounded knowledge (the gamut of scholarly endeavors running from anthropology and art history to zoology), and, aware of the certain incompleteness of the list, allow only such novel entities as can be justified through accepted methods of rigorous inquiry. Appeals to divine will, to a realm of values, to faculties of ethical perception and "pure practical reason," have to go. (Kitcher 2011, 3–4)

Kitcher seeks to offer not a definite account of how morality and moral progress actually came into being but rather what he calls "a 'how possibly' explanation" of the occurrence of morality and moral progress in the lives of humans as evolved creatures, i.e., he seeks to lay out one way in which what we—rightly —perceive of as moral progress could have come about given our evolutionary history—not given any objective domain of moral values (Kitcher 2011, 12). He views his approach as being pragmatist—in the tradition of Dewey (Dewey and Tufts 1908) and James (James 1978)—as he construes morality as a tool for social living and sees progress as prior to truth (Kitcher 2011, 210). Moral progress produces truth in that descriptive restatements of norms and rules are true when those rules have stood the test of time. To Kitcher, achieved agreement does not underscore an argument for realism—it just is what moral truth comes to. Moral truths exist only where moral rules are unlikely to be improved further. "Ethical progress is prior to ethical truth," and "truth is attained at the limit of progressive transitions" (Kitcher 2011, 210). Moral progress is not the accumulation of truth but rather the increase of functioning, and in that, it is more like technical progress than like scientific progress (Kitcher 2011, 7). Morality, to Kitcher, is a system of rules that give "normative guidance" to humans as social beings (Kitcher 2011, 74–75). These rules were "invented" (Kitcher 2011, 7), and their purpose is to remedy "altruism failures":

altruism failures, are constituted by occasions on which an animal A, belonging to the same social group as an animal B toward whom A is in other contexts inclined to make an altruistic response, fails to respond altruistically to B, either forming no altruistic preference at all or acting on the basis of a selfish desire that overrides what ever altruistic wishes are present (Kitcher 2011, 74)

Failures to behave altruistically disturb the group life of social beings, and for Kitcher, moral progress began when "[t]he first ethicists overcame some of the problems by agreeing on rules for conduct, rules remedying altruism failures that had plagued their group life" (Kitcher 2011, 7). Moral progress continues as systems of normative guidance are refined and adapted to changing outer circumstances with the goal of developing "strategies of socialization for eliciting preferred behavior on as many occasions as possible" (Kitcher 2011, 131). Kitcher only talks about moral truths reluctantly:

In fact, it is better to approach these decisions by starting with the notion of progress rather than that of truth. Thinking in terms of truth narrows the focus. For truth applies to statements, so we are led to conceive the decision as one about descriptive counterparts of rules of the alternative code. There are other components of ethical codes— concepts, exemplars, habits, emotions, modes of inducing compliance— and improvements to our own practice could occur in each of these respects. A rival code whose rules agree with ours might do substantially better at preventing relevant forms of blindness. Thinking in terms of progress responds more directly to the practical choices we face. (Kitcher 2011, 211)

To him, truth is secondary to progress. In giving an account of how progress can yield "truth" that does not correspond to an independent reality, Kitcher offers an anti-realist, functionalist reconstruction of moral progress; he offers what is sought by those who view it as a problem that moral realism seems to force itself onto us when we consider moral progress—he offers an account of moral progress in terms of a "human construction grounded in evolutionary history" (Jamieson 2002, 320). The phenomenon of achieved agreement itself is not explained by reference to the approximation of truth, but by the fact that it testifies to the increased functioning of the system of moral rules which is in place to secure social harmony. What is the explanandum for the realist (who infers from achieved agreement to moral facts by reasoning that moral agent's must have found or come closer to the truth) simply is the ground of any talk of moral truth to Kitcher. No appeal to an objective moral reality is needed.What explains apparent progress is morality's social function. Does this picture do justice to the phenomenon of moral progress?

Kitcher gives a functionalist analysis of both how morality came about and what morality basically is, i.e., he espouses both "etiological functionalism" and "constitutive functionalism" (Buchanan and Powell 2018, 78). Buchanan and Powell identify the latter as a fundamental problem with his view. By characterizing morality in terms of its original function (remedying altruism failures within a group), Kitcher's account fails to do justice to what they call the "the openended normativity of the ethical" (Buchanan and Powell 2018, 85, emphasis in original): the fact and the requirement that moral agents (should) subject norms that have manifested in the dischargement of some original function to scrutiny (Buchanan and Powell 2018, 85–86). Because the original function Kitcher assigns morality – remedying altruism failures among humans—cannot be overridden (Kitcher 2011, 247), there is no conceptual room for a thorough revision of moral objectives—i.e., for thoroughgoing transformative progress. This manifests, inter alia, in the fact that a functionalist approach such as Kitcher's cannot naturally account for "inclusivist anomalies" (see § 3.1.) such as the moral consideration of nonhuman animals (Buchanan and Powell 2018, 89).

So, this antirealist explanation of the appearance of moral progress becomes unsatisfying once we accept that it is part of the explanandum that we can—and should—revise our moral objectives in the course of moral progress, i.e., that moral progress is transformative improvement (see § 2.1.).

#### 3.5.5 Deliberative Commitments and Deliberative Duties

The main question in this section is neither whether moral progress licenses an abductive inference to moral realism nor whether there are plausible anti-realist theories of moral progress but rather whether our very concept of moral progress is so crucially affected by a commitment to realism that we must include a reference to objective values (or any other defining feature of the relevant type of realism) in our definition of moral progress. The question is whether we may, alternatively, safely set aside the metaethical problem—denying that there is a progress-realism tie-up or, if there is one, claiming that it should not be seen as affecting the core of our understanding of moral progress.

A final consideration in favor of interpreting moral progress as carrying a conceptual commitment to moral realism could be inspired by Enoch's defense of moral realism. Enoch takes moral facts to be "deliberatively" rather than explanatorily "indispensable" (Enoch 2011, 197–198). He describes deliberation in general, and thus moral deliberation, as a rationally "non-optional project" for creatures endowed with reason (Enoch 2011, 70–71), one that is "intrinsically indispensable" (Enoch 2011, 83). We cannot rationally opt out of deliberating (Enoch 2011, 70). In deliberation, we aim at making "the decision it makes most sense for us to make" (Enoch 2011, 73), and this requires us to believe that there are objective reasons that make it the case that there is a best decision. "Deliberation—unlike mere picking—is an attempt to eliminate arbitrariness by discovering (normative) reasons, and it is impossible in a believed absence of such reasons to be discovered" (Enoch 2011, 74). Belief in the existence of normative reasons, i.e., objective normative truths, is justified, according to Enoch, since it is instrumentally indispensable to the "intrinsically indispensable" project of deliberation (Enoch 2011, 83), just as inferences to the best explanation are justified because they infer to the existence of something from that thing's indispensability for "the explanatory project" (Enoch 2011, 70). Furthermore, justified belief in normative truths also licenses one to "stand one's ground" in cases of moral truth. Unless we think there is an objectively correct resolution of the conflict, we ought not to seek to advance our own view, but rather seek an impartial solution that accommodates all views and preferences (Enoch 2011, 19).

Without going into the details of Enoch's argument, we can see how his ideas could support the assumption that in explicating moral progress, we have to accommodate an irreducible realist underpinning. Just as deliberation differs from mere picking, moral progress differs from mere change. In issuing a judgment to the effect that moral progress has occurred, we articulate the result of a deliberation about the comparative value or worth of two scenarios. If we accepted that in engaging in moral discourse and deliberation, we commit ourselves to there being objective standards that are in principle the ultimate arbiters of our moral questions and conflicts, it might be assumed that this commitment is inscribed into all moral language and thought: that moral terms and moral concepts have an irreducible realist content because that simply is what we (should) commit ourselves to in using them. On this view, the concept of moral progress would not be a special case compared to other moral concepts, but it still would have to be given a definition that mentioned the items posited by moral realism.

What casts doubt on this idea is the fact that even if we were committed to objective moral truths in our deliberative practices, we would, at the same time, generally have to suspend this commitment to some degree, at least in interpersonal deliberation. Melis Erdur argues, partly in response to Enoch, that it is difficult to conceive of an objective moral reality that was "the source of all rightness and wrongness" as a morally neutral metaphysical framework (Erdur 2016, 600 – 601). In other words, the facts that make it the case that our actions are right or wrong are plausibly to be seen as having moral content, and moral realism is therefore not a normatively neutral theory but rather best interpreted as a moral theory in its own right, i.e., a theory about what ultimately makes actions wrong or right, outcomes good, or agents morally excellent (Erdur 2016, § 2). Now, Erdur points out that in giving justifications and exchanging arguments, we should never appeal to moral facts as the ultimate truth-makers of our moral judgments. In arguing that torturing babies for fun is wrong, we should stop our argument prior to making references to an independent moral reality, because such references would actually weaken, not strengthen our case:

when we arrive at a point where anything further we say seems less plausible than what we have said so far, the best thing to do, from a moral perspective, is to dispense with the implausible additions and stop at the point where our case is strongest. (Erdur 2016, 601)

Instead of a deliberative commitment to realism, Erdur identifies a deliberative duty to abstain from appealing to realism. She backs up her claim about the weakening effect of appeals to an independent morality by pointing to the problem, in substance, but not in these terms (Erdur 2016, 597), which is usually thought to plague only a special kind of realism, divine command theory: what Richard Joyce calls the "Modal Vulnerability Problem" (Joyce 2002). The problem is that it is conceivable that an objective moral reality made it the case that practices that appear to us to be extremely evil—such as genocide or slavery— were in fact right. If this fact were somehow revealed to us, it would not—and presumably should not—move us to accept that such evil practices were in fact right.

Erdur concludes that both moral realism and anti-realism ought to be rejected—leaving us with an agnostic stance on moral ontology. If we look at actual, functioning moral discourse, it seems that moral agents engaged in it routinely conform to Erdur's recommendation and refrain from appealing to the existence or absence of moral facts in making their moral case. This diagnosis fits with realists' own downplaying of the importance of metaethical questions for moral discourse. Sayre-McCord, for instance, laments the fact that a commitment to realism or anti-realism has "disappointingly little to do with" first order moral discourse (Sayre-McCord 1988, 2). George Edward Moore makes a remark to the same effect early on in the Principia Ethica (Moore 2004, 5). All this suggests that opting for omitting references to the entities posited by moral realism in an explication of moral progress would be conducive rather than a cost to the concept's fruitfulness.

While talk of "moral progress" does highlight the question about the truthmakers of moral judgments, we can nevertheless bracket these issues in an explication of moral progress. The debate over moral realims and anti-realism is unlikely to be settled anytime soon and evaluating in yet more detail what exactly the significance of the idea of moral progress for moral realism is (and vice versa) leads us too far astray from the present project, which is to clarify the concept of moral progress (even absent a resolution of the disagreement about moral ontology). Still, the question of whether in making moral progress we approximate an objective moral truth or not may well be seen as decisive for the individuation of concepts of moral progress. It may be the case that realists and antirealists have, after all, dramatically different concepts of moral progress—i.e., that they employ dramatically different ways of thinking about moral improvement. The moral historiographies that they write appealing to their respective concepts of moral progress make fundamentally different comparative judgments about the described moral communities (see § 2.3.). However, it seems, the fundamental difference should not be expected to show in their respective evaluations. They can still communicate successfully about moral progress on the normative level on which questions about what actual developments are changes for the better are debated and where answers are justified. This is so because it seems to be correct that moral realism (or anti-realism) cannot be appealed to in justifying moral progress-judgments. In this regard, these judgments are not different from other moral judgments. Anti-realists' and realists' concepts of moral progress seem to overlap enough for them to be engaging in discourse over moral progress, so they should also be able to find common ground on the adequate characterization of a more general concept of moral progress which omits reference to the metaethical grounds of moral improvement. Explicating the agnostic's concept that can be extracted from this overlap is precisely what will be done in Chapter 4. In so doing, I will sometimes use realist-sounding language, e. g., I will speak of "the fact" that an action is good or right. This should be taken as a façon de parler that anti-realists may analyze away. The next chapter will approach the "phenomenon" of moral progress irrespective of whether a vindicating or debunking metaethical story will have to be told about any apparent objectivity of moral progress.

## 4 The Phenomenon of Moral Progress

Apparently, explicating a concept of moral progress is a process burdened by different and partly incompatible theoretical expectations. Producing a generally acceptable account might be impossible, but having a clear concept of moral progress that accommodates as many points of views as possible would be an obvious benefit. The compelling force of some of the most agreeable examples for apparent moral progress only takes us so far.When we are witnessing change for the better, we can reasonably wonder whether we are in the presence of moral progress—or merely some other kind of welcome change. We cannot even begin to reflect on the relative significance of moral progress compared to such other kinds of change unless we are able to distinguish it from other improvements. This chapter seeks to make such a distinction possible.

When we are presented with examples such as the abandonment of the slave trade, we are led to think of moral progress as manifesting in official, historical turning points, accompanied by legal rituals, marked by public proclamations and registered in headlines. But for any such disruption, there is usually a change preceding it, making it possible. It is in the processes preparing a disruptive legal act (which outlaws a practice or establishes a new set of rights), where potential moral progress resides.

The symptoms of these processes may often lack the kind of gravitas that surrounds the idea of moral progress. It may be the most mundane, seemingly banal occurrences that make the best test cases for developing our understanding of what moral progress should be thought to be. For instance, what do changing eating habits reveal about our moral development? There has been a surge in sales of animal flesh-free meat products—with demand sometimes outstripping supply (Valinsky and Wiener-Bronner 2019) and global fast food chains incorporating flesh-free meat into their menus (Popper 2019). While plant-based meat has been around for quite some time, new products are now designed to mimic the taste and texture of animal flesh-based products more closely than ever, and they target consumers for whom animal flesh has remained the default ingredient. Instead of advertising the reduction or termination of the consumption of animal flesh, some of the producers of plant-based meat are seeking to work toward the same effect by making available products that provide a perfectly equivalent gustatory experience. Witnessing their economic success, we might wonder: is it also a moral success? Those who agree that the termination of the consumption of animals would be a morally good thing and that, as a matter of fact, the provision of animal-free meat products brings us closer to it could still disagree about whether this kind of change would constitute moral progress. Disagreement of this kind comes down to disagreement over the conceptual question: what does a change have to be like in order for it to count as moral progress?

In the previous chapters, we have clarified basic preexisting conceptual commitments with respect to the constituents of the concept moral progress (Chapter 2) and reviewed the most important ways in which standard ethical theories could inform our understanding of moral progress (Chapter 3). The discussion has cast light on the difficulty of providing an informative account of the concept of moral progress that will be acceptable from these different theoretical standpoints. If we think of moral progress as a moral phenomenon about which any ethical theory should have something to say (cf. Chapter 3.), it is to be expected that sometimes one theory will explain away an aspect of the supposed phenomenon which another theory will regard as central or at least nonomittable. Some theories will have to refuse recognition of some purported examples as instantiations of moral progress. In many cases, apparent instances of moral progress will be rejected, not because it cannot be agreed that the changes in question exemplify a change for the better on some level, but because it is not regarded the right level. Moreover, from the theory-laden perspective of different observers, some supposedly uncontroversial prima facie cases of moral progress will not even appear to be such candidates for moral progress to begin with. There is probably no shared perception of the phenomenon of moral progress. Some of the disagreement might indeed be verbal—and this is not necessarily the less interesting or less important kind of disagreement.

Asking for a characterization of "the phenomenon" of moral progress sounds suspiciously like a request for a conceptual analysis—an analysis which trades on an expected agreement in intuitions about the phenomenon and seeks to preserve a supposed conceptual unity. However, neither is presupposed in the investigation in this chapter. The proposal to be made here will not be that of an account of a preexisting tacit understanding of what the phenomenon of moral progress is, but a proposal of how we should come to think of different types of moral progress, based on considerations that go beyond our initial conceptual intuitions. Again, there is a difference in the spirit in which a definition is offered as a product of a conceptual analysis vis-à-vis when it is offered as the product of an explication. The idea is not that the contours of the concept merely need to be "uncovered" that there is a tacit understanding of these contours that is already shared—but rather that a redrawing of these contours is needed to facilitate theoretical progress. The investigation will make extensive use of descriptions of cases. That does not mean that it relies on them in the same fashion as would occur in the course of a conceptual analysis. The purpose of the case descriptions is not to elicit intuitions in order to balance them with a preconceived definition of moral progress. Rather, it is to illustrate the options for and consequences of conceiving of moral progress in one way or another. The weighing of these options is not purported to be used for the purpose of preserving an already grasped definition. The three concepts produced by the explication will be proposed as a useful differentiation—useful for further investigation into what the phenomenon of moral progress is.

In the following section, I will briefly discuss possible approaches to delineating moral progress from other morally desirable changes that are theoretically wanting in one way or another. I will also address the problem of aggregating judgments about different aspects of morality in an evaluation of moral progress and use the range of possible aspects to select a conceptual cornerstone for the present investigation: the dichotomy between theory and practice as it applies to morality. Based on this selection, I will, in § 4.2., proceed with the explication of moral progress. § 4.3. will evaluate the result in terms of its compatibility with different perspectives on the social and temporal scale of moral progress addressed in § 4.1. § 4.5. will summarize.

#### 4.1 Dimensioning Moral Progress

When talking about moral progress, some authors immediately make it clear (or it soon transpires) that they restrict the scope of the term to, for instance, changes taking place on a societal rather than an individual level or on a larger historical scale rather than in smaller temporal intervals. The difference between individual progress and progress on a group level is often assumed to be expressed in the terms 'moral development' and 'moral progress,' respectively, with the latter being reserved for group-level change. Decisions with respect to the spatiotemporal dimensions of moral progress are often cast in the language of "global" versus "local" moral progress, but the "locality" of moral progress can also be interpreted as pertaining to aspects of morality: moral progress might be said to occur "locally" not only in terms of its temporal expanse, but also because it might only be identifiable as it affects one particular value or one particular context of action rather than morality at large.

In this section, I will briefly address the three aspects in which moral progress is often characterized in terms of its "dimension": 1) the distinction between individual and social change, 2) the differentiation in terms of the historical or spatiotemporal scale of moral progress, and 3) the distinction of different aspects or "domains" of morality with respect to which progress could be made. I will treat all of these approaches to dimensioning moral progress as valuable pointers to challenges for the application of the concept. However, at least with respect to the first two approaches to dimensioning moral progress, I assume that no particular interest in either of the alternative ways of applying the concept outlined below justifies a definitional restriction. Nonetheless, it will prove pragmatically beneficial to the argument to focus on moral progress made by individuals for the most part. This will happen with an eye to the possibility of eventually applying the so explicated concept of moral progress to supra-individual changes.

#### 4.1.1 The Social Scale of Moral Progress

First, let us consider how the distinction between different social scales on which moral change can occur might affect the concept of moral progress. The term 'moral progress' might be used to denote either development in individual moral agents or on a higher level of social ontology, in groups, societies or even institutions. Moral progress as individual development takes center stage, e. g., in virtue ethics. Its modern subsidiary, the ethics of care, has been a reaction to a psychological model of individual moral development, which implied that increasing orientation at abstract principles was a change for the morally better (Kohlberg 1973, 633), i.e., presumably a kind of moral progress. Individual moral development is also the central concern when "moral education" is discussed, and it is important to any professional ethics (such as medical ethics or research ethics) that seeks to bring individuals to act in accordance with a professional ethical code.

However, when the topic of moral progress as such is concerned, it is often changes on a higher social scale that are of prime interest, and this focus is sometimes even reflected in how moral progress is defined. A terminological distinction is sometimes drawn, according to which 'moral progress' denotes a social phenomenon, whereas 'moral development' refers to changes in individuals. The distinction may be purported to reflect a shared conceptual commitment or established merely as a constraint on the topic at hand, i.e., it may be the claim of a conceptual analyst, a stipulation or a selection out of various presumed meanings to proceed with in an explication (cf. Macklin 1977, 370, n. 1).

However, it is not clear that the distinction between individual and collective change is the best candidate for being marked by the different terms 'development' and 'progress,' respectively. There are other possible purposes that the terminological distinction between 'moral development' and 'moral progress' could serve. Besides the distinction between (1) individual and collective change, it could also be used to differentiate between other types of changes. For instance, it might be used to mark the difference between (2) descriptive (change) and normative conceptions of the change in question (i.e., between change and improvement) or between (3) unreflective change and change induced by reason. Following up on this distinction, we could further want to differentiate unreflective change into (3.a) change that is normal or regular in the sense of following a preexisting pattern of change—either natural or culturally determined (i.e., moral maturation)—and change that is transformative relative to known standards (moral evolution). Analogously in the case of change that is induced by reason, we could draw a distinction between normal and transformative types of change: we might want to distinguish between (3.b) moral development in terms of perfection, where this is understood as a way of becoming "morally better, assuming that we have a stable idea of what better would amount to" (Hämäläinen 2017, 48, n. 2), and moral progress as a more pervasive moral change that includes the revision of the very ideals toward which we try to develop. There does not seem to be one phenomenon of moral progress, but different progressive sorts of developments which might be called moral progress. Which of the distinctions just mentioned ought to be referred to by the terms 'moral development' and 'moral progress' can at least not be determined by reference to natural language.

The most obvious competitor of (1) individual and social change for being marked by the terminological distinction between 'development' and 'progress' is the difference (2) between descriptive and normative concepts of moral change. In fact, we have already established (Chapter 2) that moral progress is to be understood as an inherently normative concept, where the constituent moral, selects the normative reading of the constituent progress. In conjunction, they focus on moral change for the better. So, our question here is whether we should take the concept to be further constrained than we have so far assumed it to be: should we take it to apply only to moral improvement in larger social entities, as suggested by distinction (1), and/or should we take it to apply to moral improvement induced by reason, as suggested by distinction (3)?

The impulse to restrict the concept of moral progress to phenomena on a larger social scale and excluding individual development might be owed to the impression that moral progress should only denote changes in morality that effect positive changes in states of affairs, as, for instance, consequentialists should want to demand. The types of changes in states of affairs that one might have in mind are often not achieved by singular individuals—individual actions at least need to accumulate to produce the morally desirable results.Veganism is a paradigmatic example. An individual who turns vegan but remains a lonely vegan in a sufficiently large society will probably not effect any change in the number of animals killed for the production of food. A critical mass of vegans has to be reached to yield this result. Seeing the connection between collective efforts and the kinds of factual changes we are typically interested in when we are concerned with the concept of moral progress, some opt to restrict at least the topic of moral progress to changes on a larger social scale, without restricting the concept itself to collective changes. For instance, Buchanan and Powell declare that they are interested in individual changes under the heading of "moral progress" "only insofar as these occur in sufficiently large number of people to effect social change" (Buchanan and Powell 2018, 47).

This way of restricting the topic at hand for a given project seems preferable to restricting the concept itself for several reasons. First, even when we are concerned with collective efforts that have been critical in effecting changes for the better, we might still want to be able to explain the occurrence of these efforts with reference to the contributions of individuals and in that, we also might want to refer to the progress these individuals have made individually. A second and related reason is that in addition to giving explanations, we are typically also interested in assigning responsibility when we are concerned with moral change (for the better or for the worse) and here again we typically care about individual contributions. In understanding and evaluating these contributions, we might often take individual moral change (for the better or the worse) into account. Once we think of these individual contributions as contributions (to a cause), the concept of individual moral progress is connected to changes in states of affairs. Third, we need to take into account the option of differentiating between development and progress in terms of distinction (3), i.e., between unreflective change and change induced by reason. This option will be discussed at length in § 4.2. At this point, we must note that if we wanted to claim that moral progress was moral change for the better that was induced by reason or that crucially involved reasoning, it might be problematic if we also determined that moral progress is not made by individuals but collectives, as this might eventually require us to rely on the notion of collective intentionality. Collective intentionality is debated within social ontology (cf., e. g., Jansen 2017), but references to this debate are remarkably absent from the contemporary discussion of moral progress. We will not enter into it here, either. The individual level is where we can allocate action, responsibility, and moral beliefs most uncontroversially. Also, changes in practices and beliefs of individuals are of interest in their own right. Since they can be changes for the better or the worse, they are aptly viewed as candidates for constituting moral progress. Individual moral agents are also individuated more easily than groups and it is more straightforwardly accepted that they persist over time. For all these reasons, the concept of moral progress will be explicated primarily with reference to changes undergone by individuals. However, the aim will be to produce an account of moral progress that will not only lend itself to attributions of moral change for the better to individuals but also to larger social entities. In applying the concept of moral progress in this way, one incurs the task of giving an account of how the relevant groups might be individuated, how they are constituted and persist. How this is to be done is a matter we will not settle here. These questions about the underlying subject of moral progress on a larger social scale transgress the present project. In the present context, when supra-individual progress is concerned, morally progressing groups will be conceived of as aggregates of individuals and the question will be whether those individuals change in a way that is adequately viewed as moral progress.

#### 4.1.2 Global and Local Progress

Another question about the dimensions of moral progress concerns its temporal expanse. As we noted earlier (§ 2.1.), progress is a process that will have to be distinguished from an event by its temporal expansion and directedness. When it comes to moral progress, our interest in durability may not only be motivated by our conceptual commitments about progress, but based on moral grounds. A certain temporal expansion is not merely a conceptual condition that gives us a criterion for detecting progress; it may also be seen as a moral desideratum. We are normatively interested in sustainable moral progress, not moral progress that relapses into its antecedent state immediately. Not only do we need the change in question to last for a considerable time to determine that a change has occurred at all; we also find that durability is what makes moral progress morally desirable and theoretically interesting.

The conceptual condition and the moral desideratum must be kept apart. The fact that we can sensibly speak of "sustainable" or "durable moral progress" points us to the separability of the morally relevant idea of durability from the idea of a process that is definitive of moral progress. Moral progress can be durable or lasting—but it does not have to be. It can be reversed and disrupted—it can be followed by moral regress (the reversal into an earlier stage) and decline (the deterioration into some other direction). Our pragmatic interest in lasting change for the better should not be built into the concept of moral progress. It might not be morally desirable that a period of moral progress be followed by an episode of moral decline, but it should be conceptually possible. And still, the concept of moral progress calls for whatever is able to fall under it to display the character of a process. A singular moral high performance—by virtue of being more of a moral event than a process—must be, in principle, distinguishable from immediately rolled-back moral progress. The way to account for this is to require some minimal duration for moral progress but not (morally desirable) permanence.

So, there is a requirement of minimal continuance on moral progress. Either the development toward the better itself has to take some time in order to be identifiable as progress or the moral views and/or actions that result from the change in question have to persist or be repeated over some time. Moral progress need not come about slowly and incrementally. It might happen abruptly, but for the change toward the better to be identifiable as moral progress in this case, it must not collapse as quickly as it came about.What is recognized as a sufficiently durable change for the better will depend, inter alia, on normative expectations, the evaluation of the capacities of the progressing (and regressing) individual (or groups), and the circumstances under which the relevant change takes place.Whether extremely short-lived progress will have to be recognized as progress will depend on the specifics of the case. The requirement of minimal continuance simply tells us that it has to persist over some time.

Commonly, when moral progress is under consideration, the envisioned time frames are anything but "minimal." For instance, when the hypothesis is that "the human race has enjoyed significant moral progress over the course of history" (Lachs 2001, 173), the focus is shifted toward world-historical dimensions. Many of those who are and have been concerned with progress are after precisely these kinds of judgments about the existence of large-scale, intergenerational progress (e.g., Condorcet 1963, Kant 1979, Ginsberg 1944, Linklater 2010, Singer 2011, Kitcher 2011, and Huemer 2016). Again, this understandable interest in large-scale progress does not require us to construe moral progress, conceptually, as being so durable. Our moral and intellectual interest in lasting and largescale moral progress does not depend on a concept of moral progress that only recognizes these types of moral progress as moral progress.

The difference between temporally large-scale changes and more circumscribed developments in smaller communities is sometimes marked as one between "global" and "local" conceptions of moral progress (Moody-Adams 2017; Jamieson 2002). Those who think that the local conception provides the preferable perspective usually argue that it is only on smaller scales that judgments of moral progress can be justified. For instance, as Godlovitch says about metaphysically laden theories of progress: "Big scale strongly teleological hypotheses such as those favored by Hegel, Marx and Spengler prove massively underdetermined by the evidence" (Godlovitch 1998, 274, emphasis added). The same concern about underdetermination—can be extended toward less metaphysically charged claims about historical moral progress in general.

On the other hand, proponents of the idea that there has been global progress sometimes turn the argument from underdetermination on its head. They, too, can be said to care about minimal continuance, after all. Their perspective is simply that changes that are worthy of being called "moral progress" must be identifiable as durable trends, and they are thus identifiable only on a larger historical scale. Social trends that might count as moral progress, like individual development, must be distinguishable from short-lived positive anomalies in shared moral practices—and the distinction can only be reliably drawn by taking larger periods of time into consideration.

What is important to note about the disagreement among proponents of local and global perspectives, is that it concerns the actual application of the concept of moral progress—not so much its content. The challenge for applying the concept of moral progress on the scale of intergenerational processes lies in the identification of a persisting subject of moral progress. For the specific comparative judgments that are moral progress-judgments to be applicable, there must be relevant substantive relations between the entities being compared (cf. § 2.3.). In the case of intergenerational progress, the problem that we might end up talking about a mere better-than relation is not so severe as in the intercultural case (cf. § 2.3.(4)), because one of the relevant relations to be identified between descendants and ancestors is, in any case, actual biological succession. By way of enculturation, there will be cultural transfer of knowledge and moral beliefs between generations. In an obvious way, they are engaged in the same "ethical project" (Kitcher 2011). Under such conditions, it might be possible to say that societies progress over generations, even though the individuals who belong to different generations might stand in mere better-than-relations to one another as an effect of intergenerational progress (unless they have also progressed individually and their respective progress can be compared).

If the social scale or moral progress need not be seen as definitive of the concept, this is also not the case for its spatiotemporal scale. In principle, the scale on which moral progress-judgments are made could be as grand as we like. Determining the right scale will mostly be a matter of pragmatic interests and epistemic limitations.

#### 4.1.3 Domains of Morality

While decisions with respect to the spatiotemporal dimensions of moral progress are often cast in the language of "global" versus "local" moral progress, the "locality" of moral progress can also be interpreted as pertaining to aspects of morality; moral progress might be said to occur "locally" not only in terms of its temporal expanse but also because it is sometimes identifiable only as it affects one particular value or a particular context of action rather than morality at large. Michele Moody-Adams, for instance, argues that "[c]hanges that are reasonably deemed to constitute moral progress occur locally, in relatively circumscribed domains of concern" (Moody-Adams 1999, 169). This suggests that locality concerns the thematic scope or comprehensiveness of progress within morality. Moody-Adams does not fully specify what exactly she means by "domains of concern." She mentions that moral progress "within one domain" may entail regress in another (Moody-Adams 1999, 170, 179). The idea can seem problematic (cf. Dixon 2005) because judgments of progress might reasonably be expected to be all-things-considered judgments (Moellendorf 2017), especially in the case of moral progress.

What might "domains of concern" be and could the individuation of such domains possibly help us in finding a starting point for an explication of moral progress? First of all, we might think of such domains as different action contexts, or descriptions of social roles moral agents can take on in different actions contexts, for instance, in private interactions, economic, public, or political contexts. These could be further specified to identify a context in which evaluations of improvement could be reliably carried out. The problem with this way of thinking about "domains of concern" is that there is hardly a one-to-one relation between action contexts and sorts of morally relevant values. So, second, we could think of such values as individuating domains of concern by themselves, e. g., the "domain" of wellbeing, equality or autonomy. Third, we could think of domains of concerns as values in specific action contexts, e. g., fairness in the context of the education system. Alternatively, one could also think of substantive and structural elements of morality as constituting domains of concern, where values such as autonomy or wellbeing would be on one side and structural features such as moral norms' universalizability or stringency on the other (though the latter could be considered "domains of concern" only from a highly theoretical perspective). Problems with these ways of approaching the concept of moral progress will be discussed shortly (§ 4.1.4.).

Eventually, Moody-Adams points to yet another way of differentiating between aspects of morality. The types of domain-specific changes she explicitly mentions are "moral progress in belief" and "moral progress in practices" (Moody-Adams 1999, 169). These very general areas of "aspects of moral functioning" (Schinkel and de Ruyter 2017) will be taken here to provide us with an idea not so much of domains of (moral) concern, but of domains or constituents of morality: theory and practice (cf. § 2.2.). This coarse-grained differentiation will be useful to make the task of explicating moral progress manageable. There is likely not one natural ontological attitude toward the individuation of parts of morality or domains of concern that cuts morality at its joints. But it is common practice to evaluate moral agents in terms of their behavior and motives, and thus with respect to the domains of practice and theory. To break down the task of assessing possible constituents of moral progress, we will take the division of morality into theory and practice as a starting point.

#### 4.1.4 Identifying the Moral Point of Departure

In proceeding with an explication of moral progress along these lines, the ensuing investigation deliberately does not start from an account of "domains of concern" in terms of a set of values or moral principles. Deliberation about whether any given change counts as moral progress ultimately requiresanormative frame of reference, but laying out a basic understanding of moral progress is a different matter. The idea here is that the task of providing the fundamental conceptual conditions that an apparent change for the better has to satisfy in order to qualify as moral progress is separable from specific ideas about what is good or right and that it should be taken on before charging the conception of moral progress with these evaluative and normative ideas. I will say a bit more on why that is before I commence with the explication of MORAL PROG-RESS in § 4.2.

In principle, when moral progress is further specified (beyond the fundamental conceptual structure that is sought here), normative conditions can be formulated in either a neutral or a directed way. For instance, a neutral formulation of the condition of inclusivity (cf. § 3.1.) would indicate that moral progress occurs when the performance of moral agents is being improved in such a way that their moral code's inclusivity is being corrected, i.e., when adjustments to the scope of moral concern are being made. A directed formulation of the same condition could state, for instance, that moral progress occurs when the performance of moral agents is being improved in such a way that their moral code's inclusivity is being increased, i.e., when the scope of moral concern is widened (cf. Singer 2011). In order to be able to operationalize our concept and specify criteria for the occurrence of moral progress, we require directed formulations of normative conditions; we can only hope to identify indications of moral progress, if we know whether we have to look, for example, for signs of increasing or decreasing inclusivity in our moral code.

Sometimes, however, material criteria of moral progress are offered not as operationalizations of, but without a specification of the normative conditions on which they are based. On their own, criteria can be only vaguely indicative of the normative assumptions behind them. For instance, if the expansion of legal protection to nonhuman animals was advanced as indicative of moral progress, it would not be obvious what it was meant to be indicative of: it could be interpreted as an instance of the expansion of moral concern (inclusivity) or the increase in the stringency of a community's moral code. Underdetermining conceptual conditions for moral progress in this way is often the point of approaches to characterizing moral progress that rely foremost on criteria, as they seek to be acceptable to different normative theorists:

Fortunately, we do not need an analysis of moral progress. For many of our purposes, it would be sufficient to specify a working definition of moral progress by identifying a 'proxy property,' a reliable correlate of moral progress that could garner overlapping consensus by most of the major normative traditions. (Evans 2017, 76)

The idea articulated in this passage is that it is preferable to keep with an agreeable criterion of moral progress rather than to risk disagreement over the conceptual conditions to which that the criterion corresponds.

When the normative underpinnings of criteria are provided, normative conditions reveal what their proponent identifies as the moral point of departure (cf. § 3.1.). The directed formulation of the condition of inclusivity, for instance, identifies an overly exclusive moral code as the problem that constitutes the point of departure for moral progress. A directed formulation of the condition of stringency—the idea that "the domain of moral duty in an improving society, is always widening" (Mill 1969, 338)—reveals that its proponent takes the preexisting moral problem to consist in an overly lax moral code that does not identify all the moral agent's actual obligations. By selecting a direction for progressive change, approaches to moral progress that rely on directed conditions and criteria paint a certain picture of what has been going wrong so far.

In this way, directed normative conditions and material criteria are contingent on the actual defective antecedent moral conditions. They are indispensable when we want to employ the idea of moral progress in an evaluation of actual historical or present processes and as a guide to practical change in the future (see § 2.3.). However, as approaches to characterizing the concept of moral progress or as starting points for theories of moral progress, they share certain problems. For ease of expression, I will refer to approaches that start with criteria as well as those that start from directed conditions as "criterial approaches" hereafter, although this is a slight simplification.

One problem for criterial approaches is moral disagreement. What seems obviously desirable to some is fundamentally misdirected to others. It is inevitable that moral disagreement bears on the identification of moral progress. However, it is desirable that verbal disagreement about the expression 'moral progress' does not render this more substantially normative disagreement impossible. Moral disagreement should bear as little as possible on our basic notion of moral progress. Criterial approaches introduce the prospect of moral disagreement right into the basic understanding of what moral progress most generally is.

Synchronic moral disagreement is not the only problem of this type for criterial approaches. They can be outpaced by ethical progress, i.e., undermined by diachronic moral disagreement. Consider the example of anthropocentrism. For some, the claim that morality's ultimate objective is to serve "human flourishing" is simply obvious (Beauchamp 2010, 43). On this view, morality is constitutively anthropocentric in a normative sense: it ought to serve the good of human beings. Truly unqualified anthropocentrism is sometimes even seen as one of the most important results of morally progressive developments, as it is an advancement over more exclusive doctrines that legitimated the disadvantageous treatment of members of other cultures or groups (cf. § 3.4.1.). However, moral anthropocentrism is not as widely accepted as it used to be. Today, many think that this supposed moral obviousness conveys a seriously misguided moral view. Sentientists argue that the restriction of moral concern to humans or the moral prioritization of humans is a fundamental flaw in contemporary moralities that needs to be overcome. The expanding circle model of moral progress (Singer 2011) claims to identify one of the most important ways in which morality could progress. Today, the sentientist idea that "what matters to sentient beings matters, because it matters to them" (Donaldson and Kymlicka 2011, 33, emphasis in original) is increasingly recognized as a new moral obviousness. But criterial approaches to moral progress that reflect the idea that morality exists primarily or exclusively to further human flourishing cannot easily incorporate this insight merely by adding an item like "legal protection for moral animals" to a list of criteria—at least they cannot do so and thereby transform into truly sentientist accounts.

The underlying problem is the rigidity of criteria and (to a lesser degree) of directed conditions of moral progress. In the course of ethical progress, both can be subjected to reevaluation. In this process, criterial approaches to moral progress can lose all plausibility. The problem can be attributed to what Buchanan and Powell identify as a prevalent flaw in many accounts of moral progress, a "lack of epistemic humility" (Buchanan and Powell 2018, 87). Approaches to moral progress that suffer from a lack of epistemic humility do not do justice to the "open-endedness normativity of the ethical" (cf. §§ 3.1., 3.5.3.). Criterial approaches seem to be guilty of this epistemic vice. It is when a list of criteria for moral progress is proposed as an account of the nature of moral progress that the charge of a failure to account for the "open-endedness of normativity" is most apt. Criterial approaches to moral progress underdetermine the concept of moral progress, are fundamentally challenged by moral disagreement and vulnerable to changes that result from transformative ethical change (cf. § 4.1.4.). Therefore, the plan for this chapter is to leave normative conditions of moral progress out of the consideration of its basic conceptual conditions—thus refraining from identifying a moral point of departure.

Yet, even without further normative input, the conceptual account if moral progress to be developed here will already provide us with a minimal normative condition that might generate a criterion of moral progress. The condition might be called "irrevocability." Morally progressive developments are irrevocable in the sense that they cannot be taken back without moral damage. Because moral progress is a change for the better, it ought not to be reversed. Of course, it can be reversed as a matter of fact—it is not irreversible. Borrowing from Mackie, we might say that moral progress has not-to-be-reversedness built into it. This condition is sometimes alluded to in the literature, albeit under the heading of 'irreversibility':

I attributed irreversibility to moral progress of a sort that ratchets us up, and thereby closes off certain past options by creating ever-new and newly obvious moral points of departure. (Godlovitch 1998, 284, emphasis added)

Similarly, Wilson speaks of moral progress as becoming the object of progressive narratives—accounts of developments that clearly mark them as changes for the better. In contrast, non-obviously progressive developments—developments that might be morally neutral or whose status is controversial—lend themselves to "bi-directional narratives," descriptions of changes that could be told in reverse and would not be recognized as accounts of a decline (Wilson 2011, 109). The condition of irrevocability likewise might be straightforwardly translated into a criterion of moral progress, i.e., perceived irrevocability.We know moral progress by sensing that the change in question ought not to be reversed (it might need to be further transformed, but it must not simply be reversed). In this way, perceived irrevocability indicates moral progress. In moral discourse, such a "perception" would need to be backed up by arguments in order to convince others that the relevant change is to be recognized as progress. At this point, normative reasons will have to be given as to why the seemingly irrevocable changes count as improvements. This normative deliberation will be more straightforward if built on a shared account of the basic structural features of moral progress. It is such an account that will be proposed in the following section.

#### 4.2 Theoretical and Practical Progress

When theory and practice are addressed as two constitutive domains of morality, these two notions are used to draw a rather basic distinction between the realm of mental states—or the psychological side of morality—on the one hand and overt behavior—or the practical side of morality—on the other. Theory, in this sense, is not supposed to imply advanced moral reasoning or a coherent set of beliefs. It simply trades on the customary dichotomy between the realm of thought and the realm of action. If we refrain from restricting the notion of theory at least to cognitive mental states or events, i.e., beliefs, intuitions or judgments, the noteworthy consequence is that the theoretical side of morality will include a variety of rather different items. There are types of mental states that are set apart from the ones just mentioned, partly, but crucially, by how they relate to action: conative states (such as desires, wishes or intentions) are crucial to action because, according to a standard view (cf. Chapters 2 and 5), only in the presence of a desire-like state can an agent be moved to act, whereas cognitive states on their own do not command action. If the motivating forces of moral action are addressed as part of theory, we are using the term "theory" in a loose sense to refer to a fairly heterogeneous set of items. In the following, when progress on the theoretical side of morality will be concerned, it will sometimes be paraphrased as ethical progress (widely construed, cf. Chapter 2).¹⁴

When I speak of the theoretical side of morality, I mean to include all those psychological parts of morality that are sometimes said to be disregarded by philosophers whose overly narrow concern is with "moral theory." Samuel Scheffler discusses the supposed problem after claiming that "[m]oral change and moral development consist in adopting new theories to replace old ones, preferably on the basis of rational considerations." He points out that

<sup>[</sup>c]ritics of this picture argue that morality does not fundamentally consist in a set of propositions, and that most people have never 'adopted' any 'moral theory.' […] Morality, insofar as it is something real and not a philosopher's fiction, is an aspect of human psychology, and not a set of propositions […]. It gets entrenched in human beings […] by answering to powerful psychological needs, engaging human emotions, and serving significant social and psychological functions. (Scheffler 1992, 50)

He goes on to say that thinking of morality in terms of a set of propositions—i.e., in terms of theory, more narrowly construed—and in terms of "an aspect of human psychology" ought to be seen as "two complementary forms of understanding" (Scheffler 1992, 50). Morality is probably an aspect not exclusively of human psychology. Otherwise, Scheffler's reconstruction is very helpful. In the following, talk of the theoretical side of morality is meant to cover the psychological aspects of morality, not only propositional thought, and the rough distinction between theory and practice is employed to reflect the distinction between psychological and behavioral parts of morality.

Another complication with the rough distinction is that it does not offer a plausible place for behavioral dispositions. Such dispositions are properties of individuals to act in specific ways (with a certain probability) under certain external conditions. They fit neither on the theoretical nor on the practical side of our dichotomy. Their role in moral progress will occupy us throughout this section.

When we are concerned with moral practice, we are concerned with actions that can be evaluated in terms of their quality (e.g., in terms of being good or bad) or deontic status, i.e., in terms of being right or wrong (or permissible or supererogatory). Some such evaluations do not merely evaluate the practical side of morality but also examine how the theoretical side is involved. For instance, some views of moral rightness require that the theoretical side of morality is adequately engaged in action in order for the moral agent to count as doing the right thing: an action that is correct in terms of its conformity to a rule might still not be judged morally right if done for wrong (e. g., self-serving) reasons. Also, when actions are judged not in deontic, but in quality terms, this might be done to express a judgment of their consequences, but also to express a judgment of the motives of the agent. In the latter case, it is not merely the practical behavioral—part of the moral action that is evaluated; the theoretical side of the action is evaluated as well. If we want to draw a neat distinction between theory and practice as two domains composing the phenomenon of morality, and if we want to test the plausibility of talking about moral progress as improvement in either one of these domains separately, we cannot, in the following, presuppose notions of good or right action that require the kind of congruency between practical and theoretical performance just outlined. Therefore, when such evaluative and normative terms are applied to practices in the following subsections, they are used—for the sake of the investigation into the necessary conditions of moral progress—in a way that allows for there to be goodness and rightness without "the right kind" of theoretical involvement. The question of whether theory and practice can and should be so disassociated in evaluating moral progress is precisely what concerns us in this chapter—therefore, our use of normative vocabulary must not make this disassociation impossible.

#### 4.2.1 Moral Progress: Hollow or Disoriented?

Could moral progress, in distinction from ethical progress, consist entirely in "progress on the ground" (Jamieson 2002, 318), where this is taken to mean either the practices of moral agents or the resulting states of affairs? Alternatively, could we be justified in counting improvements in moral beliefs as a form of moral progress in their own right? So much is presupposed when it is argued that "moral progress in beliefs" and "moral progress in practices" can sometimes—or will even typically—outrun one another (Moody-Adams 1999, 183). Contrary to this liberal way of speaking of moral progress when there is change in just one domain, we might feel that we should rather withhold a judgment to the effect that moral progress has occurred until one has caught up with the other, i.e., until the theoretical side and the practical side of morality are in harmony. Initially we might assume that "ethical progress without practical change is hollow; practical change without ethical progress is disoriented." If we want moral progress to be effective or useful in terms of involving a change of practices (or even in terms of being conducive to axiological progress) and if we want it to be, at least in some sense, informed and morally deliberate rather than inadvertent, then, presumably, we need to include both a theory- and a practical change requirement in our account of moral progress. These requirements would demand, roughly, that moral progress involve a change in moral views and a change in moral practices. Possibly, they even should demand that the latter is caused by the former.

The present section (§ 4.2.) will vet this starting idea about moral progress, about the requiredness of both practical and theoretical change. The case descriptions relied on here will focus on individual agents and the question of whether they individually should be said to have made moral progress. The possible consequences for a picture of moral progress on a social and even intergenerational level (cf. § 4.1.) will be discussed in § 4.3. Of course, from the viewpoint of particular normative theories, some of the example cases will not be recognizable even as prima facie examples or proper test cases for our concept of moral progress. For example, in the case of the Unwilling Meat-Eater (see § 4.2.3.) who judges it to be wrong to eat meat but fails to act on this judgment on any given occasion, from a Kantian perspective, it may not seem apt to even consider the question of whether she has made moral progress merely by forming a judgment. As she routinely acts on passion rather than from duty, she clearly fails as a moral agent and may not even be ascribed the supposed moral judgment about the wrongness of meat-eating in the fullest sense to begin with. However, I take it that the characters described in this section to illustrate the consequences of incorporating or omitting different requirements in a concept of moral progress are moral phenotypes readily conceivable from the standpoint of ordinary morality and that, in any case, their consideration provides appropriate occasion for contemplating whether we should adopt the perspective of a specific normative theory in our account of moral progress.

The objective of this section is (a) to determine the extent to which the moral agent, in order to count as progressing morally, must be "intellectually involved" (i.e., "ethically involved") in the changes her moral actions undergo, (b) to find out whether she must change practically at all, and also (c) whether these changes must have positive effects outside of themselves. Closely related to these three issues is the question of whether agents must be intellectually involved in their moral progress in a yet more ambitious sense, i.e., in the sense of making a conscious effort to progress. We will address this question first (§ 4.2.2.).

#### 4.2.2 Must Moral Progress Be "Made?"

In the contemporary debate on moral progress, it has been suggested that moral agents can be thought of as becoming morally better in the way "the world" might be thought to become morally better, where this is first and foremost supposed to mean: through no effort of their own (Rønnow-Rasmussen 2017, 147– 148). In a paper on the value ascriptions involved in moral progress-judgments, Toni Rønnow-Rasmussen (2017) considers and rejects the idea that for the concept of moral progress to be applicable, there is an agency-requirement to be fulfilled: "that in order to have made moral progress, a person must somehow have been actively involved in this change" (Rønnow-Rasmussen 2017, 146). Before examining the reasoning underlying his rejection of the agency-requirement, let us consider the different ways in which such a requirement might be interpreted at the outset.

(1) One might think that an agency-requirement is a requirement of change in moral views that ensures that a behavioral change is accompanied by a rise in moral awareness. As changes in her practices occur, the moral agent must be properly intellectually involved and that requires that a relevant ethical progress accompanies the practical change. An agent who starts buying fair trade coffee instead of the cheapest brand in the store cannot be said to have made moral progress, if the change in her grocery shopping pattern occurs only because, after tasting the fair trade brand at a friend's, she prefers it for purely gustatory reasons. Only if she also comes to acknowledge that her new choice is the morally better one, which we can simply assume it is, does she qualify as having made moral progress. In version (1), the agency-requirement only requires the co-occurrence of relevant ethical progress and improvements in practices. On this interpretation, the agent can be said to be "actively involved" if her cognitive moral capacities are "activated."

(2) The agency-requirement could also be a conscious effort-requirement, according to which the relevant change in moral views must not have been "handed" to the agent, but she must have reasoned her way through the moral problem she confronts and actively gained a new moral insight, i.e., she must have made ethical progress in the sophisticated sense (see § 2.2.3.). On this view, the moral agent would not have made moral progress if she had come to see that the fair trade option is morally better because her friend had just told her so. Instead, she would have had to have engaged in some minimal form of deliberation before changing her shopping habits, grounding the new habit on a moral insight for which she herself has striven intellectually.

(3) The agency-requirement could be an initiative-requirement. Interpreted in this way, the agency-requirement would exclude from the extension of moral progress the case in which the moral agent did engage in some deliberation, but only after being prompted to do so by her friend's remark at the coffee table that she really ought to start buying fair trade coffee. Instead, the agent would have been required to have initiated the crucial deliberation herself. The initiative-requirement could be interpreted as either ruling out only improvements prompted by other moral agents or also all improvements that do not critically depend on an agent's preexisting commitment to self-improvement. The weaker former requirement seems to collapse into a conscious effort-requirement. On the distinct (more extreme) latter view, moral agents could not be said to have "made" moral progress if the change in question occurred due to any kind of moral epiphany or pointer.

(4) Finally, one could interpret the agency-requirement as a requirement of retrospective¹⁵ endorsement. This would mean that in order to make moral progress, the agent would at least have to intellectually approve of the change that she has undergone. The endorsement-requirement is not a substitute for, but presupposes requirement (1); it is a weaker substitute for requirements (2) and (3). The idea here is that if the moral agent does not reason her way to her new moral insight (requirement (2)), let alone initiate her ethical progress (requirement (3)), then she should at least eventually come to recognize that she has undergone an ethical improvement. The endorsement-requirement calls for the moral agent to recognize the "better-than"-relation between her former views and her new views (and conduct)—a relation she would have had to recognize earlier had she reasoned her way to the new view (requirement (2)) and she would have actively sought, if moved by a desire for self-improvement (requirement (3)). On this view, the moral agent who starts buying the fair trade coffee can be said to have made moral progress only once it occurs to her that by coming to

The endorsement would be retrospective regarding the initial change but not with respect to the later continuation of the new practice.

think that she morally ought to buy this coffee, she has improved her moral views and thus her conduct.

Of these possible interpretations, (1) is obviously the least demanding. In the upcoming subsections (especially § 4.2.3.), it will be investigated further, though it will not be referred to as an agency-requirement, but as a "theory-requirement" or "requirement of theoretical change." The more demanding requirements (2)- (4) are more aptly viewed as possible interpretations of an "agency-requirement." They all serve to make the change in question attributable to the moral agent undergoing it in a more thorough way than simply by "locating" the change in the agent or allowing that her mental moral capacities be "activated." They require the agent to be the author of her change in some more meaningful way. This does have some intuitive appeal and connects well at least to one set of crucial functions of the concept of moral progress: facilitating recognition and praise (see § 2.3.). It is when we are crediting a moral agent with or praising her for having made moral progress that we assume her to be properly "actively involved" in the improvement we have in mind. Requirements (2)-(4) secure such an active involvement. But the question is whether this pragmatic interest is so crucial that the requirements that answer to it must be inscribed in our concept of moral progress. In the following, I will reject this restrictive idea of moral progress. In so doing, I will discuss Rønnow-Rasmussen's argument only in the context of the rejection of requirement (2), as this is the kind of interpretation he gives to the agency-requirement. Let us consider requirements (3) and (4) first.

Requirements (3) and (4) introduce a quite highly intellectualized picture of the moral agent who is capable of making moral progress. They portray her as someone who is either driven by a desire for moral self-improvement (requirement (3)) or at least attentive to the fact that the change she has made is an improvement (requirement (4)), i.e., she is prone to making judgments of moral progress (or regress) about herself. Earlier (§ 2.2), we rejected the claim that an individual incapable of fulfilling the requirement of retrospective endorsement (4) lacks a level of ethical reflection that is required for genuine moral agency to begin with. This rejection does not settle the question of whether this could be a requirement for moral progress. Maybe not all moral agents have the capacity for making moral progress. The following three ideas about moral agency and moral progress would be compatible: a) the capacity to make moral progress is not a condition for moral agency; b) some moral agents are capable of making moral progress; c) in order for moral agents to be capable of making moral progress, they must be capable of judging themselves to have made moral progress. In other words, refuting an intellectually demanding threshold for moral agency is not yet to refute an intellectually demanding threshold for moral progress.

However, requirements (3) and (4) simply are objectionably restrictive. They exclude all cases of moral improvement from being cases of moral progress in which agents come to appreciate the morally relevant facts better and come to hold more adequate views without reflecting on their improvement as an improvement or without having pursued it as an exercise in moral improvement. At this point, the consideration against these requirements as formulating a threshold for moral progress is of the same kind as in the case of the supposed threshold for moral agency: the concern is that they exclude too much. Those who have a fundamentally more demanding views of moral action and improvement will naturally disagree. I can only try to give presumptive evidence for the fruitfulness of a less restrictive concept of moral progress.¹⁶ Requirement (3) will be rendered implausible a fortiori, if the assessment can be supported that requirement (2) is too demanding. Inasmuch as requirement (4) can be conceived of as an afterthought to the considerations expressed in (2) and (3)—a backup substitute for these requirements—, the idea of a need for this kind of substitute will be weakened accordingly, if the case against (2) and (3) is convincing.

The conscious effort-requirement (2) seems to demand too much in that it rules out that a better view may simply "occur" to an agent. Consider the following excerpt from an interview with Jeff McMahan:

Lurking behind [his] formal discussion of killing humanely-reared animals lies McMahan's decision to go from being a hunter in South Carolina to a committed vegetarian. He made

There are two non-decisive concerns about requirements (3) and (4). The initiative requirement (3) demands that the morally progressing agent be motivated in a way that could be seen as problematic in its own right. The objection here could be that being driven by a desire for self-improvement renders the moral agent objectionably disconnected from the aims of her moral actions and objectionably focused on herself. Ultimately, in Chapter 5, I will offer a defense of the desire for improvement against the allegation that such a desire shows the agent to be objectionably self-involved. However, this defense will also show the kind of motivation to be probably somewhat unusual.

Requirement (4) introduces the oddity that a change in practices and in moral views is only retrospectively turned into moral progress once the moral agent realizes she has undergone a change for the better. Progress becomes progress only after the basic change has occurred to the agent. If the moral agent realizes that buying the fair trade coffee brand is morally better only once she is half a year into her new shopping habit, there isaconsiderable time lag within her moral progress. She may not have reflected while buying the new brand of coffee on the fact that this is an improvement over her old habit. But the salient change—the switch from buying one brand of coffee to buying another—has happened months ago at this point. The temporal dissociation at the very least is a conceptually awkward consequence as far as the conceptual component "progress" is concerned. But it may not be sufficient to disqualify the endorsement-requirement on its own.

the choice as a teenager, after he saw a man wound a bird on a dove shoot. 'I remember this man walking in such a leisurely way towards this bird that was flapping across the ground trying to evade him,' he recalled. And so McMahan sold his gun and stopped hunting. (Goldhill 2017)

Apparently, this is the very brief description of a behavioral change pointed by a moral epiphany. Should we accept that this change does not qualify as moral progress, because it came about effortlessly?

Requirement (2) rules out any moral improvement counting as moral progress that comes about by a genuinely sudden insight. More generally, it is incompatible with any non-rationalist moral epistemology: it rules out moral improvement that is prompted by moral emotion or moral intuition. It could imply that many of the supposed paradigmatic instances of moral progress do not qualify as actual moral progress, because those involved were not sufficiently ethically involved.

It is noteworthy how adopting this requirement would position us with respect to the paradoxes of progress to which Mill points us (§ 3.4.2). Mill's considerations of intellectual progress reveal a tension between the dissemination of truth and the intellectual vividness with which true moral beliefs are held the tension that forms the theoretical background for the paradox of fallible progress. Requirement (2) resolves this conflict not in favor of the proliferation of truth, but in favor of intellectual vividness. The requirement would rule out intergenerational moral progress that builds on "moral obviousnesses." Subscribing to requirement (2) would mean that, instead of saying that when better moral beliefs are adopted without a conscious effort, the progress that is made is somewhat smaller than in the case where such an effort is present, we would have to deny that moral progress has occurred at all. However, the former assessment—while also in keeping with Mill's resolution of the analogous problem for intellectual progress—seems to be the more nuanced one. Requirement (2) would also entail that those resisting changes for the better longer qualify as making moral progress whereas those who accept new moral ideas right away do not. We may consider the case of the abolition of slavery here: those who accepted the abolitionists' case as it was presented to them would, according to requirement (2), not qualify as having made moral progress. Those who resisted it and had to be convinced through elongated argument would have proven to have made a conscious effort in grasping the new, better moral view through their resistance.

In his treatment of the agency-requirement, Rønnow-Rasmussen also rejects what is, in substance, a conscious effort-requirement, but his argument is based on the consideration of yet different types of cases and on the discussion of the sense in which "the world" can be said to make moral progress. The cases he considers are these:

Suppose someone is born with a mutation making him in an unparalleled way sensitive to other people's suffering. Or suppose someone, unknowingly to another person Geert, has tinkered with Geert's brain to the effect that he now sincerely renounces his racist views and actions from the past. (Rønnow-Rasmussen 2017, 147)

Note that the first case prima facie lacks a procedural character at least as far as the intrapersonal evaluation is concerned: since the person is simply born with the mutation, she could only be said to be "more progressed" vis-à-vis other moral agents—and this borders on using the concept of moral progress in the sense of a mere comparative judgment concept (cf. § 2.3.(4)).

Regarding the case of Geert, note that he is not merely said to have had racist views but to have performed racist actions as well. So, the change to be envisioned here fulfills requirement (1), there has been a change in moral views, but this is not the only change that has occurred: there has been practical change as well.

Rønnow-Rasmussen then goes on to argue that whatever we might say about Geert himself, the change he undergoes may be enough to say that "the world" he inhabits makes moral progress:

we might want to say that Geert's change made the world better in some sense, and that this is enough to say that at least the world has morally progressed. I am inclined, albeit for reasons of a purely intuitive nature, to say the latter. (Rønnow-Rasmussen 2017, 147)

Through the change in Geert, the suggestion goes, there has been valuable change in the world. Rønnow-Rasmussen distinguishes "positive change" from "progress" by requiring the latter to involve a "correction" over and above a valuable change. The change undergone by Geert, he argues, makes the world progress morally, because it is a valuable change that occurs in this world and involves a correction (from racist views to the rejection of racist views) (Rønnow-Rasmussen 2017, 147). This way of attributing moral progress to something, he observes, is obviously possible only based on a "less demanding" idea of what it is to "make moral progress" than the agency-requirement would allow (Rønnow-Rasmussen 2017, 147–148). The next move in the argument is this:

But if I am ready to use this attenuated sense of moral progress about the world, what logically or even substantively prevents me in the first place from applying it to Geert? As far as I can see, nothing. […] a person who has undergone some valuable changes […] because of favourable circumstances out of his control […] might then qualify as having made moral progress in this latter sense in which the world has made moral progress. (Rønnow-Rasmussen 2017, 148)

The idea here is that once we have brought ourselves to saying that something that is not properly seen as an agent in the first place—viz., the world—has "made moral progress," we can see how this way of making progress could be said of a moral agent as well.

I think this detour weakens rather than strengthens the case for rejecting an agency-requirement in terms of a conscious effort-requirement. That is because it relies on a metaphorical way of using "making moral progress" in a context where this is applied to an entity that certainly "is not actively doing anything" (Rønnow-Rasmussen 2017, 148)—the world. The world is not simply "not doing anything," it is not capable of doing anything—it is not an agent. Saying that the world "makes" moral progress is not merely a very loose way of talking—it is better seen as metaphorical. The world, interpreted as the spatiotemporal configuration of entities among which the proper subjects of moral progress—moral agents—make or undergo moral progress, could not even "undergo" moral progress. By saying that "the world morally progresses" one has only established a figurative way of saying that someone in the world makes or undergoes moral progress (i.e., let us agree, a valuable change involving a correction). Rather than using this figurative speech and then applying it in the sphere of proper agents, it would be more straightforward to claim that, intuitively, moral agents who undergo a correction of their views and actions do not need to be more involved than they are by somehow acquiring new views in order to qualify as making moral progress. After all, the entire argument still depends on what goes on within moral agent. The way the world can be said to "make moral progress," according to Rønnow-Rasmussen, is still firmly anchored in changes in the mind of the moral agent: what makes the valuable change an instance of progress is that it involves a correction. Given that the specific correction envisioned here is a renouncement of racist beliefs (which might be straightforwardly moral beliefs, but also empirical ones), it is not clear whether the correction is supposed to be a moral or an epistemic one. Yet, either way, it would require a moral or epistemic subject. "The world" itself cannot be corrected in the relevant way. It can simply be the stage of a correction that occurs in an agent's belief system. So, the detour via the metaphorical way in which the world might be said to make moral progress eventually even seems unnecessary.

However, the case of Geert might be seen as more helpful for testing the conscious effort-requirement than the case of the agent to whom a moral insight simply "occurs." In the latter case, one might still be led to think that the moral agent would at least have to accept the new moral insight, which still required a minimal effort, therefore, the case was not a good illustration of what would be ruled out by the conscious effort-requirement to begin with. Geert, on the other hand, did not even have to accept a new insight. What the conscious effort-requirement is supposed to rule out were such cases of sudden, passive conversion as illustrated by the case of Geert.

Now, instead of taking the detour via a metaphorical sense of moral progress made by the world, we can simply recognize that our pragmatic interest in the concept of moral progress does not command us to incorporate an agency-requirement (interpreted as a conscious effort-requirement) into our concept of moral progress (thus ruling out a case like Geert's). In fact, for functions other than praise and recognition (see § 2.3.), none of the requirements (2)-(4) are obviously necessary, and even in the case of praise and recognition, it is not clear why those must be restricted to changes for the better that the agent has actively reasoned through, willfully pursued or retrospectively endorsed. On the contrary, we may praise an agent precisely because she has not yet appreciated her own development as progressive herself. And even if we choose not to find the moral agent with a moral epiphany praiseworthy for the change she undergoes following her conversion, we could still recognize that she has undergone a change for the better and express that by saying that she has made moral progress. The relevant differences between these changes simply need not be interpreted as marking the difference between the occurrence or absence of moral progress.We have plenty of ways to draw distinctions among phenomena of moral progress to mark different degrees of agential involvement. In case there is pragmatic need to imply more agential involvement, we can satisfy this need either by fashioning an explication tailored to the specific purpose at hand or by relying on specific terms referring to intentions and the moral agent when characterizing how some moral progress has come about: we can speak of moral progress being "made," "pursued," "brought about" or "initiated." By rejecting requirements (2)-(4), we preserve the option of saying that moral progress can simply "occur" and be "undergone," whenever such agential efforts are not involved in the change in question.

We can passively strengthen the omission of an agency-requirement by refuting arguments in support of the requirement. Two further considerations in favor of an agency-requirement shall be dealt with here. First, it could be argued that the requirement is needed to rule out the possibility that we "simply breed our way to moral progress," (Jamieson 2002, 332, n. 40), i.e., that moral progress comes about as an effect of mere demographic growth. Let us call this the "demographic concern." Second, it could be claimed that the way we judge cases of supposed moral regress or decline reveals a commitment to an agency-requirement. Therefore, the same type of agency should also be required for morel progress. Let us call this the "symmetry concern."

The demographic concern could be spelled out in two different ways. In the first version, we look at the case of the person "born with a mutation making him in an unparalleled way sensitive to other people's suffering" (Rønnow-Rasmussen 2017, 147). This person's condition enables her to perform better as a moral agent than other agents before her. If her offspring inherited her morally advantageous mutation, eventually spreading it throughout the population, creating new generations of better moral agents, it seems as if we are forced to call this moral progress, if there is no agency-requirement in place that demands more than evolutionary forces to be at work in the betterment of mankind. The second version of the demographic concern is an adaptation from the following consideration put forward by Dale Jamieson (which in turn is a variation of Parfit's "mere addition paradox"; Parfit 1984, 419):

Suppose that all the same people exist at the same levels of happiness in state of affairs A and its consequent, state of affairs B, but that there is an additional slightly happy person who also exists in B. This would not be sufficient for supposing that the transition from A to B constitutes moral progress. The moral of the story is that we cannot simply breed our way to moral progress. (Jamieson 2002, 332, n. 40)

The change that Jamieson discusses as a possible instance of moral progress is a mere change in axiological value: an increase in happiness. However, if we focused on the people living in A and B not in terms of their experience of axiological value, but in terms of their moral agency, we could say that: if all the same people exist and do morally well in state of affairs A and its consequent, state of affairs B, but that there is an additional well performing moral agent who also exists in B, and if there is no agency-requirement in place for moral progress, we are forced to accept the succession of A by B as moral progress. If moral progress does not have to be achieved, the mere multiplication of moral agents who are doing well seems to suffice for there to be moral progress. The demographic concern is that this cannot be correct, just as it cannot be correct that moral progress manifests in the spread of moral mutations.

The reply to version one of the concern should be that the moral agent's genetic make-up is not all that moral progress factually depends on. There is still the possibility of a "practical gap" (see § 4.2.4.), i.e., a gap between the agent's sensitivity and her acting in accordance with this sensitivity. There is still the possibility of moral failure in terms of nonconformity to the moral provisions the moral agent is able to detect through her relatively higher sensitivity. The fact that evolution continues to influence the preconditions of moral action does not pose more of a threat to the existence of moral progress than it does to the existence of moral action more generally. It is at the very least not obvious that better action that is facilitated by naturally evolving capacities is not to be recognized as moral progress. If this were the case, any moral progress-judgment would prima facie be undermined, as evolution has created the preconditions of moral action in general. To show how further moral progress is possible despite these evolutionary set preconditions is one of the main concerns in the major contributions to the contemporary debate on moral progress (Singer 2011, Kitcher 2011, and Buchanan and Powell 2018). However, they all must assume that moral progress is, in one sense, only possible because of the workings of evolution.

Addressing all the implications of the second version of the demographic concern would require more comprehensive normative theorizing about moral progress than this investigation will even attempt to offer. However, we can note at this point that if we wanted to guard against the mere multiplication of good moral agents to count as moral progress, we would not need to rely on the agency-requirement. The deterrent for applying the concept of moral progress in the second demographic growth scenario is not that no change is actively made. Rather, it is the fact that no moral change is undergone by any moral agents whatsoever. There is neither intrapersonal change nor even a better-than-relation between the additional moral agent and the rest of the moral community in scenario B. If we want to exclude such cases from our concept of moral progress, we can do so by specifying that the change in question cannot merely be a quantitative change in good actions being done but requires a relative improvement of moral views or practices in the performance of a significant portion of the population of scenario B compared to scenario A, i.e., that moral agents had to do better individually or become collectively more efficient in the fulfillment of moral requirements. No agency-requirement in terms of a conscious effort-requirement would be needed.¹⁷

Ultimately, it is, however, not obvious that we even should excludeascenario of mere demographic growth from counting as an occurrence of moral progress. When there is an additional agent who is doing as well as (and not individually better than) the other moral agents, her presence does not increase the share of right acts over wrong acts, but, if everyone in the moral community is doing well rather than badly overall, her presence also adds to the creation of value. Furthermore, it is her mere presence that could make it the case that the entire moral community could come to do better morally by becoming more efficient: her additional moral workforce could improve the workings of institutions. In that way, there could in fact be a relative improvement within the moral community brought about by the mere addition of moral agents (who are, individually, not better than average moral agents). I will get back to the demographic concern in § 4.2.6., but even then, it will only be possible to demonstrate that potential moral progress by mere demographic growth is not obviously a problem and that its evaluation depends on farreaching normative considerations. The upshot of this observation is that it will be those norma-

The symmetry concern centers on the assumption that our judgments of moral progress and moral regress should be symmetrical and that whenever we do not judge instances of apparent moral improvement and apparent moral deterioration symmetrically, the underlying reason is often an asymmetry in the capacity to exercise agency in terms of making a conscious effort. For instance, as children grow up, they also mature morally, and this often involves their conscious effort or an increase in their capacity to make conscious efforts to reach better moral conclusions. Therefore, we might be prepared to call the development during which they increasingly acquire justified moral beliefs "moral progress." Contrast this with the case of a dementia patient who, due to her condition, exhibitsadecrease of prosocial behavior—becoming ever less interested in others, less cooperative, and more aggressive. Would we charge her with "moral regress?" According to the symmetry concern, we should not, and this reflects an underlying asymmetry in these cases: children increasingly exercise agency as they acquire new moral beliefs and manage to act in accordance with them. Only insofar as they are actively involved in their moral maturing by making a conscious effort for moral learning are they said to make moral progress. The dementia patient lacks the counterpart to the children's conscious effort-based achievement: she does not fail to make an effort to be a better person—she does not fail to exercise moral agency by neglecting the maintenance of her moral status quo. Because she undergoes a diminishment of the capacities required to make conscious efforts, the decline in her moral performance is not attributable to her in the way the children's moral improvement is attributable to them as growing persons. If the dementia patient's condition had the effect that she became a nicer and more pleasant person to be around, we would also not attribute moral progress to her; because a change in her character traits simply happens to the patient, she does not make any effort to become a nicer person. So, our judgments of moral progress and regress are symmetrical, because they track the involvement of the agent in the relevant change in terms of the conscious effort-requirement.

The response to the symmetry concern is that it is likewise anything but obvious that we should judge the cases as suggested and treat moral progress and moral regress as symmetrical because our judgments track conscious efforts. If dementia made a person aggressive and more difficult to get along with than she used to be, we might well think that it is correct to attribute a moral regress to her, if we think that this judgment does not amount to charging her with a

tive considerations, not the inclusion or exclusion of an agency-requirement in the concept of moral progress, that will have to settle the matter.

reproachable failure to make a moral effort. The pragmatic function that is served by the inclusion of a conscious effort-requirement in the assessment of moral progress is the one to convey praise and blame. If moral progress necessarily involves a conscious effort, then acknowledging that moral progress has occurred is, inherently, to acknowledge a moral agent's achievement. However, this pragmatic function is neither obviously more important than other functions, nor is it dependent on moral progress being defined in a way that involves the conscious effort-requirement. Again, we can mark the distinction between moral progress in virtue of which the moral agent deserves moral recognition or praise and moral progress without this characteristic in other ways: we can differentiate between moral progress that is "made," moral progress that is "undergone," and possibly even moral progress that simply "occurs."

In the case in which the dementia patient becomes an overall nicer person than she used to be, we could say that, paradoxically, the disease caused her to undergo moral progress. And in case we were reluctant to issue this judgment, the reason for this reluctance need not be that this case of moral progress is lacking in conscious efforts, but rather our recognition of the fact that the patient has lost the grasp of the relevant moral beliefs. The reluctance to ascribe moral progress would then be grounded in the patient's failure to fulfill the agency-requirement interpreted in terms of (1), a requirement of change in moral views consistent with progress. No reference to conscious efforts or the lack thereof is needed to make sense of attributions of moral progress and cases where this attribution is withheld. Likewise, it is not so clear that our evaluation of morally maturing children should be based on attributions of conscious efforts to them. The moral maturing of children probably often does not involve conscious striving for moral improvement. Nevertheless, we could sensibly speak of them as undergoing moral progress—even when they are not making conscious efforts to make moral progress.

A concept of moral progress that does not require conscious effort or the other forms of agency captured by requirements (3) and (4) is more flexible when it comes to the adequate classification of cases like these. What we have not ruled out here is that moral agents might have to be somehow involved in their moral progress with what belongs to the theoretical side of morality. But requiring that this involvement must take the form of conscious efforts, ethical initiative or retrospective endorsement results in an overly restrictive concept of moral progress. We should leave it open how the moral development in the described cases should ultimately be characterized based on a specification of a more modest theory-requirement (e. g., in terms of a requirement of change in moral views) and how this requirement connects with other dimensions of moral evaluation. Waiving the agency-requirement does not seem to involve any severe detriment to conceptual clarity and function. Rather, doing away with this requirement should render the concept of moral progress more fruitful overall.

#### 4.2.3 Uncertainty Regarding the Roles of Theory and Practice

Deciding whether moral progress must include—or could even consist entirely in —a change on the theoretical side of morality bears on many different areas of interest, marked by reasonable disagreement. For instance, a tendency toward a consequentialist, a Kantian or virtue ethical approach would make, in that order, the priority of practice, the priority of theory or a requirement for interaction between both seem more plausible. Which of these we regard to be crucial for moral progress also bears on the plausibility of the claim that moral progress has occurred in the past and can be pursued in the future. It also has implications for the kind of evidence that could be presented in order to back up such claims and the means by which moral progress is to be pursued. Furthermore, it determines the range of possible downsides to any seemingly progressive development: if moral progress requires theoretical resources, any prima facie improvement in practices that comes with a disadvantage in terms of moral awareness might not be deemed an instance of moral progress overall. This is the concern Mill raises with respect to the proliferation of true beliefs (see § 3.4); as a true belief becomes widely accepted, people's grasp of it decreases because they are no longer challenged to defend it. The same concern could be raised about moral change. For instance, the proliferation of human rights could make it the case that moral agents' understanding of the "moral rights doctrine" abates. Even if the development were to be judged as progressive all-things-considered, should the decline in understanding (i.e., the regress on the theoretical side of morality) be counted as diminishing the amount of moral progress that has been made? Finally, clarifying the role of the theoretical side of morality (conceived as relating to morally charged mental states in distinction from practices) for the attribution of moral progress also bears on the question of what moral philosophy has to contribute to moral progress. Some authors who take part in the contemporary debate on moral progress are quite optimistic about ethics' relevance to moral progress (Nussbaum 2007; Elzanowski 2013), but the extent to which we can expect moral theory (in the narrow sense of systematic rational engagement with questions of and about morality) to contribute to moral progress depends on whether we are ready to recognize the kind of improvement it can involve as constitutive for moral progress. If moral progress consisted entirely in changes in moral views, the potential contributions of moral philosophy to moral progress would be indefinitely more substantial than if moral progress was construed as requiring changes in practices as well.

The question of whether we should incorporate a requirement for practical change in our account of moral progress or whether moral progress might be made merely by achieving improvements in one's moral views will be taken on in the following subsection. Depending on how we frame this question, the idea that theoretical improvements might suffice for making moral progress can seem outright absurd. The question "Might moral progress be made 'in theory?'" expresses a somewhat repulsive contemplation. Moral improvements that would occur only "in theory" appear to be severely flawed. Morality "seeks" to guide action, to be practically relevant, regulate social interaction. In other words, action guidance and practical relevance more generally are what we want from moral judgments. Improvements that remain theoretical fail to attain moral objectives. As 'moral progress' is, inter alia, a laudatory term (cf. § 2.3), it would seem at least premature to apply it to merely theoretical improvements that have yet to be put into practice.

In the contemporary debate on moral progress, the question about the respective significance of theoretical and practical change for moral progress is anything but settled. Some do hold that progress in either domain suffices to justify speaking of moral progress. For instance, Moody-Adams distinguishes "moral progress in belief" from "moral progress in practices," recognizing both types, without qualification, as moral progress (Moody-Adams 1999, 169).

Moral progress in belief, for instance, is progress in grasping what Mark Platts calls the "semantic depth" of particular moral concepts (1988, 287–88, 298–99). This involves coming to appreciate more fully the richness and the range of application of a particular moral concept (or a linked set of concepts), as well as understanding how some newly deepened account of a moral concept—some new moral conception—more adequately captures features of experience which the concept aims to pick out. (Moody-Adams 1999, 169)

This characterization of "moral progress in belief" involves notable references to the practical domain in that it mentions experience and the application of moral concepts. It seems that Moody-Adams subsumes under theoretical progress not only increases in the ability to apply moral concepts, when the moral agent is struck by a relevant feature, but even the ability to identify the relevant feature in the first place. The connection between theory and practice—the grasp of moral concepts and moral action—is pictured as being even deeper in other passages:

newly deepened moral understandings can be widely "disseminated" only if engaged social critics and political actors can get others to confront and reject their shallow grasp of moral concepts, and then to contemplate ways of embodying some deeper understanding in everyday experience (Moody-Adams 1999, 171, emphasis added)

Still, the characterization of "moral progress in belief" is properly viewed as a characterization of purely theoretical progress: it does not require the improved understanding of moral concepts to become practically relevant, yet it acknowledges this kind of change as "moral progress," with no reservation. Practical progress, in contrast, is described as follows: "Moral progress in practices results when some newly deepened moral understanding is concretely realized in individual behavior or social institutions" (Moody-Adams 1999, 169). Here, it appears that practical moral progress relies on moral progress in theory as improved behavior and institutions are said to realize a previously deepened moral understanding. But since Moody-Adams also allows moral progress in practices to precede moral progress in beliefs (see § 4.2.5.), she clearly acknowledges both changes in theory and in practices as types of moral progress in their own right.

In many cases, it is not quite clear whether an author want to allow for merely theoretical improvement to be counted as moral progress or not. For instance, Allen Buchanan and Russell Powell engage with Moody-Adams's view when delineating their own account and they accommodate her view to some extent. They declare that "we agree that improvement in moral concepts is one important kind of moral progress" (Buchanan and Powell 2018, 61); however, their own account does not easily allow for this agreement. On their view, 'moral progress'—in the fullest sense—refers to "changes that either involve improvements in moral capacities or come about through the exercise of those capacities." (Buchanan and Powell 2018, 46) "Moral capacities" or "moral powers" (Buchanan and Powell 2018, 51) are moral agents' "abilities to make moral judgments, engage in moral reasoning, employ moral concepts, and experience moral emotions such as sympathy and indignation at injustice" (Buchanan and Powell 2018, 28). These are items which fall squarely into the theory-domain of morality as described earlier. So, their account of moral progress involves the idea that in its strongest sense, moral progress involves the theoretical side of morality. The "changes" of which it says that they must "involve" or "come about through the exercise of [theoretical moral] capacities" are, however, practical changes: changes in states of affairs brought about by moral agents' practices. Now, the weaker senses of 'moral progress' are not those in which the theoretical side of morality remains involved, effecting lesser and lesser change "on the ground." On the contrary, the weaker senses of moral progress refer to changes that are less and less connected to or reliant on the theoretical side of morality:

moral progress in the most full-bodied sense is not simply change that is desirable from a moral point of view but also must involve the exercise of or improvements in the moral powers. The second and weaker understanding allows changes that are improvements from a moral point of view to count as moral progress even if they came about through self-interested, prudential, or other non-moral motivations […] The third and weakest understanding of moral progress would equate it with changes that are desirable from a moral point of view, without requiring that any human motivational capacities be involved. (Buchanan and Powell 2018, 51)

What disappears from the scene as the senses of 'moral progress' are weakened is the theoretical side, not the practical side of morality. However, in agreeing with Moody-Adams in accepting "improvement in moral concepts" as moral progress, Buchanan and Powell affirm a view of moral progress that does not require the change in question to have any practical effects. They seem even more determined to make room for moral progress in terms of mere change on the theoretical side of morality when they say that

[b]y coming to have [a better understanding of why people, all people, have rights], humans have improved their ability to conceptualize morality and to reason more skillfully about some of its most important features. That is moral progress. (Buchanan and Powell 2018, 297)

Also, when they discuss recent proposals for understanding the concept of moral progress, they particularly criticize those accounts that identify moral progress with increased compliance with a presently identifiable set of valid moral norms for being unable to account for moral progress in terms of theoretical improvements. With respect to the work of Macklin (who proposes the material criteria of "humaneness" and "humanity" for characterizing moral progress, see § 2.3.) and Singer (see § 3.1.), they say that

[their] views are reductionist in that they hold that moral progress consists wholly in increased compliance with some valid norm or set of valid norms. Improvements in moral reasoning, motivation, moral concepts, or conceptions of morality itself are all said to be progressive only insofar as they contribute to better compliance with the identified substantive norm or norms. Such views are of determinate fixed content because they hold that all valid moral norms […] can be identified at the present time and will remain valid in all future social contexts. (Buchanan and Powell 2018, 71, emphasis in original)

Part of the problem they identify with views such as Macklin's and Singer's is the presumably preposterous idea that the relevant moral norms are presently identifiable (and are in fact identified by those authors). But the other, and indeed the first part of the problem lies in those views' commitment to a strong requirement of practical change. They are said to be (objectionably) reductionist in that they reduce theoretical improvements to being instrumentally valuable for practical change. Buchanan and Powell find views like Singer's and Macklin's "problematic for several reasons" (Buchanan and Powell 2018, 71), one of which simply is this reductionism. They elaborate on this critique in the following way:

Some moral improvements in motivation appear to be morally progressive independently of whether they contribute to better compliance with valid moral norms. For example, suppose that people become more strongly motivated to combat certain injustices because they perceive them to be injustices but that, due to some contingency beyond their control, their efforts are frustrated and no progress is made toward remedying the injustices. Presumably this improvement in motivation would constitute moral progress even though it did not result in better compliance with the relevant norms of justice: individuals who developed this motivation would be morally better people, and people becoming morally better is in itself a form of moral progress. So, while it is true that better moral motivations may increase the chances of norm compliance when institutional contexts change, it is mistaken to characterize improvements in moral motivation as moral progress solely by virtue of this contingent, instrumental connection. (Buchanan and Powell 2018, 74)

The decisive factor for affirming that there is moral progress in this case remains somewhat unclear. The explicit answer to this question seems to be: "improvement in motivation." However, it is not clear whether this motivational improvement is to count as moral progress only because it results in an overriding motivation which is practically effective in leading the moral agents to engage in efforts to combat injustice. In the end, Buchanan and Powell only speak, in a general way, about "improvements in motivation." This would allow an alternative reading according to which we should grant that the same moral agents had made (smaller) moral progress if they had only been slightly more but not overridingly motivated to combat injustices, thus failing to make any actual efforts (instead of simply "failing" to make efforts successful). Given Buchanan and Powell's readiness to count improvements in moral concepts as moral progress, it seems that they should even grant that the moral agents would have made moral progress if they had simply gotten better at "perceiving" injustices, i.e., in the application of their moral concept injustice, even if this had failed to make a positive impact on their motivation. This is at least a conceivable option: while, in principle, we would expect moral agents who had improved their ability to apply the concept injustice to also undergo an according change in motivation (because they should be motivated to act in accordance with their new judgments about the existence of injustice), this need not be so. Perhaps upon "perceiving" new instances of injustice, moral agents start to feel overwhelmed and fall victim to weakness of will with respect to acting on their injustice-judgments. In this case, we would be faced with a theoretical improvement that did not entail a motivational improvement. How should we evaluate this case based on Buchanan and Powell's reasoning? If we are not supposed to evaluate improvements in motivation solely in terms of the "contingent, instrumental connection" to practical outcomes, why should the same not be said of motivationally inert theoretical improvements?

The example is also confusing in another respect. It is supposed to show that there is no "better compliance with valid moral norms" in spite of the improvement in motivation. But this simply does not seem to be the case. By making an effort to "combat injustices," the moral agents of which the example speaks attempt to improve the state of affairs, the problem with which is constituted by other moral agents' moral performance. The situation in which they act is non-ideal in that some moral agents do not comply with the norms of justice. It is because other moral agents do not comply with valid norms and create injustice that the group of moral agents that is salient in the example has to act on a valid moral rule: the rule that is part of a non-ideal theory of justice, which orders them to "combat injustice." Before the salient group of moral agents had improved their concept of injustice, they did not act in accordance with this rule. Through an improvement of their concept and an according improvement in their motivation, they become overridingly motivated to follow the rule; thus, there is "better compliance with a valid moral norm." What they do not achieve is to effect moral progress of this sort in other moral agents—those responsible for the original injustice. Consequently, they fail to effect a change in states of affairs: they fail to remedy injustice. But in the example as it stands, there is a) theoretical improvement that includes b) a motivational improvement, and there is c) an improvement in practices, i.e., in compliance with a norm for non-ideal circumstances ("combat injustice!"). The way in which they evaluate the example suggests that Buchanan and Powell only want to rule out the requirement that moral progress involve d) an improvement with respect to compliance with the rules of justice as given by ideal theory. Alternatively, their decision can be described as excluding e) a requirement of a change in states of affairs (an exclusion of something wholly external to the moral agents who improve their practices). The way the example is presented and discussed suggests that a)—c) are viewed as necessary conditions for moral progress, but this does not seem to reflect Buchanan and Powell's definition of moral progress. Given their other remarks indicating the sufficiency of (non-motivational) theoretical change for moral progress, it is unclear whether and why the cut-off point for requirements for moral progress should really be between c) and d)/e) and not between b) and c) or even a) and b).

In other places, Buchanan and Powell stay closer to their explication of moral progress, which prioritizes practical change, when they suggest that even types of moral progress such as "better moral reasoning" are to be understood not as changes that exhaustively consist in changes in the realm of theory ("improvements in moral capacities"), but practical changes that come with or happen due to changes in the realm of moral theorizing. For instance, when they cite "the transition from a crude medical paternalism to a more nuanced view of the professional obligations of physicians" as an example for better moral reasoning (Buchanan and Powell 2018, 56), they are apparently concerned with improved reasoning that has practical ramifications (in this case: how patients are being treated). It seems that only because of these ramifications does the cited example qualify as an instance of moral progress. Also, when defending the inclusion of improvements with respect to "theoretical conceptions of morality" among the types of moral progress, they stress that even these conceptions can have a "practical effect on human well-being" (Buchanan and Powell 2018, 58). On the other hand, most types of moral progress on their list explicitly name only changes in moral views, not practices. The list includes, inter alia, "Better moral concepts" (Buchanan and Powell 2018, 54), "Better understandings of the virtues" (Buchanan and Powell 2018, 55), "Better moral reasoning" (Buchanan and Powell 2018, 55), "Better understandings of moral standing and moral statuses" (Buchanan and Powell 2018, 57), "Improvements in understandings of the nature of morality" (Buchanan and Powell 2018, 57) and "Better understandings of justice" (Buchanan and Powell 2018, 59). Potentially, the list is to be read as a list of the theorical grounds of improvements in moral practices, rather than of theoretical improvements that constitute moral progress on their own.

As this brief discussion shows, it can be difficult to pinpoint conceptual commitments with respect to the theoretical and the practical side of moral progress. In the next section, we shall work through some cases in order to illustrate and evaluate the alternatives of different theoretical and practical requirements for our account of moral progress. The cases revolve around the evaluation of "moral performance" and are supposed to aid the decision what to recognize as a crucial part of moral performance: moral views, moral action, or both?

A clarification is required regarding the evaluation of moral performance in terms of relations of "betterness." Moral beliefs and moral actions could both be evaluated either in deontic terms—i.e., in terms of being right, wrong or permissible—or in terms of moral quality—i.e., in terms of being good or bad, better or worse. The former way only lends itself to a categorical evaluation, the latter to a gradual evaluation. Deontic terms basically refer to an action's conformity or non-conformity to a valid norm, whereas terms of moral quality usually indicate how an action or underlying state of mind is to be evaluated with regard to its effects or motives. They can also be used to rank actions or express uncertainty about their status (cf. Birnbacher 2013, 44). However, since we are interested here in moral agents' development over time, we can leave aside the problem of deciding whether actions or beliefs are ultimately to be evaluated in terms of their deontic status or their quality. In our depiction of potentially morally progressive developments we can focus on evaluations in terms of quality and take those to cover evaluations that would be issued in deontic terms: saying that a moral agent's performance (her actions and/or beliefs) becomes "better" over time could then be read as covering the assessment that the share of her right actions increases over time (or as conveying some other more comprehensive assessment of her performance which could still crucially rely on registering actions' deontic status).

The idealized picture of moral progress from which the ensuing investigation departs is that of a synchronic improvement of the theoretical and the practical side of morality which succeeds in effecting improvements in states of affairs. This idealized picture is shown in Figure 1. For the sake of simplification, the illustration ignores a possible delay in effects.

Fig. 1: Idealized moral progress with congruent change in moral practices, moral views and states of affairs.

The rest of this chapter is devoted to settling the question of which deviations from this idealized picture are compatible with categorizing a process as a case of moral progress.

#### 4.2.4 Moral Progress in Theory?

The first idea to be evaluated is that for the concept of moral progress to be applicable, there is only a requirement of ethical progress (in the broad sense), i.e., that the moral agent needs to improve with respect to her moral beliefs, intuitions or other items in the theory-domain of morality, but she need not change her practices. The idea is that there is no requirement of practical change—no practical requirement—for moral progress to occur.

The moral phenotype with regard to whom we could evaluate the plausibilty of this proposal is a type of unwilling wrongdoer. An unwilling wrongdoer is someone who is convinced that φ-ing is wrong and who has an according desire not to φ but who cannot actually bring herself to refrain from φ-ing. A stock example of this type of moral agent is a character whom we could call the Unwilling Meat-Eater, a moral agent who believes that consuming meat (or animal products more generally) is wrong but who cannot abandon the habit of eating meat. The conflict between stated moral convictions and behavior when it comes to the consumption of nonhuman animals has prompted plenty of empirical research into what has been termed the "meat paradox," the conflict between "people's concern for animal welfare [and] their culinary behavior" (Bastian et al. 2012, 247). Philosophically, it has been interpreted as a case in point regarding the paradoxical phenomenon of weakness of will or akrasia—the choice of an option the agent deems bad. In this context, Elisa Aaltola speaks of the "omnivore's akrasia" as "a state within which one voluntarily consumes products which one deems to have been produced with immoral means." (Aaltola 2015, 35) The basic phenomenon, though, appears in many other forms and it need not revolve around action against a moral judgment, but can be concerned with acting against a prudential judgment, such as the choice of unhealthy food options or of a less demanding, but less rewarding career path. In the moral case, the Unwilling Meat-Eater has lately received the most attention. The term "unwilling" is freely adapted from Harry Frankfurt's distinction between the willing and the unwilling addict, i.e., addicts with and without a desire opposing their addiction (Frankfurt 1971). In a deviation from Frankfurt's construal of the case of addicts, however, the Unwilling Meat-Eater should be imagined as being morally responsible for her actions. The case of the Unwilling Meat-Eater is construed as involving a person who has a desire that opposes the consumption of animals—a desire that is not acted upon on any given occasion. In order to be relevant for the question at hand—which is whether making theoretical progress in the realm of moral views might suffice for achieving moral progress—our unwilling wrongdoer cannot merely be any kind of unwilling wrongdoer but must be someone who has undergone a change in her moral views, i.e., someone who has not always been but has rather become unwilling.

Unwilling Meat-Eater: Over the course of several years, a moral agent has first acquired an uneasiness with the consumption of animals and eventually formed the belief that eating meat is actually morally wrong. Yet, she cannot bring herself to act in accordance with her judgment. She regularly purchases or accepts animal-flesh based products offered to her and consumes them. She has acquired a belief that what she is doing is wrong and a desire to refrain from doing it. But her desire to refrain from eating meat will remain ineffective throughout her life.

The Unwilling Meat-Eater could provoke several reactions: first, the assessment that she has made moral progress (implying that there can be practically irrelevant moral progress) or second, the assessment that whatever change for the better the Unwilling Meat-Eater has undergone, it does not amount to moral progress due to the lack of corresponding action. The reaction that the example is supposed to provoke is the latter. What those experiencing the desired intuition might find disturbing in the case of the Unwilling Meat-Eater is the practical gap that is created when she becomes unwilling. She acquires the belief that she ought not to eat meat and she acquires an according desire, yet it fails to change her ways. The change in her moral views—her ethical progress—is ineffective, leaving a gap between her moral beliefs and practices. Her moral development remains disintegrated.

Fig. 2: Ineffective ethical progress creating a practical gap, i. e., a gap between theory and practice, with practices remaining on a lower level of moral quality.

It is not merely a concern about congruence between theory and practice and, therefore, about the form of the development that is relevant here. The objection to calling this type of change moral progress is not an aesthetic objection based on the perception of disorder. It rather turns on our idea of the function of the theoretical side of morality: the idea that the theoretical part of morality is supposed to guide, regulate, and motivate action. The theoretical accomplishment of the Unwilling Meat-Eater falls short of meeting this objective. The practical gap isamorally repulsive gap—its existence means that the theoretical change that has occurred fails to realize its instrumental value.

The idea that morality ought to be practically relevant is shared by different normative theories, although it takes on different forms in each. From a consequentialist point of view, the practical gap is objectionable because it hinders the realization of good outcomes; from a Kantian perspective, it amounts to a failure to act from duty and, therefore, a failure to exercise moral agency; from a virtue ethical perspective, it can be described as the manifestation of a vice; for a pragmatist, it means that the moral insight is dysfunctional.

Still, one could hold that there is, after all, a difference between being true or being motivating (cf. Parfit 1984, 23) that can be recognized from an ethical standpoint, and maybe a certain value accrues to the recognition of the truth of a moral view even when it is not acted upon. In contrast, it might as well be claimed that in this context, knowledge does not qualify as an intrinsic value, at least not as one that can outweigh the instrumental disvalue of the knowledge being ineffective knowledge. Godlovitch, for instance, claims that if we want our past and current moral failings to be excused from a more progressed standpoint, it is "[b]etter to plead ignorance than incompetence" (Godlovitch 1998, 285). By embarking on a similar detour as Rønnow-Rasmussen does regarding the value of "active involvement" in one's moral progress, we could ask: "What would be better: a world in which evil is done unknowingly or one in which it is done knowingly?" Without taking a detour via the moral evaluation of "the world," we can ask directly: "Who is better, the moral agent who acts wrongly unknowingly or the one who does so knowingly?" Admittedly, this is not the same as asking whether the agent who has become not simply aware but also unwilling is better than the morally unaware agent. But in our case, in which we stipulated that the unwillingness-constituting desire remained completely practically ineffective, the difference is as small as it can possibly be.

The idea that merely theoretical improvements should not be recognized as moral progress is supported by the fact that such merely theoretical changes can be singled out with great terminological ease by drawing on the wide notion of ethical progress we established earlier. The practical function of moral beliefs creates a presumption in favor of reserving the term 'moral progress' for practically effective change. There is no need to overturn this presumption by relying on the concept of moral progress for appreciating the intellectual development of the moral agent who does not put her moral insight into practice. We can describe her as "ethically progressed" instead of describing her as having made moral progress. For those who find our notion of ethical progress to be too wide after all and would prefer to reserve this term for theoretical improvement that satisfied some stronger version of the agency-requirement discussed earlier, there would still be the option of saying that the moral agent has "progressed in her moral views," but "has not made moral progress." If the pragmatic need for recognition of the improvement that has in fact occurred can be satisfied in a reasonable way without invoking the notion of moral progress, we should make the worthwhile distinction between practically effective and practically ineffective change by reserving the term 'moral progress' for change in the realm of morality that involves practical change.

A commitment to the idea that morality (metaphorically speaking) seeks to be practically relevant suggests that the practical gap is a sufficiently severe flaw to justify withholding the attribution of moral progress to the ethically progressed moral agent. The problem is that this seems to be a trivial result. It seems we would not have needed the preceding considerations to simply decide at this point that because moral insights ought to be acted upon, those who merely have more moral insights, but do not act on them, do not count as morally progressed. Of course, if we are committed to the significance of morality's practical relevance, then we will not accept mere ethical progress as progress.

But the matter is not so trivial after all. By considering two more cases of apparently merely ethical change, I hope to show that the commitment to practical relevance still ought to be qualified. Ultimately, I propose to give several different explications of moral progress based on different conceptions of practical relevance. The initial bifurcation of concepts will be proposed at the end of this subsection. First, I attempt to provide some support for the idea that in some cases, the need for a change in practices can in fact be denied. To this end, I return to the case of the converted racist we encountered in the previous subsection.

Recall Rønnow-Rasmussen's description of Geert, whose brain had been "tinkered with […] to the effect that he now sincerely renounces his racist views and actions from the past." (Rønnow-Rasmussen 2017, 147) Because we are still interested in whether theoretical change alone can ever suffice for moral progress, we must amend Geert's case by eliminating the reference to racist "actions."

Conversion of the Strategically Silent Racist: Imagine a non-racist society in which stereotypes concerning people with certain physiological characteristics, the disadvantageous treatment of members of certain ethnic groups and the like have become entirely socially unacceptable. A single racist in this society has good reason to believe that there is no "silent majority" secretly harboring such stereotypes or desires to treat members of certain groups disadvantageously. She, however, holds these stereotypes and desires. She believes that she belongs to a biologically real race that is superior to others, that she is therefore entitled to preferential treatment by public institutions and private citizens, and that she has a moral claim to treating members of other races in ways that would be disadvantageous to them. But her private convictions are highly stigmatized, and she is aware of that fact. So, she remains a Strategically Silent Racist: engaging a lot of self-control, she rigorously refrains from displaying her racist attitudes, she does not harm members of those groups she secretly despises, and she engages in multiple cooperative schemes with them out of self-serving considerations. For instance, she is friendly with her neighbor, who, according to her racist beliefs, is actually "beneath her," because she knows that friendly relationships with neighbors make life easier in many situations. Above all, she does not want to be outed as a racist. Where self-serving considerations do not apply—in singular, non-public encounters with members of the despised group—she still treats the people she looks down on as she would treat anyone else out of respect for the law, which tells her that all people are created equal and that she must not discriminate.

Subsequently, the Strategically Silent Racist is converted. Her conversion might come about in a science fiction-like scenario as in the case of Geert, through a moral epiphany or even through moral debate: for instance, absent any real debate about racism in her non-racist society, our Silent Racist could be converted by confronting racism in fiction. Maybe she volunteers to play a racist character in an amateur dramatic performance and comes to reconsider her position based on the criticism her character faces in the play or through engagement with the views of other cast members. Either way, she comes to renounce her former racist beliefs and attitudes, seriously, though silently. She regrets them, fully internalizes the non-racist socially acceptable point of view that she merely faked before. Her conversion does not affect her practices, because she already treated people friendly, professionally, and fairly, when she still secretly despised them. Her friendship with her neighbor does not grow any deeper, because she thinks that while friendly relationships with neighbors are good to have, it is not wise to let the relationship become too close. So, the former Strategically Silent Racist's change of heart remains entirely inconsequential.

The case of the Converted Strategically Silent Racist is crucially different from the Unwilling Meat-Eater's case: in both cases, only theory changes, but in the Racist's case, no practices could have possibly changed accordingly. So, should we not credit the Converted Racist with having made moral progress because in her case, it was only ever her beliefs, not her actions, that were in the focus of our (negative) moral evaluation? The Silent Racist's practices have always been "good" before: at the very least, they were always in conformity with what we shall assume are valid moral norms; what changed was the moral agent's mindset, transforming "mere conformity" into "genuine compliance […] where this means conformity to valid moral norms for the right reasons" (Buchanan and Powell 2018, 68).Whereas the Unwilling Meat-Eater creates a practical gap, the Silent Racist closes her ethical gap: she attunes her moral views to her already morally right actions; see Figure 3.

Fig. 3: The closing of an ethical gap, i. e., the overcoming of a divergence of theory and practice through the elevation of moral views to the higher moral level of practices.

Should we not view an overcoming of a divergence of theory and practice through the elevation of moral views to the already higher moral level of practices as a way in which people can become "morally better?" When Buchanan and Powell say that "people becoming morally better is in itself a form of moral progress" (Buchanan and Powell 2018, 74), this seems to be a highly agreeable statement. What does the case of the socially compliant Silent Racist tell us about ways in which people can become morally better in the relevant way? If the agent's entire potential for improvement lies in the realm of her moral views, not practices, and if she manages to make or happens to undergo this improvement, so that as a result, both her moral views and her practices ought to be evaluated favorably, should we not really recognize this development as moral progress?

In contrast, one could insist that since no practical change occurred, we could perfectly well honor the change undergone by the Strategically Silent Racist by recognizing it as a form of ethical—but not moral—progress. Certainly, her ethical progress was ineffective in a different way than the Unwilling Meat-Eater's theoretical change because there was no good effect on practices to be called for to begin with; nevertheless, it was ineffective. If we care about moral progress due to its practical ramifications, then we cannot grant the Converted Silent Racist the status of a morally progressed agent—notwithstanding her appreciable ethical progress. If the Silent Racist's conversion does not compel us to retract our commitment to practical relevance, it might at least move us to qualify it—or rather, to clarify it. To this end, we need to contrast the case of the Strategically Silent Racist with her historical and, in some sense, ideological counterpart: the Inversely Akratic Racist.

"Inverse akrasia" is the name that has been given to the "phenomenon in which an agent does the right thing, but does so against her best judgment" (Arpaly 2002, 227). An inversely "akratic course of action is superior to the course of action recommended by the agent's best judgment" (Arpaly and Schroeder 1999, 162). In a way, the "inversely akratic" agent is merely akratic: she chooses the worse course of action, by her point of view. But the name "inverse akrasia" has been given to this phenomenon "[b]ecause these cases reverse our usual expectations from akratic action" (Arpaly and Schroeder 1999, 162). Descriptions of common akratic acts typically trade on our agreement with the agent that the chosen course of action is inferior: choosing the unhealthy food option, taking the line of least resistance, buying a luxury item instead of donating money to a good cause—these are cases we find exemplify choosing the worse option. This is different in inversely akratic acts, which are described in a way that makes the chosen course of action seem better to us even if it contradicts the agent's own best judgment. Accordingly, in cases of inverse akrasia, the "failure" of the agent to act on her conscious evaluation of her options is attributed not to self-serving reasons, but to a responsiveness to the right reasons. The stock example in the debate is the case of Mark Twain's Huckleberry Finn, discussed originally by Jonathan Bennett, who presents it as an example of a clash between "sympathy" and "bad morality" (Bennett 1974). In the context of the debate about inverse akrasia, we find the following description of the case:

As Huckleberry becomes the friend of Jim, a runaway slave, his conventional southern moral convictions tell him clearly that he should proceed to return the slave to his lawful owner. He knows, so he believes, what the right thing to do is. To his embarrassment, however, Huckleberry finds himself psychologically incapable of doing what he believes to be the right thing. When an opportunity comes to turn Jim in, he feels too sick at heart, and displays what he takes to be weakness of will. He just cannot do it. Eventually he completely gives up the idea of turning Jim in, and consequently decides that he is a weak, bad boy, and that being moral is far too hard and thankless a task. (Arpaly and Schroeder 1999, 162)

Huckleberry is in some ways the counterpart to our Strategically Silent Racist: he is a racist in terms of the doctrine he officially endorses, but on some level, he obviously harborsadissenting non-racist "conviction." Different from the Strategically Silent Racist, though, he acts on his private—even if subconscious conviction rather than on the official doctrine or on considerations of what is socially acceptable. Precisely because of these differences, it is worth contrasting the theoretical development an "akratic official racist" like him could make with the Conversion of the Strategically Silent Racist:

Enlightenment of the Inversely Akratic Racist: The Inversely Akratic Racist lives in a racist society whose openly racist beliefs and policies she consciously endorses. However, on any given occasion, she finds herself psychologically incapable of participating in racially discriminatory practices herself. In her encounters with members of the oppressed group, she cannot bring herself to treat them any differently from members of her own ethnic group. Because of this inability, she sometimes violates legal codes.

One day, the Inversely Akratic Racist has a moral epiphany, enlightening her about the wrongness of her society's official racist doctrine. She now consciously rejects her former beliefs, but her behavior toward others is not changed by her moral insight. She does not become a political activist, for instance. She only, as before, personally treats members of the oppressed group as equal citizens on any given occasion.

Comparing the two cases, what should we say about the development each agent makes? We could say, for instance, that both made ethical progress, that the Silent Racist made more ethical progress than the Akratic Racist, but that neither made moral progress. Why would we be inclined to say that the Silent Racist made the greater ethical progress? Because she had to revise a more pronounced racist mindset? This is one plausible description that saves the idea that no moral progress need be ascribed to either agent. But the point of invoking both cases is to show that sometimes it makes sense to talk of moral progress where there is no practical change. The idea turns on the effects of changes in moral views not on how moral agents do act, but on how they would act given a change in circumstances.

If, prior to her conversion, the Strategically Silent Racist were in the position of the Akratic—a society like the one depicted by Mark Twain—i.e., if ordinary morality and the law had been in accordance with her own deeply held beliefs and attitudes, she would not have been a silent but an outspoken, practicing racist. In this counterfactual setting, she would have behaved in racist ways with impunity. If we take the circumstances of each character's action into consideration, we can see that there is more that has changed for the Converted Silent Racist than for the Enlightened Inverse Akratic. The Silent Racist's good practices depended on circumstances that were extremely conducive to the occurrence of these good practices. The Inverse Akratic managed to act well under extremely adverse, compromising circumstances. And the conditions in the former case would not have had to be dramatically different for the Silent Racist to start acting on her actual moral beliefs—they would not have to have closely resembled the world of Huckleberry Finn. Under slightly different conditions, her racism would have been evident: if public opinion had not been as uniform, if there had been legal loopholes for discrimination, etc. If and insofar as these considerations move us toward granting that the Converted Silent Racist has indeed made more than merely ethical progress, if we are willing to affirm that she has also made moral progress in some sense, it is because of how she would have acted given different circumstances. Unlike the case involving inverse akrasia, the Strategically Silent Racist's conversion involved the reversal of the agent's dispositions.

If behavioral dispositions matter, what does that say about the respective significance of changes on the theoretical and the practical side of morality? Behavioral dispositions do not fit into this distinction neatly. Certainly, moral beliefs are relevant to—are co-constitutive of—dispositions, but dispositions are not purely theory-like entities. On the other hand, a disposition to act in a certain way is not yet a practice. The discussion of cases so far suggests dispositional change is relevant to moral progress. It suggests that the less impact any "mere" theoretical change has on the dispositional constitution of the agent, i.e., the more it remains "purely theoretical," the less it counts toward moral progress. Our reservation against moral progress occurring solely "in theory" still stands.

The three cases of the Unwilling Meat-Eater, the Conversion of the Strategically Silent Racist and the Enlightened Inversely Akratic Racist support the reservation against the idea of "moral progress in theory" and illustrate what a commitment to the practical relevance of morality and thus to moral progress can entail. The cases reveal differences between several developments in moral views that apparently do not become practically relevant. These differences are grounded in the dependency of practical relevance on outer circumstances. The practical inertness of a change in moral views was, in the case of the Strategically Silent Racist, crucially dependent on contingent outer circumstances. Her initial (bad) moral outlook was prevented from having an effect on her practices by the societal conditions in which she found herself. Given these contingent initial circumstances, the ethical progress she made could not effect an improvement in practices (which were already conforming to moral requirements). This case was different from the closing of the ethical gap in the Inversely Akratic Racist. Whereas for the Strategically Silent Racist, the change in views increased the functional value of her overall moral outlook in that it worked toward a protection for less conducive circumstances, the Inverse Akratic had already proven to be prepared for adverse circumstances, as she lived in such circumstances and nevertheless performed well morally. Clearly, then, the theoretical change in the two cases must be said to have different practical relevance, even if neither has practical effects.

The practical relevance that we now seek is the practical relevance of dispositions. The preceding considerations suggest that if we are generally interested in moral progress involving practically relevant change, but if we also want to be able to distinguish change that is practically ineffective due to contingent external factors from change that is ineffective due to reasons inherent to the moral agent, it seems that we must distinguish a concept of moral progress in terms of actual moral progress from a concept of dispositional moral progress.

Dispositional moral progress <sup>1</sup> Moral progress disp is the improvement in a moral agent's moral views that increases her disposition to perform morally well.

This first attempt at a definition of dispositional moral progress renders the concept too wide, however, because it includes the case we had already decided to exclude from any concept of moral progress; certainly, the Unwilling Meat-Eater underwent a change in her moral dispositions as well. Remember that we called her "unwilling" to indicate that she had in fact formed a desire not to eat animal flesh—a desire that opposed her habit. Clearly, by integrating a habit-opposing desire, her motivation also changed. If we accept any dispositional improvement as moral progress, are we not forced to reconceptualize our Unwilling Meat-Eater as an "Almost-Vegan?" If circumstances changed, there would come a point at which her desire not to eat meat would gain the upper hand and she would indeed quit consuming flesh or animal products generally. That point may come when a whole range of animal-free products that perfectly resemble corresponding animal products will become just as readily available as animal products are so far. If we still do not want to count the Unwilling Meat-Eater as morally progressed, this suggests that we need to establish a threshold for sufficient dispositional change:

Dispositional moral progress <sup>2</sup> Moral progress disp is the improvement in the moral agent's moral views that significantly increases her disposition to perform well morally.

Note that this characterization of dispositional moral progress does not require the agent to perform perfectly as a result of her change (though it may require perfect performance in a subset of potential circumstances in which compliance with a newly accepted norm is extremely easy).

The reason we would not grant the Unwilling Meat-Eater the status of a morally (and not merely ethically) progressed agent is not necessarily that she has not yet become a vegetarian; it could merely be that she never acts on her acquired moral belief about the wrongness of meat-eating in actual circumstances. What counts as performing well in varying circumstances, and therefore what counts as a significant change with respect to good performance (thus, the very distinction between change that is practically ineffective due to contingent external factors and change that is ineffective due to reasons internal to the moral agent), depends on moral assessments of reasonability and demandingness. It is because we deem the requirement of refraining from eating meat reasonable in the moral agent's circumstances as they are that we do not recognize her dispositional change as significant. The change she has made does not have any impact on performance in actual circumstances in which better performance can be normatively expected (at least occasionally) not solely of morally perfect agents but of anyone who is to be counted as minimally morally—and not merely ethically—progressed with respect to the issue at hand. "Significant" change toward better performance should be interpreted as requiring the reaching of a baseline and allowing for gradual improvement beyond that baseline: after having come to comply with the relevant norm at least occasionally, any further increase in compliance should probably be seen as further progress. The motivating thought for introducing a threshold of "significant" dispositional improvement was that our Unwilling Meat-Eater was to be evaluated differently than the Converted Racist, who revises her dispositions with the result that she could now act in non-racist ways in circumstances much less conducive to nonracist action. In contrast, the Unwilling Meat-Eater would only be able to act in accordance with her moral belief in circumstances that became much more conducive to avoiding meat. If those more advantageous conditions came about, she might even attain perfect conformity to the moral requirement of abstaining from meat-eating. But since she never conforms to the relevant moral norm in actual circumstances, she has not only made significantly less dispositional moral progress than the Converted Racist, the dispositional change that she has undergone does not reach the threshold necessary for being counted as moral progress at all. The idea underlying this proposed assessment of the cases is that for an apt attribution of moral progress to the moral agent, the change the agent herself undergoes must make a significant contribution to the realization of better moral performance vis-à-vis a variation in outer circumstances.

The present suggestion is that the concept of dispositional moral progress should be reserved for changes in moral views that reach a certain baseline of potential practical relevance, where this baseline is established based on profoundly normative considerations of what is to be expected of moral agents in different actual and counterfactual circumstances.

Should we also factor in the likelihood of the counterfactual circumstances? Possibly, the Converted Racist's progress would not have counted as moral progress if it had occurred in a world that had not only overcome racism but also represented a case in which racism's revival seemed to be overwhelmingly unlikely. In that world, the risk that the Silent Racist would have ever become an outspoken, active racist would have been vanishingly small, i.e., it would have been a "merely theoretical" option. The elimination of this insignificant option by a change in the Racist's moral views would therefore itself be a merely theoretical enterprise.

This consideration speaks to our nevertheless practical interest even in merely dispositional moral progress. To accommodate it, we need to add a requirement of probability regarding the circumstances under which the moral agent's relevant disposition would be actualized:

Dispositional moral progress <sup>3</sup> Moral progress disp is an improvement in a moral agent's moral views that significantly increases her disposition to perform morally well in sufficiently probable circumstances.

Establishing how significant the moral agent's change would have to be and how probable the relevant circumstances need to be is the task of a more comprehensive theory of moral progress that draws on many more normative presuppositions than have a place in an explication of moral progress. While the further specifications of dispositional moral progress transcend the present project, an explication of moral progress surely must account for progress that actually shows in the practices of moral agents. In defining actual moral progress, we need to include a condition mentioning the temporal expanse of the change in question in order to incorporate the requirement of minimal continuance for progress (§ 4.1.). Whereas dispositional moral progress must be possible even when the disposition is only actualized in an isolated instance, actual moral progress cannot come down to a singular good act.


ment in the moral agent's moral performance, lasting for a certain period of time.

Saying of an agent who has improved morally in terms of her dispositions that she has not morally progressed means that one is relying on the concept of actual moral progress. From the point of view of someone who was only interested in using the concept of moral progress for actual moral progress, what we have conceptualized as dispositional moral progress could be deemed "blameless moral stagnation." An evaluation based on the concept of actual moral progress would deny the moral agent the status ofamorally progressed agent when the circumstances under which she would enact her new moral outlook simply do not obtain. By adding that the moral agent is "blameless" for failing to progress morally, the evaluator could try to mitigate the implication of moral luck in her judgment. Moral luck obtains when moral agents are being praised, blamed, or held accountable for actions or their results when relevant conditions are not under the agent's control (Nagel 1979). We subject moral agents to moral luck if we to some extent suspend the assumption that they cannot be blamed for what happens due to factors beyond their control. Withholding the judgment that an agent has made moral progress when the reason for her failure to actualize better practices is circumstantial—the fact that the occasion for this actualization is lacking—seems to entail subjecting the moral agent to "circumstantial moral luck" (Nagel 1979, Chapter 3). Moral luck is commonly explicated in terms of blaming or praising practices. Blaming the moral agent, however, is simply not what we are doing when we are thinking of her development in terms of "blameless moral stagnation." We are merely withholding recognition. Whatever valuable change she has undergone, when we hold on to a concept of actual progress, we are not recognizing her change as moral progress.

In theory, the concept of dispositional moral progress makes the recognition of the moral agent less dependent on circumstances than the concept of actual moral progress, subjecting them to moral luck less frequently. Because it selects a different condition for moral progress to occur (an improvement in dispositions rather than in actual practices), it could, in principle, feature in moral progressjudgments that could not be issued based on the concept of actual moral progress. In theory, the dispositional concept allows to attribute moral progress to agents who simply do not encounter the right conditions to activate their advanced moral beliefs—though the question remains based on what beliefs such judgments could be made. We could say that such agents live in "moral exile": a situation in which a moral agent subscribes to a system of beliefs in moral requirements that circumstances keep from being activated. A moral exile is constituted by circumstances that are unfavorable to moral action in the sense that they render actions impossible that would otherwise be obligatory and that the moral agent would be willing to perform. With regard to most moral belief systems, a complete moral exile would be a far-out hypothetical. A moral agent who believed in a sentientist ethics and who lived on an island that was uninhabited by any sentient being and surrounded by likewise inhabited waters, cut off from communication with any other moral agent, would be in complete moral exile; her moral beliefs concerning duties to other sentient beings would not be activated, absent any sentient beings in her environment. In more mundane situations, moral agents live in moral exile with respect to those of her moral views that are systematically not activated due to the lack of relevant circumstances. In moral exile, a moral agent could acquire a lot of correct moral beliefs about what she is and is not momentarily required to do (and act in accordance with them), but also acquire many correct views about (and be in principle ready to act in accordance with) moral norms that would be activated only in different circumstances. When these circumstances come into being, they activate the formerly inert moral beliefs: they (practically) allow the agent to act on these beliefs—thereby (normatively) demanding her to act on them. Upon encountering these new circumstances, the agent who has truly made dispositional moral progress should act on her intermediately acquired beliefs.

Let us consider the case in which, once circumstances change, the agent fails to act on the beliefs she has acquired in moral exile. In these circumstances, a moral agent could find acting morally well to be exhausting and—contrary to her prior disposition—could fail to comply or gradually decrease her compliance with the newly activated moral requirements. To illustrate the moral expatriate and the potential moral progress she makes or fails to make, consider the following case:

Overwhelmed Grandmother: A moral agent might come to believe that if she had grandchildren, she should engage in some cultural activity with them every other weekend. The moral agent is at an age where she might be a grandmother, and as she acquires the belief about what she should do as a grandmother, she becomes fully prepared to act in accordance with that belief. Absent any grandchildren, the moral agent lives in moral exile with respect to her grandparental duties.

A few years later, her son has a daughter.When the girl is at an age when she can partake in the envisioned cultural activities, the grandmother still believes she should take her out to museums and such every other weekend. However, she finds herself burdened by the task and, after only a few museum visits, fails to continue the planned cultural activities with her grandchild.

If the case description did not assure us that the moral agent had previously been fully prepared to meet the then-hypothetical obligations at the time she adopted the belief in them, we might be tempted to judge that she had only ever made merely ethical progress in her moral exile. This assessment is the more likely the sooner her compliance with the newly activated moral norms declined because the more isolated initial acts of compliance with the newly activated moral requirements remain, the more they appear to be instances of acting "out of character" instead of activations of previously acquired dispositions that are simply worn down over time. It will merely seem as if the moral agent had not acquired the right kind of dispositions to begin with. Such an evaluation would still operate on a concept of dispositional moral progress.

Even though the dispositional account is, in terms of the condition for moral progress it states, suited in principle to acknowledging the moral change undergone in moral exile as moral progress, it faces a diagnostic problem: behavioral dispositions are detected via their (repeated) actualization. As a criterion for attributing moral progress, it will usually use the same criterion as the account of actual moral progress: practices. When the right circumstances for the actualization of the disposition do not obtain, the dispositional account can only be made to produce a moral progress-judgment that is correct by its own lights in thought experiments in which dispositional improvement is merely stipulated. This shows that moral progress-judgments that are issued in the face of actual practical improvements are ambiguous—they might express recognition of the actual observable improvements or the underlying dispositional change which enables it.

The fact that different case descriptions seem to activate different concepts of moral progress, in conjunction with the fact that we detect moral progress by evaluating acts and omissions over time, also means that we can miss moral progress if it occurs in a seldom activated realm. This is a final complication for a requirement of practical relevance for moral progress to be considered here. To generate an example of a rare moral challenge, we can draw on the class of seemingly heroic acts, for instance, cases of risky (in contrast to easy) rescues. While high risk rescues are the types of cases that are quite plausible to occur only seldom for average moral agents, they are at the same time not easily conceptualized as occasions for undergoing profound moral change. Instead, they seem to prey on preexisting moral dispositions. The hero who saves the child from the burning house rarely has time to deliberate her course of action or form a new disposition on the spot prior to running inside the building. Still, this type of case could be made to work as an illustration of the present problem by introducing a delay between the decision to engage in the high-risk rescue and the performance of the rescue. We could, for instance, amend the famous pond case in which the moral agent is faced with the task of performing an easy rescue by pulling a drowning child from a shallow pond (Singer 1972, 231) and transform it into a case of an especially demanding risky rescue that allows for preceding ethical deliberation.

The Daring Rescuer: A person hiking by a remote frozen lake suddenly spots a deer in the middle of the lake that has fallen through the ice and cannot get out of the cold water on her own. By the lakeside is a motorized boat that is big and heavy enough to break the ice and might just remain stable on the water in a potential rescue endeavor. But the hiker is worried that if she went out onto the lake and tried to pull the deer into the boat, she might tumble and end up in the water herself. She calls the fire department but is told that a rescue team will need well over half an hour to reach the lake. Meanwhile, the deer's movements are becoming slower.Waiting for help and faced with the deer's plight, the hiker has time and reason to reconsider her decision to wait for assistance. She deliberates about her moral choice, weighing the risks and potential benefits, reconsidering her beliefs about what can be asked of people in situations like these, eventually changing her belief about what can and cannot be demanded of her. She concludes that she ought to save the deer. Her new moral belief—that a high-risk rescue such as this is required of her—is an improvement over her previous belief that high-risk rescues are never required. She goes out onto the lake and saves the deer, just in time, as it would seem. Never before had she encountered a situation remotely like this, but the experience gives her confidence that she can perform risky rescues in the future. However, she will never face a challenge like this again: rescuing the deer is the only risky rescue she ever has to perform in her life.

The Daring Rescuer, we have stipulated, has made theoretical progress: she has improved her moral belief system by updating her views on duties of assistance.¹⁸ This progress has facilitated better moral action than her previous beliefs had commanded. So, there is certainly a positive practical effect of her theoretical progress. Yet, this positive effect does not seem to warrant a judgment of actual progress, because the "practice" that has been changed was an unprecedent and never iterated kind of action, i.e., the minimal continuance requirement is not satisfied; see Figure 4.

Unlike in the case of the Converted Silent Racist, it is not because her practices have been good all along that the Daring Rescuer's theoretical change did not have the practical effect that would warrant a judgment of actual moral progress. The judgment of actual moral progress is blocked because her "practices" did not exist before and ceased to exist soon after the change in her moral views occurred. The relevant effect was a singular act. By improving her moral views and holding on to them after the rescue at hand is completed, she has moved into moral exile: her newly accepted duties of assistance are never

Note that for the change in her moral beliefs to have been an improvement, it is enough that she has gained the insight she has managed to gain—she need not have perfected her moral outlook. If morality demanded her not only to take high risks, but also to generally invest more resources into assistance to others than she believes she must, her new moral insight would still be an improvement over her previous lack of recognition of either of these supposed moral demands.

Fig. 4: A successful moral deliberation in the face of a rare moral challenge results in an improvement in the moral agent's views, allowing her to perform well in a singular act.

again activated. So, if any judgment of moral progress is apt in this case, it must be a judgment of dispositional moral progress—even though the disposition has in fact once been actualized.

The presumably disturbing fact that this case is supposed to illustrate is that by allowing for mere dispositional change to count as moral progress, we have committed to including any successful moral deliberation that results in a singular right moral act as moral progress. The issue one could take with this result is that it shows our concept of (dispositional) moral progress to be too wide. It cannot be the case that any successful moral decision whatsoever entails moral progress.

In response, we must point out that not any successful moral decision-making process could result in moral progress; it would still have to have a lasting effect on the moral agent's disposition. If moral deliberation where exhausted by the gathering of factual information and applying pre-held moral principles, no relevant change would have occurred. If we think that moral decisions ought to be reached by bringing our normative and empirical beliefs into reflective equilibrium, of course, any factual case-specific information can have an effect on moral dispositions. But given the requirement of significant change in the characterization of dispositional moral progress, we have the conceptual tools to rule out marginal effects of gathering situational information for everyday moral decision tasks as contributing decisively to moral progress.

On the other hand, it would be odd if novel moral challenges with their potential to provoke transformative moral deliberation—moral decision-making that includes a moral reorientation—could not be readily seen as occasions for making moral progress. Right decisions made in the face of novel challenges rather seem like prime examples of moral progress. If our considerations entail that any successful moral decision can be like the mastering of a novel moral challenge in that way, we should simply accept that. This means that moral progress can be made every day, incrementally, through conscientious moral deliberation—even if everyday moral life does not present the occasion for improved action repeatedly.

What is noteworthy about the Daring Rescuer's case is that we should judge her to have made moral progress by relying on the dispositional rather than the actual moral progress-concept, despite the fact that she very well enacted her moral decision. Again, judging a case like this in reality will be a quite different issue. Committing conceptually to the existence of moral progress (by tying it to dispositional change) and determining that there is moral progress in reality are two different things. The criterion for asserting a specific behavioral disposition usually is repeated actualization on occasion. However, in a case like this, the occasion remains a singular one. The fact that the act in question would likely be seen as heroic does not mean that the agent will be given the benefit of the doubt as far as the stability of her readiness to perform risky rescues is concerned. The more outstanding the act, the less justified seems the inference to dispositional change. If it is not only unprecedented in the moral agent's life but also a rare type of act in the moral community, we do not dare to ascribe moral progress even based on the evidence provided by the act itself. This reluctance would reveal a commitment to the concept of dispositional moral progress and the impression that an according moral progress-judgment is underdetermined by the evidence.

Changes in moral views constitute ethical progress. When these changes have significant impact on the moral agent's disposition to act in accordance with her new, improved views, they amount to dispositional progress. By acknowledging this type of progress, we acknowledge the dependence of changes in practices on circumstances that provide occasions for these practices. Actual moral progress—the improvement in the moral agent's actual performance—depends on occasions to perform, the occurrence of which is beyond the agent's control. Acknowledging that an agent can make moral progress in the absence of these occasions for better performance makes sense in order to free the practical relevance we ascribe to the theoretical side of morality of the dependency on contingent factors. We now have to return to the question of whether the theoretical side of morality always has to be involved in moral progress to begin with. Above (§ 4.2.2), the idea was rejected that moral progress always has to be "made" (where this was supposed to convey a requirement of a conscious effort, ethical initiative or retrospective endorsement). The claim was that moral progress should not be viewed as requiring agency in any of these senses. The question of whether moral progress (now to be specified as actual moral progress) at least required a change in practices to be accompanied by a change in moral views was left open. This question—concerning a potential theory-requirement for moral progress—is addressed next.

#### 4.2.5 Progress by a Fluke?

Many of those who engage with the concept of moral progress in ways that seek to be uncommitted to any particular moral theory (including consequentialism) are very clear in requiring moral progress to be practically relevant. As was pointed out before, the naïve conception, for example, bespeaks an interest in moral progress as something that involves change "on the ground": "change in states of affairs" and "the prevalence of right actions" refer to, in reverse order, the practices of moral agents and their impact on the world. We have rejected the idea that changes in states of affairs suffice for moral progress in § 3.1 and problematized the reference to right acts in § 3.2. The previous subsection introduced a version of the concept of moral progress that even replaces references to right actions with a reference to mere disposition. We now return to the original idea of moral progress involving practical change. The question is: if moral progress comes in the form of actual, not merely dispositional change, does it even need to be accompanied by a change in moral views—progress on the theoretical side of morality?

In one obvious way, the possibility of moral progress without improvements in moral views would be an auspicious prospect. Based on her distinction between "moral progress in beliefs" and "moral progress in practices" mentioned earlier, Moody-Adams addresses the fact that the possible dissociation of the two is a reason for hope. She argues that the fact that sometimes prudential, rather than moral purposes drive moral improvement

suggests that it is possible to provide incentives to accept morally progressive practices without first deepening moral understandings. Moreover, when this possibility is realized in practice, the pace of moral progress in practices will sometimes outstrip the pace of moral progress in beliefs. But this is not a reason for moral pessimism. On the contrary —given how easy it is to ignore the moral demands of self-scrutiny—the fact that progress in individual beliefs may be a consequence of prior progress in social practices is the most compelling reason we have to be optimistic about the possibility of moral progress. (Moody-Adams 1999, 183)

In the scenario Moody-Adams describes, practical changes do not appear to be completely independent from changes in moral views. Practical changes outrun theoretical ones in portions of the moral community, but for these behavioral changes to be incentivized it might seem that at least some members of the community must have advanced their moral outlook, viz. those who provide the incentives. However, this is not necessarily so: those incentivizing morally better behavior could themselves do so for entirely self-interested (e. g., economic) reasons. Similarly to Moody-Adams, Buchanan and Powell acknowledge that an improvement in practices might plant the seed of moral progress (Buchanan and Powell 2018, 50 –51) and they ultimately even recognize it as a kind of moral progress in and of itself—even if only in a weak sense (Buchanan and Powell 2018, 52).

What could be registered as a positive about a concept of moral progress that covered practical improvement in the absence of any moral insight—beside the fact that it makes room for incentivizing moral progress—is that it certainly is not guilty of enshrining an overly intellectualistic picture of moral progress. But of course, the concern is that the case is precisely the opposite. Even if the prospect of dispositional moral progress convinces us that ethical progress without practical progress is not necessarily hollow, the question still remains: is practical progress without ethical progress not disoriented? If the moral agent did not need to improve her moral outlook in any way (if she did not need to intensify her grasp, sense, or feeling of anything) in order to count as having made moral progress, would that not mean that moral progress could come about by a fluke?

Let us start the investigation of this question with this extreme suggestion: that moral progress could come about without the moral agent even being minimally morally aware. An improvement in moral performance of this sort would be an improvement in a moral agent's actions where these would be properly regarded as subject to moral evaluations but not subjected to moral evaluation by the moral agent herself. A moral agent's choice of a means of transportation might serve as an example: it is suitable for moral evaluation, but the moral agent herself could base her choice exclusively on considerations of personal comfort and efficiency, leaving moral considerations completely aside. The moral agent might start off by relying on her car as her default means of transportation. However, more traffic and cheaper train tickets let her choose the train over her car on more and more occasions, thus causing her to make the supposedly morally better choice increasingly often over the course of time—not merely without realizing that it is the better option and approving of it in these terms¹⁹

This is what the agency-requirement interpreted as a requirement of endorsement, (4), would have demanded.

but also without recognizing it as a morally good option,²⁰ because she fails to think of her choice in moral terms at all. The Prudent Train Passenger thinks of car rides and train rides exclusively in terms of her schedule, her comfort, and monetary costs. The calculation is a purely prudential one. The moral agent is not thinking about emissions, accident rates or use of space. She makes what appears to be moral progress by a fluke.

Moral progress by a fluke results in inadvertent rightdoing: the moral agent does the right thing, but without being in any way responsive to this fact. The moral agent who progresses by a fluke is not to be thought of as resembling the Inverse Akratic: unlike the latter, she is not responsive to the moral relevance of the facts on any level. Her improvement, then, opens up a steep ethical gap: a divergence between moral practices and moral theory with the theoretical side of morality falling way behind; see Figure 5.

Fig. 5: Moral progress by a fluke, resulting in inadvertent rightdoing and the creation of an ethical gap.

Figure 5, like the others, does not include a scale, because it serves mere illustrative purposes, but the fact that the theory-line is not set at what could be assumed to be "zero" might be taken to indicate that there is some moral awareness with regard to the issue at hand, after all. Instead, the illustration should be viewed as part of the overall evaluation of a moral agent who does have

This is what the agency-requirement in terms of a requirement of a change in moral views, (1), i.e., a theoretical requirement, would demand.

some correct views and is morally aware with respect to some issues but whose improvement with respect to a certain practice is not accompanied by a relevant moral insight. Such an agent might lack moral awareness with respect to the practical issue at hand altogether. Among the correct moral beliefs of such an unaware agent, there may well be some that would in principle apply to the situation at hand. For instance, an agent pondering the choice between taking the train and taking the car could subscribe to the principle "in everything you do, minimize harm to others," but if she fails to see her choice of a means of transportation as a case of application of that—or any other—principle, we should say that she lacks moral awareness and a moral improvement in her ultimate choices would be a fluke.

The conceptual question regarding the explication of moral progress is whether such a fluke should count as an improvement in "moral performance," which is mentioned in the definition of actual moral progress. Does an improvement in moral performance require an improvement in moral views, or could it consist in a mere increase in inadvertent rightdoing?

It was assumed earlier that the evaluation of moral performance over time should be cast in the language of moral quality. However, at this point, it seems apt to shift to the language of deontic classification, because it is required to clearly express that the potential result of moral progress under consideration here is "mere conformity" to valid moral norms. The vocabulary of gradual evaluation of quality could in principle be used to make a similar point—we could talk of "inadvertent betterdoing"—but it would invite confusion, because quality evaluations (of action) might be mistaken for aretaic evaluations (of agents) that turn on agents' motives for action.

The phenomenon of what I have called "inadvertent rightdoing" is usually discussed in terms of "accidental" rightdoing or "accidental rightness" (cf. van Zyl 2009) and this might be assumed to be the more fitting way of framing the issue, because, after all, it is the contingency of the right act that appears to be the source of irritation. But the term "accidental" is unfortunately broad and covers too many ways of failing to do the right thing with the right mindset. In trying to find the baseline of ethical involvement for attributions of moral progress, it is useful to start by considering cases in which there is a complete lack of moral awareness. Accidental rightdoing, however, could occur when agents are, in some sense, morally aware. Accidental rightdoing could be rightdoing with a moral motive, but not the correct one. It would then clearly not be "inadvertent" rightdoing. For example, a moral agent may be wholly moved by anthropocentric concern for the environment to avoid animal products but lack any kind of awareness of the nonhuman suffering connected to the consumption of animals. In reality, we might suspect that moral agents like this Environmentally Concerned Vegan are always, unbeknownst to themselves, somehow 'responsive to the right reasons' (like Huckleberry Finn, the Inverse Akratic, cf. Arpaly 2002). In a thought experiment, we could stipulate that our Environmentally Concerned Vegan is not. Still, unlike the Prudent Train Passenger, she aims to do the right thing and she is still ethically involved: her choices are influenced by moral considerations. She even has the correct sense of the valence of possible courses of actions, though she has a wrong theory about what that is grounded in. Her type of accidental rightdoing might even be considered the one more aptly termed a "fluke"; here, the agent has at least aimed for success—inadvertent rightdoing does not even have the feature of aiming for (moral) success. I will still continue referring to progress that results in inadvertent rightdoing as progress by a "fluke"; whereas accidental rightdoing isalucky strike from the agent's point of view, inadvertent rightdoing could be considered a lucky strike from morality's (or the observer's) point of view. Inadvertent rightdoing can be seen as an extreme case of accidental rightdoing, i.e., completely oblivious rightdoing in which the rightdoer does not only fail to see what makes the act right, but even that it is morally right.

One might object at this point that the portrayal of moral progress by a fluke is already a caricature of moral action, a strawman; surely, what should interest us is not whether completely random improvements could count as moral progress—they cannot—but how much or little intellectual involvement to demand. However, excluding progress by a fluke from the concept of moral progress is not trivial. As debates in normative ethics show, inadvertent rightdoing cannot be dismissed as a form of rightdoing out of hand. Given that we started with the idea that moral progress could be evaluated in terms of the increase of the share and significance of right actions over time (i.e., in terms of the naïve conception's second conjunct which mentions the prevalence of right acts), considering inadvertently right acts as potentially counting toward moral progress is not outright absurd. Also, what is evaluated relying on a concept of moral progress that allowed for flukes would still be moral agents and their actions, and these would still be evaluated in moral terms—the requirement of relatedness to action (§ 2.2.) would still be fulfilled for the resulting judgments of moral progress. The only thing that is missing from cases of inadvertent rightdoing is the moral agent's own moral perspective. This is how the issue turns on the question of proper agential involvement—on our first interpretation of the agency-requirement for moral progress in terms of a requirement of change in moral views (§ 4.2.2.). The question is whether in order for the agent to qualify as making moral progress she must undergo a change in her moral views, i.e., be ethically involved in her moral progress.

Inadvertent rightdoing is a problem as long as we are torn between two modes of evaluating actions: evaluating them with respect to features such as actions' outcomes, that are "external to agency" ("evaluational externalism"), and with respect to features such as the agent's motive, which are "internal to agency" ("evaluational internalism") (Driver 2012, 121). As long as we want both agency-external and agency-internal features to be honored in the evaluation of moral action, we will frequently encounter "evaluative conflict" (Louise 2006) with respect to the two realms of evaluands, because deeds that are good in terms of their effects can be done for bad reasons and vice versa. Different ethical theories typically privilege one of these types of features: virtue ethicists and Kantians focus on internal features; consequentialists and deontologists focus on external features, viz., outcomes and conformity to rules or principles, respectively (van Zyl 2009). The latter therefore allow for the phenomenon of accidental, and, therefore, inadvertent rightdoing: they allow acts which bring about the best consequences or which are in accordance with moral norms to count as right even if done for reasons that are, in one way or another, "wrong." This is a problem just to the extent that we strive for "evaluative consistency" (Louise 2006), i.e., only insofar as we do not want internal aspects of actions to be left out of consideration in moral evaluation. However, it is not trivial to establish that we should not conceive of evaluative conflict as a problem, i.e., that we should not strive for "evaluative consistency" in the first place. On the contrary, different ethical theories' respective ability to each accommodate one of these sorts of intuitions about the relevant features of moral action must be seen as a major factor in the competition among them. It is their respective focus on consequences, motives and character traits, respectively, which earns major strands of ethical theories "plausibility points" (Enoch 2011, 14), and it is their lack of appreciation of the respective other elements of moral evaluation that renders them implausible.

The conflict that arises in the case of inadvertent rightdoing can also be described as a mismatch between the act's deontic classification (as being right) and aretaic evaluations, i.e., the evaluation of the agent and her involvement in bringing about rightness. When the problem of evaluative conflict or the failure of normative theories to rule out accidental rightness is concerned, the focus usually is not on cases in which the deontic classification of an act as right is simply not complemented with a likewise positive evaluation of the agent (in which the agent fares merely neutrally) but on cases of conflict with a particularly bad aretaic evaluation. The cases often revolve around what we might call "tainted" or "vicious rightdoing," i.e., doing the right thing while being aware of its rightness but not because of its rightness or because of concern for the features that make it right, but for a shady or questionable reason, for instance, because one wants to be admired (cf. van Zyl 2009, 92). In order to pursue a course of action for the sake of being admired rationally, one has to have identified it as being right and therefore capable of serving as the grounds of admiration. In this sense, vicious rightdoing is fundamentally different from inadvertent rightdoing. The vicious rightdoer is aware of the value of the action but pursues it for the wrong motivating reasons. It must be noted, therefore, that vicious rightdoing is not the focus here.²¹

So, could moral progress occur by a fluke? I will argue that we should preserve the evaluation that practical change without ethical change is disoriented and we should therefore not count progress by a fluke as moral progress. There is, in other words, a theory-requirement for moral progress.

What may encourage the idea that practical improvements that result in inadvertent rightdoing "deserve" to be acknowledged as moral progress is their potential to have a positive impact on the world. The moral agent who changes her ways for the morally better without knowing it still may, in some small way, make the world a better place. Should that not be enough to acknowledge her improvement as moral progress? In order to give more weight to the factor of positive impact, we can, for once, entertain an example for change on the collective, rather than the individual level. For instance, a singular dietary vegan, who avoids products made from nonhuman animals for entirely prudential reasons, e. g., thinking that a vegan diet is healthiest (cf. Christian 2018), does not have much influence on the morally relevant area of states of affairs related to her dietary choices (the wellbeing or suffering of nonhuman animals). In order to include the factor of positive impact, we can imagine an entire society turning toward dietary veganism, but lacking any moral awareness regarding the new practice whatsoever:

Economic Decline of the Dairy Industry: Over the course of a few months, a virus renders an entire moral community lactose-intolerant. At the same time, naturally lactose-free plant milk becomes much more easily available than cow's milk from which lactose has been extracted. The dairy industry undergoes complete economic decline, eliminating the affliction of suffering on bovines used in the production of cow's milk. This factual development is not accompanied by any insight into the cruelty that had been involved in the perished industry. Human members of the now dairy-free society remain completely unaware of the moral value of their new dietary practice. There is a lasting and complete lack of awareness that the consumption of cows' milk ever was a morally relevant practice.

The proposal that will be made with regard to the requirement of ethical involvement in moral progress will entail that the moral agent who comes to be a vicious rightdoer after being a wrongdoer can, in principle, be thought of as making some kind of moral progress.

This case of a change toward inadvertent rightdoing on a societal level—presumably—involves massively morally desirable outcomes. From a consequentialist perspective, it would be these positive outcomes that would bestow on the changed actions the deontic status of being right; from a deontological perspective they can be seen as right by virtue of conforming to a principle like "do no harm." The positive outcomes and the conformity to an important moral norm are brought about by a change in the actions of moral agents—undertaken for purely prudential reasons, certainly, but undertaken by moral agents nonetheless.

What does the absence or presence of moral insight matter? Focusing on the outcomes of the practical change: what does it matter for a cow that is never born into a life in which her interests are violated that she is spared suffering only thanks to prudential human choices, not moral ones? It is not at all obvious that the change toward the avoidance of suffering, that would otherwise have occurred, should not count as moral progress, even if it came about with no accompanying moral insight in the moral evil that this suffering would have been. If there is, even for contingent reasons, no practical difference between the behavior adopted by someone who comes to avoid dairy products for moral reasons and someone who does so for other—morally contingent—reasons, why not acknowledge both practical outcomes as outcomes of moral progress? The case description is supposed to put moral pressure on the intention to reserve the term 'moral progress' for only one of these improvements.What could the overwhelming pragmatic interest in upholding this conceptual distinction be?

I propose that the answer to this question should be: to distinguish moral progress from the accumulation of morally desirable events. The idea here is that the emerging inadvertently right acts in a case of progress by a fluke are, from the point of view from which they are evaluated as morally desirable, more like events than like actions. They are actions, but they are, in a sense, not moral actions.

The concept of moral progress by a fluke does not obviously establish an entirely different subject matter than that picked out by a more restricted concept of moral progress. However, we can decide to treat change toward inadvertent rightdoing as a different subject matter than moral progress, based on an interest in carving out worthwhile conceptual distinctions. We want to differentiate types of phenomena to be covered and not covered by the concept of moral progress based on significant differences. One significant difference between practical improvements by a fluke and moral progress accompanied by some moral awareness concerns the attributability of the practical change to the moral agent. In an improvement in practices that results in inadvertent rightdoing, moral agents are involved qua agents, but they are not involved qua moral agents. On the contrary, these improvements are morally desirable results of the failure of moral agents to act as moral agents. This failure is constituted by the complete lack of becoming in any degree morally aware. If the result of the action were bad or the action would not happen to conform to a moral norm, the moral agent would be judged as performing badly because of her lack of ethical involvement. In the case in which the result is good or the action is inadvertently right, the agent appears to perform well despite her lack of ethical involvement. The practical changes made by the moral agents are attributable to them in terms of their agency, but they are not attributable to them in terms of moral agency. The lack of moral awareness renders the relevant changes in practices more similar to morally desirable non-agential changes than to changes undergone with some minimal form of moral awareness. Nothing morally noncontingent relates them to their outcomes or underlies the relation of conformity to the relevant moral norm. In this way, improvements of practices lacking any moral awareness are fundamentally different from changes undergone with some moral awareness and share with the results of mere events the feature of being not morally willed, approved or even registered as morally relevant by a moral agent. While observers might register the moral relevance of the change and evaluate it as morally good, there is no responsiveness to the grounds of this evaluation in the ones involved in the change. This fundamental difference between an agent whose improvement is a fluke and an agent who is at least vaguely aware of the valence of her practical change constitutes a plausible ground for a conceptual distinction, i.e., a separation of different types of phenomena. Moral progress belongs on one side of this distinction, while morally desirable practical change belongs on the other side alongside the axiological progress that it entails. In the example of the Decline of the Dairy Industry, the switch to plant milks constitutes a morally desirable practical change, while the resulting decline of suffering in cows constitutes axiological progress (a betterment in states of affairs). The agents involved are not aware of the desirability of either development. Therefore, they should not be said to make or undergo moral progress.

There are alternative ways of supporting the decision for a conceptual distinction not chosen here. In order not to set the bar for the moral agent's intellectual involvement too high, the argument does not invoke the idea that the role of ethical progress is to provide a kind of theoretical warrant for the practical change in moral progress. The idea here is not that for moral progress to occur, the moral agent must have made such sophisticated ethical progress that the change in her views amounts to a justification of the change in her practices. The theoretical change that we require can only be seen as a warrant in a very weak sense. The moral awareness we require is only a potential security mechanism—it allows the moral agent to register when her non-moral motivation is not in accordance with the moral requirement anymore. What is not demanded is for the moral agent to elaborate her moral theory so much that it could be regarded a genuine warrant for her moral intuitions or emotions. Also, the argument for a requirement of moral awareness does not rely on the notion that moral views are necessarily the best stabilizers of right moral action. The importance of moral awareness is that it could make the moral agent recruit other motivational resources when her prior non-moral motivation leads her into a morally bad direction. Moral awareness is a potential stabilizer for morally right action. But in any given case, a practice might be better stabilized by some non-moral motivation than by a progress in moral views. For some moral agents, concern for one's health might happen to be a more successful stabilizer for veganism than moral considerations. Moral insight can have important instrumental value by working to the same effect, but moral views will not necessarily be superior in this regard than any nonmoral motivation.

The argument for a requirement of moral awareness can be read as articulating a Kantian reservation (§ 3.2.) about attributions of moral progress to changes in moral agents who lack the right kind of ethical involvement; but it must be noted that what is articulated here is only a Kantian reservation in a very weak form. By starting from minimal moral awareness, we are only considering a very weak sense of acting in view of the right reasons—not to be confused with acting for the right reasons in the demanding sense of being motivated solely or overridingly by moral considerations.We are not requiring the moral agent to act from duty. For many non-Kantians, requiring the moral agent to be overridingly motivated by moral concerns in order to grant her having made moral progress would be a break from how we assess (non-progressive, but good) moral performance in general. This sentiment is captured in the following defense of utilitarianism by Mill against the objection that the theory required an overly demanding motivating mindset from moral agents:

It is the business of ethics to tell us what our duties are, or by what test we may know them, but no system of ethics requires that the sole motivation of all we do shall be a feeling of duty; on the contrary, ninety-nine hundredths of all our actions are done from other motives, and rightly so done, if the rule of duty does not condemn them. (Mill 2008b, 174)

The view of moral progress defended in the present section sides with Mill's permissive view of moral motivation up to a point. It allows non-moral reasons to play a key role in the occurrence of right action. However, it requires that at least some moral awareness be present in order for a moral agent to make moral progress. This requirement corresponds to a very weak version of a Kantian reservation about morally desirable practical changes. In view of the demandingness and restrictiveness of real Kantian constraints on moral motivation, the proposal here is much more limited. The fundamental interest here is in treating the presence or absence of moral awareness as important for the grouping of different kinds of morally desirable changes.

Once we make the conceptual decision to count mere practical change, unaccompanied by any moral awareness whatsoever, as a different kind of phenomenon than moral progress, the next question is: can we—or should we quantify how much moral insight is required for moral progress? Is moral awareness all that is required for moral progress to be possible? Can we further specify how sophisticated or motivationally important the moral insight accompanying a change in practices has to be? The answer to be given in the remainder of this subsection will be that we cannot establish a lower bound of ethical involvement below which moral awareness does not count as facilitating moral progress. Moral awareness, however minimal, of the valence of the new practice should be seen as the threshold for moral progress.

Consider, once again, the Unwilling Meat-Eater and the potential progress that she could ultimately make. Imagine that she eventually came to close the practical gap created by her failure to act in accordance with her moral judgment, i.e., she stopped eating animals; see Figure 6.

Fig. 6: Closing of the practical gap: the assimilation of moral practices to moral views without further ethical progress.

There are different ways in which the closing of the practical gap in the case of the Unwilling Meat-Eater could come about. The first type of change we might consider is that her moral conviction and according motivation could grow stronger—which could mean that she became more certain of the correctness of her moral assessment over time or that she became more devoted to acting on moral rather than prudential grounds generally. However, this change would, by all means, probably have to be counted as an improvement in moral views (an improvement on the theoretical side of morality) and is therefore not what is captured in Figure 6 and in the very idea of a mere closing of the practical gap. The change envisioned here is practical change related to "catching up" with a prior change in moral views—but not because of further improvement in those moral views. It is the kind of change that would occur if the moral agent's moral motivation only grew relatively stronger in that the active change occurred in her opposing non-moral desires: the moral agent's culinary desire for meat could simply decrease over time. Alternatively, the relevant change could also occur if the moral agent acquired a new belief that provided her with a new prudential reason to refrain from eating meat, that is, if she acquired a new non-moral desire that would point her toward the same direction as her moral desire. For instance, she might come to believe that stress hormones produced in animals' bodies during slaughter negatively influence human health when consumed. Based on this belief, she could form a prudential desire to refrain from eating meat. Let us focus on this latter case.

The new belief could work in favor of conformity to the agent's moral belief in two different ways. In one scenario, (1), it could give rise to a corresponding desire to refrain from eating meat that would make a critical contribution to the agent's overall motivation. In this case, the moral and the prudential motivating thoughts would both be necessary and would jointly suffice to motivate the agent to refrain from meat-eating. It would be the thought that the meat is supposedly stress hormone-laden together with the thought that eating meat is morally wrong that would allow the agent to refrain from eating meat. In another scenario, (2), the prudential motivation alone could suffice to secure the change in practices: the thought of the hormone-laden meat alone would be enough to put the moral agent off meat. In this case, the moral agent is slightly motivationally overdetermined to refrain from meat eating: not in the sense that each desire, the prudential and the moral one, could motivate action on its own (only the prudential desire, not the moral one, would suffice by itself), but to the extent that the prudential and the moral desire overlap, i.e., to the extent that the prudential desire could decrease and the moral desire would fill in the motivational gap. Even if the desire to avoid hormone-laden meat would decrease over time, this change would not discontinue the new better practice as long as the standing moral desire to avoid meat could compensate for the loss of prudential motivation. But as long as the motivational overlap is there, the moral agent does not need to draw on her moral motivation at all—she could suspend the moral thought and act on her prudential reasons alone. Even if both motivating thoughts are somehow present to her, this would not make a difference with respect to her practices compared to the case in which she suspended the moral thought or did not have it at all.

The case of the Unwilling Meat-Eater who has become the Newly Practicing Vegetarian illustrates an improvement in moral practices that is not a fluke in the sense of being unaccompanied by a moral insight. So, the Newly Practicing Vegetarian must be seen as a potential undergoer of moral progress. The change she makes is, however, still fortunate in that it ultimately occurs not (just) based on her prior ethical progress, but only once that progress is complemented by a contingent change in the moral agent's non-moral views. Prima facie, the Newly Practicing Vegetarian qualifies as making actual moral progress; but should we, in light of the contingency of her practical change, qualify our moral progress-judgment about her in any way? Should we even say that she does not make moral progress, after all, because her ethical involvement is insufficient?

Not crediting the Newly Practicing Vegetarian with having made moral progress in the full sense because her moral motivation happened to play only such a contingent role—in scenario (1)—or even no role at all—in scenario (2)—is intended to articulate the intuition that the theory-requirement associated with actual moral progress has not been fulfilled—that the agent's improvement does not reach an important threshold. We have already said of one possible threshold that the then Unwilling Meat-Eater did not reach it: she did not meet the requirements for making dispositional moral progress. In the search for a threshold for ethical involvement in actual moral progress, we might have found a natural candidate: should we not demand that actual moral progress was accompanied or had to emerge from dispositional moral progress? Should we say that actual moral progress needs to be actualized dispositional moral progress? This seems to be a plausible interpretation of the reservations we might have against attributing moral progress in the case of the Newly Practicing Vegetarian. My proposal here is to suspend such a reservation. While actual moral progress must not come about by a fluke, it need not, I suggest, occur alongside dispositional moral progress, either.

Recall our earlier effort to distinguish the merely theoretical change undergone by the Unwilling Meat-Eater (who had acquired a practically ineffective desire to refrain from eating meat) from the change undergone by the Converted Racist and the decision to credit the latter, but not the former, with having made dispositional moral progress. According to the definition tailored to this pragmatic need, in order to make dispositional moral progress, the moral agent has to undergo a change in moral views that would significantly improve her disposition to perform well in sufficiently probable circumstances. We recognized that in the evaluation of what could count as a good moral performance, considerations of demandingness would come into play. Given that the Unwilling Meat-Eater was not sufficiently motivated in actual circumstances to act in accordance with the moral requirement in which she believed and provided that the requirement would be evaluated as not excessively demanding, her failure to act in accordance with her moral belief would warrant an unfavorable evaluation of her performance and in effect, the judgment that she had not made dispositional moral progress. The effect the change in the moral agent's views had on her disposition to perform well was not recognized as significant.

Now, the Newly Practicing Vegetarian's accomplishment to close the practical gap between her beliefs and actions confronts us with a puzzle. Consider scenario (1), in which the practical gap is closed because the agent's insufficient moral motivation is supplemented by a prudential motivation. In this scenario, it appears that the moral thought becomes, in one sense, a significant component of the moral agent's disposition to perform well after all, but only because of its interaction with a separate, non-moral thought that is contingently acquired later on. The puzzle is that the moral motivation's contribution seemed insignificant when the moral agent was still the Unwilling Meat-Eater, but now that she has become the Newly Practicing Vegetarian, it seems significant, precisely because it now makes a relevant contribution which in turn depends on the contingent contribution of a new prudential motivation. The influence of such contingent contributions was precisely what the concept of dispositional moral progress was supposed to delimit. If it turned out that we would have to acknowledge, after all, that the Unwilling Meat-Eater had made dispositional moral progress because contingent changes rendered her formerly insufficient motivation sufficient, something would be wrong.

Fortunately, this is not what we should acknowledge at this point because it is not true, after all, that the moral agent's belief in the wrongness of meat-eating becomes significant with respect to her disposition to perform well as she acquires her new prudential belief. It becomes significant for her motivation in those new actual circumstances. It does not become significant for her disposition to do well, i.e., to how she would be motivated in varying circumstances. In different circumstances, she would acquire a different prudential motivation in conjunction with which the moral motivation would not suffice. So, it remains true that we cannot credit the moral agent who is motivationally positioned as the Newly Practicing Vegetarian with having made dispositional moral progress. And it might seem to us that this is the reason why we should also not credit her with making actual moral progress. When we think of the instrumental value ethical progress can have for the occurrence of better for moral practices, dispositional moral progress simply gives us the threshold for reaching the realization of this value: dispositional moral progress is attributed when the ethical change is such that it would stabilize morally right action across different contexts. The moral agent who does not reach this threshold makes potentially instable moral progress.

Yet, the somewhat morally aware moral agent is fundamentally different from the moral agent who improves her practices by a fluke. And this is true of her—with her consciously adopted belief in the wrongness of meat-eating as it would be of any moral agent with some minimal awareness of the moral valence of her chosen course of action. Any moral awareness, however minimal, distinguishes the moral agent categorically from the moral agent who lacks moral awareness altogether. (a) In combination with another (in itself only barely insufficient) motivation, moral awareness can contribute to bringing about right action. (b) In being morally aware, whether she performs well or badly, the moral agent acts as a moral agent; someone who is lacking moral awareness entirely acts as an agent but fails to act as a moral agent. (c) At the very least, moral awareness allows the moral agent to register when her non-moral motivation ceases to point her into the direction in which her moral awareness points her, enabling her to recruit any motivational resources available to secure that the right course of action is taken—in this limited sense, any moral awareness, however minimal, can act as a warrant for non-moral motivation and has the basic potential to be a stabilizer for good moral action. In this way, minimal moral awareness keeps practical change from being disoriented.

So, the theoretical side of morality is required to be involved in moral progress merely by contributing some minimal moral awareness. Dispositional moral progress is not a prerequisite for actual moral progress. It requires more of the moral agents than actual moral progress in terms of the agent's ethical development. Considering how contingent factors can facilitate practical changes that the moral agent could have been expected to bring about on her own, we come across a sense in which moral agents can be said to make merely actual progress. This result is surprising when seen against the backdrop of the idea we started with: that moral progress had to be practically relevant—which suggests that actual moral progress was the more important and more fundamental sort of moral progress, because it is what constitutes "change on the ground." We had previously acknowledged the existence of dispositional moral progress only because it was a way to honor potential improvement in practices. Eventually, it appears to be the more fundamental notion of moral progress.

'Actual moral progress,' therefore, remains to some extent ambiguous.When we judge that someone has made actual moral progress, we can mean that she has made actualized dispositional moral progress or that she has made the kind of fortunate actual moral progress achieved by those whose dispositions have not improved significantly enough for dispositional progress, but who are helped out by favorable non-moral changes. The ambiguity in the term reflects the fact that practical change underdetermines the attribution of dispositional change and our judgments of others' moral improvement are often ambiguous themselves.

This ambiguity of our attributions of moral progress plays out as a certain flexibility in our judgments in light of new evidence. Consider the idea of progress as a historical necessity and how especially the history of the 20th century is perceived asadefeater of this idea. In particular, the atrocities of the Nazi era are commonly named as conclusive evidence against the idea that progress necessarily occurs in human history. The crimes committed during this time seem to fundamentally discredit any grand progress narrative. They can be thought to do so in one of two ways. They may either be seen as staggering manifestations of moral decline—or as revealing that all presumed prior progress was not of the kind it had previously been thought to be. Hannah Arendt issues a judgment of the former kind when she concludes the inspection of evidence for the motivations of the conspirators of the 20 July plot by saying:

From the accumulated evidence one can only conclude that conscience as such had apparently got lost in Germany, and this to a point where people hardly remembered it and had ceased to realize that the surprising "new set of German values" was not shared by the outside world. This, to be sure, is not the entire truth. For there were individuals in Germany who from the very beginning of the regime and without ever wavering were opposed to Hitler; no one knows how many there were of them—perhaps a hundred thousand, perhaps many more, perhaps many fewer—for their voices where never heard. (Arendt 2006, 103)

In this passage, Arendt talks about the population's moral dispositions—attributing the actual development to a resolution of conscience, a large-scale decline in what we have categorized as the theoretical side of morality. Arendt's diagnosis comes to that of a gigantic moral decline—more specifically, the actualization of an underlying dispositional moral decline.

In contrast, one could also claim that the same factual development amounted to actual moral regress, where this is not to be read as actualized dispositional moral regress. This evaluation would imply that previously, moral dispositions had not ever become better than those that were, as it were, merely actualized in Nazi Germany. On this view, the crimes committed during that time are evidence of a merely actual moral regress—a change in practices that constitutes a deterioration into a prior state of development (not a new direction) and that reveals that moral agents' underlying dispositions had never changed for the better to begin with. This evaluation would fit with how the moral agents of this era are sometimes referred to in the literature on moral luck. In this context, the moral evaluation of Germans living during the Nazi regime often considers them as moral agents facing bad circumstantial luck (Nagel 1979); they are, according to proponents of moral luck, of course, to be blamed for their complicity, of whatever magnitude, in the crimes perpetrated at the time, but they are also to be considered merely unlucky compared to moral agents in other times, given that the circumstances they faced were unusually adverse to performing better morally. Many of them would have acted morally better under better circumstances, and many people who are so lucky as to live in those better circumstances might not have managed to avoid complicity in Nazi crimes, either. Yet, we rightly do not judge would-be collaborators as we judge actual collaborators. Proponents of moral luck should say that the moral deterioration that occurred during the Nazi regime was a case of merely actual moral regress: Germans during the Nazi regime simply happened to have the same kind of dispositions actualized that prior generations possessed as well and which they had not managed to improve upon. Whatever progress people had made previously was not a kind of dispositional progress, but merely actual moral progress. The case illustratesamore general point: given new evidence of moral underperformance, our judgments of moral decline or moral regress can prompt clarifications in our evaluations of prior moral progress. When a significantly bad moral performance calls into question prior moral progress, what once seemed to have been actualized dispositional moral progress, turns out to have been merely actual progress followed by moral regress. Differentiating between dispositional and actual moral progress provides the conceptual framework to make sense of moral reassessments in the face of volatility in moral performance.

The account proposed here still leaves us free to spell out a theory of moral progress that determines the degree to which moral motivation influences the extent of accomplished moral progress. One might, for instance, think that the agent is all the more praiseworthy the more her actual moral progress is an actualization of dispositional moral progress and that the agent being more praiseworthy should lead us to judge the moral progress she has made to be greater. Greater intellectual involvement may well be a normative or strategic desideratum, we have only rejected it here as a requirement for talk of moral progress to be warranted at all.

The answer to the question for this subsection comes to this: moral progress cannot come about by a full-blown fluke, it cannot result in fully inadvertent rightdoing, but it can be fortunate in that non-moral motivation can play a vital role in completing ethical progress practically. If we are permissive with regard to the grounds of motivation for moral action—if our Kantian reservations against moral progress by a fluke do not go so far as to demand moral performance to be moral action entirely out of the right moral reasons—then the door to fortunate moral progress is open. The distinction between dispositional and actual moral progress allows us to differentiate between moral improvements that are attributable to moral agents qua moral agents to different degrees.

#### 4.2.6 Improvement in Moral Performance

The preceding discussion does come to a specification of what the term 'moral performance,' as mentioned in our explications of moral progress, signifies. The relevant specification is that minimally good moral performance has a theoretical component in the wide sense: it requires the presence of moral awareness, not merely the awareness of the moral relevance but the moral valence of the act. The moral agent who can be said to perform morally well (or better than before) must be in some minimal way aware of the moral value of her action.

This specification will be rejected by some, which is to be seen as a downside, but it is an inevitable downside given the possibility of reasonable disagreement about the most important functions of the concept. The requirement of minimal moral awareness might be a problem for consequentialists, even though we have argued for it partly based on consequentialist reasoning about the possible consequences of a lack of moral awareness. Consequentialists can allow for the intuition that moral awareness matters, but not categorically. They can agree that moral awareness is relevant in the evaluation of moral progress on the grounds that it is generally more likely to figure in the production of best consequences than is the lack of moral awareness. They can also construe the required level of awareness in some intellectually undemanding terms such as by reference to a "responsiveness to the right reasons" (Driver 2012, 109; Driver 2003, Chapter 5), but this remains a substantive theoretical investment consequentialists need not make qua consequentialists. They might as well only care about an increase in right action, where rightness is evaluated based on consequences.

The basic division that could arise over the concept of moral progress turns precisely on this issue. The concept of moral progress explicated here requires an evaluation not merely of the moral desirability of the change in moral agents' practices but, owing to what we have labelled a (weak) Kantian reservation, also an evaluation of the moral worth the action has by virtue of the moral agent's mindset. Establishing that moral progress has occurred thus requires a more comprehensive assessment than the mere registering of either good consequences or good intentions. It leaves open the possibility of evaluating moral desirability in purely consequentialist terms, but it establishes a nontrivial restriction of the concept of moral progress in order for it to pick out a distinct kind of morally desirable change—change that is accompanied by moral awareness.

By introducing a requirement of moral awareness, we affirmed something similar to what was the first, least demanding interpretation of a potential agency-requirement for moral progress (see § 4.2.2.), the requirement of change in moral views. The requirement of change in moral views would be too specific, though. This is already suggested by the case of the Unwilling Meat-Eater's eventual transition to vegetarianism. At the time when the relevant change in practices is accomplished, it counts as an instance of actual moral progress due to a moral insight that predates the practical change.While this case might still make it seem as if a change in views was required—though theoretical and practical change could be temporally dissociated—we could amend the case to eliminate change in moral views as far as possible. We could think of an agent who had been introduced to meat-eating not as an infant, but at some later time in her life and who had always, for as long as she can remember, experienced a moral unwillingness to participate in the practice, but who had done so nonetheless, maybe because it was promoted to her as a special culinary pleasure and she had grown accustomed to it over time. In such a case of "Early Onset-Uneasiness" (and there may be more mundane examples of this that do not involve the artificial delay of a practice typically adopted early in childhood), we would still want to credit the Uneasy Meat-Eater with making (mere) actual moral progress once she is able to quit the silently rejected practice. If this is so, what is required for moral progress is not a change in moral views, but the presence of the right kind of views—or some whatever minimal moral awareness. The necessary awareness must not simply be of the moral relevance of the situation and her actions, i.e., of the fact that they are to be evaluated morally. In order to be able to act well, the moral agent must also have a sense of the moral valence: a sense of which are the good and which are the bad courses of action. This recognition need not be the relatively strongest motivational source for her according action. The agent may be drawing on any motivational resource whatsoever. So, the way the theoretical side of morality needs to be involved in a practical change for the better for this change to qualify as moral progress, comes to the following requirement:

Theory-requirement For an improvement in the practices of a moral agent to count as an improvement in moral performance, it must involve an improvement in relevant moral views or be a change toward accordance with a preexisting moral in-

sight, where both a moral view or insight might be given by mere moral awareness of the relevant practices' valence.

The theory-requirement specifies moral performance as involving both a practical and a theoretical side, widely construed. Let me indicate how it differs from the one relied on by Buchanan and Powell, whose stance on the characterization of moral progress in terms of changes in theory, practice and its factual results we have dealt with earlier in this section (§ 4.2.3.). Buchanan and Powell restrict their use of 'moral progress' to the stronger two of three senses of the term they distinguish (Buchanan and Powell 2018, 52, 53), excluding "changes that are desirable from a moral point of view, without requiring that any human motivational capacities be involved" (Buchanan and Powell 2018, 51). In other words, they exclude from the concept of moral progress axiologically valuable changes in states of affairs that are not, at least not willingly, brought about by moral agents. In the first, strongest sense the term 'moral progress' refers to "changes that either involve improvements in moral capacities or come about through the exercise of those capacities" (Buchanan and Powell 2018, 46). In this version, their theory-requirement comes to the demand of either theoretical change or the pursuit of practical change based on theory, i.e., a strong agency-requirement. In the second sense, the term 'moral progress' allows "changes that are improvements from a moral point of view to count as moral progress even if they came about through self-interested, prudential, or other non-moral motivations" (Buchanan and Powell 2018, 51). On this reading, there is no theory-requirement in the moral sense. For there to be moral progress, humans' mental, but not moral capacities must be involved. As we saw earlier, they draw a distinction between "mere conformity," which occurs when moral progress in the second, weaker sense occurs, and "genuine compliance" (Buchanan and Powell 2018, 68), which characterizes moral progress "in the most full-bodied sense" (Buchanan and Powell 2018, 51). Regarding one type of moral progress in their list, "better compliance with valid moral norms," they clarify:

"compliance" is not to be understood in a purely behavioral sense—that is to say, conformity to the norms in question cannot result solely from external forces that incentivize behavior. It must, rather, involve some exercise of or improvement in the moral capacities if it is to count as moral progress in the robust sense. (Buchanan and Powell 2018, 54)

Progress in terms of mere increased conformity is what they say "resulted from the introduction of institutionalized incentives that aligned self-interested action with valid moral norms" (Buchanan and Powell 2018, 50). Recalling our introductory example—the introduction of the fleshless burger patty as a possible manifestation or indicator of moral progress—we might think of this innovation as the provision of an "institutionalized incentive" which works to align customers' self-interest better with the valid moral norm of veganism. Under this description, the growing share of plant-based meat in human consumption would be covered by Buchanan and Powell's weaker, second sense of moral progress.

In contrast, on the view advocated here, whether or not a growing share of fleshless products in the entirety of consumed foods (with its positive effects in terms of a reduction of nonhuman suffering) is to be seen is indicative of moral progress, will depend on whether the providers of legume-based burger patties cater to former Unwilling Meat-Eaters or to Newly-Turned Dietary Vegans. The theory-requirement that I propose to include in an account of moral progress is weaker than the theory-requirement in Buchanan and Powell's first sense of moral progress (it does not involve an agency-requirement), but it is a theory-requirement in the moral sense, and in that way, stronger than the one in their second sense of moral progress. It counts the morally aware Unwilling Meat-Eater who switches to fleshless products only once they come to be readily available as an undergoer of moral progress, but denies this status to the fast food chain customer who has heard that plant-based food (even fast food) is healthier than animal products and, upon the introduction of flesh-equivalent products, turns into a Dietary Vegan.

With respect to the case that ethical progress is part of the process, the theory-requirement states that the views undergoing improvement must be relevant to the practical change (this rules out making moral progress by becoming a Dietary Vegan and simultaneously dropping a racist belief). Deciding that a belief and a practice that are undergoing change are fairly unrelated implies that the processes that affect them should be individuated with reference to the belief and the practice. If progress in a belief and progress in a prima facie unrelated practice do not belong to one and the same progressive process under any reasonable description, they do not constitute moral progress. The theory-requirement rules out moral progress by a fluke and it provides a specification that is absent from the second conjunct of the naïve conception: a mere change in the prevalence of right acts, absent any clarification of the motivational grounds of those acts, does not count as moral progress.

The concept of moral performance obviously requires further clarification to settle questions about the applicability of the concept of moral progress in various circumstances. But any further specification quickly forces us to make additional profound normative judgments—it requires more genuinely normative theorizing than an explication of moral progress should involve. Any decision made in the course of such normative considerations can induce a bifurcation of concepts of moral progress. The argument here has been taken only so far as to produce a concept that could lend itself to different further developments of this sort.

#### 4.2.7 Moral Progress with an Impact

The idea that whatever counts as moral progress must be practically relevant is included in both concepts of moral progress proposed so far. In dispositional moral progress, the requirement of practical relevance is honored because theoretical change is required to impact moral performance as the occasion for the relevant type of action arises. For actual moral progress, the change in moral performance itself is picked out by the very concept. However, we have thus far not included a requirement to the effect that actual moral progress must have an impact on states of affairs. Actual moral progress deviates from our initial idealized picture of moral progress according to which moral progress occurs as moral views, moral practices and states of affairs improve. Not only does our concept of moral progress not require moral views to change (as long as moral awareness is present), there is also no mention of positive effects on the world. Change in states of affairs is the ultimate "change on the ground" that usually prompts the judgment that moral progress has occurred. If the abolition of the slave-trade and the abolition of the institution of slavery would not appear to have ultimately bettered the situation of those who had formerly suffered the injustice of slavery in any way whatsoever, the case would probably not be treated as a paradigmatic example of moral progress. Interest in moral progress typically involves interest in improvements in states of affairs. Our explications of moral progress so far do not honor this fact. In this subsection, a supplementary explication is offered to compensate for this deficit. It also shows how a unifying thought behind our different concepts of moral progress could be spelled out.

To see clearly in what way the types of moral progress we have dealt with so far deviate from paradigmatic examples of moral progress, recall the case described by Buchanan and Powell of moral agents whose efforts to change an unjust situation are being thwarted. Above (§ 4.3.2.), it was proposed that in a case like this, moral progress does occur because activities are undertaken to combat injustice. These activities themselves constitute an increase in compliance with norms of justice, viz., norms for non-ideal circumstances. So, there is a relevant change in practices in this case. It is a case of actual moral progress, but it deviates from our ideal picture in that it lacks the feature of an improvement in states of affairs: because the moral agents fail to remove the injustice they are fighting, their moral progress is inconsequential moral progress; see Figure 7.

By not mentioning effects on states of affairs, we have defined actual progress as potentially inconsequential. This seems right if we agree with Buchanan and Powell that failing activists can still be credited with having made moral progress (if we disagree with their reconstruction of the case as far as practical change is concerned). Failing activists make inconsequential actual moral progress.

What we lack is a concept that includes the stricter practical requirement of an impact on states of affairs. Therefore, we have to introduce a third concept of moral progress:

Impactful moral progress Impactful moral progress is actual moral progress that brings about an improvement in states of affairs.

Moral agents who come to recognize an injustice and become activists make impactful moral progress, if they succeed in combatting the injustice; if they fail, they merely make actual moral progress.

Note that if the activists in the example make impactful moral progress, their success need not consist in bringing about impactful moral progress in other moral agents. The original problem their actions target, was—by virtue of being an injustice—a product of moral underperformance of other moral agents. This may make it seem as if the solution of this problem needed to be an improvement in moral performance, i.e., at least actual moral progress. However, the injustice could also be remedied because legislation is being passed that prohibits the practices constituting the injustice and neither lawmakers nor the perpetrators of injustice, whose practices are outlawed when activists succeed, need to share the activists' sense of the moral valence of the situation. They might all uphold their moral assessment that the practices in question never constituted an injustice. Lawmakers might acquiesce to the activists' pressure only because the activists happen to be an economically powerful group and citizens might obey the new laws grudgingly and without ever coming to endorse their moral point. Our explications of moral progress tell us that in a case like this, the passing of the crucial legislation is to be viewed as the outcome of moral progress only because it occurred due to an improvement in the activists' views and practices. In relation to lawmakers' crucial decision for this outcome and the subsequent change in practices of the wider population, the change for the better does not exhibit all the characteristics of moral progress we required. Moral agents' practices were changed for the better, but absent (the right kind of) underlying moral awareness, we cannot count this improvement as one in moral performance.

The present investigation into the concept of moral progress began by locating the relevant improvement in the performance of moral agents—differentiating moral progress from axiological progress, i.e., improvements in states of affairs. Yet, in a more comprehensive theory of moral progress, one could rely on the explications proposed here and still select impactful moral progress as the most important kind of moral progress, for instance by adding the thought that this is the "strongest" sense of moral progress, after all (cf. Buchanan and Powell 2018, 51). Looking at moral progress from the point of view of impactful moral progress in this way means to introduce a consequentialist evaluation of the moral agent's performance. Different ethical theories will want to privilege different concepts. And the concepts proposed here will probably not be acceptable in "as-is"-condition to just any ethical theory given its respective particular inherent view of moral progress. On the one hand, the present proposal of explications does aim to be conducive to theorizing moral progress from different normative views—where this may mean that the proposed concepts would be further modified. On the other hand, it is put forward as a genuine attempt to give an account of what moral progress should be conceived to be. Therefore, it accepts the anticipated need for further modifications somewhat reluctantly. Talk of conduciveness to different perspectives is not meant to convey that the picture of moral progress painted here is not a genuine attempt to capture what it would make most sense to take the phenomenon of moral progress to be.

One possible way to express how the different concepts of moral progress thus explicated are thought to cohere with one another would be to say that jointly, they produce an idealized account of moral progress according to which moral progress is (a) durable change for the better in moral performance, (b) on given occasions, (c) that is sufficiently suited to effect change for the better in states of affairs. According to this characterization, (a) is supposed to cover the concept of actual moral progress, (b) hints at the possibility that moral progress might be made by changing dispositions that remain unactualized, and (c) indicates that moral progress is expected to have positive outcomes.

#### 4.3 Dimensioning Reconsidered

We have addressed the question of the social (§ 4.1.1.) and temporal scale (§ 4.1.2.), along with the problem of individuating "domains" of morality beyond the coarse distinction between theory and practice (§ 4.1.3.) as a problem of application for the finished concept of moral progress. While identifying and evaluating the significance of different "domains" of morality does require more extensive reasoning on the level of normative ethics than belongs within this explicatory project, we shall now briefly revisit the question of the scale of moral progress.

In the course of explicating the concept of moral progress in the preceding section, we have predominantly dealt with individual moral agents and their individual progress, though in the cases of the Failing Activists (§ 4.2.6.) and the scenarios that did illustrate changes on the group or societal level (e. g., the Decline of the Dairy Industry, § 4.2.5.) were somewhat simplified in that the social moral change that occurred in them was stipulated to be homogenous; while none of the members of the society in which the dairy industry went out of business were said to have progressed morally, all of the activists who failed to remedy an injustice (§ 2.3.) were said to have made moral progress. In reality, contributions to factual change, levels of individual practical change, and theoretical involvement are usually much more diverse—as are individual starting points for the progress we might want to say has occurred on the group level. The definitions of dispositional and actual moral progress given so far mention single moral agents, but they can be made to work for the social case by making only slight modifications—omitting the reference to a single agent's moral performance and speaking of moral views and moral performance instead.

Dispositional moral progress social Moral progress disp, social is the improvement in moral views that significantly increases the disposition to perform


The definition of impactful moral progress can remain in the form in which it was introduced.While direct references to single moral agents are omitted, questions remain as to who can be the believer in moral views, the bearer of behavioral dispositions and who can be said to give a moral performance. As indicated in § 4.1., the aim here is not to present an account of collective action to accommodate judgments about the moral progress presumably made by some group, nor to argue for a specific way in which mental and agency-related concepts ought to be applied to supposed collective entities. The most important thing to point out with regard to these problems is that the concepts proposed here prima facie lend themselves to different ways of accounting for progress on higher social levels.

In the context of the discussion of the potential agency-requirement for moral progress, we noted that speaking of an agent with a mutation that renders her more sensitive to others' needs as "progressed" vis-à-vis her ancestors borders on using "morally progressed" synonymously with "better than" (§ 4.2.2., cf. § 2.3.(4)). This remark should not be seen as discarding the idea of intergenerational moral progress. On the contrary, the weak theory-requirement that has been assumed here facilitates thinking of intergenerational moral progress as being driven, inter alia, by moral mutations. Comparative judgments about individuals belonging to a community and judgments about the community itself must be considered differently. Comparing an individual with a moral mutation such as the one described above to her ancestors and casting the evaluation in the language of progress is in fact a stretch of the concept of moral progress, at best. The descendant is better than her ancestors, but since she has not changed, she cannot literally be said to have made moral progress. However, as far as the moral community to which both individuals belong is concerned, we may well make sense of the idea that it morally progresses through the occurrence of moral mutations. If there is a plausible way to individuate the moral community such that it remains in a relevant sense the same as new individuals are born into it, we can say that the moral community changes for the better as the moral mutation spreads in the population, causing an increasingly greater share of individuals to have the more prosocial moral mindset and behavior. Like in the individual moral agent who has a moral epiphany, a new moral view (or, in this case, sensitivity) and corresponding behavior simply occurs in the moral community. The more it spreads and shapes the predominant moral views and practices, the more the moral community can be said to progress morally. Based on the requirements of ethical involvement in moral progress, which was defended here with individual moral agents in mind, we can allow for this communal change to count as moral progress even if no moral agents work for the change—and more so: even when no moral agents change individually. If new moral agents are simply born with a different moral mindset vis-à-vis their ancestors, for them to change the larger community, it is not necessary that they themselves undergo change. This is an advantage of the explications proposed in the previous section. They allow us to account for the fact that moral progress on a large scale may occur incrementally through "mere" mutations. The weak theory-requirement also renders it possible for moral progress to occur in cases in which better performance—given minimal moral awareness—is facilitated by favorable institutional circumstances and that moral progress can be "incentivized" (cf. Moody-Adams 1999, 183). The concepts proposed here therefore stand a good chance of actually applying to real improvements in the world. In Chapter 5, we will see how they can even be seen as indicating something for which the moral agent could rightfully strive.

That moral progress could really have the kind of motivational significance that has been attributed to it (§ 2.3.), might seem questionable—given it is a highly abstract concept, could it provide the moral agent with an actual moral objective? And would it be morally desirable for moral agents to desire "moral progress?" These are the questions addressed in the final chapter.

## 5 Moral Progress and Moral Motivation: Improvement as a Fetish?

Not only is the abolition of chattel slavery in the 19th century univocally treated as a paradigmatic example of moral progress, contributors to the contemporary debate about this concept also commonly point to one specific figure involved in this process, British abolitionist William Wilberforce, as an exemplary promoter of moral progress (e.g., Godlovitch 1998, 280, Jamieson 2017, 178, and Stokes 2017, 1828). Wilberforce, who lived to see his efforts for ending the slave trade succeed, died only shortly before slavery was outlawed in Britain and received the honor of a burial at Westminster Abbey. The inscription on his statue only a few feet from his burial site includes the following lines:

His name will ever be specially identified with those exertions which, by the blessing of god, removed from England the guilt of the African slave trade, and prepared the way for the abolition of slavery in every colony of the empire.²²

While the epitaph correctly identifies the accomplishment for which Wilberforce continues to be remembered, it is noteworthy how this accomplishment is being characterized: it is the removal of guilt from England with which Wilberforce is being credited—whereas the contemporary visitor to Westminster Abbey will probably rather think of this episode of history as the ending (if that) of immense suffering and injustice (and might legitimately question whether this came with the removal of guilt for this suffering and injustice at all).

It has been argued that invoking a sense of national honor was in fact decisive for the success of British abolitionism (Appiah 2010), so the epitaph would seem unfortunately apropos.²³ This explanation has not been unopposed. Buchanan and Powell, for instance, argue that genuinely moral reasoning was allowed to succeed by conducive circumstances that had only just come about, citing, among others, a reduction in the risk of death by violence and infectious disease, flourishing markets and "advances in communication due to the spread of literacy" (Buchanan and Powell 2018, 319–320). Irrespective of which is the

Open Access. © 2022 the author(s), published by De Gruyter. This work is licensed under the Creative Commons Attribution 4.0 International License. https://doi.org/10.1515/9783110798913-007

https://www.westminster-abbey.org/abbey-commemorations/commemorations/william-wil berforce-family#i15160, last accessed June 28, 2022.

Though, with respect to Wilberforce in particular, Appiah notes that while his mentions of 'honor' were not mere "rhetorical flourish" (Appiah 2010, 116), his notion of honor was a "highly moralized" (Appiah 2010, 118) one, leaving "no place to honor independent of religion and morality" (Appiah 2010, 116).

more empirically adequate view, what the controversy shows is that the moral agent's concern with the wrong kind of values can cast doubt on his praiseworthiness for an accomplished change for the better. In this case, one might say that at the very least the author of the epitaph demonstrated an obsessive concern with the moral standing of England, misdirecting his attention to the nation's supposedly restored innocence, while failing to address the actual value that lay in the legal change Wilberforce helped to bring about—the termination of an injustice and the suffering it had brought with it. Thus, the inscription on the statue that honors his accomplishments seems to reveal a moral misorientation—a false emphasis in the appreciation of past moral progress.

A very similar charge might in fact be leveled against a certain kind of prospective use of the concept of moral progress itself. As we have seen (§ 2.3.), the concept of moral progress is said to be relevant to moral philosophy not least because it is motivationally relevant to the moral agent. Against the backdrop of the clarifications reached in the past chapters, we can now take a closer look at this suggestion. One form this suggestion might take is the conjecture that the idea of moral progress presents a proper goal for which the moral agent might strive. There is a possible critical response to this suggestion that will be explored in this chapter. The objection is that striving for moral progress—moral progress as such, i.e., an abstract goal—would be a flawed way to relate to moral value in some of the same ways in which it is wrong to think of Wilberforce's achievements foremost as a service to the moral standing of England. Just as the moral agent should not be occupied with honor, where cruelty and injustice are the real concern, one might insist, she should also not to be concerned with an abstract idea of moral progress where particular moral problems need to be overcome. The objection to concern for moral progress as such, which will be spelled out and motivated by drawing on a famous argument in the metaethical discourse about moral motivation (Smith 1994), would be that it renders moral progress a "fetish." This chapter scrutinizes this argument, which goes against some visions of the motivational relevance of the concept of moral progress.

I will begin (in § 5.1.) by differentiating ways in which the idea of moral progress is said or might be said to be morally relevant and singling out the motivational role that will be of concern in this chapter, i.e., that of an abstract moral desire. § 5.2. will sketch the argument that inspires the fetish objection to desiring moral progress, which is Michael Smith's "fetish argument" against motivational externalism (Smith 1994, 1996). I discuss how the original argument relates to other claims about supposedly problematic relationships with moral value and show how it gives rise to the suspicion that a desire for moral progress would be objectionable on very similar grounds as the object of Smith's criticism and I consider which of the responses that have been given to Smith might bear on the present issue.§5.3. modifies the fetish objection to specifically tackle the desire for moral progress. The argument identifies three flaws in a desire for moral progress: that it is objectionably abstract, that it is, at the same time, too specific, and that it is overly modest. I refute the modified fetish objection by offering replies to each of these three concerns. Based on the preceding discussion, § 5.4. sums up the conditions that a defensible desire for moral progress has to meet and examines its particular positive value.

## 5.1 On the Motivational Relevance of the Idea of Moral Progress

As we have already seen in § 2.3., investigations into the idea of moral progress are sometimes associated with emphatic affirmations of its motivational relevance—exemplified by Nagel's proclamation that "the idea of the possibility of progress is an essential condition of moral progress" (Nagel 1986, 186) or Godlovitch's claim that the hypothesis of a goal or "end-state" of moral progress is necessary for actual change for the morally better to occur (Godlovitch 1998, 278). Moody-Adams even goes so far as to declare "[t]he idea of moral progress […] a necessary presupposition of action for beings like us." (Moody-Adams 2017, 153) These views of the motivational relevance of the concept of moral progress align with the way modern (post-enlightenment) authors tend to treat the more general idea of "progress" (where this is not specified as concerning moral progress). Here, it is common to highlight the idea's motivational importance as well, usually while distinguishing one's outlook from a naïve, obsolete belief in the necessity of progress. One insists that "progress is still a viable possibility or prospect" (Niiniluoto 2011, 97) and emphasizes that "[t]o abandon belief in progress as a historical necessity is not to abandon work for progress as a task" (von Wright 1993, 227). When progress is specified as moral progress, it is also sometimes addressed as a task, e.g., by Buchanan and Powell, who motivate their topic ("The evolution of moral progress") by saying: "If moral progress is possible, so far as one cares about morality, one needs to know how to achieve it" (Buchanan and Powell 2018, 20).

Here, two distinct roles for the concepts of progress and moral progress emerge. Depending on whether progress is thought of as a possibility or a task, the idea of moral progress may figure in different ways in the moral agent's motivation. On the one hand, the idea of moral progress could figure in the agent's motivational setup as the content of a belief—the belief in the possibility of moral progress. On the other hand, the idea of moral progress may be the content of a desire-like rather than a belief-like state—the desire for moral progress or the desire to bring about moral progress. It is this latter idea of moral progress as the object of the moral agent's desire that will prove to be in need of defense in this chapter.

However, it is possible that few mentions of "progress" or "moral progress" are meant to refer to the abstract idea in the first place. Maybe it is not moral progress as such that is recommended as something to strive for but always particular improvements and realizations of moral value that are simply summarized as "moral progress." With respect to proclamations like the ones cited above, it is not always clear at which level of abstraction the relevant notion of progress actually is located: are all of these claims about progress as such, or is the notion of (moral) progress to be read as a mere placeholder for particular developments? It might be the case that when authors talk of "moral progress"—especially when they are concerned with progress as an objective (e.g., Buchanan and Powell)—they often use the abstract notion of moral progress to refer to concrete, specific changes. In this sense, Jamieson points to "the particular" as "the locus of judgments about moral progress, and the source of our motivation" (Jamieson 2002, 338), limiting the motivational relevance of moral progress to the level of concern for concrete instances of (pending) moral progress.

Yet, at least in some cases, it is made explicit that it is indeed the abstract "idea of moral progress" which is envisioned as having some kind of motivational relevance (e. g., Moody-Adams 2017). Furthermore, when it comes to the belief in progress, it is in fact the belief in the possibility of progress in general which does seem relevant to moral motivation. The moral agent who faces the choice to devote resources to trying to bring about a particular improvement better believe that her actions, in principle, stand a chance of being successful because progress is something that is, in general, possible. If any efforts for improvements are to be rendered rational, it seems that belief in the possibility of success, and thus, a general belief in the possibility of moral progress needs to be present. Should the conscientious moral agent not also develop an according general desire to realize what she believes to be possible? Even if this did not happen for the actual average moral agent, a general desire for moral progress could be the asset of a special kind of moral exemplar.

As far as a general belief in the possibility of progress is concerned, however, it may be doubted that it does in fact play a crucial role in motivating the moral agent. Something less may appear to be sufficient. Potentially, the moral agent need not harbor the belief in the possibility of moral progress, but lacking the converse belief—that moral progress is generally impossible—may be enough to motivate rational efforts for the concrete cause. Agnosticism about the prospects of overall moral progress may seem to suffice. However, when we consider more closely why efforts for particular improvements should at least not be paired with disbelief in the possibility of moral progress generally, the absence of this disbelief does not appear to be enough for defensible pursuits of improvements. A moral agent who firmly believed moral progress to be generally impossible would neither be rational nor praiseworthy if she nevertheless acted to promote a particular improvement at hand. She would either have to believe that the particular improvement could not be brought about or that it would not count toward general moral progress, because good developments are inevitably cancelled out by bad ones. If she nevertheless proceeded to further the specific cause at hand, she would have to be characterized as either irrational (if she were pessimistic about the cause at hand) or partisan and reckless (if she held the second pessimistic view) because of her willingness to accept the downstream negative effects of her efforts to further a specific cause (i.e., the negative developments that would offset her specific accomplishment).²⁴ At the very least, advancing one cause while fully recognizing that whatever success this yields will be offset by a negative development does not look like the project of a morally exemplary agent. If this is so, agnosticism about moral progress seems to be problematic as well, because the agnostic moral agent would act from a position in which the impossibility of moral progress is not ruled out and thus offsetting positive effects by bad ones is a chance one takes in furthering a particular cause at hand. The agnostic agent does not act as recklessly as the one who pursues "progress" against her own better judgment (that progress is impossible), but her actions are not as fully rationally supported as those of the believer in the possibility of progress. At the very least, the claim that the belief in moral progress is relevant to rational efforts for moral progress is certainly plausible.²⁵

In case we come to believe in the possibility of moral progress, it seems a natural move to also make it the object of our wishing, desiring, and our prac-

It might be thought that working for progress while lacking a belief in its feasibility can sometimes be morally desirable and therefore rational. Moral agents tilting at windmills might at least manage to avoid moral regress. However, if avoiding regress is the only reasonable expectation in a given situation, the rational desire to ascribe to the moral agent who is fully aware of this fact has to be a desire to maintain the status quo, even if by means of "aiming for more," where this would, however, be merely a hyperbolic formulation of her actual aim. Depending on the situation, maintaining the status quo can be a substantial, even heroic achievement in itself. We must not be taken in by the idea that bringing about change for the better is the only way of performing well or exhibiting moral excellence.

As far as possible roles of the idea of moral progress in motivation are concerned, even simply withholding the pessimistic judgment would still be based on a consideration of the optimistic alternative, such that the idea of moral progress would still be vital for motivation.

tical efforts. Whether this is a morally required move is a separate question that we can leave aside here. For now, we can simply assume that the belief that moral progress is possible may well result in the moral agent desiring to bring moral progress about. But what exactly would be the content of such a desire? Based on the explication in § 4.2., we have three basic options available for specifying what the moral agent with a desire to bring about moral progress might wish for: dispositional, actual, and impactful moral progress. In principle, desiring to bring about (mere) dispositional moral progress is a reasonable option: the moral agent might wish to prepare the moral community for an eventuality. The desire for merely actual moral progress—in contrast to impactful moral progress—appears more problematic: why ought the promoter of moral progress settle for changes in practices that do not affect states of affairs? Initially, impactful moral progress seems to be the most worthwhile object of a moral agent's desire. More specifically, the kind of progress responsible for the positive impact on the world should, prima facie, be actualized dispositional moral progress. Impactful actualized dispositional moral progress emerges as a plausible candidate for being the ideal type of moral progress from the preceding discussion. However, the exact content of a defensible or even praiseworthy desire for moral progress will have to be settled in the course of the argument. It depends on the rational grounds of the moral agent's belief in the necessity of progress. In any case, the desire to bring about moral progress is not to be identified with a general desire to "make the world a better place," but rather amounts to something like a desire to make the world a better place by making one's fellow moral agents perform better (where moral performance is supposed to include a theory-dimension as laid out in §§ 4.2.5. and 4.2.6). An improvement in behavior counts as improvement of moral performance only if it is accompanied by new moral insight or amounts to better consistency with a preexisting moral view. The desire for moral progress relates to such kinds of improvements.

In the following, I will sometimes speak loosely of a "desire" or a "concern for moral progress," "wishing for moral progress" and the like when what I really mean is a "desire to bring about moral progress," i.e., a desire on the part of the moral agent to initiate changes that qualify as instances of moral progress. There is certainly a difference between simply wishing for something to happen and the desire or intention to bring it about. The passive wish for moral progress to occur is not what is considered here as a worthwhile object of scrutiny. It would not be analogous to the desire to "make the world a better place," the articulation of which we come across more often than the expression of a desire for moral progress and of which the latter may be considered a variation. A mere wish for moral progress to occur is also certainly not what anyone has in mind who defends the idea that the concept of moral progress matters due to its motivational relevance: a mere passive wish for moral progress would be a pious hope that does not become motivationally efficacious.

The idea that moral progress may have a much more crucial role to play in moral motivation collides with the fact (noted in the introduction) that the term 'moral progress' is relatively absent from moral discourse.Why is it that we rarely ever hear anyone proclaim a desire to bring about moral progress, whereas it is far from unheard of for individuals to aim at "making the world a better place?" Against this striking factual backdrop, the proposal that the idea of moral progress is motivationally relevant in that it may figure as the object of a moral agent's desire seems highly implausible. At least, moral agents seldom articulate such a desire. Possibly, the apparent descriptive inadequacy is to be explained by a perceived prescriptive inadequacy: the apparent irrelevance of the concept of moral progress in everyday discourse must possibly be traced back not only to the empirical fact that moral progress does not play the role of an end that moral agents aim at, but also to the normative assumption that it should not play such a role. Whether moral progress actually frequently occurs as the content of a desire-like state is an empirical question that will only be touched upon in the final section of the chapter (§ 5.4.).What will be addressed is the normative question of whether the concept of moral progress qualifies as a proper content of a moral agent's desire. Even if a desire for moral progress is not common, would it not be a defensible type of motivation? Should we possibly view it as something that would render a moral agent especially praiseworthy? Might such an agent qualify as a special type of moral exemplar? Or should we, after all, view her as harboring a concern that is irrelevant at best and misguided at worst?

#### 5.2 The Fetish Objection to Concern for Morality

The view that a general desire for moral progress would be misguided attributes to the moral agent who has such a general, abstract desire a misconception of what is really morally desirable—of what is a worthy object of moral motivation. This criticism would be in good company in normative ethics. Many objections to particular normative theories either explicitly are or could readily be seen as arguments about the moral agent's proper relation to moral and non-moral value. For instance, the claim that utilitarianism generally renders individuals "value receptacles" (Regan 2004, 205–211, 236)—a criticism that can be levelled against other ethical theories as well (Stocker 1976, 460)—the charge against rule utilitarianism that it amounts to a kind of "rule worship" (Smart 1973, 10) or the complaint against virtue ethics that it is wrongly focused on the moral agent's own good (Nagel 1986, 197; Prichard 1912, 34) are all expressions of the idea that a certain normative theory paints a flawed picture of how the moral agent ought to relate to what is actually morally relevant. The criticism of a failure to provide a proper relation to moral and non-moral value is particularly pointed in the charge of self-effacingness against theories in normative ethics. Theories are charged with being self-effacing if whatever they identify as the ground for justifying an action "had better not be the agent's motive for doing it" (Hursthouse and Pettigrove 2018, § 3.). This problem is referred to by Michael Stocker as the "schizophrenia" of egoism, deontology and utilitarianism, as these theories allegedly disrupt the continuity between an agent's reasons and her motives (Stocker 1976). The reasons for action they identify should not be on the moral agent's mind as she pursues the right action. For instance, the person who visits a friend at the hospital because it is his duty seems to miss the point of a friendly visit. When the moral agent's motivation relies on these ethical theories, "the wrong sort of thing is said to be the proper motive" (Stocker 1976, 463). The term "self-effacingness" conveys the idea that if, in order to avoid giving an improper motive, a moral theory "told everyone to cause himself to believe some other theory"; the moral theory "would then be self-effacing" in that it would "remove itself from the scene" (Parfit 1984, 23).

The proponent of a theory may endorse its self-effacingness, as Henry Sidgwick is taken to do (Driver 2014, § 3) when he claims that

a Utilitarian may reasonably desire, on Utilitarian principles, […] that the vulgar should keep aloof from his system as a whole, in so far as the inevitable indefiniteness and complexity of its calculations render it likely to lead to bad results in their hands (Sidgwick 1907, 490)

Here, it is recommended to let the average moral agent, for strategic reasons, rely on a different way to reach moral conclusions than by utilitarian reasoning, because reliance on complex utilitarian calculations renders successful moral action improbable. Usually, though, self-effacingness is seen as a problem. Michael Stocker's criticism receives its force from the idea that there are certain values such as "love, friendship, affection" (Stocker 1976, 456) that any ethical theory should and does recognize, but that these same theories cannot guide the agent in realizing these values because of an exclusive focus on "rightness" (Stocker 1976, 455), which renders them unable to accommodate the value of engagement with others (and oneself) for the sake of the other person:

What is lacking in these theories is simply—or not so simply—the person. […] The person not merely the person's general values nor even the person-qua-producer-or-possessor-ofgeneral-values—must be valued. (Stocker 1976, 459)

Stocker sees it as a fundamental flaw of normative theories when they cannot be lived by because "to the extent that you live the theory directly, to that extent you will fail to achieve its goods" (Stocker 1976, 461).

The self-effacingness objection comes to the charge that if an agent is motivated by the ground of an action's justification as characterized by a certain normative theory, this fact works as a defeater for the realization of the value the action seeks to realize. It is to be distinguished from a charge of self-defeat against the theory, in that the theory could still be correct about what constitutes right actions (Parfit 1984, 23). The objection with which we will be dealing in the remainder of the chapter is yet less severe than the charge of self-effacingness. However, the two are closely related. The complaint we will scrutinize will not be that the pursuit of moral progress defeats its own purpose, but that it is still an unsuitable motivation for the moral agent. The moral agent pursuing moral progress might succeed in working for improvement, but, according to the objection, harboring an abstract desire for moral progress detracts from the moral agent's praiseworthiness or shows her to be less virtuous than her pursuit of moral improvements would otherwise indicate she is.

#### 5.2.1 The Right Thing as a Fetish

The particular argument that most directly inspires the objection to a desire for moral progress is Michael Smith's fetish argument against motivational externalism. This argument states that morality is made into a "fetish" (Smith 1994, 75) when the moral agent's ultimate source of moral motivation is identified with a general, abstract desire to do the right thing, i.e., to do whatever the right thing in any given situation may be. Of all the arguments concerned with the picture that some ethical theories paint of the moral agent's motivational make-up, the fetish argument is most relevant to the problem of a possible desire for moral progress for two reasons. First, it is not directed against any particular normative theory or group of normative theories, but it originates in a metaethical debate and is supposed to be directed against the most general and abstract moral motivation conceivable. The desire for moral progress, which will concern us here, is to be imagined as similarly general—not further specified by any particular normative theory about what would constitute progress. Second, the fetish argument does not rely on the claim that the abstract desire to do the right thing makes it impossible to in fact act rightly but rather on the claim that the agent who acts on this desire does so in a deeply flawed way.

The background for the original fetish objection is the metaethical trilemma, i.e., the incompatibility of three presumably widely shared convictions about the nature of moral judgment and thus, the nature of morality more generally. The trilemma holds among the ideas that (1) moral judgments express beliefs (cognitivism), (2) motivation requires a relevant desire, as beliefs by themselves do not motivate (the Humean theory of motivation), and (3) moral judgments necessarily motivate (motivational internalism) (cf. Fisher 2011, 138–139).

The first, cognitivist conviction concerns moral psychology and moral semantics. It comprises cognitivism in the narrower sense, i.e., the view that making a moral judgment consists in undergoingabelief-state, and the descriptivist thesis that uttered moral judgments not only superficially appear to be but really are to be interpreted as truth-evaluable propositions expressing those beliefstates.

The second conviction is the endorsement of the Humean theory of motivation, i.e., the view that beliefs alone do not motivate, but must be accompanied by a relevant desire. One way in which this can be spelled out is by referring to beliefs and desires as having a different direction of fit to the world (Smith 1994, § 4.6): beliefs represent the world and aim to fit the way the world actually is, desires "try to get the world to fit them" (Fisher 2011, 135). In desiring something, we want the world to be a certain way. Therefore, desires are "[w]hat injects the dynamism into beliefs" (Fisher 2011, 135), i.e., what contributes the motivational force. For instance, one might believe that in taking an umbrella, one would be prepared to prevent oneself from getting wet when it starts raining, but unless one desires to keep dry, the belief about the instrumental value of the umbrella is motivationally inert. In order to be motivated, the agent must have a desire (e. g., to keep dry) and a relevant means-end belief (e.g., that in using an umbrella, one would manage to keep dry) (Smith 1994, 92–93). The Humean view of motivation is traced back to Hume's exposition of the difference between beliefs and desires (Stahl 2013, 41–43):

Since morals, therefore, have an influence on the actions and affections, it follows, that they cannot be deriv'd from reason; […] Morals excite passions, and produce or prevent actions. Reason of itself is utterly impotent in this particular. […] Reason is the discovery of truth or falshood. Truth or falshood consists in an agreement or disagreement either to the real relations of ideas, or to real existence and matter of fact. […] Now tis evident our passions, volitions, and actions, are not susceptible of any such agreement or disagreement; […] Reason is wholly inactive, and can never be the source of so active a principle as conscience, or a sense of morals. (Hume 1960, 457–458)²⁶

However, it has been argued that Hume himself was not a Humean about motivation (Millgram 1995; cf. Fisher 2011, 134).

The third, internalist, conviction that co-constitutes the metaethical trilemma concerns a supposed conceptual necessity. It is the commitment that in order for a judgment to count as a moral judgment, it must be the case that the moral agent making the judgment is in some—at least minimal—way motivated to act in accordance with the judgment. The idea is that genuine moral judgments require that the ones making them recognize, feel or are to some whatever small extent subject to their prescriptive force. Thus, internalism can be thought of as a way of accounting for morality's supposed practical relevance (cf. Stahl 2013, §§ 1, 3). Internalists insist that failing to be motivated in accordance with a moral judgment means that one has not really made a moral judgment at all. They insist on reserving the term 'moral judgment' for those kinds of judgments about what is good, bad, right or wrong that come with some however small degree of according motivation.

While all three positions are supposed to capture widely held convictions about the nature of morality, all taken together are incompatible. If moral judgments are beliefs (cognitivism) and beliefs alone cannot motivate (the Humean theory of motivation), then moral judgments do not motivate by virtue of conceptual necessity (rejection of motivational internalism, thus, motivational externalism). If they do so motivate (motivational internalism) and motivation requires a desire (the Humean theory of motivation), moral judgments cannot be (mere) beliefs (rejection of cognitivism). Finally, if moral judgments are beliefs (cognitivism) and necessarily motivate us (motivational internalism), the Humean theory of motivation must be rejected.

One way to resolve the metaethical trilemma is to reject internalism and be an externalist about moral motivation. Externalists with respect to moral motivation deny that the sincere issuing of a moral judgment necessarily involves being to some (however small) degree motivated to act in accordance with that judgment, where the necessity in question is conceptual necessity. They deny that it is part of the meaning of 'making a moral judgment' that one be in an according conative state. Instead, they accept that the conative contribution to moral motivation to act stems from a desire that is external to the moral judgment, which is a proper belief (e. g., Svavarsdóttir 1999). The fetish argument is directed against the externalist view of moral motivation.

Externalists do not require us to restrict the attribution of a moral judgment to agents who are in fact moved by their sincere judgments. They, too, however, acknowledge that moral agents typically are so moved. The supposed fact "that moral motivation is a strikingly regular and reliable phenomenon […] that moral judgments dependably, if not unfailingly, motivate, that they effectively influence and guide how people feel and act" is widely accepted (Rosati 2016, § 1, emphasis in original). The dispute between internalists and externalists with respect to moral motivation revolves around the question how to adequately characterize what goes on in a moral agent who is motivated in accordance with her moral judgments. It is the fact that moral judgment and motivation regularly are congruent and that motivation changes in accordance with changing judgments that Smith has deemed "the striking fact," which any account of moral motivation must explain (Smith 1994, 71). Smith thinks that the externalist way to fulfill this requirement is deeply problematic.

His argument begins with the observation that a source of motivation that is external to the moral judgment must be such that it explains the reliably occurring congruence between judgment and motivation (Smith 1994, 73). The externalist, Smith argues, must posit a general desire to "do what is right" (whatever the right thing to do may be in any given situation) that is capable of motivating the moral agent when combined with different beliefs about what is in fact right (Smith 1994, 74). However, he claims that the desire to "do what is right" would be a "fetish or a moral vice" (Smith 1994, 75) i.e., it would make "the right thing to do" into a fetish—the undeserving object of a misdirected desire. It seems that if there is something to this criticism, a desire as abstract as one for moral progress (in whatsoever it may consist) would be equally under pressure. In the remainder of this section, we will seek a clearer understanding of Smith's original fetish argument against externalism and the replies that have been given to it, before applying the core concern—about the proper relation to moral value and morally relevant values—to the idea that moral progress should figure as the object of the moral agent's desire.

It has been claimed that the fetish argument actually contains several related objections to externalism. Some have, for instance, stressed the separability of Smith's charges of descriptive and normative inadequacy (e. g., Kauppinen 2015, 252). That there is an objection to externalism's descriptive adequacy next to the objection to its normative plausibility comes out most clearly in a later restatement of the fetish argument, where Smith relates the example of his college friend's conversion from maximizing act-utilitarianism (with its commitment to impartiality) to a constrained version of consequentialism (that licenses bestowing special benefits upon one's loved-ones). With respect to this case, Smith alleges that describing his friend's change of heart in externalist terms would be highly implausible as an empirical assertion.

When he was a beginning graduate student he came under the influence of Peter Singer's writings and converted to utilitarianism. He believed that it is always right to maximize happiness and minimize suffering […] Over the years, however, via a process of argument, he came to change his mind. He now believes that it is sometimes right to give extra benefits to his family and friends, even when doing so cannot be given a utilitarian justification. […] His moral motivations have followed reliably in the wake of his new judgment […]. (Smith 1996, 180)

The way in which externalists can make sense of his friend's development is problematic in the following sense:

externalists tell us that he must really only have had an instrumental desire to maximize happiness and minimize suffering in the first place, and that he must really only have acquired instrumental desires to confer special benefits on his family and friends, at least in the first instance. These desires can all only have been instrumental because […] they must all have been derived from the one and only non-instrumental desire […] to do what is right that leads him to have various instrumental desires […] when combined with his earlier and later means-end beliefs about how to do what is right. But this seems to me to be manifestly false, an utterly theory-driven redescription of my friend's psychological change. (Smith 1996, 181)

When the externalist account is criticized on the grounds that it provides a "false, an utterly theory-driven redescription" of the change in motivation (Smith 1996, 181), it seems that Smith is in fact primarily concerned with empirical matters (cf. Kauppinen 2015, 252–253), whereas the alleged (normative) problem of moral fetishism is not reflected in these empirical matters. In other places, the charge of descriptive inadequacy is more intertwined with what has been taken to be the core of the fetish argument, the charge of normative inadequacy against the externalist picture of motivation.When Smith claims that externalism givesaflawed account of how "good" people are motivated, he speaks about how moral agents we identify as good or as moral exemplars in fact are motivated as well as about how moral agents in general should be motivated:

Good people care non-derivatively about honesty, […] the well-being of their fellows, […] justice, equality, and the like, not just one thing: doing what they believe to be right, where this is read de dicto and not de re. (Smith 1994, 75)

It has been observed that there is some degree of variation in different presentations of the normative argument (e. g., Svavarsdottir 1999, 194). That the fetish argument is a multi-pronged objection even in its core normative form already emerges within the passage quoted above, where Smith makes a point about moral concern having to be "non-derivative" but also, apparently, pluralistic. In a later defense of his argument, Smith connects the idea thatamoral agent's motivating desires should be underived from a more general desire to do what is right to the idea that they should be non-instrumental to such a general desire, and he reformulates his critique of externalism in terms of the picture it paints not of just any "good" person who performs decently as a moral agent but of moral perfection. Externalism according to Smith suggests that

[morally perfect people], even though they believe it right to, say, look after the well-being of their family and friends, […] do not desire non-instrumentally to look after the well-being of their family and friends. Their desire to look after the well-being of their family and friends is only an instrumental desire because it must have been derived from their non-instrumental desire to do the right thing together with their true means-end belief that they can do the right thing by looking after the well-being of their family and friends. (Smith 1996, 182)

Just as externalism gives a flawed account of motivational change, it also gives a flawed account of the motivation of moral agents we recognize as exceptionally good:

But this externalist picture of what morally perfect people are like is surely quite incredible. We normally assume that such people are moved by the very features of their acts which make them right. That is, if what makes a morally perfect person's act right is the fact that it serves the well-being of their family and friends, then we normally assume that what moves them is that very fact: that is, the fact that their act serves the well-being of their family and friends. This is part of what makes them morally perfect: morally perfect people are moved by right-making features. But this is the internalists' picture of things, not the externalists' […] (Smith 1996, 182)

Smith then takes his criticism of the alternative, externalist account one step further, insinuating that in virtue of caring about rightness instead of right-making features, moral agents as depicted by externalism are eventually overly concerned with themselves.

On the externalists' picture […] [morally perfect people], [i]n desiring to do what is right for the sake of its being the right thing to do, rather than for the sake of the feature that makes it the right thing to do, they desire something that is not of primary moral importance. They seem precious, overly concerned with the moral standing of their acts when they should instead be concerned with the feature in virtue of which their acts have the moral standing that they have. (Smith 1996, 182–183)

The last part at least seems to insinuate that there is a further layer of criticism directed against the externalist picture of motivation. Elsewhere, Smith calls a desire to do the right thing a "self-consciously moral motive" (Smith 1994, 74). The problem, it appears, is that the moral agent is concerned with her acts and therefore ultimately with herself (rather than with what is of moral concern). Searching for a possible ground for the fetish charge, Vanessa Carbonell reads Smith in a similar way and argues that misplaced egocentrism cannot be the reason one may find a desire to do what is right objectionable, because egocentrism is not inherent to this desire:

Why would it be a fetish or moral vice to care only about doing "what one believes to be right," de dicto? As I discussed earlier, one reason is that what one believes to be right might diverge from what is in fact right. Another reason is that a person who is more concerned with the fact that her commitment is described as "morally right" than with the content of the commitment itself might be simply keeping up appearances. But these explanations are unsatisfactory. They simply point to some other, preexisting flaw in the moral agent, such as false beliefs or an exaggerated concern with her self-image. What we need is an explanation of why de dicto motivation is fetishistic that doesn't rely on contingent associations with other moral or epistemic character flaws. (Carbonell 2013, 464)

So, if the charge of egocentrism has been part of the fetish objection, this part of it needs to be suspended. With the concern of egocentrism being dismissed as a mere afterthought, is the label of fetishism possibly not warranted anymore? Sigrún Svavarsdóttir argues that "fetish" is a misnomer for Smith's objection on yet different grounds:

Webster's New Dictionary of Synonyms comments that 'fetish' "[i]n extended use … may be applied to whatever is unreasonably or irrationally regarded as sacred or sacrosanct." This seems to corroborate my understanding of what moral fetishism would be like. It would be the characteristic of holding oneself and others to very rigorous moral standards, while being completely unwilling to entertain any reflective question about their nature or grounds. It would be accompanied by a fear of any skeptical questions about morality […] But certainly, the desire to be moral will not alone yield such a character trait. (Svavarsdóttir 1999, 200)

There is room for reasonable disagreement with Svavarsdóttir's conclusion. The phenomenon she describes seems to be more aptly termed moral "dogmatism." At the very least, the dictionary definition of the term 'fetish' that she cites also aligns well with the core of the fetish concern: the assessment that the moral desire in question is a misdirected desire. The key charge is that externalist motivation makes "the right thing to do" an "unreasonably or irrationally" valued end, that is in actuality best seen as merely an abstract placeholder for the things that are really worthy of being valued. Because of the fetishization of rightness, moral concern is misdirected in that it only reaches the ultimately valuable thing via a detour, i.e., via concern for the abstract category under which the thing falls. Thus clarified, the fetish objection against the desire to "do the right thing" carries over to the first of the three threads of the objection against a desire for moral progress that are to be spelled out in the next section, viz., the charge that moral progress likewise is too abstract to be a proper object of moral concern.

According to Smith, the externalist's story about the moral agent's psychology which centers on the idea of a standing desire to do whatever is right is both empirically implausible and morally repugnant. It is this moral repugnancy that is highlighted by the criticism of moral fetishism. This criticism is spelled out in several slightly different ways. For one thing, there is the concern about particular motivations being "derivative" of a general, abstract desire. Here, the worry is being framed in terms of the logical hierarchy of apparent immediate and ultimate motivation, where the ultimate source of motivation is always the maximally abstract desire to do what is right. Smith identifies morally relevant things (such as the wellbeing of one's family) with which the moral agent is supposed to be directly concerned. Externalism is said to make this direct concern impossible, because it requires the agent to derive her motivation from a much more general desire (to do whatever is right). Externalism allegedly portrays the moral agent as being occupied with the fact that something matters, rather than with what matters. Smith phrases the problem as one of alienation "from the ends at which morality properly aims." To highlight the requirements of a proper relation to what is morally relevant, he draws a parallel to the case of love: "Just as it is constitutive of being a good lover that you have direct concern for the person you love, so it is constitutive of being a morally good person" that she directly cares about the particular things that are right to do, rather than the fact that they are right do to (Smith 1994, 76). Here, the close relation to Stocker's self-effacingness—or "indirection"—charge against normative theory shows. Stocker's point likewise is that theories such as utilitarianism do not allow the moral agent who lives by the moral theory to have a proper relation to what should be valuable to her. He illustrates this in very similar terms as Smith uses to illustrate the alienation from the proper ends of moral action caused by externalism. Stocker says: "Indeed, a person who values and aims at simply love, that is, love-in-general or even love-in-general-exemplified-by-this-person 'misses' the intended beloved" (Stocker 1976, 459). For Smith, though, the case of love is an analogy that illustrates a slightly different point. He is not saying that the person with an abstract desire to do what is right could not succeed at that, but that she would not count as a "morally good person." In effect, because he demands attention for right-making features instead of rightness, the criticism Smith puts forward against externalism amounts to an objection against placing the focus on moral value (here, rightness) at all. The idea seems to be that the good moral agent's focus should instead be entirely on the features on which moral value (merely) supervenes.

Then there is also the complaint about rendering particular motivations merely "instrumental," which frames the crucial worry in terms of a concern about the means-end-relation between motivation for particular actions and the general desire to do what is right. Externalism is now said to reduce all the moral agent's motivations to particular actions to means in the pursuit of rightness itself. As far as the ends of moral action are intrinsic axiological values such as wellbeing, the externalist view reduces them to mere instrumental values that serve the production of moral value (rightness). Because of this means-end-relation, moral agents lack proper appreciation of intrinsically valuable features. They cannot appreciate that which matters first and foremost.

In the original presentation of the argument, the distinction between de dicto and de re motivation features prominently. Of the kind of motivation the externalist has to posit to make sense of "the striking fact" Smith says: "the only motivational content capable of playing this role, it seems to be, is a motivation to do the right thing, where this is now read de dicto and not de re" (Smith 1994, 74). Crucially, the desire to do what is right must be read de dicto if it is to fit the externalist's explanation of the reliable change in motivation in the wake of change in moral beliefs. A de dicto desire to do what is right is the desire to do whatever falls under the description of being the right thing to do, where the agent need not be aware of any of the actions actually falling under this description. A de re desire to do what is right is the desire for particular actions that are in fact right (but of which the agent need not think as being right). For instance, if we say of an agent that she desires to bring about moral progress, on a de dicto reading, this would mean that she desires to bring about whatever would be morally progressive, whereas on a de re reading, it would mean that there are particular changes in the world that would constitute moral progress and the agent desires to bring about those very changes (cf. Dreier 2000, 621–622). Only on a de dicto reading can the desire to do the right thing work in the externalist's favor, because only when it is left open what will be the right thing to do can this desire reliably motivate the agent who changes her views about what is in fact right.

The distinction between de dicto and de re desires is helpful to characterize the controversial desire to do what is right and it might also be helpful in characterizing the desire to bring about moral progress. Yet, the distinction itself does not serve the differentiation of laudable and flawed moral motivation. The supposed problem with the de dicto desire to do what is right is not that it is a de dicto desire (notably, Smith does not use this differentiation in the later restatement of his argument; 1996) but rather this de dicto desire's level of generality, which allows it to have the fundamental function it is alleged to have for moral motivation: ensuring that a moral agent will be motivated in accordance with any (new) moral judgment. It is the abstractness of a desire to do whatever is right, which follows from its level of generality, that provokes the fetish accusation. But de dicto desires are not maximally abstract per se. The description that provides the content of a given de dicto desire may well mention concrete objects —in the case of moral concern, it could mention particular individuals. The desire "to bring about whatever furthers Fred the pig's wellbeing" would be a de dicto desire, indeed, but a relatively specific one, and it would mention in the apparently right manner exactly what Smith takes to be a prime example for an item deserving concern: a concrete individual and his wellbeing. It seems that what Smith is or at least should be saying is that moral agents ought to desire de re the ultimately valuable thing and they should desire in a de dicto fashion the unknown means to furthering or bringing about that which is of ultimate value. For instance, when we think of Smith's college friend, it seems that his desire to "maximize happiness and minimize suffering" must be taken to be a de dicto desire itself, given that Smith's friend is hardly familiar with all the manifestations of happiness and suffering in order to relate to them in a way that would give rise to corresponding de re desires. On the standard view, de dicto desires are ascribed to account for the fact that an agent may not be familiar with an object that falls within the extension of her desire (or may not realize that an object she is familiar with does fall within this extension). Very likely, Smith's friend was ignorant of many of the ways in which he could have maximized happiness or minimized suffering. He could not have formed desires to perform acts that he did not know would realize these values (i.e., maximized happiness and minimized suffering).

If Smith's friend's judgment that he ought to maximize happiness as well as his later judgment that he ought to bestow special benefits upon his family are to be counted as moral judgments, they must, on Smith's own internalist account, necessarily entail a corresponding desire. This is also what Smith claims they did. But being the result of exposure to "the influence of Peter Singer's writings" and "a process of argument," respectively, rather than exposure to all the suffering that he could curb and all the happiness that he could further, his friend's desires were abstract desires. His abstract moral judgments, derived from abstract reasoning, must have yielded abstract desires, i.e., theoretical views about what is "the right thing to do." It seems that even after his conversion to the view that he ought to bestow special benefits upon friends and family with whom he was, of course, acquainted—Smith's friend should, on some level, still have been motivated by de dicto desires. If he had, for instance, figured out that he should buy his sister an expensive gift in order to increase her wellbeing, it seems that it is also only under the description that the gift will make his sister happy that he ought to desire buying it. If he is to be properly connected to the ultimately valuable end of his actions, his sister's wellbeing, then he should have retained the de dicto desire to do what makes her happy. He ought not be moved solely by a de re desire to buy the item he has chosen but (also) by a de dicto desire to do what makes his sister happy. If in order to avoid moral fetishism, we should always be motivationally anchored in that which is ultimately morally relevant or ultimately valuable—in this case, the sister's wellbeing—, then the de dicto desires for whatever is instrumental to the ultimately valuable thing should remain in place. In the case of Smith's friend, the de dicto desire to do what makes the sister happy must not be suspended when entering the store; rather, it must be in place even while buying the specific gift.

According to the fetish argument, good moral agents have to acquire specific motivational connections to the worthy objects of moral concern that externalism allegedly does not provide. However, the worthy objects of moral concern are hardly the actions that merely further items like wellbeing, but those items themselves. Determining the right course of action always moves the focus from the end the act is supposed to further to the act itself. The issue on which the fetish argument turns is that the actions of a good moral agent require certain descriptions that link the agent appropriately to the worthy objects.What is important in order to produce a description which renders the agent a good moral agent is that it must not mention a too general desire as the ultimate source of motivation—the mistake made by externalists. So, the fact that it identifies as the moral agent's guiding desire a de dicto desire is not in itself what counts against externalism; rather, the problem is this desire's level of generality, which prohibits a proper motivational relation to what is ultimately valuable.

Of the many responses that have been given to Smith's argument, the following will be of some relevance to evaluating the fetish-problem that seems to arise for a desire for moral progress. First of all, it has repeatedly been pointed out that the kind of de dicto desire that Smith thinks the externalist must ascribe to the moral agent could well coexist with other, more concrete, i.e., de re desires (cf., e. g., Copp 1997, 49–50, Svavarsdottir 1999, 205, and Carbonell 2013, 465), resulting in a kind of "motivational overdetermination" (Olson 2002, 91). Olson illustrates this point in the following way:

Consider a person who is strongly engaged in the animal rights movement. She would perhaps have the de re concern for animals (the desire to minimize their suffering, say), but she may also desire to engage in such issues because it is right. That it is right might be part of her reason to desire to engage in these things. It is not clear to me whether Smith would say that this de dicto concern would somehow reduce the moral status of her engagement. In any case, I do think that such cases of motivational overdetermination are quite common, and it does not seem platitudinous that in each and every case it would be morally wrong to be motivated (at least in part) by a de dicto desire to do what is right. (Olson 2002, 91)

At the very least, in cases like these, a desire to do the right thing does not seem to make it the case that the moral agent fetishizes morality.

Second, with respect to the version of Smith's complaint that centers on externalism not allowing the moral agent to care non-derivatively about what ultimately is valuable, it has been pointed out that the criticism is unclear. Moral agents could "care non-derivatively" by having a desire that they acquired in another way than by deriving it from the desire to do the right thing, i.e., by having a desire that is "acquisition non-derivative," or, alternatively, they could care in a nonderivative manner by having a moral desire that has become "self-standing," i.e., whose "role in the psychology of its possessor at that time is not limited to motivating him to do something he believes is needed in order to satisfy another desire" (Svavarsdóttir 1999, 206, n. 66). The argument here is that even derived desires can become non-instrumental desires, so that Smith's way of treating the two as one and the same seems inapt. If the crucial issue is that the proper moral desire ought not to be instrumental, even derived desires can fulfill this condition. As Svavarsdóttir argues,

[f]or Smith's argument to go through the desire to φ must be acquisition-nonderivative in the good person. But I see no merit in that claim, although I agree that the psychology of the good person has to be such that it is possible for the desire to become nonderivative in the sense of self-standing. (Svavarsdóttir 1999, 206, n. 66)

Third, the desire to do what is right has been attributed instrumental value because it can be a backup for more specific moral motivation. The de dicto desire can be something to "fall back on" (Carbonell 2013, 469)—a "motivational substitute" (471) or "safety device" (Olson 2002, 92)—for when de re motivation fails to override non-moral motivation. When the moral agent is lead into temptation to choose a morally inferior course of action, having a desire to do what is right in addition to the failing de re desire for the specific morally right option can help her to actually do the right thing. Therefore, it has been argued that a reliable de dicto desire should even be regarded as a characteristic of a type of moral saint (Carbonell 2013, 469).

Fourth, it has been argued that a desire to do what is right can have instrumental value as a transitional solution (Svavarsdóttir 1999, 205–206). Precisely in cases of changes in moral views (the explanation of which precisely is externalism's downfall, according to Smith), it is unreasonable to expect the moral agent to form a de re desire for his newly adopted moral ends on the spot. Relying on de dicto desire after moral change is forgivable (Lillehammer 1997, 191–192).

A given thought or type of motivation could be criticizable or undesirable when it plays a role in a fresh moral decision without thereby being equally criticizable or undesirable when it plays a role in the endgame following a change of mind. For changing one's mind is not merely an instance of making up one's mind. Intuitively, it seems it would be more difficult to muster direct motivation in line with a new judgment that comes from a change of mind than one that comes from simply having made a decision (Carbonell 2013, 466)

Carbonell argues that being moved by a de dicto desire to do the right thing is not merely forgivable in the wake of ethical change but "the most we can ask of people" (Carbonell 2013, 468, emphasis in original). "De dicto moral motivation as a legitimate way for a good moral agent to transition to a new behavior" (Carbonell 2013, 468).

A desire to do what is right could have instrumental value in yet another way that is related to it being a backup and a transitional solution. The desire to do what is right could facilitate recognition of the fact that a practice the moral agent has become engaged in because she thought it was right is no longer supported by the evidence as being the right practice. A practice adopted for moral reasons takes on its own dynamics—partly because de re moral motivation attaches to it—and it may be difficult to realize when one ceases to be justified in believing that it is still the morally right course of action. A standing desire to do what is right—whatever that may be—can be a due corrective to a moral habit, because it motivates occasional reevaluations. In other words, a standing desire to do what is right can guard against the kind of practical paradox of progress discussed by Mill (§ 3.4.1).

In the following, the arguments that focus on the instrumental value of an abstract de re desire and the etiology of moral motivation will be most relevant for our argument concerning the desire for moral progress. Though they will not be straightforwardly transferable to that argument, they can be taken as hints toward a relevant issue.

#### 5.2.2 Morality as a Fetish

We have noted that the fetish argument, although it originated in a metaethical debate, can be placed within a wider context of normative criticism of certain contents of moral motivation. Smith himself contextualizes his argument in this way when he suggests that it is of the same kind as Bernard Williams's "one-thought-too-many"-argument, which is directed against the commitment to impartiality. Using the example of a man who is faced with the choice of either saving his drowning wife or a drowning stranger, Williams attempts a reductio of the idea that the man's favoring of his wife would require an impartial justification:

it might have been hoped by some […] that his motivating thought, fully spelled out, would be the thought that it was his wife, not that it was his wife and that in situations of this kind it is permissible to save one's wife. (Williams 1981, 18)

According to Williams, including a consideration of the impartial justifiability of choosing to safe one's wife ascribes to the moral agent facing this choice "one thought too many." Being concerned with what is "permissible in situations of this kind" is apparently inadequate for the moral agent who should have known all he needed to know once he realized that "it was his wife" who required assistance. Adherence to impartiality in a situation like this would undercut the "deep attachment" that characterizes relationships like that between the rescuer and his wife—thereby undermining what gives "substance" to life to begin with. Such "substance" is needed "if anything is to have sense, including adherence to the impartial system," so that considerations of impartiality cannot be given prime importance (Williams 1981, 18). This line of thought places the argument in the camp of self-effacingness objections.²⁷ Smith identifies a parallel between his concern and that of Williams:

in taking that a good person is motivated to do what she believes right, where this is read de dicto and not de re, externalists too provide the morally good person with 'one thought too many.' They alienate her from the ends at which morality properly aims. (Smith 1994, 76–77)

It has been noted that the parallel does not actually hold, given Smith's account of externalist moral motivation:

Michael Smith has just posited that externalism entails motivation de dicto but not de re. […] But this doesn't seem to be exactly what is going on in Williams' drowning example. […] if the thought "it was his wife" is here meant to be analogous to a de re concern for doing the right thing (that is, a concern directly for the wife), and the thought "in situations of this kind it is permissible to save one's wife" is here meant to be analogous to a de dicto concern for doing the right thing, then what we have is a person motivated to do what he believes to be right de dicto and de re, not one who is motivated de dicto but not de re. (Carbonell 2013, 465, emphasis in original)

Scheffler (1986, 532–533) claims that Williams' objection to impartiality considerations can also be seen as being concerned with the demandingness of a requirement to be motivated by impartial considerations.

Observing that the force of Smith's criticism of externalism relies on his insistence that it ascribes to the moral agent only one—de dicto—motivating thought, Carbonell emphasizes that the presumed analogy to Williams's concern would "entail [that Smith] thinks the good person ought to have zero thoughts" (Carbonell 2013, 465, n. 3). What the reference to Williams at least indicates is that Smith's main concern is about the moral agent's proper relation to what bears intrinsic value. The parallel to Williams is that both take issue with categories of moral value blocking a direct connection to that which is valuable.

As Carbonell observes (Carbonell 2013, 460), rather than being of the same kind as Williams's concern, the attitude that is expressed in the charge of fetishism more closely resembles the one that drives Susan Wolf's critique of "moral saints," i.e., "person[s] whose every action is as morally good as possible" moral agents who are "as morally worthy as can be" (Wolf 1982, 419). Both the fetish argument and Wolf's complaint about moral saints are concerned with a misdirection of moral attention to the maximally general, abstract level. While the fetish argument is directed against a desire to do what is right, Wolf frames the problem as "morality itself" being the undeserving object of moral desire.

[T]here is something odd about the idea of morality itself, or moral goodness, serving as the object of a dominant passion in the way that a more concrete and specific vision of a goal (even a concrete moral goal) might be imagined to serve. Morality itself does not seem to be a suitable object of passion. (Wolf 1982, 424)

Here we come very close to a version of the fetish objection that would directly target the desire to bring about moral progress. "Concrete moral goals"—such as ending animal experimentation, human trafficking, or establishing "equal pay for equal work"—are precisely what could be expected to be the more likely and more deserving objects of the moral agent's desire—compared to moral progress as such. Recalling Buchanan and Powell's claim that "[i]f moral progress is possible, so far as one cares about morality, one needs to know how to achieve it" (Buchanan and Powell 2018, 20), we can now add a caveat: in case one should not "care about morality" to begin with, one should also not pursue moral progress as such. The moral agent who desires to bring about moral progress rather than to achieve concrete moral goals seems to turn moral progress into a fetish. At the very least, the fetish argument against externalism gives rise to the suspicion that a desire for moral progress would turn moral progress into a fetish. This suspicion can be developed into a distinct three-pronged objection.

#### 5.3 The Fetish Objection to Desiring Moral Progress

Prima facie, if the concern about the desire to do what is right is that it is too general and abstract, the desire to bring about moral progress looks similarly troubling. The question is whether the replies that can be given to the fetish objection, if successful, simultaneously rid the desire to bring about moral progress from the suspicion of being fetishistic. It might look as if a desire to bring about moral progress will be either beyond saving (if the attempted refutations of the fetish argument fail) or it will be safe anyway so that nothing specific can be said about it beyond what has been said in defense of the desire to do what is right.

However, the fate of a desire to bring about moral progress is not bound up with the fate of externalism. On the one hand, if the fetish objection against the externalist picture of moral motivation was convincing, the desire for moral progress would still be worth investigating, as it is not constitutively connected to motivational externalism: moral agents can be thought to acquire a desire to bring about moral progress if the internalist picture of moral motivation is correct. On the other hand, successfully defending externalism against the fetish objection does not amount to ridding the desire for moral progress from a fetish suspicion. "Bringing about moral progress" is certainly not equivalent to the desire do "do the right thing." While we are explicitly dealing with concern for moral progress in an abstract form, i.e., with a de dicto, not a de re desire for moral progress, the desire to bring about moral progress is undeniably more specific than the original target of the fetish objection. If there is, after all, no problem with a desire to do what is right, the same need not be true of any abstract motivation for moral action. The case of the desire for moral progress thus cannot be closed even if defenses of motivational externalism succeed. Nonetheless, the cases are similar enough to think of them as congeneric problems.

It is with respect to the level of abstractness of the desire in question that the original fetish objection and the one against the desire for moral progress are most closely related. In addition to being objectionably abstract, the desire for moral progress can also be charged with being, at the same time, too specific, and finally, with being overly modest. Each of these charges will be spelled out and refuted in turn.

#### 5.3.1 A Reasonably Abstract Desire

The problem of abstraction arises when we think about how the desire to bring about moral progress can be rational. A rational desire to bring about moral progress should be based on a justified belief that bringing about moral progress is right or good.²⁸

If we speak of bringing about moral progress as the right thing to do, we are involving the idea of moral progress in moral motivation in a way that renders making moral progress obligatory. This seems to conflict with ordinary morality, which, it might be suggested, tends to classify paradigmatic efforts for moral progress as supererogatory, i.e., beyond what is required from the agent. The paradigmatic promoter of moral progress seems to go "beyond the call of duty." This way of thinking about the moral agent who actually promotes a good cause is noteworthy, but unproblematic. First of all, the moral agent might just plainly disagree with ordinary morality. After all, there must be some difference between an agent who happens to be motivated to initiate moral change and what is supposed to be the average moral agent, who is not. The promoter of moral progress might well think of herself as fulfilling a moral requirement. Second, the reconstruction of the view ordinary morality takes on the promotion of moral progress might simply be incorrect: furthering moral change may not usually be regarded optional. Third, the moral agent who in fact desires to bring about moral progress might well agree with the average moral agent in classifying the pursuit of moral progress as supererogatory but nevertheless be guided by a more ambitious general desire. She may thus believe that there are morally good things to do beyond what one is required to do, and she may desire to do these things. She might harbor a desire to do "what is morally best" or "whatever is morally good" instead of a desire for (mere) "rightness." This possibility nothwithstanding, the argument from here on will deal with the belief that bringing about moral progress is the right thing to do.

For the desire to bring about moral progress to be a sincere moral concern rather than an idée fixe, it must be based on a belief which is simultaneously an empirical and a normative belief; it is the belief that there is room for moral improvement and that moral improvement ought to be brought about.

While this possibility needs to be acknowledged, it is also worth noting that the fact that not all moral agents need to have such ambitious desires can underly an argument against motivational internalism. Alfred Archer notes that judgments to the effect that an act is good, but not required, pose a problem for the idea that moral judgments necessarily motivate, since "we do not expect people to be motivated by a judgment that an act is supererogatory" (Archer 2016, 606). He argues that a "Planet Duty," where no-one is ever motivated by supererogation judgments but only by judgments that acts are morally required, is easily conceivable (Archer 2016, 621), i.e., "'supererogation amoralists' are quite possible" (Archer 2016, 607). While Planet Duty may well be conceivable, this does not mean that on planet Earth there could not be moral agents with a standing desire to do more than their mere duty.

To be a justified empirical belief, it must be based on empirical evidence which supports it—indicators of actual moral problems and of the possibility of change for the better. Now, acquiring the belief that one ought to bring about moral progress as such requires abstracting from this evidence: abstracting away from the actual problems that need to be overcome and the concrete states of affairs one envisions could be realized. If forming the abstract de dicto belief that bringing about moral progress is the right thing to do requires abstracting from the evaluations of particular needs for improvement, it involves thinking about actual moral problems—actual injustice or suffering—in objectionably abstract terms. It involves, for example, beginning to think of the need to end modern slavery as a (mere) manifestation of the need to bring about moral progress. It involves beginning to think of the plight of nonhuman animals suffering and dying in laboratories or of the suffering of an underaged sex slave as mere occasions for moral improvement. This move toward a general desire for moral progress looks dubious—possibly on its own, but at the very least against the backdrop of the debate about motivational externalism.

Just like the desire to do what is right, the abstract desire for moral progress seems to alienate the moral agent from the proper ends of her efforts. And it does so in an even worse way than in the case targeted by Smith's argument. The desire to do what is right at least has an enabling function for moral action: it is what moves a moral agent to act rightly who would, ex hypothesi, otherwise lack any motivational connection to the worthy objects of moral concern. The (rational) desire for moral progress, by contrast, only removes the moral agent from the worthy objects of her moral concern. The problem with its derivation is converse to the problem in the case of externalist motivation. In the context of externalism, what was supposed to be problematic was the fact that de re moral motivation to do the right thing needed to be derived from the de dicto desire to do whatever is right. In the case of a desire for moral progress, the problem seems to be that this abstract de dicto desire must be derived from more specific desires to remedy observed moral problems. In fact, thinking about how a rational desire for moral progress could be formed makes the way moral motivation is formed on the externalist picture seem praiseworthy by comparison. On the externalist account, a moral agent moves from an abstract desire toward the connection with the actually valuable objects of proper moral concern. For example, a moral agent who met a specific pig, Fred, at a sanctuary, could, given that she had a desire to do what is right, come to believe that the right thing for her to do was to campaign to end animal farming. From her desire to do the right thing she moves to a desire to "change the practices that cause Fred the pig's conspecifics' suffering." Her moral motivation is formed by moving toward the worthy goals of moral action. The desire for moral progress, in contrast, is formed by a move in precisely the opposite direction—away from the concrete moral ends. If the moral agent were to come to form a desire for moral progress based on her experience at the sanctuary, she would move from a desire to change the practices that cause suffering for Fred's conspecifics to an abstract desire for moral change. Beginning to think of the specific moral cause she faces in terms of moral progress seems to alienate her from the ends of her moral efforts.

The resulting abstract desire for moral progress is at best a) practically redundant, because the actual moral problems are there to be solved—and they have been recognized by the moral agent. They are the evidence that she has for the belief that moral progress needs to occur. At worst, the desire is b) distracting, blocking the route to actual progress; certainly, the supposed moral exemplar with a desire for moral progress will not be more likely to achieve actual improvements (ending animal farming) by focusing on the abstract notion of moral progress instead of on the concrete problems that need to be overcome. In any case, because of its original intimate link to these problems as its empirical evidence, the rational abstract desire for moral progress is c) objectionably "out of touch" with what really matters morally. Objections a) and b) represent arguments from the lack of instrumental value and from instrumental disvalue, respectively, while objection c) is concerned with moral worth—it is a complaint about the moral agent and her relation to value.

The concern about abstraction forms the first version of the fetish objection to a desire for moral progress. This version is most closely related to the original fetish argument. In defending the desire for moral progress against the argument from abstraction, the first step is to question the portrayal of the genesis of a desire to bring about moral progress. If the desire to bring about moral progress were, in actuality, formed based on a mere step of abstraction—for abstraction's sake—this would in fact look dubious. It would indeed seem to alienate the moral agent from the proper objects of her moral concern.

However, when abstraction is the effect of extrapolating from the observed cases of needs for moral improvement, the case appears to be different. When the agent, observing particular moral shortcomings, factors in a recognition of the general fallibility of human moral agents, including her own propensity to overlook morally significant situations and her own limited experience of the world inhabited by these fallible moral agents, she may well come to believe that there are other moral problems that need solving, beyond the ones she has become aware of so far. In fact, this is what a moral agent should come to believe by applying, in an especially careful way, precisely the kind of moral reasoning that is required for resolving any moral problem and figuring out what is morally relevant and what is to be done in more circumscribed situations.Whenever engaging in moral reasoning because the right course of action is not obvious, the moral agent must seek out the relevant facts. The careful, conscientious moral agent will not rest content with the assumption that any given description of a situation of moral choice is correct, complete, or makes all the relevant fact salient. She will strive to be morally alert to the possibility that some morally relevant factors might have escaped her so far. This mindset is also relevant for detecting morally significant situations in the first place. For example, because of an alertness to morally relevant facts that are not made salient in a given description, moral agents can detect the moral relevance of the question of "what to have for dinner." In attending to factors beyond the ones immediately given by the description "having chicken for dinner," they can reconceptualize the situation as "having chicken," "having a chicken," "eating a sentient being" and so on. The moral attention of alert moral agents responds to factors that were already there but not highlighted by a given description of the situation. Experiences like these can inform a rational abstract desire for moral progress: in a conscientious moral agent, experiences with formerly overlooked moral problems can give rise to a desire to confront moral problems of which one has not yet become aware. This desire is also informed by recognition of the fact that there are moral problems that one has not overlooked but never encountered.

So, the desire for moral progress in a conscientious moral agent is acquired by extrapolating from already recognized moral problems to unobserved or unrecognized needs for moral improvement. If this is what makes it defensible against the charge of fetishizing abstract moral progress, this defense rests on the fact that a rational desire for moral progress—though "self-standing"—is decidedly not "acquisition non-derivative" (Svavarsdóttir 1999, 206). What makes it defensible is the fact that it is based on moral experience and transcends, rather than merely abstracts from, this experience.

Having clarified what etiology a defensible desire moral progress must have, we can show how this—unobjectionably acquired—desire is instrumentally valuable. Two kinds of arguments from the externalism-debate are relevant here, one being more decisive than the other.

The less decisive consideration is that the desire for moral progress might well be considered a backup solution (Lillehammer 1997, 192; Olson 2002, 92) for failures of specific motivations, just as the desire to do the right thing might compensate for weak specific desires to be honest, faithful, and the like. However, whether the desire to bring about moral progress will in fact be an efficient substitute motivation is debatable. The moral agent who believed it was right for her to bring about moral progress could also "fall back on" her desire to do what is right in case her motivation to work for a particular cause ever became ineffective. It is not obvious why, in falling back on a more abstract desire, she would need to rely on the desire to bring about moral progress (rather than the desire to do the right thing).

The more decisive argument takes its cue from the claim that the desire to do the right thing has instrumental value in terms of being a transitional solution in times of changing moral beliefs (Lillehammer 1997, 191–192, Svavarsdóttir 1999, 205–206, and Carbonell 2013, 468). The idea there was that relying on an abstract desire for rightness might be forgivable or "the most we can ask" of her when the moral agent has just gone through a change of mind regarding what is in fact right. Likewise, the desire for moral progress will, with respect to certain problems, eventually be superseded by more specific desires to solve those problems. However, in the meantime, the desire for moral progress is not only a useful supplement but rather the only possible motivation for moral change. In the case of moral progress, the abstract desire is not merely a useful transitional solution but an indispensable one. Absent any idea of what the necessary moral improvements might actually turn out to be, the desire for moral progress (which entails a desire to seek out moral problems in the first place), is not merely a transitional supplement to compensate for an only slowly forming emotional commitment to a newly accepted moral end but the only conceivable solution in a situation in which the new moral end remains unknown. It provides motivation "under uncertainty," or rather, under limited familiarity with the moral problems that require solving. The moral agent operates on reasonable assumptions about the prevalence of moral problems and the typical latency of moral problems, i.e., the proneness of moral agents to overlooking them. Her uncertainty is not complete; she can reasonably assume that there are moral problems that are unknown to her—but she does not know what precisely they are.

Furthermore, the need for an abstract desire for moral progress in the sense just spelled out is permanent. Limited epistemic access to moral problems is a lasting condition. It would not be reasonable for the moral agent to assume that once she has substituted a number of specific desires for particular moral improvements for her abstract desire for moral progress, there are no undetected moral problems left so that there is no change to be covered by an abstract desire for moral progress.When a moral agent adopts new specific desires to contribute to ending modern slavery or animal experimentation, she cannot assume that there now remain no further moral problems for her to discover. The desire for moral progress is a permanent transitional solution.

At the center of the fetish objection against abstract moral desires was the concern about the moral agent's disconnectedness from the morally relevant features of things in her life. In the present case, disconnectedness is not to be seen as a problem, not only because disconnecting oneself from immediate situations is necessary for reevaluating them in the first place (thus, gathering the type of information on the basis of which a belief in the need for moral progress is formed by extrapolation), but, more importantly, being disconnected from moral problems one has not yet even encountered is not a problem that is caused by an abstract moral desire. It is an inevitable fact to which the desire to bring about moral progress is the only adequate response. The argument here is similar to a reply that has been given to the objection against utilitarianism cited earlier, namely, that it renders individuals "value-receptacles" (Regan 2004; cf. Stocker 1976). Confronting the objection as put forward by Rawls (1999, 24), Richard Yetter Chappell argues for a "token-pluralistic utilitarianism, on which each particular person's interests are (separately) accorded final value," so that utilitarian moral agents may also, on utilitarian grounds, form "particularized desires" for the particular other person's individual welfare. He discusses a situation in which, nevertheless, unparticularized desires about aggregate welfare seem acceptable—or rather, even required. This discussion is relevant to the present issue—a situation in which one does not know the particular person whose welfare is concerned.

Consider some particular unknown person, Harry. Our utilitarian cannot have a particularized desire for Harry's welfare, since she cannot even refer to Harry in particular. […] it seems that we need something like a generic desire for aggregate welfare to step in and fill the gap. […] It doesn't seem so objectionable to treat people you've never even heard of as faceless members of the aggregate. (Chappell 2015, 329)

Just as it "doesn't seem so objectionable to treat people you've never even heard of as faceless members of the aggregate," it is not objectionable to respond to one's factual disconnectedness to the bulk of moral problems in the world with an abstract desire for moral problems. And the problem of anonymity within the domain of moral concern is amplified along the morally progressive development that Singer calls the "expanding circle." The more kinds of individuals that we recognize as entitled to moral consideration, the less we can expect to be able to take them into consideration in a "particularized" way. So, in the course of one type of moral progress, the abstract desire for moral progress seems to be all the more necessary.

#### 5.3.2 Moral Progress as the Right Thing to Pursue

The defense against the objection from abstractness leads to a second problem with the desire to bring about moral progress. There is a tension between, on the one hand, the idea that the abstractness of the desire for moral progress is not a problem, because such an abstract desire is the adequate response to uncertainty about what specific states of affairs need to be ameliorated, and, on the other hand, the idea that moral progress is in fact the means which would have to be employed²⁹ in order to bring about the relevant amelioration. If we do not even know what types of changes are needed in order to bring about improvements in the world (axiological progress), how can we be confident, at the outset, that moral progress will be the way to achieve them? Once moral progress is distinguished from axiological progress in the way it has been in the preceding chapters, pursuing the former emerges as a means to achieving the latter and the question arises on what grounds we can legitimately settle for the specific means "moral progress," before we even know the end this means would have to serve (i.e., before we know what specific improvement in states of affairs needs to be brought about). From the perspective of any moral system that assigns some moral relevance to axiological value, which at least ordinary morality arguably does, this is a genuine problem. The fact that we do hear people say that they want to make the world a better place, but rarely (if ever) hear anyone say that they want to bring about moral progress, might be an expression of the underlying issue: the right thing to do might be to strive for axiological progress instead of moral progress. The desire to bring about moral progress could stem from a desire to "make the world a better place" (when making the world better is decided to be the right thing to do) and the belief that moral progress is the preferable mode in which this improvement is to be brought about. But for any particular need for improvement that might come up, it might be the case that inducing moral progress, i.e., improving (others') moral performance, could be less efficient than other ways of bringing about change for the better —and it is because of this possibility that "bringing about moral progress" comes to look likeafetishistic desire. It favors improvements in moral performance over possibly more efficient other types of change for the (morally relevant) better.

The desire for moral progress is then too specific to point us toward what is right. The agent who is motivated by it seems to be preoccupied with "advancing morality, wherever it may lead." The question she faces is: why morality? Why not the ends of morality (where these may include, but not be entirely constituted by, improvements in moral performance)? Advancing morality looks like a fetishization of the means, blocking a proper concern for the ends. Ironically, from this

Talk of moral progress as a means of improving states of affairs relies on the concept of actual moral progress. Alternatively, relying on the concept of impactful moral progress, we could speak of moral progress being the mode in which improvement in states of affairs is to be achieved.

perspective, the desire to bring about moral progress is seen as fetishistic not due to its generality but due to its specificity.

In response to this charge of fetishizing the means to the proper ends, one might of course claim that the moral value that is actualized when moral progress is made (the value of increasing right action) compensates for any potential loss in the realization of morally relevant value. However, this amounts to a mere rejection of the (consequentialist) premise of the question. The problem with the specificity of a desire for moral progress arises against the backdrop of a moral outlook that recognizes the realization of axiological value as an important moral aim. The problem is not solved by simply positing that the relative importance of the realization of moral value (e. g., the value of rightness) and nonmoral (but morally relevant) value (e. g., wellbeing) will tip the scales in favor of the realization of moral progress over axiological progress. The problem with specificity can be solved only to the extent that it can be shown that striving for moral progress is in fact a reasonable way of striving for axiological progress even from a perspective that accepts the consequentialist premise of the problem.

For those who share the consequentialist commitment to the realization of axiological value, whether the desire for moral progress is defensible—and even possibly something for which a moral agent may be especially praiseworthy—depends on the reasonability of the belief that there are undetected moral problems that a) will best be solved by improving moral performance (rather than ameliorating the effects of continued moral underperformance), i.e., there are moral problems that are to be solved by inducing moral progress. Furthermore, it also depends on the reasonability of the belief that these b) will be the most urgent or important ones among all the moral problems. And finally, it also depends on the reasonability of the belief c) that these problems are also the ones on which the moral agent will most likely have a positive impact—the ones she will be competent to help solve. The overlap between problems that meet these three conditions—and are therefore the targets of a rational desire for moral progress—is illustrated in Figure 8.

Fig. 8: Problems targeted by a rational desire for moral progress.

The concern about the specificity of the desire to generate moral progress pertains to the question of why the moral agent should not aim, instead, to facilitate the amelioration of any of the most important moral problems (cf. Figure 9), whether or not the way of accomplishing that task is to induce moral progress.

Fig. 9: Problems targeted by a rational desire for axiological progress.

In addressing the specificity concern, we must first refer back to the etiology of the rational desire for moral progress as it was characterized in the previous section. In order for the desire for moral progress to be rationally defensible, it must be based on the extrapolation from moral problems the moral agent is familiar with. Recalling the differentiation between two senses of saying that something "ought" to be the case mentioned in § 2.2.1., we can distinguish between two different ways in which a problem might be said to be a moral problem. As we noted above, moral discourse includes "ought-to-be"-judgments (Broad 1967) that present us with an "axiological ideality" (Pereboom 2014, 139) that is evaluated independently from any moral agent's capacity to bring this state about. For instance, the "ought-to-be"-judgment that "every sentient being ought to be happy" presents us with a state of affairs in which a morally relevant value is optimized but that no moral agent can bring about. We said earlier that such states might serve as ideals moral progress could aim at, but the theoretical considerations in Chapter 3 and our explication of moral progress in Chapter 4 have also established that the mere realization of such an axiologically ideal state (or a state that is closer to the ideal than a preceding one) should not count as moral progress. Moral progress involves actions (or dispositions for action) of moral agents. So, while it might be the case that there is a sense in which a state of affairs that ought not to be the case (in the ought-to-be sense of 'ought') is a moral problem in a wide sense, problems of this kind are not the ones that moral progress can directly aim to remedy. Moral progress targets directly only those kinds of problems in which moral agents are involved—that are constituted by moral agents' underperformance. Such moral problems in the narrow sense are situations in which the other types of ought-clauses apply, i.e., situations in which moral agents do what they ought not to do. These are the problems that a desire for moral progress can target—and the kinds of problems in which the extrapolation that gives rise to a rational desire to bring about moral progress is grounded. These types of problems are moral problems proper (cf. § 2.2.1.), whereas problems that are constituted by states that ought not to be can be more adequately labelled "axiological problems." Possibly, the term 'problem' itself will be seen as implying too much in terms of agential access to and power over the situation. However, I will assume that such situations of axiological disvalue can be adequately labelled problems given that moral agents have at least evaluative access to them—they can perceive them as problems.

In order for an axiological problem to become the direct target of any moral agent's effort, it must become a moral problem proper, i.e., it must become practically addressable by a moral agent. For instance, the fact that there is suffering in nature due to predation is an axiological problem. One might issue the oughtto-be-judgment that "every sentient being ought to be herbivorous and die a painless natural death at old age." If the biological condition of predation were a moral problem, then it would only have become such a problem when moral agents, who were able to diagnose the axiological problem and could strive to find solutions or ways to mitigate the problem, decided against this, e. g., based on a dismissal of nonhuman suffering (cf. Horta 2010).

How we can clearly differentiate between axiological and moral problems in a world that has very much become accessible to human agency is a worthwhile question that we cannot hope to answer here. Likewise, we cannot and need not address the question of whether we might have an obligation to turn axiological problems into moral problems (bring them within our moral jurisdiction). We simply have to note that moral problems, properly speaking, are co-constituted by the performance of moral agents and that at least any axiological problem that we recognize and for which we find a way to address it is thereby turned into a moral problem (if there is a prospect of finding a solution but we fail to try to find that solution, this failure itself becomes the moral problem).

The upshot of these considerations is that the moral agent who desires to bring about (impactful) moral progress desires to solve moral problems proper. The object of her desire is the solution of problems that are co-constituted by the way moral agents act, i.e., by moral agents' underperformance. There are three basic ways in which the moral agent who notices these problems could go about the task of solving them: (1) by inducing moral progress, (2) by inducing inadvertent rightdoing, i.e., flukes, (3) by discharging the negative effects of (what otherwise would be) continued wrongdoing. Inadvertent rightdoing, (2), could be induced, e. g., by providing economic incentives to change a morally damaging behavior. An example of discharging the negative effects of moral underperformance, (3), could be the use of technological tools such as CO<sup>2</sup> reservoirs that mitigate the effects of the use of fossil fuels. The specificity objection against the desire for moral progress is that the moral agent should desire to rely on all three of these strategies in solving moral problems. The desire she should form based on the observation of concrete moral problems is not the desire to bring about moral progress but the desire to identify and solve moral problems, not merely by inducing moral progress but also by inducing inadvertent rightdoing and "inconsequential wrongdoing."

The distinction between problem-solving options (1) and (2) is that inducing moral progress requires the presence, inducing inadvertent rightdoing the absence of moral awareness. The distinction between both of these options and option (3) is yet different. Discharging the negative effects of continued behavior (3) may well be seen as basically changing the nature of the action. Rather than inducing "inconsequential wrongdoing," the moral agent would be asked to change what an apt description of the behavior in question would be, where this description would not mention any moral harms at all anymore: the action "use of fossil fuels" has different apt descriptions when emissions are no longer an issue than when they are still a problem.

In response to the specificity objection, we must consider what type of moral progress would be the object of the moral agent's desire. Apparently, the desire for moral progress ought to be the desire for impactful moral progress. This could be impactful actualized (i.e., dispositional) moral progress or (mere) actual moral progress with an impact. The specificity objection has some force against the first, the more ambitious version of the desire, the object of which is a practical improvement routed in significant improvements in the moral agent's overall motivation. However, the objection loses plausibility if we think of the relevant desire as a desire for merely actual moral progress with an impact, i.e., axiological progress that is brought about through change in practices, which is accompanied by merely some minimal level of moral awareness. When minimal moral awareness is present, any practical progress will not count as a fluke resulting in inadvertent rightdoing, even when better behavior depends decisively on (non-moral) incentives. Any positive change in practices in the presence of moral awareness counts as actual moral progress and, given it has a positive effect, as impactful progress. What the objector seems to suggest is that the moral agent ought to replace—or rather complement—her desire for moral progress with a desire for practical improvement without moral awareness. However, unless the moral agent has reasons to suspect that the presence of moral awareness will actually hinder practical solutions—unless she must expect her fellow moral agents to be "moral subversives, people who intentionally and knowingly pursue what they acknowledge to be morally wrong or bad, and do so for that very reason" (Svavarsdóttir 1999, 161)—it is hard to see why she should not wish for the presence of moral awareness.

The desire to generate impactful actualized dispositional moral progress seems to be the desire for the morally ideal solution of moral problems proper. It is ideal in that it strives for the realization of what is morally desirable and has some moral worth. This ideal solution also has instrumental value, because it is the solution that improves moral agents' disposition beyond a threshold so that they are prepared for performing well in different scenarios (with or without incentives, with or without advantageous conditions for good moral performance). The desire for impactful moral progress in the actual, not actualized, sense is the next best desire in terms of the moral quality of the aimed at solution. Creating and preserving moral awareness is hardly an objective worthy of criticism. On the contrary, raising moral awareness is important for moving other moral agents to contribute to those kinds of solutions to moral problems which pursue the strategy of discharging negative effects or inducing genuine flukes.

The availability of the problem-solving option of discharging the negative effects of continued behavior cannot discredit the desire for moral progress because of the way in which moral and technical solutions to moral progress are interrelated. Even morally non-ideal, merely technical solutions to moral problems require some moral agents to be aware of the problems and to be willing to solve them. In any given case in which technical solutions outperform moral progress, some prior moral progress is still required—in the ones developing those technical solutions. Technical approaches such as the development of storage technologies for CO<sup>2</sup> are not something that any individual moral agent could simply pull off on her own. Some of the developers of fleshless meat alternatives might be examples of moral agents who have been recruited by promoters of moral progress to solve a moral problem in a (merely) technical way: their technological contribution trails in the wake of the moral progress they have been induced to make by the moral arguments of animal advocates. The moral progress of some moral agents can be a necessary condition for pursuing the technical solution of moral problems in the absence of moral progress on a larger scale.

Against this backdrop, reconsider the overlap that Figure 8 shows between problems that the moral agent is competent to help solve with problems that are best solved by moral progress. The problems within this overlap initially appeared to be the more worthy objects of a general abstract moral desire. The desirability of their solution seemed to discredit the desire for moral progress as too specific. However, in thinking about problems that are "best solved by bringing about moral progress" we must not think exclusively of problems that are conclusively solved by moral progress alone—and neither should the moral agent with a desire for moral progress. Rather, she should be concerned with those problems the solution of which requires moral progress as a non-sufficient, but necessary condition. In the case of problems which are most efficiently solved by employing technical solutions (to discharge negative effects), moral progress is still—in most cases one could think of—a necessary condition in that at least some other moral agents need to be moved to wanting to contribute to the technical solutions. By raising awareness, the moral agent in effect expands the realm of problems she is competent to help solve, because she turns others into potential contributors to solutions that she could not fashion herself.

Thus, the desire for moral progress is not fetishistic if it is a desire for impactful actualized moral progress as the ideal mode of the solution to moral problems and if the moral agent with this desire stays alert to the possibility that she might need to downgrade her desires for specific changes to desires for impactful moral progress in the non-actualized sense once she has become familiar with the previously unknown moral problems that were covered by her abstract desire for moral progress.

The objections from abstractness and specificity motivate a clarification in the form of a rationality constraint on a desire for moral progress. The constraint requires that a rational desire for moral progress (a) stems from a justified belief in the need for improvement in moral performance, where (b) this belief is to be based on a process of reasonable extrapolation from known moral problems to the assumption of the existence of unknown moral problems.

Condition (a) expresses the basic requirement of rationality:adesire for moral progress ought to be based on a justified belief that moral progress is the right thing to pursue. The phrase "need for improvement in moral performance" is supposed to encode the moral requiredness of moral progress, which presupposes that moral progress is in fact both possible and the right means to remedy existing moral problems. Condition (a) thus comprises the conditions of reasonability of the moral agent's belief, the practical possibility of moral progress as well as its suitability or appropriateness for realizing the ends of morality. So, it answers to both the abstraction concern and the specificity concern.

Condition (a) on its own, however, does not fully address the objection from abstraction. It does not yet ensure that the desire for moral progress has the right etiology. The original worry about a desire for moral progress being abstract was that if it was based on mere abstraction from desires for particular changes, it would be redundant or distracting and, above all, fetishistic, because it would be out of touch with real moral issues. The problem of epistemic justification addressed in condition (a) does not yet answer this concern. The belief that progress (in general) is required would be perfectly justified if it were the result of a mere abstraction from particular justified desires for specific moral changes. What is needed to address the core of the fetish concern is a requirement of substantiality that applies to the belief in the need for moral progress on which the desire for it is to be based. Condition (b) accommodates this requirement. It can in fact be seen as a concession to the objector who is not convinced by the refutation of the abstraction concern which borrowed arguments from the original fetishism-debate. Condition (b) is only required if one believes that the instrumental value the abstract desire for moral progress has as a permanent transitional solution or the fact that it may figure in a situation of motivational overdetermination on their own do not diffuse the worry about abstraction. If the fetish concern about a general desire for moral progress is not dismissed on similar grounds as the fetish argument against externalist motivation, the more comprehensive rationality constraint—which includes (b)—can be invoked to specify how the moral agent should come to desire moral progress in order to avoid the fetish objection. The entire rationality constraint then continues to relate to both the epistemic and the moral defensibility of the desire for moral progress.

#### 5.3.3 Settling for Progress

So, the desire for moral progress can be defended against objections relating to its abstractness and specificity. It can be shown to be a rational desire for a moral agent to have. The question remains whether a moral agent with this desire could even be considered praiseworthy in virtue of it. One might want to deny this on the grounds that the desire for mere progress is objectionably modest. Absent any specification of the amount of improvement that is desired, any degree of improvement whatsoever could count as the objective of our supposed moral exemplar who is moved by a desire for moral progress. This cannot be right—it would be objectionably unambitious.

In fact, aiming for mere progress—whatever threshold might be added concerning the degree of improvement—might look objectionable anyway. It seems that what is actually (morally and rationally) desirable is the solution of problems, the elimination of preventable suffering, injustice and the like—not their mere incremental amelioration. When trying to bring about any morally relevant change, settling for mere betterment instead of resolutions seems objectionably relaxed. So, the third version of the fetish objection is that the desire for moral progress has an undeserving object in that it aims too low. It fetishizes mere change for the better, where change for the best is what is really desirable.

To avoid the charge of objectionable modesty, the rationality constraint on a desire for moral progress must be augmented by an optimality clause: a defensible desire for moral progress is a desire for the greatest possible progress.

The optimality clause itself is not merely a moral addendum to the rationality constraint, added to ensure that the moral agent may be considered praiseworthy in virtue of her desire for moral progress. It is itself part of the rationality constraint.What is morally desirable is not in and of itself rationally desirable for the moral agent. There are different modes of being morally desirable—expressed in different types of statements about what ought to be the case. The solution of moral problems, the eradication of injustice and suffering is morally desirable in one sense but not in a sense that makes it rationally desirable for the moral agent as her moral aim. The moral agent may well think that "injustice ought to be eradicated," but the "ought" in this thought cannot be an "oughtto-do" that may give rise to the thought that it ought to be her, the moral agent, who should take on the task of eradicating injustice. Utopian ideals like these are, by definition, not proper practical aims. With respect to particular problems, it may well be in the moral agent's power to solve them by effecting moral progress (though even these occasions might be rare). These cases are covered by a desire for the greatest possible progress, because the greatest possible progress, in any case, might actually be full success. However, on the abstract level on which a desire for moral progress operates, whose scope is the entirety of moral issues, only progress is rationally desirable. The moral agent cannot rationally hope to solve all moral problems conclusively.

In response, it might now be objected that the optimality clause, in fact, requires too much. There could, for instance, be satisficing accounts of justifiable desires for moral progress as there are satisficing accounts of acting rightly. Requiring that the moral agent desire "the greatest possible progress" in order to be praiseworthy (or even have a rational desire) at all might therefore be too demanding and possibly even indicative of an implicit commitment to an even more demanding and restrictive concept of moral progress: one that requires a maximizing account for the goodness that is implied in the definition of progress as a special kind of "change for the better." This would result in a more exclusive concept of moral progress: with this concept, the recognition of any progress as progress would be restricted to changes that manifest the respectively greatest possible progress (where this possibility is relative to the agent who effects the change). An agent who effected only minor change would then not even be credited with having effected moral progress at all. The objection to the optimality clause here is not that there is no problem with complacency in an unspecified desire for moral progress whatsoever—the objector can still require some measure of sufficient progress to be a worthy object of the moral agent's desire, but she would want to lower the bar. The idea is that there need not be a problem with aiming only for some significant, even if it is not the greatest possible improvement.

Depending on what exactly underlies the objection, the optimality clause could be made to work for the objector as well. Insofar as the general sentiment underlying the objection is a commitment to safeguarding the moral agent's freedom to pursue other goals than the greatest possible progress, this sentiment can be accommodated by the optimality clause.While the formulation as an optimality clause might not lend itself to a satisficing construal of a justifiable amount of progress as the object of the moral agent's desire, it still lends itself to construals of the desirable amount of progress that take the moral agent's commitment to other projects besides the realization of moral progress into account. The question is about the kind of possibility to which the optimality clause speaks. On the reading targeted by the objection, it is bare practical possibility. The moral agent is supposed to desire the most progress she could actually bring about. The objector could introduce a more constrained notion of possibility. A substantive theory of moral progress could specify that the greatest possible progress (to figure in a desire for moral progress) would be the greatest progress compatible with the moral agent's pursuit of various personal non-moral projects.

Also, one could take the concern about optimality on board by weakening the role of optimality and employing it only as a sufficient condition of a justifiable desire for moral progress. Those who want to allow for desires for less than greatest possible progress could then specify alternative conditions that cover measures of sufficient amounts of progress in a rational desire for moral progress.

Basically, the question of whether the moral agent needs to aim for maximally achievable progress in order to be praiseworthy for her motivation is a substantive normative question that simultaneously touches upon the nature of morality and how it is constrained by other considerations, an isssue on which there is reasonable disagreement. Ideally, we should avoid excluding any view at this point. Even if we therefore abandoned optimality as a threshold, at the very least, it should be more agreeable that the more change the agent desires to bring about (within the limits of what she can reasonably expect to achieve), the more praiseworthy she can be in virtue of this desire (as far as the content of the desire, not its motivational force or anything else about it is concerned). So, the optimality clause at least specifies what would be the best desire for moral progress (in terms of its content) the moral agent could have. Read in this way, it states a debatable desideratum, not a non-negotiable requirement.

The concerns about optimality point us toward a final amendment to the rationality constraint: an efficiency clause. The desire to bring about whatever little change should happen to be effected by whatever great effort is being made seems to be, after all, at least borderline irrational. So, an efficiency clause is needed to ensure that the change the moral agent wishes to effect is the greatest possible progress as determined by the circumstances in which it is to be brought about and the amount of effort that the moral agent is prepared to invest. Irrespective of whether or not the moral agent's desire for moral progress licenses her to pursue various non-moral projects that detract significantly from the amount of effort with which the project of moral progress is pursued, for whatever amount of work she is prepared to put into the pursuit of moral progress, she ought to expect the greatest possible effects. So, the optimality clause can at the very least be read as an efficiency clause.

#### 5.4 Motivational Relevance Reconsidered

In responding to the three prongs of the fetish objection, we have established several conditions on a rational desire for moral progress: the rationality constraint covered the reasonableness of the belief in the requiredness of improvement—and thereby the possibility and appropriateness of moral progress—and was augmented by a requirement of efficiency. We have also added the desideratum of optimality. We have thereby argued that the desire for moral progress that meets these conditions is not inherently flawed—it does not make progress into a fetish—and we have taken steps toward explaining how it may be praiseworthy. In the course of the argument, it was claimed that an abstract desire for moral progress is instrumentally valuable and—as a specific kind of desire to "make the world a better place"—even indispensable. In this final section, the following remaining questions shall be addressed. (1) What could the practical relevance of a desire for moral progress be, when moral agents are more generally motivated to do the right thing? (2) How is the pursuit of moral progress even conceivable? If only specific causes can ever be practically pursued—what is the practical role of a desire for moral progress "as such?" (3) What would a moral agent be like who actually adopted an abstract de dicto desire for moral progress? Why should she be considered a moral exemplar in virtue of this desire?

1) Is the desire to do what is right enough, after all?—In defense of the desire to do the right thing, it has been argued that this desire provides the moral agent with the motivation to be "especially morally diligent," i.e., "extra scrupulous," where this is supposed to mean that an agent would not act "unreflectively" on any given de re desire for what seems to her to be the right thing to do (Carbonell 2013, 466). The de dicto desire to do what is right is also said to be "helpful" when the agent does "not (yet) know exactly what is right":

For example, suppose that I want to do what is right with respect to the question of whether to raise the minimum wage. Suppose also that I have learned that prominent economists who care about helping the poor disagree about whether raising the minimum wage is morally good. Since I don't know what course of action is right with respect to this question, I don't have any direct motivation one way or the other, but nonetheless I have a standing motivation to do whatever happens to be right in the end. This is not the same as simply being indifferent about what to do. Whereas the indifferent person might not care which way she votes on a ballot measure to raise the minimum wage, the person with the standing de dicto motivation to do what is right cares a great deal. Indeed, it is because she cares about doing the right thing that she will seek out information and deliberate about how to vote. Once her epistemic gap has been closed and she has arrived at a judgment, the de dicto motivation will move her to act even if she has not yet acquired an original desire to vote one way or the other. (Carbonell 2013, 471, emphasis in original)

The very last sentence articulates the—presumably uncontroversial—role of the de dicto desire as an "epistemic stopgap" (Carbonell 2013, 471) or transitional solution mentioned above (§ 5.2.). The preceding passage, however, portrays a very different role for the desire to do what is right. On the view articulated there, the de dicto desire to do what is right moves the agent to improve her epistemic situation in order to be able to act rightly.³⁰ In effect, the de dicto desire to do what

In a similar vein, Ron Aboodi argues that de dicto moral motivation is required in situations

is right is thought to be enough to make the agent strive to make ethical progress with respect to a particular moral problem at hand, thus, facilitating better action. Notably, there is no mention of a desire to make moral progress with respect to the problem in such situations. Rather, it is assumed that merely desiring to do what is right will be enough to induce moral (self‐)improvement—the only alternative being a state of indifference.

Yet, other than undertaking efforts to improve her moral outlook or remaining indifferent, the moral agent could also retain her desire to do the right thing and remain puzzled as to what the right thing is. Similarly, the moral agent who has already favored a course of action need not reconsider it even if she consciously reconsults her desire to do what is right.

"Desiring to do what is right" is ambiguous. As Lillehammer observes, it is ambiguous between "desiring to do what is [in fact] right" and "desiring to do what one believes to be right." He discusses this with respect to the de re/de dicto distinction, pointing out that one can only desire de re what one believes to be right (Lillehammer 1997, 189–190). We can, however, also take his observation as putting a question mark behind the idea that a desire to do what is right suffices to achieve moral progress. What marks the "extra scrupulous" moral agent is an awareness that what she believes to be right may not coincide with what is in fact right. The fact that the moral agent has a de dicto desire to do what is in fact right tells us nothing about whether she has such an awareness—nothing about the carefulness with which the so motivated agent determines what is the right thing to do on any given occasion. Similarly, the moral agent who is puzzled as to what course of action to choose may or may not go on to take further steps to resolve the issue, even while holding on to her desire to do what is right in each case. But if she does go on to take these steps and in case she does so ready not merely to collect factual information but also to reconsider her moral stance (i.e., to "seek out information and deliberate," Carbonell 2013, 471, emphasis added), the desire she will have derived from her desire to do what is right will be the one to improve her moral outlook, i.e., to make ethical progress in order to facilitate better moral action. She will thus have acted on a desire to make (individual) dispositional moral progress.

The moral agent who wholeheartedly cares about doing what is in fact right and is aware of her own fallibility with respect to identifying the right thing to do will desire to make moral progress in order to do what is right. The fact that she wholeheartedly cares and is self-conscious with respect to her fallibility is an ad-

of uncertainty for pointing the moral agent to undertake a moral "inquiry" in order to figure out the morally best course of action (Aboodi 2017, § 1).

ditional fact about her psychology that is not built into the general desire to do "what is right." This additional psychological condition may not often or even typically condense into a conscious (de dicto) desire to make moral progress, but in substance, the "extra scrupulous" or conscientious moral agent is motivated do to precisely that. The mere desire to do what is right cannot, on its own, stand in for this type of motivation. Redescribing the mindset of the moral agent with reference to "moral progress" can bring out a praiseworthy quality and thus be fruitful. More so than in the case of the extra diligent agent who secondguesses a moral choice or in that of the puzzled moral agent facing a limited number of options (Carbonell 2013), this is evident in the case of the moral agent who is not facing a choice between several courses of action but who is nevertheless clueless concerning how to generate moral progress with respect to a particular moral problem.

In this kind of situation, a de dicto desire to do whatever furthers moral progress with respect to that issue is what the moral agent should have. Compare the motivational setup of an ethical vegan with respect to animal suffering in agriculture with her motives as far as animal suffering in nature is concerned. She may desire to bring about moral progress with respect to the problem of animal suffering in agriculture and have determined that in order to do so she must campaign to spread sentientism and veganism. She need not rely on an abstract desire for moral progress, nor even on the desire for progress with respect to the issue, but can be moved by her concrete goals with respect to the specific change in moral views (acceptance of sentientism) and behavior (veganism) she would like to induce. In comparison, she might be genuinely clueless concerning what to do to effect improvements with respect to unalleviated suffering of nonhuman animals in nature. She might therefore come to the preliminary conclusion that the best thing for her to do is to advocate for progress.Arecommendation to initiate theoretical progress in order to facilitate moral progress with respect to this issue is exemplified in the following statement:

Promoting debate on this issue, doing research on it and questioning speciesism appear to be the most important ways in which we all can work today in order to reduce the immense amount of suffering and death that exists in the world. (Horta 2010a, 88)

The recommendation to "question speciesism" is quite substantively morally committed as it relates to the pre-identified problem that suffering of animals in nature is often illegitimately dismissed as a moral challenge for humans in the first place due to bias against nonhumans (Horta 2010a). However, the other two recommendations—to promote debate and to conduct research—urge epistemic and theoretical progress with respect to the moral problem in order to enable moral agents to take moral action in the first place. Now, in substance, this amounts to the recommendation to make open-ended ethical progress in order to facilitate moral progress. In the absence of a more practical approach to solving the problem—or even a conclusive description of what the moral problem consists in—, what the moral agent ought to want to induce is moral progress, whatever that may be, with regard to the problem at hand, and she must do so first by initiating ethical progress.

The virtue of this type of response to an unresolved moral problem is its genuine openness as far as the solution is concerned—which reveals, one could say, a genuine, unprejudiced concern about the problem. The more the moral agent is aware of the fact that in order to find the unknown solutions to roughly identified problems she needs to make theoretical efforts first or the more a moral agent is aware of the need to reconsider her own previous judgments or undertake further consideration to come to a decision—and the more she is committed to doing all of that—the more we have reason to attribute to her a desire to make moral progress which complements the desire to do what is right. Consequently, we can find the moral agent praiseworthy for her commitment to finding the really right solution—or, with a different emphasis, finding the right solution.

The less we know about how to confront a moral problem—and possibly even how to frame it exactly—the more all we can do to address it is to seek to make progress in understanding the problem and finding possible solutions. Insofar as this theoretical progress is sought in order to facilitate better moral action, the underlying motivation is, in substance, to make moral progress. The desire for moral progress emerges as a moral faculty to cope with a type of moral ignorance. The mere desire to do what is right does not suffice.

2) How does one "pursue moral progress?"—The preceding discussion has already brought us closer to answering our second question, which is how we can even conceive of the desire for moral progress to become motivationally effective. The answer provided here is that moral progress is pursued via the pursuit of individual ethical progress.

Even those who argue that the idea of moral progress is motivationally relevant tend to think of it as relevant short of being action-guiding. As mentioned at the beginning of this chapter, Moody-Adams, for instance, characterizes the idea of moral progress as "a necessary presupposition of action for beings like us." Yet, at the same time, she also insists that it does not provide action-guidance (Moody-Adams 2017, 153). If this were so, the idea that the concept of moral progress could not only be relevant to a belief in the possibility of improvement, but could also furnish a relevant moral desire, would be absurd, because such a desire would be condemned to be idle. No moral agent, it would seem, could ever act on a desire for moral progress. But the preceding discussion has shown that this claim is much too strong. What the moral agent with a de dicto desire for moral progress certainly cannot do—based on this desire alone—is take measures to initiate the practical changes that will ultimately qualify as (the completion of) moral progress. Her idea of moral progress being abstract, she does not yet know what types of changes these would be. Once she knows what specific changes ought to be generated, however, it might not (or not only) be her desire for moral progress driving her to induce those changes but rather her newly acquired desire for those changes themselves. However, we have seen that this does not mean that a desire for moral progress could not prompt the agent to do anything at all. In order to be able to transition to the pursuit of practical changes, she must first find out which changes are required. Desiring improvement—in whatsoever change it may consist—first and foremost prompts the moral agent to gather and evaluate information as to what would in fact constitute improvement. The abstract desire to make the world a better place by improving moral performance can only be actively pursued by aiming to learn what it is that would actually make the world better and what concrete behavioral changes are needed to realize it. The only thing that the moral agent who desires to bring about moral progress can immediately pursue is ethical progress. Only based on these improved views can she pursue change in practices, but once she has arrived at this point, she has moved outside the realm of moral action for which an abstract desire for moral progress is the only or major driving force.

This result does not contradict the earlier rejection of an (over‐)intellectualized picture of what it is to make moral progress. The relatively weak theory-requirement for moral progress allows for moral progress to consist in "mere" changes in practices—provided the moral agent is minimally morally aware of the valence of that change. For moral progress to occur, not much is required agentially on the theoretical side of morality. This view of the sufficient conditions of moral progress is not in conflict with the view of what is involved in the pursuit of moral progress. While moral progress can occur where the progressing agent is not progressing in her moral views, acting on a de dicto desire for moral progress is to pursue ethical progress (first).While moral progress need not have to be pursued in this way (i.e., it need not involve ethical progress) in order to come about, when it is pursued, it does involve ethical progress.

3) What kind of moral agent is the moral exemplar with a desire for moral progress?—We now have an understanding of how the idea of moral progress could feature in the moral agent's motivation as the object of a moral desire. The question remains whether the moral agent who actually has this kind of motivation could be viewed as a moral exemplar in virtue of having this motivation (leaving aside her actual efforts to initiate change and their success).What would this moral agent be like? How would we know that someone was motivated in this way? How would the desire for moral progress reveal itself? Again: any specific efforts the moral agent ends up undertaking to bring about specific changes only reveal determination to effect those specific changes. How do we tell the moral agent who supports some specific "progressive" cause from the moral saint (if she were rightly so conceived) who desires moral progress as such i.e., moral progress, in whatsoever change it may consist?

In dealing with the abstraction version of the fetish objection to the desire for moral progress, we have already gestured at a characterization of the mindset of the moral agent with a desire for moral progress. The desire to bring about moral progress would be the desire to identify and solve problems of which one has not yet even become aware. It calls for making progress first with respect to moral views and more basically yet, with respect to moral awareness. We said that the moral agent who will be able to develop a belief in the need for (and a desire for) moral progress maintains a critical stance even when she does not seem to be confronted with immediate situations of moral choice.While thinking of situations in the circumscribed terms any completed description of them offers is necessary to make moral decisions manageable, for the careful moral agent, this necessary simplification comes with a readiness to evaluate. No description of a given situation of moral choice may ever be complete—preconditions and downstream effects of any course of action can only be considered up to a certain point. Our supposed moral exemplar with a desire for moral progress would be one who not only retains an awareness of this fact, but also remains ready to look beyond the individual case. In this way, she proceeds from viewing her task as a moral agent as doing the right thing in any situation that presents itself to viewing it as that of seeking out morally significant situations that do not "present themselves"—based on an experience of the latency of moral problems and on the recognition of limited access even to situations where moral relevance would be salient, once encountered. She proceeds from viewing herself as charged with acting rightly in particular circumscribed situations to viewing herself as morally challenged even once an obvious situation of moral choice has been handled.

A desire for moral progress prompts the moral agent to look beneath the surface of apparently benign situations, to eschew the silencing of doubts about normal-seeming practices. It requires cultivating a critical mindset and the courage to seek out—rather than avoid—problems—in order to solve them. A moral agent with a desire for moral progress is one who is genuinely open-minded and conscientious—someone who acts based on an attitude that is the opposite of mere, arbitrary solidarity with a cause she has happened to have come across. An abstract desire for moral progress rather equips the moral agent with a unique resource for critical thinking, since it prompts her to conceptualize her own moral awareness as a resource that has to be carefully allocated, based on the recognition that the moral issues she has happened to have encountered might not be the only, nor the most important ones. The moral agent with a desire for moral progress might eventually, given the right circumstances, be distinguishable from the moral agent with merely specific desires to further particular ends in the following way: by her responsiveness to evidence of other problems than the ones with which she is already familiar. The praiseworthiness of a desire for moral progress thus attaches to the alertness and courage to seek out moral problems in which it manifests.

There is a counterpart to the concern about modesty discussed above that needs to be considered when evaluating the moral agent's praiseworthiness for being motivated to bring about moral progress: this motivation might be dangerously ambitious. Not every "success of morality" must take the form of something that triggers intuitions of "progress"—maintaining the status quo (preventing worse things from happening) might often be equally as important or the only thing for which we can hope. These desirable accomplishments are presented in a negative light when we give special praise to the moral agent who wants to improve the state of the world rather than prevent it from deteriorating.

The rationality constraint spelled out in the preceding section deals with this concern about moral zeal. The desire for moral progress will only be rational, and could therefore only possibly render the moral agent praiseworthy, if it is based on the justified belief that initiating moral progress really is what is being called for.Amoral agent who neglects her moral obligations in the pursuit of moral progress will not count as performing morally well. A desire for moral progress that obeys the rationality constraint, in contrast, looks like something in virtue of which the moral agent would well be morally praiseworthy. The ethical progress that is pursued based on this desire also enables a better-informed evaluation of where currently recognized obligations rank among the agent's moral tasks (if they continue to be obligations at all). The moral agent with a desire for moral progress does not exhibit hubris but rather a humble recognition of her own moral ignorance and a readiness to overcome it.

An agent with a mindset that facilitates a desire for moral progress is someone who is not prone to what we might call "affected pessimism." Affected pessimism is something like the willfully adopted version of the moral agent's learned helplessness (the failure to recognize and act on the possibility to avoid an averse stimulus, based on past experiences of a lack of control). An affected pessimist will insist that there is no point in trying to "make the world a better place" despite evidence that this is not impossible. She will fail to form a rational desire to effect change for the better. One type of affected pessimism would be what Peter Unger calls "futility thinking," i.e., the idea that one's contribution to the solution of a problem will be so negligibly small that it is entirely futile and therefore, need not be made (Unger 1996, 75–76). One could also be an affected pessimist by making oneself believe that more powerful agents will purposefully and successfully thwart one's plans to effect change for the better, that there are forces that necessitate some negative development for any positive one etc.With respect to moral progress as such, the affected pessimist may insist that people simply cannot change for the better. All these are strategies to fend off obligations—or merely to waive options—to resolve moral problems, adopted in view of the demandingness of these tasks. Like learned helplessness, affected pessimism is a manifestation of the failure to recognize options for exercising agency, and like "affected ignorance" (discussed as a hindrance for moral progress, for example, in Moody-Adams 1994 and Wieland 2017), it is adopted in a self-serving manner, functioning as a rationalization for passivity. A rational desire for moral progress is the polar opposite of this morally impeding mindset.

## Conclusion

We are all aware of many different ways in which the world could be a better place.We even know the solutions to some of the most pressing problems caused by our own conduct and we struggle with the difficulty of putting this knowledge into practice. In other cases, we have yet to find out how we should remedy a bad situation. And we can be certain that our moral learning process will continue to reveal moral issues to us that we have hitherto not even identified as such. We know we need to improve our moral outlook and our actions, and we pride ourselves with having achieved such improvements in the past. Because we are imperfect, fallible morally aware beings, opportunities for moral improvement abound. Because of our impact on a world inhabited by morally considerable beings, how we do morally matters a great deal for the state of the world—even if only during a small slice of its history. Insofar as we care about all this, we care about moral progress.

Despite the common interest of many moral agents in achieving—both individually and collectively—particular improvements, they rarely talk about "moral progress." One reason for this—among others—might be that it is not so clear what "moral progress" is supposed to be. Clearly, not just any change for the better counts as moral progress. Upon identifying any (aimed at or achieved) change for the better, we can meaningfully ask: "This is an improvement but is it moral progress?" Sometimes the question of what could count as moral progress is addressed by giving paradigmatic examples of moral progress or listing criteria for what would make a development a change for the better. The preceding investigation took a different route. It was based on the idea that the identification of instances of moral progress should follow the statement of a clear account of the very concept of moral progress. The answer to the question of what we should take moral progress to be was given in the form of a threefold explication differentiating dispositional, actual and impactful moral progress improvements in moral agents' propensity to act morally well, their actions, and practical changes that succeed in influencing the world in a positive way.

Aside from uncertainty about what 'moral progress' actually means, we also looked at another potential reason for why there is comparatively little talk of "moral progress" in moral discourse. It might be thought that—given its abstractness—moral progress as such is not a proper goal for moral agents to pursue and thus, at best irrelevant to moral motivation. In opposition to this assessment, it has been argued here that a concern for moral progress—in whatsoever change it may consist—can be shown to be a hallmark of the conscientious, morally alert moral agent. The concept of moral progress constitutes a way of thinking about moral change wherein the moral agent takes her own uncertainty and fallibility into account and that prompts her to seek out opportunities to improve her own moral outlook.

The initial step in addressing the question about the very concept of moral progress was to select the mode in which any definition of moral progress would be offered. It was clarified that when talking about the concept of moral progress, the subject matter was taken to be a way of thinking about moral change expressed via use of the term 'moral progress.' The definitions offered here were proposed in the spirit of offering an explication rather than a conceptual analysis of moral progress: a proposal of what it would be useful to take moral progress to be rather than the discovery of a conceptual truth or an elucidation of a preexisting, shared tacit understanding.

However, certain conceptual commitments presupposed in this endeavor had to be clarified first. Chapter 2 distinguished among different notions of progress and marked transformative change and improvement as particularly relevant to the topic at hand. As far as the component moral in the compound moral progress is concerned, some clarifications of the presuppositions about morality that went into addressing this concept had to be laid out. The very capacity to improve morally is sometimes seen as a condition of morality. In opposition to this demanding view of moral agency, morality was characterized here in terms of moral awareness and moral considerability as complimentary preconditions of moral action. Morality takes hold where morally aware beings coexist with morally considerable beings. The resulting picture of morality establishes a low threshold for moral agency but does not forestall any decisions about the requirements of moral progress.

Having made these methodological and conceptual determinations, we then considered how the characteristic perspectives of major theories in normative ethics could inform the ensuing explication despite the fact—or rather, in view of the fact—that the account of moral progress to be proposed here was not to be deduced from any one of these theories. The discussion of consequentialism and consequentialist appeals to moral progress allowed us to establish a first restriction on our concept of moral progress: based on a critical consideration of global consequentialism as well as progressive consequentialism, we determined that moral progress had to be distinguished from progress with respect to the realization of axiological value by locating it in the realm of moral agency rather than identifying it as the result of moral action or as change in states of affairs more generally. The consideration of Kantian views on the idea of moral progress showed that this restriction could be built on with respect to the distinction of moral desirability and moral worth. The emerging question to be borne in mind in our explication concerned how comprehensive moral progress-judgments are thought to be—whether they include an evaluation of the moral agent's motives or not. The perspective of virtue ethics was considered but rejected as a supposedly natural approach to the concept of moral progress, although it was recognized that moral progress could certainly receive an explication in terms of a meta-virtue concept in the less specific sense of virtue theory (while it was not considered obvious at the outset that moral progress would have to be viewed this way, one of the explicata ultimately proposed here was indeed given in terms of a change in dispositions). Broadening our perspective beyond normative ethics toward classical political philosophy proved helpful to understand the potential inherent evaluative conflicts involved in moral progress-judgments. John Stuart Mill's remarks on the inherent tension within intellectual progress (in terms of the proliferation of truth) translated into the paradoxical implications of a potential requirement of theoretical—i.e., ethical progress as an element of moral progress. Whenever the moral agent's performance in both realms, the realm of ethical beliefs and the realm of moral action, is taken into account in evaluation moral progress, there is room for evaluative conflict. And following Mill, we might fear that whenever better actions become more uniformly endorsed, there might be an accommodating regress in the realm of moral beliefs, in that moral agents lose their grasp of the moral reasons for their morally better action. The threefold explication of moral progress that has been given here will leave the weighting of these potential inherent setbacks in certain stages of moral progress to a more comprehensive theory of moral progress. The final part of Chapter 3 addressed the supposed metaphysical baggage of the idea of moral progress—the alleged tie-up between moral progress and moral realism. Neither the (dim) prospects of a transcendental argument from progress to realism nor the existence of abductive arguments from progress support the idea that the concept of moral progress has such a special liability to a realist perspective that its explication would have to make reference to the metaphysical underpinnings of moral progress. Taking into consideration the arguments made against an appeal to metaphysical commitments in moral discourse in general, we moved on to an explication of moral progress that would not incorporate any realist commitment into the concept.

The explication proceeded from the idea that the rough distinction between a theoretical (psychological or ethical in the wide sense) and a practical (behavioral) side of morality would give a useful structure to the investigation into the basic conceptual components of moral progress. It was argued that even though only moral agents can be said to make moral progress, it is not a plausible requirement for moral progress that the agent must be actively ethically involved in accomplishing moral progress in the way of pursuing, initiating or endorsing her own improvement as an improvement, i.e., there is no agencyrequirement of this sort for moral progress. In search of the necessary conditions of moral progress in terms of ethical and practical progress the idea was elaborated that moral progress cannot be made "in theory" or by a fluke, i.e., with no moral insight whatsoever, and it was recognized that outward circumstances condition the ways in which moral agents can succeed as moral agents. Against this backdrop, the distinction between three concepts of moral progress serves to recognize different types of moral accomplishments. Dispositional moral progress is the improvement in a moral agent's moral views that significantly increases her disposition to perform well morally in sufficiently probable circumstances. The concept of dispositional moral progress can be relied on to acknowledge significant changes in moral agents' propensity to act morally well that enable them to do morally better in varying circumstances, while this propensity can also—depending on circumstances—remain unactualized. The concept dispositional moral progress answers to the need to differentiate moral progress from mere ethical progress, i.e., merely theoretical progress, by establishing a threshold for the motivational impact that any change in a moral agent's ethical outlook must have in order to count toward moral progress. At the same time, it addresses the need to account for the fact that outward circumstances may fail to give the moral agent occasions to actualize her capacity to do better.

Actual moral progress, in contrast, is the improvement in the moral agent's moral performance over a certain period of time. The concept of actual moral progress serves to highlight improvement in practices that are accompanied by an according improvement in moral views or that bring the agent's practices into greater harmony with a preexisting moral view. The threshold for the required ethical involvement in actual moral progress was deliberately set low: actual moral progress requires merely a minimal moral awareness of the new practices' moral valence. This requirement serves to distinguish moral progress from an improvement in practices that results in inadvertent rightdoing. One surprising result of the investigation was that although there is a strong commitment to requiring moral progress to be practically relevant, theoretical progress can be viewed in one sense as superior to practical change—when it has sufficient impact on the moral agent's dispositions. When a moral agent adopts a new practice while knowing that this is what she should do, but only manages to do so thanks to more advantageous conditions, we may say that she makes merely actual moral progress. Attributions of moral progress based on observed moral performance should be seen as ambiguous between attributing actual moral progress and actualized (dispositional) moral progress. Actual moral progress covers changes for the better that are neither random—in that the agent lacks awareness of the moral significance of her action—nor attributable to a significant contribution of the agent's moral motivation. Judgments of actual moral progress acknowledge changes for the better that are facilitated by advantageous circumstances but that are not mere flukes because the agent is aware of the quality of her improved practices.

Finally, impactful moral progress is actual moral progress that brings about an improvement in states of affairs. Impactful moral progress complements the conceptual account of moral progress by covering cases in which moral agents' changes in practice (accompanied by some relevant moral insight) effect an improvement in states of affairs, i.e., axiological progress. Impactful moral progress might be conceived of as the ideal version of moral progress. However, when it comes to acknowledging the fact that a moral agent has progressed morally, we will not require her to act ideally and we will not want to hold her hostage to the circumstances on which success in effecting positive change depends, either. Therefore, other concepts of moral progress must be in place to acknowledge changes for the better that fail to effect axiological progress or that are not expressible in terms of axiological progress. The resulting picture of moral progress is thus not tied to a consequentialist perspective. On the contrary, the requirement of minimal moral awareness represents a requirement that may not be accounted for in purely consequentialist terms.

The concepts of moral progress established here cover several ways in which moral agents may be plausibly said to progress while preventing changes that are in a specifiable sense merely theoretical and those that come about as moral lucky strikes from counting as moral progress. The account explicated here gives clearer content to the intuition that "ethical progress without practical change is hollow; practical change without ethical progress is disoriented"—by specifying and interrogating the particular requirements of ethical involvement and practical relevance. It thereby improves on existing accounts of moral progress, which often include contradictory hints toward the relative relevance of ethical progress and practical change. The proposed concepts lend themselves to further specification through the addition of normative conditions of moral progress.

The final chapter addressed the question of whether it would be morally objectionable or desirable that moral agents were motivated to bring about moral progress, where 'moral progress' was to be read in a de dicto rather than de re sense. Drawing on the debate about the moral worth of a de dicto desire "to do what is right"—a metaethical debate revolving around thoroughly normative issues—an objection to a desire was developed that stated that desiring to bring about moral progress was to treat moral progress as a fetish. The three threads of the argument were that a desire for moral progress was 1) objectionably abstract, because it was directed at the idea of progress rather than concrete moral problems, that it was 2) objectionably specific compared to the desire to improve the world in any way possible, and that it was 3) objectionably modest, because it aimed at mere progress rather than at the solution of problems. The fetish objection to the desire for moral progress can be refuted once we consider how a rational desire for moral progress is based on the (morally desirable) extrapolation from observed moral problems, when we recognize how it aims not at objectionably specific but rather at desirable ideal solutions to moral problems, and once we notice that it is adequately modest in that, on an abstract level, the moral agent can only rationally hope to effect progress—not to solve all moral problems. A moral agent who had actually cultivated a desire for moral progress would avoid the vice of affected pessimism and be marked by a comprehensive moral outlook and exemplary moral alertness—desirable features for moral agents to possess.

Moral progress might appear in many different forms, and it may be more or less actively pursued and ethically considered. On one end of the spectrum, it can appear in the form of a non-morally incentivized behavioral change that only qualifies as moral progress because the changing agents are vaguely aware that their new course of action is not only prudentially advantageous but also morally good. On the other end of the spectrum, it can be related to hard-earned ethical insights prompting morally motivated modifications of practices in order to "make the world a better place." One of the bright sides of this view of moral progress is that it renders moral progress quite possible—and a perfectly desirable mode of improving the world.

## Bibliography


Adams, Robert M. (1976): "Motive Utilitarianism." In: Journal of Philosophy 73, 467–481.

Adorno, Theodor W. (2003): "Progress." In: Tiedemann, Rolf (Ed.): Can One Live after Auschwitz? A Philosophical Reader. Stanford: Stanford University Press, 126–145.

Albersmeier, Frauke (2020): "Analysis, Explication, and the Nature of Concepts." In: Logical Analysis and History of Philosophy 22, 180–201.

Anderson-Gold, Sharon (2001): Unnecessary Evil: History and Moral Progress in the Philosophy of Immanuel Kant. Albany: State University of New York Press.

Anderson, Elizabeth (2014): "Social Movements, Experiments in Living, and Moral Progress: Case Studies from Britain's Abolition of Slavery." The Lindley Lecture. Feb. 11, 2014. The University of Kansas.

Annas, George J. (2002): "Moral Progress, Mental Retardation, and the Death Penalty." In: The New England Journal of Medicine 347. No. 22, 1814–1818.

Anscombe, Gertrude Elizabeth Margaret (1958): "Modern Moral Philosophy." In: Philosophy 33, 1–19.

Appiah, Kwame Anthony (2010): The Honor Code. How Moral Revolutions Happen. New Haven and London: W. W. Norton and Company.


Arruda, Caroline T. (2017): "The Varieties of Moral Improvement, or Why Metaethical Constructivism Must Explain Moral Progress." In: Ethical Theory and Moral Practice 20. No. 1, 17–38.

Baier, Annette C. (1991): A Progress of Sentiments. Reflections on Hume's Treatise. Cambridge: Harvard University Press.

Baier, Kurt (1958): The Moral Point of View. Ithaca: Cornell University Press.

Baker, Alan (2016): "Simplicity." In: Zalta, Edward N. (Ed.): The Stanford Encyclopedia of Philosophy. https://plato.stanford.edu/entries/simplicity/, last accessed June 28, 2022.


Bealer, George (1998): "Intuition and the Autonomy of Philosophy." In: DePaul, Michael and Ramsey, William (Eds.): Rethinking Intuition: The Psychology of Intuition and Its Role in Philosophical Inquiry. Lanham: Rowman and Littlefield.

Beaney, Michael (2018): "Analysis." In: Zalta, Edward N. (Ed.): The Stanford Encyclopedia of Philosophy. https://plato.stanford.edu/archives/sum2018/entries/analysis/, last accessed June 28, 2022.

Beauchamp, Tom L. (2010): "The Four Principles Approach to Health Care Ethics." In: Beauchamp, Tom L.: Standing on Principles. Oxford: Oxford University Press, 35–49.

Beauchamp, Tom L., and Childress, James F. (2001): Principles of Biomedical Ethics. 5 th ed. Oxford and New York: Oxford University Press.

Bellantoni, Lisa (2004): "Future Ethics: MacIntyre and Whitehead on Moral Progress." In: Polanowski, Janusz A. and Sherburne, Donald W. (Eds.): Whitehead's Philosophy. Points of Connection. Albany: State University of New York Press, 103–126.

Bennett, Jonathan (1974): "The Conscience of Huckleberry Finn." In: Philosophy 49. No. 188, 123–134.

Bentham, Jeremy (2007): An Introduction to the Principles of Morals and Legislation. Mineola: Dover Publications.

Birnbacher, Dieter (2013): Analytische Einführung in die Ethik. 3 rd ed. Berlin: De Gruyter.


Brink, David O. (2013): Mill's Progressive Principles. Oxford: Clarendon Press.

Broad, Charlie D. (1952): "Determinism, Indeterminism, and Libertarianism." In: Broad, Charlie D.: Ethics and the History of Philosophy. Selected Essays. New York: Humanities Press, 195–217.

Broad, Charlie D. (1967): Five Types of Ethical Theory. 9 th ed. Oxford: Routledge and Kegan Paul.

Brun, Georg (2016): "Explication as a Method of Conceptual Re-Engineering." In: Erkenntnis 81. No. 6, 1211–1241.

Buchanan, Allen (2012): "The Open-Ended Normativity of the Ethical." In: Analyse und Kritik 2012. No. 1, 81–94.

Buchanan, Allen, and Powell, Russell (2018): The Evolution of Moral Progress. Oxford and New York: Oxford University Press.

Bury, John B. (1920): The Idea of Progress: An Inquiry into Its Origins and Growth. New York: Dover.

Calhoun, Cheshire (1989): "Responsibility and Reproach." In: Ethics 99. No. 2, 389–406.

Cappelen, Herman (2012): Philosophy without Intuitions. Oxford: Oxford University Press.

Cappelen, Herman (2018: Fixing Language: An Essay on Conceptual Engineering. Oxford: Oxford University Press.

Carbonell, Vanessa (2013): "De Dicto Desires and Morality as Fetish." In: Philosophical Studies 163. No. 2, 459–477.

Carnap, Rudolf (1962): Logical Foundations of Probability. Chicago: The University of Chicago Press.

Carnap, Rudolf (1963): "Replies and Systematic Expositions. III. 19. P. F. Strawson on Linguistic Naturalism." In: Schilpp, Paul A. (Ed.): The Philosophy of Rudolf Carnap. Chicago: Open Court, 933–939.

Carus, André W. (2007): Carnap and Twentieth-Century Thought. Explication as Enlightenment. Cambridge: Cambridge University Press.

Chappell, Richard Yetter (2015): "Value Receptacles." In: Noûs 49. No. 2, 322–332.

Chappell, Sophie Grace (2016): "How Encounters with Values Generate Moral Demandingness." In: Ackeren, Marcel van and Kühler, Michael (Eds.): The Limits of Moral Obligation. Moral Demandingness and Ought Implies Can. New York: Routledge, 84–99.

Christian, Alexander (2018): "Vegan Lifestyle." In: Thompson, Paul B. and Kaplan, David M. (Eds.): Encyclopedia of Food and Agricultural Ethics. 2 nd ed. Dordrecht: Springer, 2396–2403.

Cohen, Alix (2000): "The Notion of Moral Progress in Hume's Philosophy: Does Hume Have a Theory of Moral Progress?" In: Hume Studies 26. No. 1, 109–128.

Cohnitz, Daniel, and Häggqvist, Sören (2010): "The Role of Intuitions in Philosophy." In: Studia Philosophica Estonica, February, 1–14.

Coliva, Annalisa, and Moruzzi, Sebastiano (2012): "Truth Relativists Can't Trump Moral Progress." In: Analytic Philosophy 53. No. 1, 48–57.

Condorcet, Jean-Antoine-Nicolas De Caritat (1963): Entwurf einer historischen Darstellung der Fortschritte des menschlichen Geistes. Frankfurt am Main: Europäische Verlagsanstalt.


Daly, Chris (2010): An Introduction to Philosophical Methods. Peterborough: Broadview Press.


Dixon, Ben (2005): "Achieving Moral Progress Despite Moral Regress." In: Social Philosophy Today 21, 157–172.

Donaldson, Sue, and Kymlicka, Will (2011): Zoopolis. A Political Theory of Animal Rights. Oxford and New York: Oxford University Press.

Driver, Julia (2012): Consequentialism. London and New York: Routledge.

Driver, Julia (2014): "The History of Utilitarianism." In: Zalta, Edward N. (Ed.): The Stanford Encyclopedia of Philosophy. https://plato.stanford.edu/archives/win2014/entries/utilita rianism-history/, last accessed June 28, 2022.

Dutilh Novaes, Catarina, and Reck, Erich (2017): "Carnapian Explication, Formalisms as Cognitive Tools, and the Paradox of Adequate Formalization." In: Synthese 194. No. 1, 195–215.

Dworkin, Ronald (1978): "Liberalism." In: Hampshire, Stuart (Ed.): Public and Private Morality. Cambridge: Cambridge University Press.

Elzanowski, Andrzej (2013): "Moral Progress: A Present-Day Perspective on the Leading Enlightenment Idea." In: Argument 3. No. 1, 9–26.

Enoch, David (2011): Taking Morality Seriously: A Defense of Robust Realism. Oxford: Oxford University Press.


Forst, Rainer (2012): "Toleration." In: Zalta, Edward N. (Ed.): The Stanford Encyclopedia of Philosophy. https://plato.stanford.edu/entries/toleration/, last accessed June 28, 2022.


Gert, Bernard, and Gert, Joshua (2016): "The Definition of Morality." In: Zalta, Edward N. (Ed.): The Stanford Encyclopedia of Philosophy. https://plato.stanford.edu/entries/mor ality-definition/, last accessed June 28, 2022.

Ginsberg, Morris (1944): Moral Progress. Glasgow: Jackson, Son and Co.

Glock, Hans-Johann (2009): "Concepts: Where Subjectivism Goes Wrong." In: Philosophy 84. No. 1, 5.

Godlovitch, Stanley (1998): "Morally We Roll Along: (Optimistic Reflections) on Moral Progress." In: Journal of Applied Philosophy 15. No. 3, 271–286.

Goldhill, Olivia (2017): "An Oxford Philosopher's Moral Crisis Can Help Us Learn to Question Our Instincts." Qz.Com. https://qz.com/1102616/an-oxford-philosophers-moral-crisis-canhelp-us-learn-to-question-our-instincts/, last accessed June 28, 2022.

Goldman, Alvin I. (2007): "Philosophical Intuitions: Their Target, Their Source, and Their Epistemic Status." In: Grazer Philosophische Studien 74, 1–26.

Gowans, Christopher (2015): "Moral Relativism." In: Zalta, Edward N. (Ed.): Stanford Encyclopedia of Philosophy. https://plato.stanford.edu/entries/moral-relativism/, last accessed June 28, 2022.

Grice, Herbert Paul (1989): Studies in the Way of Words. Cambridge: Harvard University Press.

Hämäläinen, Nora (2017): "Three Metaphors Toward a Conception of Moral Change." In: Nordic Wittgenstein Review 6. No. 2, 47–69.


Horgan, Terry, and Timmons, Mark (2008): "Analytic Moral Functionalism Meets Moral Twin Earth." In: Ravenscroft, Ian: Minds, Ethics, and Conditionals. Themes from the Philosophy of Frank Jackson. Oxford: Clarendon Press, 221–236.

Horta, Oscar (2010): "Debunking the Idyllic View of Natural Processes: Population Dynamics and Suffering in the Wild." In: Télos 17, 73–88.

Huemer, Michael (2016): "A Liberal Realist Answer to Debunking Skeptics: The Empirical Case for Realism." In: Philosophical Studies 173. No. 7, 1983–2010.


Hursthouse, Rosalind, and Pettigrove, Glen (2018): "Virtue Ethics." In: Zalta, Edward N. (Ed.): The Stanford Encyclopedia of Philosophy. https://plato.stanford.edu/archives/win2018/ entries/ethics-virtue/, last accessed June 28, 2022.


Moody-Adams, Michele M. (2017): "Moral Progress and Human Agency." In: Ethical Theory and Moral Practice 20. No. 1, 153–168.

Moore, George Edward (2004): Principia Ethica. Mineola: Dover Publications.

Moran, Kate A. (2012): Community and Progress in Kant's Moral Philosophy. Washington, DC: Catholic University of America Press.

Nagel, Thomas (1979): Mortal Questions. Cambridge: Cambridge University Press.

Nagel, Thomas (1986): The View From Nowhere. New York and Oxford: Oxford University Press.

Niiniluoto, Ilkka (2011): "Progress by Enlightenment: Fact or Fiction?" In: Temenos: Nordic Journal of Comparative Religion 47. No. 1, 97–110.

Nimtz, Christian (2012): "Begriffsanalyse heute." In: Zeitschrift für philosophische Forschung 66. No. 2, 218–247.

Nussbaum, Martha Craven (2007): "On Moral Progress: A Response to Richard Rorty." In: The University of Chicago Law Review 74 (Cambridge 1970), 939–960.


Peirce, Charles Sanders (1966a): "Issues of Pragmaticism." In: Peirce, Charles Sanders: Charles S. Peirce: Selected Writings. Values in a Universe of Chance. Philip P. Wiener (Ed.). New York: Dover Publications, 203–223.

Peirce, Charles Sanders (1966b): "What Pragmatism Is." In: Peirce, Charles Sanders: Charles S. Peirce: Selected Writings. Values in a Universe of Chance. Philip P. Wiener (Ed.). New York: Dover Publications, 180–202.

Pereboom, Derk (2014): Free Will, Agency, and Meaning in Life. Oxford: Oxford University Press.

Pettit, Philip and Smith, Michael (2000): "Global Consequentialism." In: Hooker, Brad, Mason, Elinor, and Miller, Dale (Eds.): Morality, Rules and Consequences: A Critical Reader. Edinburgh: Edinburgh University Press, 121–133.

Pinder, Mark (2017): "Does Experimental Philosophy Have a Role to Play in Carnapian Explication?" In: Ratio 30. No. 4, 443–461.

Pinker, Steven (2018): Enlightenment Now. The Case for Reason, Science, Humanism, and Progress. New York: Viking.

Platts, Mark (1988): "Moral Reality." In: Sayre-McCord, Geoffrey (Ed.): Essays on Moral Realism. Ithaca: Cornell University Press, 282–300.

Pokrovsky, Nikita Evgenevich (2001): "Material Progress and the New Morality: Russia as Probing Ground." In: The Journal of Speculative Philosophy 15. No. 3, 201–213.

Pölzler, Thomas (2016): "Revisiting Folk Moral Realism." In: Review of Philosophy and Psychology 8, 455–476. https://doi.org/10.1007/s13164-016-0300-9, last accessed June 28, 2022.

Popper, Nathaniel (2019): "Behold the Beefless 'Impossible Whopper.'" In: The New York Times. https://www.nytimes.com/2019/04/01/technology/burger-king-impossible-whop per.html, last accessed June 28, 2022.


Quine, Willard V. O. (1976): "Carnap and Logical Truth." In: Quine, Willard V. O.: Ways of Paradox and Other Essays. 2 nd ed. Cambridge: Harvard University Press.


Scheffler, Samuel (1992): Human Morality. New York and Oxford: Oxford University Press.


Shepherd, Joshua, and Justus, James (2015): "X-Phi and Carnapian Explication." In: Erkenntnis 80. No. 2, 381–402.


Valinsky, Jordan and Wiener-Bronner, Danielle (2019): "America Is Running out of Impossible Burgers." CNN Business. https://edition.cnn.com/2019/04/30/business/impossiblemeat-shortage/index.html, last accessed June 28, 2022.


## Index

abolition 7–9, 44, 86, 119, 174, 180 affected pessimism 227 f, 234 agency-requirement 115–117, 120–126, 138, 154f., 157, 171–173, 178, 2312 akrasia 135, 141, 143 – inverse 141, 143, 155, 157 animals 35, 45, 60, 71, 94, 98, 102, 108, 110, 118, 135f., 156, 159, 163, 198, 205, 223 anti-realism 4, 81 f., 86, 96f. axiological progress 34, 56, 114, 161, 176, 210–212, 215, 233 Buchanan, Allen 3, 8, 25, 28, 44f., 53, 60, 85 f., 94, 103, 110, 124, 129–133, 140, 154, 172–176, 180, 182f., 202 Carbonell, Vanessa 193f., 198–202, 208, 221–223 Carnap, Rudolf 14–17, 19 f., 22 character 17, 31, 33, 44, 52, 61, 63–65, 104, 120, 125, 139, 142, 149, 158, 194 clarification 3, 7, 13f., 16, 19–21, 29, 133, 173, 216 conceptual analysis 5, 8–20, 99, 230 conceptual engineering 17, 22 conscientious moral agents 152, 183, 207, 223, 226, 229 conscious effort-requirement 115 f., 118f., 121 f., 124f. consequentialism 47, 51–57, 63, 153, 191, 230 – global consequentialism 52, 230 – progressive consequentialism 47, 55–57, 230 continuity 4, 23, 187 cultivation 63–65 de dicto desires 192, 194, 196–201, 203, 205, 221–223, 225, 233 de re desires 192, 196–201, 203, 205, 221 f., 233 deontology 63, 187

Descartes, René 74f. descriptive theory of meaning 12, 31 descriptivism 31 f., 80, 84 disagreement 4, 11, 13, 18, 40, 45 f., 72, 97, 99, 106, 109f., 127, 170, 189, 194, 220 dispositions 47, 52, 64–66, 80f., 113, 143–145, 147, 149, 151, 167f., 177f., 213, 231 f. domains of concern 107 Enlightenment 2, 142 Enoch, David 4, 84, 90f., 94 f., 158 Erdur, Melis 95 f. ethical gap 140, 143, 155 ethical progress 6, 39, 51, 77, 110, 112f., 115 f., 135–138, 140, 142f., 148, 152, 154, 161, 163, 165 f., 169, 173, 222, 224f., 227, 232f. ethical regress 66, 68, 77 f. Expanding Circle 3, 47, 50, 59, 82 explanation 3 f., 14, 36, 80, 86, 90f., 94 f., 180, 194, 196, 199 explication 5, 14–23, 32, 34, 43, 47–49, 58, 63, 68, 72, 78, 85, 87, 96 f., 99– 101, 107, 122, 132, 146, 156, 174, 185, 213, 229–231 fetish argument 181, 188, 190–192, 198, 200, 202f., 206, 217 fetish objection 181, 186, 188, 194, 202f., 206, 208, 217f., 220, 226, 234 Godlovitch, Stan 8, 27 f., 39f., 44 f., 63f., 73–75, 78, 105, 111, 137, 180, 182 Grice, Herbert Paul 9 f., 13, 20 Huemer, Michael 4, 66f., 82, 90, 105 Humean theory of motivation 83, 189f. ignorance 4, 40, 42, 137, 224, 227f. impartiality 32, 49–51, 191, 200f.

Open Access. © 2022 the author(s), published by De Gruyter. This work is licensed under the Creative Commons Attribution 4.0 International License. https://doi.org/10.1515/9783110798913-010

inadvertent rightdoing 155–160, 169, 214f., 232 inclusivity 47, 49–51, 57, 60, 108f. individual progress 100, 104, 177 initiative-requirement 116f. Jamieson, Dale 4, 9, 41, 44, 47, 49, 53–55, 66, 80, 82, 85, 93, 105, 113, 122f., 180, 183 Kant 4, 47, 58–63, 105 Kitcher, Philip 3, 38, 44, 47, 92–94, 105f., 124 liberalism 1, 4, 30, 48, 66–70, 72, 75f., 114 Mackie, John Leslie 33, 82–84, 111 Macklin, Ruth 4, 25, 28, 41 f., 44, 101, 130f. making the world a better place 58, 159, 185 f., 210, 221, 225, 227, 234 meat paradox 135 metaethical trilemma 83, 188–190 Mill, John Stuart 48 f., 55f., 67–73, 75–78, 109, 119, 127, 162, 200, 231 minimal continuance 105f., 146, 150 Moody-Adams, Michele 2, 4, 27, 39 f., 42, 64, 77, 105, 107, 114, 128–130, 153f., 179, 182f., 224, 228 moral awareness 5, 35–37, 42, 73, 77f., 115, 127, 155f., 159–163, 167, 170–172, 174, 176, 179, 214f., 226f., 230, 232f. – minimal moral awareness 162, 167, 215 moral considerability 5, 36f., 230 moral decline 29, 104, 122, 168f. moral development 40, 59, 64, 98, 100– 102, 112, 126, 136 moral obviousness 44, 73, 110 moral ontology 23, 79, 81, 85f., 96 f. moral performance 125, 132f., 145f., 154, 156, 162, 165, 169–175, 177f., 185, 210f., 215–217, 225, 232 moral phenomenon 5, 46, 99

moral progress – actual 144, 146f., 149f., 152f., 156, 165– 167, 169–171, 174 f., 177, 185, 210, 214f., 232f. – by a fluke 153–157, 159–161, 165, 167, 169, 173, 215, 232 – dispositional 144–149, 151 f., 154, 165– 167, 169, 174, 185, 215, 222, 232 – impactful 175 f., 178, 185, 210, 214–216, 229, 233 – in theory 128, 143, 232 – incentivized 154, 179, 234 – inconsequential 175 moral realism 4f., 40 f., 48, 67, 79–85, 87–89, 90–97, 231 moral regress 4, 104, 125, 168, 184 moral saint 199, 226 moral worth 5 f., 62 f., 170, 206, 215, 230, 233 motivational externalism 13, 158, 181, 188, 190–193, 195, 198f., 201–203, 205, 207 naïve conception 4, 9, 47, 53 f., 56f., 153, 157, 173 objectivism 19f. open-endedness 27f., 94, 110, 224 Peirce, Charles Sanders 74, 78 Powell, Russell 3, 8, 25, 28, 44f., 53, 60, 85 f., 94, 103, 110, 124, 129–133, 140, 154, 172–176, 180, 182f., 202 practical gap 123, 136–138, 140, 163, 166 practical relevance 6, 128, 138, 141, 143– 145, 149, 152, 190, 221, 233 Prinz, Jesse 4, 80f., 89 progress challenge 81 progress-realism tie-up 48, 79, 81, 85, 92, 94, 231 reflective equilibrium 9f., 151 relativism 4, 81, 89 requirement of retrospective endorsement 126, 152 responsiveness 35, 37, 44, 141, 161, 170, 227

right motives 62f.

right reasons 140f., 157, 162, 170 Rønnow-Rasmussen, Toni 41, 53, 115, 117, 119–121, 123, 137f.

self-effacingness 65, 187f., 195, 201


theory-requirement 117, 126, 153, 159, 165, 172f., 178, 225 tolerance 69 transcendental argument 79f., 87, 90, 231 transformative progress 28, 59f., 68, 72, 94, 102, 111, 151, 230 unwilling wrongdoer 71, 135–137, 144, 194 utilitarianism 49, 51, 54, 56f., 162, 186, 191, 195, 209 virtue ethics 48, 54, 63–65, 101, 186, 231

weakness of will see akrasia Wilberforce, William 180f.

virtues 3, 63–66, 133