Prospects for pure procedural moral progress

ABSTRACT

Issues of methodology are central to the philosophy of moral progress. However, the idea that effective moral methodology, as well as being instrumental to progress, might also constitute progress has not been adequately explored. This paper will critically assess the merits of this idea – what I call ‘pure proceduralism about moral progress’ – taking Philip Kitcher's recent theory of ‘democratic contractualism’ (2021) as a test case. An epistemology of pure procedural moral progress will be sketched: namely, a naturalised epistemology of paradigmatic instances of moral progress. Assuming an ideal procedure exists and has been at least approximately instantiated (an empirical claim), we learn about its form by studying paradigms of historical moral progress. From this epistemological method will be derived theoretical constraints on what can count as a theory of moral progress per se: it must capture the core semantic constraints on the term ‘moral progress,’ those features without which we would not recognise a change as moral progress. While this poses a tougher challenge for pure proceduralism than for more traditional approaches to moral methodology, it will be concluded that the potential of pure proceduralism as a viable metaethics of moral progress remains an exciting open question.

KEYWORDS:

• Moral progress
• metaethics
• Philip Kitcher
• democratic contractualism
• pragmatic naturalism

Issues of methodology occupy a central place in the literature on moral progress. In chapter 3 of On Liberty ([1859] 1985), J.S. Mill advocated an experimental methodology for moral improvement. More recently, philosophers have much discussed the power of effective methodology to overcome the epistemic biases (Buchanan 2004; Anderson 2016) and affective limitations (Buchanan and Powell 2018, especially part II; Sauer 2019) that frustrate moral progress. However, the idea that the right moral methodology, as well as being conducive to and causative of moral progress, might in some sense also be constitutive of progress has not been explored in great depth.

The idea that effective methodology could all by itself constitute some moral progress might strike one as a non-starter. It seems a conceptual truth that progress involves moving closer to a goal, and that a methodology does a better or worse job at achieving that goal. However, moral philosophers should at least entertain the idea of methodological constitutivism about moral progress for several reasons. First, if some instances of moral progress are constituted by the methodology by which they come about, and if we were able to identify such a constitutive methodology, we would possess an action-guiding theory of moral progress that doesn't depend on a murky epistemology of ultimate moral goals. Second, if we were able to identify such a methodology, we would possess an explanatory schema in terms of which we could explain the successes (and failures) of historical attempts at making moral progress in a philosophically illuminating and non-debunking way; the successes (or failures) of certain progressive movements in history are explained by the extent to which they embodied the constitutive methodology. Third, methodological constitutivism about moral progress is more at home within a naturalistic worldview than traditional teleological conceptions of moral progress. Even if the idea seems superficially implausible, these are three good reasons for keeping an open mind.

In a recent contribution, Kitcher (2021) can be interpreted as making the case that a transformation of moral practice qualifies as an instance of moral progress by virtue of the form of the conversation that produces it. The form of moral deliberation advocated by Kitcher as constitutive of moral progress – what he calls ‘democratic contractualism’ (2021, 38ff) – is intended to cohere with and follow from Kitcher's metaethical view of what moral progress is, what he calls ‘pragmatic naturalism’ (2011, 3ff). However, this interpretation of democratic contractualism, within the framework of pragmatic naturalism, is just one instance of a more general approach to moral progress that one might aptly call ‘pure proceduralism’ about moral progress.

This paper will introduce and critically assess the idea of pure proceduralism about moral progress, taking the pure proceduralist interpretation of democratic contractualism as a test case. It will introduce the notion of pure procedural moral progress by comparison with Rawls's notion of pure procedural justice. Traditionally, methodologies of moral progress aim to produce changes that qualify as moral progress by virtue of independent criteria of moral progressiveness. On the other hand, pure proceduralism about moral progress assumes only that there is some constitutive methodology, such that any change qualifies as progressive solely by virtue of its being the product of that methodology.

A brief sketch of an epistemology of pure procedural moral progress will be offered: namely, a naturalised epistemology of paradigmatic instances of moral progress from history, similar to that endorsed by Lowe (2019). Assuming a constitutive methodology exists and has been historically instantiated (an empirical claim), we learn about its form by studying paradigmatic instances of moral progress from history. From this epistemological method, I will derive theoretical constraints on what can count as a theory of moral progress per se: it must capture the core semantic constraints on the term ‘moral progress,’ those features without which we would not recognise a change as moral progress at all.

Specifically, I will argue that any pure proceduralist theory of moral progress must: (i) adequately capture the obligatoriness of making moral progress; (ii) make moral progress matter in itself, and not as a means to a non-moral end. These two proposed desiderata are especially challenging for pure proceduralism about moral progress, because they must be met without appealing to anything beyond the ideal procedure. A prima facie plausible pure proceduralist framework capable of accommodating these features will be presented, and it will be demonstrated how exactly democratic contractualism fits into it. The ability of democratic contractualism to meet these desiderata will be assessed. Subsequently, I will reflect on the prospects of any form of pure proceduralism about moral progress.

It will be concluded that, while proceduralism about moral progress bears a heavier theoretical burden than more traditional approaches to moral methodology, the potential of pure proceduralism as a viable position in the metaethics of moral progress remains an exciting open question.

Defining pure procedural progress

There is a closeness between the concepts of justice and moral progress. Analytically, any change that achieves a state of greater justice from one of injustice also achieves moral progress. Many (though not all) instances of moral progress inhere wholly in the achievement of greater justice. This conceptual closeness licenses us to think about progress and justice similarly. In particular, it licenses us to think about the possible roles we might ascribe to procedure in achieving moral progress along the lines of the possible roles we might ascribe to procedure in achieving justice.

In A Theory of Justice (1972, §14), John Rawls distinguishes three kinds of procedural justice:

1. Perfect procedural justice: a procedure that is guaranteed to output a just end state.

2. Imperfect procedural justice: a procedure that is more-or-less likely to output a just end state.

3. Pure procedural justice: a just end state qualifies as such solely by virtue of the procedure that produced it.

Perfect and imperfect procedural justice assume independent criteria for what is to count as a just end state defined separately from and prior to the procedure which is to be followed. A procedure does a better or worse job at outputting just states of affairs, thusly construed. In contrast, pure procedural justice assumes only that there is, in a given context, a constitutive procedure such that the outcome is just, whatever it is, provided that the procedure has been properly followed. The procedure consists of a set of methodological rules and principles that constitute justice in that context.

By analogy, one might distinguish between three kinds of procedural progress:

1. Perfect procedural progress: a procedure that is guaranteed to output a progressive change.

2. Imperfect procedural progress: a procedure that is more-or-less likely to output a progressive change.

3. Pure procedural progress: a progressive change qualifies as such solely by virtue of the procedure that produced it.

Similarly, perfect and imperfect procedural progress assume independent criteria for what is to count as a progressive change, defined separately from and prior to the procedure which is to be followed. A procedure does a better or worse job at outputting progress, thusly construed. In contrast, pure procedural progress assumes only that there is, in a given context, a constitutive procedure such that the outcome will be a progressive change, whatever it is, provided that the procedure has been properly followed. The procedure consists of a set of methodological rules and principles that constitute progress in that context.

Conceptual closeness is not identity; there are dissimilarities. I will mention what I take to be the most important two. First, while a procedure for justice usually (though not always) terminates, a procedure for moral progress can in principle continue indefinitely. A particular legal trial, for instance, conducted wholly in accordance with the principles of procedural justice, terminates in a just ruling. Progressive reform of the legal system, on the other hand, will always be possible and worthwhile; progress in this context need not result in nor even assume the existence of an optimal arrangement. Therefore, a procedure for moral progress should be thought of as a processual analogue of a procedure for justice. Second, while justice concerns only the basic structure of society and not individual choices within that basic structure¹, moral progress can occur both at the level of social institutions and at the level of individual choice.

Democratic contractualism as pure procedural progress

Democratic contractualism construes the ideal procedure as a conversation, proceeding according to a certain agenda or programme, conducted under certain constraints.² Put this way, the process of moral inquiry strongly resembles a well-run meeting – not coincidentally, as it is intended to approximate a ‘meeting of the tribe.’ The constraints go that:

• All potentially affected individuals³ participate in the conversation.

• All participants in the conversation are optimally factually informed.

• All participants regard one another with mutual sympathy and as equals.

The order of the conversation runs roughly as follows (Kitcher 2021, 33–40). The first task is to identify the prima facie problematic situations into which moral inquiry is urgently needed. A situation is prima facie problematic if individuals or groups find themselves seeking relief from it; they put forward a moral claim as an ‘order of business’ within the context of the ideal conversation. The situation is justifiably regarded as urgently problematic if a process of deliberation conducted under the ideal constraints (i)-(iii) would endorse it as such; proposed orders of business are ‘seconded.’ Prima facie problematic situations endorsed as urgent become ‘action items’ – topics worthy of initiating moral inquiry. Moral inquiry itself involves making, assessing, and trying out proposals for responding to an urgently problematic situation, again under the constraints (i)-(iii). A particular moral inquiry justifiably terminates – an action point is marked off as completed – when the proposed response would be endorsed by all potentially affected individuals, again, under the constraints (i)-(iii).

Kitcher writes that justification ‘emerges from the process of moral inquiry’ (36) and that progress ‘happens when justification sticks. … A change in moral practice is progressive just in case it would be retained in an indefinitely proceeding sequence of justified resolutions’ (39, his emphasis). From this conjunction, and on the basis of the definition of pure proceduralism about moral progress outlined above, it seems that Kitcher intends democratic contractualism as a form of pure proceduralism about moral progress. The procedure alone licenses us to speak of its outputs as morally progressive; to be moral progress is simply for that license never to be revoked.

Kitcher's understanding of our epistemic access to the maxims of democratic contractualism differs from Rawls's understanding of our epistemic access to the principles of justice as fairness. Democratic contractualism, unlike justice as fairness, is underpinned by a ‘naturalised epistemology’ of moral progress. Naturalised epistemology involves studying the actual practices and processes of epistemic justification in a domain and extracting from them the forms and possibilities of epistemic justification in that domain generally. The domain most associated with naturalised epistemology is the philosophy of science (Quine 1969; Kitcher 1992). The practice of studying the actual process of epistemic justification in the sciences is a methodological through-line connecting Thomas Kuhn's The Structure of Scientific Revolutions (1962) to Kitcher's own The Advancement of Science (1993), and many other classic works besides (e.g. Lakatos 1970; Laudan 1978). It is no surprise then that Kitcher favours the approach for studying progress in the moral domain as well.

By studying the history of moral progress, Kitcher believes we can identify methodological features likely to underpin any future attempts at making moral progress. He argues that while the history of ethics is not a story of cumulative discovery of moral facts, it is a story of tentative and fallible (re)discovery of the form of the ideal moral conversation. One might object that history is a descriptive exercise that tells us nothing about how progress ought to go. However, if we make the methodological choice to investigate only ‘a certain set of beliefs which we independently judge to deserve our epistemic confidence’ then the project of naturalised moral epistemology becomes ‘the empirically informed study of how certain beliefs – specifically, beliefs which are clear cases of successful knowing – are actually formed’ (Lowe 2019, 4–5). By studying instances of moral progress about which we are particularly epistemically confident or even certain⁴ – the ‘paradigmatic’ cases – we can teach ourselves the constitutive rules of making moral progress: ‘compress the history, collapse the stereotypes, focus only on the heroes, and an approximation to the ideal conversation will emerge’ (Kitcher 2021, 44).

It is important to note that pure proceduralism about moral progress rests on a fallible thesis: that paradigmatic instances of moral progress throughout history are each imperfect implementations of a single constitutive procedure for moral progress. It might turn out that paradigmatic instances of moral progress were in fact irreconcilably methodologically heterogenous (Eriksen 2019), or that the history of moral progress in fact involved the discovery of mind-independent moral facts after all (Huemer 2015), or that the history of moral progress in fact debunks our belief in objective progress altogether by explaining the progression of moral history in terms of non-moral causes or motivations (Hopster 2019; Cofnas 2020). This is largely an empirical question. The purpose of this article is to offer a preliminary assessment of the theoretical viability of pure proceduralism about moral progress. This task will occupy the remainder of the article.

Semantic constraints on the term ‘moral progress’

Well-conducted historical-philosophical investigation into paradigmatic instances of moral progress may challenge some of our assumptions about, and thereby improve our understanding of, the concept of moral progress itself. In this sense our search for moral progress is reflexive; we might arrive at a more refined idea of what we are looking into by looking into our initially naïve idea of it. But any account of moral progress ought to be capable of accommodating our very strongest intuitions about what can and what cannot count as moral progress – those features without which we simply wouldn't recognise something as an instance of moral progress at all. As Lowe puts it, ‘[a]ll inquiry presupposes some minimal conception of the object of inquiry, so that we might recognize it when we go looking for it’ (2019, 5, n. 2).

Some of our intuitions about moral progress are sufficiently strong to present themselves as semantic constraints on the use of the term ‘progress’ – someone who acts as if certain obviously inapplicable changes might be progressive calls into question their grasp of the meaning of the term ‘progress.’ Philipa Foot once suggested that the ‘internal relations’ of a moral predicate to its objects places constraints on the sorts of moral evaluations a speaker can make while making sense (1959, 86ff). For example, argues Foot, it is ‘surely clear that moral virtues must be connected with human good and harm, and that it is quite impossible to call anything you like good or harm’ (94). Something similar can be said about moral progress.

Sometimes, a particular instance of moral change can rise to the status of an ‘exemplar’ or ‘paradigm’ of moral progress: an archetype by analogy with which other judgements of progress are made. Grasping the paradigms of a concept is at least partially constitutive of grasping the concept itself, a precondition of its use in practice – in other words, a semantic constraint. Kitcher takes as his paradigms of moral progress the abolitionist movements of the 18th and 19th centuries, the feminist movements of the twentieth century, and the LGBT + movements the late twentieth century.

It is a desideratum for any methodology for moral progress that it accommodates (at least) the semantic constraints on moral progress, whatever they are, including its paradigms; nothing should follow from the methodology that violates them.⁵ Preferably, it should tell us much more about moral progress besides – but if it fails to achieve even this minimal desideratum, it fails to be a methodology for moral progress at all.

First challenge: the obligatoriness of moral progress

For pure proceduralists, one's preferred procedure for moral progress creates, in the sense of constitutes, the opportunity for certain progressive changes to occur. However, making moral progress is widely regarded as obligatory – it is natural to think that if we find ourselves with an opportunity to make even some moral progress without incurring morally significant costs, but don't, then we are doing something wrong. One challenge for a pure procedural account of moral progress is to capture this ‘something wrong’ in a way that doesn't appeal to extra-procedural criteria of the morally true or the morally ideal.

To test the ability of pure proceduralism to meet this challenge, I will explore a particularly difficult instance of it. The difficulty is to obligate moral progress under conditions of what Kitcher suggestively calls ‘false consciousness’:

False consciousness is the product of situations in which the community's moral practice declares that some ideals of the self are appropriate for one group within the community but not for the rest, when some members of the latter group could in principle profitably pursue the ideals denied to them, and when at least some members of that group acquiesce in the moral claim that those ideals are not for them. The extreme case of false consciousness occurs when all members of the community agree with this aspect of the moral practice. When false consciousness prevails to this extent, there will be no complaints about the restriction of the ideals. (2021, 66–67)

Several things are relevant for making diagnoses of false consciousness. At play are various sociological measures such as the degree of norm tightness or looseness⁶, the balance between achieved and ascribed status, and equality of opportunity. In cultures with a high degree of norm tightness, expectations of conformity to norms are high, and enforcement of norms is strict. In cultures for which ascribed status outweighs achieved status, individuals have little choice or influence over the expectations and norms that apply to them. Where opportunity for choice is unequal, the balance between ascribed and achieved status differs between individuals and between groups.

Kitcher recognises that some limitation is characteristic of socialisation, that humans are social beings, and therefore that ‘[c]onfinement is likely to be a persistent feature of the human condition’ (2017, 59). It is implausible that every human preference could in principle be satisfied; even if it could, it is even more improbable that all conflicts between preferences can be rationally resolved in a way that is attractive to all affected. ‘Some frustration – even some imposed frustration – of some human preferences must be legitimate and unexceptionable’ (Geuss 1981, 16). The limitations relevant to judgements of false consciousness must therefore be in some sense illegitimate.

These four features – norm tightness, ascribed status, inequality, and illegitimacy – connect false consciousness with the concept of Herrschaft (Geuss 1981, 16–18) – the power of a dominant group to impose ‘surplus repression’ on a subordinate group. What is crucial to false consciousness, however, is acceptance of or resignation to the limited choices available to all or some individuals in a society, by many or all of those very individuals whose choices are limited – what has been called ‘adaptive preference’.⁷

Kitcher chooses the (still ongoing) conversations surrounding the ‘proper roles’ of women in society as his case study for false consciousness as an obstacle to moral progress. ‘Under false consciousness … women do not protest their situation, and sympathetic men, committed to mutual engagement, can understandably infer the permissibility of a patriarchal status quo’ (2021, 49). Thus, a conversation conducted in accordance with the methodological maxims of democratic contractualism may still overlook the moral obligation to dismantle the patriarchal status quo and to expand and facilitate the choices for living available to women. In extreme cases, the situation simply never arises as an order of business in the ideal conversation; in less extreme cases, it arises but it fails to be endorsed as urgent or justified.

Kitcher views false consciousness as a result of past and present moral discourse straying from the path of democratic contractualism (2021, 55–59). A woman's false consciousness originates in a regressive culture of acquiescence to patriarchal moral-epistemic authority, starkly opposed to the sort of moral-epistemic egalitarianism on which democratic contractualism is premised. To counteract this, Kitcher prescribes that.

[e]ven in the absence of challenges, societies should periodically assess whether restrictions on the appropriateness of ideals of the self for some subgroups can be justified. (2021, 67)

In other words, we have an obligation to scrutinise and improve our beliefs about the (in)appropriateness of certain ideals of the self for particular ‘kinds of people,’⁸ even – or perhaps especially – when we, or our society, or those we regard as moral-epistemic authorities, perceive the (in)appropriateness of those ideals for those kinds of people as obvious or natural. Somehow this obligation must be a product of the ideal procedure.

I am most interested in the extreme cases of false consciousness. To my eyes, the garden variety version of false consciousness, in which some of the constrained group do protest but are not listened to, is a special instance of conversational exclusion, not a distinct problem in its own right. Democratic contractualism seems well equipped to deal with problems of this sort. With Kitcher, I regard it as an unexceptionable methodological virtue that we ought to think very hard indeed about the reasons we might take ourselves to have to exclude certain claims made by individuals or groups from moral consideration, especially when those are grounded by largely unexamined ideals of the self. What I doubt is whether, in the absence of challenges, we have good reason to reassess anything at all.

False consciousness and moral-epistemic failure

To adequately grasp false consciousness, we must understand what is meant by a form of consciousness in general, false or otherwise. A form of consciousness in the general sense might aptly be described as a pre-cognitive orientation, or a Weltanschauung, or world-picture. A form of consciousness understood in these terms consists of both cognitive and non-cognitive elements lying behind an individual's understanding of and participation in their world. It will contain a backdrop of basic beliefs, a network of concepts, a suite of patterns of inference, stances of expectation, readiness, and preparedness for certain events, reactive dispositions, tastes and preferences, a system of values and motivations – and much else. These elements set an individual up to engage with their world in a particular way, usually in coordination with others in their society with whom they share a form of consciousness. A form of moral consciousness is a component of a general form of consciousness that specifically influences thought and action in the moral sphere (however this sphere is demarcated).

A form of consciousness might be defective in some sense and thereby liable to critique. Given Kitcher's evident recognition of the non-doxastic dimensions to moral practice and moral progress, spotlighting as he does the development of moral ‘sensitivities’ and ‘capacities’ as modes of progress (2021, 83f; 2015, 129–132), and his emphatic rejection of the ‘Discovery View’ of moral truth (2011, 138–170; 2021, 15–20), his choice to characterise a form of moral consciousness specifically as false seems doubly odd. It is odd because many of the elements of our moral consciousness are non-propositional; they are not even capable of being false. Self-defeating or maladaptive perhaps – but not false. It is doubly odd because even those moral beliefs that are capable of being true or false are not made so by their correspondence with objective moral reality; per Kitcher, we are licensed to speak of them as such on the basis of how they fare across successive applications of the ideal procedure (I will examine this picture of ethical truth in the penultimate section), but it is unclear how we can call a form of consciousness false ahead of time.

In what sense, then, is false consciousness a moral-epistemic failure? The notion of false consciousness has its roots in the Marxist idea of ‘ideology.’ The phrase ‘false consciousness’ itself belongs not to Marx but to Engels, who wrote to Franz Mehring in 1893 that ‘[i]deology is a process accomplished by the so-called thinker consciously, but with false consciousness. The real motive forces impelling him remain unknown to him’ (Marx, Engels, and Torr 1936). For Engels, false consciousness was indeed a form of ignorance – namely, ignorance of the ‘real motive forces’ lying being one's thought and action.

We can make some sense of Kitcher's choice to characterise a form of moral consciousness specifically as false by supposing that the agent under false conscious is similarly ignorant. The form of moral consciousness in question contains false non-moral beliefs that influence moral thought and practice, or one has false non-moral beliefs about one's form of moral consciousness, for example about its ‘genealogy.’ A few examples:

1. A social phenomenon or institution (such as chattel slavery or patriarchy) might mistakenly be believed to be ‘natural,’ ‘necessary,’ or ‘inevitable,’ and thereby erroneously exempted from moral criticism (cf. Haskell 1985; Pleasants 2010).

2. The hegemonic interests of a portion of society (such as the bourgeoisie or white males) might mistakenly be believed to be or to represent the interest of society as a whole, and thereby erroneously endorsed (cf. Davis 1987).

3. A self-validating belief (such as the belief that Africans are too indolent to work freely, or that women are not intelligent enough to be educated) might mistakenly be regarded as a belief that is not self-validating, and thereby erroneously assumed (Mill 1869).

4. A form of moral consciousness might be regarded as ‘self-evident,’ and therefore as ‘justified,’ when in fact our confidence is the product of a non-ideal formative process, and therefore ‘unjustified’ (Street 2006; Nietzsche 1887; Sauer 2018).

Examples (i)-(iii) seem to be adequately handled by the democratic contractualist constraint that deliberation occurs under conditions of ideal information. Optimal information presumably includes all the relevant and available information regarding the necessity, inevitability, or naturalness of a particular phenomenon or institution (however that is understood), the ‘actual’ interests and needs of all portions of society (however that is understood), and the causal or logical loops that lead to the self-validation of certain beliefs. The democratic contractualist prescription to reassess whether limitations on our existing ideals of the self are justified even in the absence of challenges seems to be intended to address beliefs about the justificatory status of a form of consciousness covered by (iv). One is under false consciousness when the limited scope of one's available ideals of the self appears to be a pure ‘product of thought,’ but is in fact a ‘reflection in thought’ of one's situatedness in a dysfunctional moral-epistemic practice.

The argument appears to be an attempt to debunk patriarchal moral consciousness, and, presumably by extension or analogy, any other possible instance of false consciousness. Our moral intuitions about the appropriateness or inappropriateness of certain ideals of the self are products of a form of moral consciousness that owes its existence to the exercise of a non-ideal moral-epistemic procedure – in this case, that of patriarchal epistemic gatekeeping. It is assumed that were its origins made known to us, we would cease to regard it as justified.

Moral-epistemic justification for pure proceduralists

But how exactly does it matter that one's acceptance of a limitation has a non-ideal genealogy? To assume that providing a non-ideal genealogy of an aspect of a form of moral consciousness automatically undermines its justificatory status is to commit the genetic fallacy. The oppressed ‘may recognise that they acquired certain beliefs or traits under conditions of coercion, but maintain that they would have acquired them anyway, even if they had been in circumstances of complete freedom’ (Geuss 1981, 89). What must be shown by the democratic contractualist, specifically, is that the origin of a form of moral consciousness in a non-ideal procedure makes a difference to its prospects under scrutiny conducted under more ideal conditions.

Here are two seemingly plausible ways genealogy might make such a difference: (i) the form of moral consciousness is the product of a formative process that does not track moral reality, and therefore has a vanishingly small probability of successfully reflecting moral reality; or (ii) the form of moral consciousness is the product of a formative process that would be morally condemned under more ideal moral-epistemic conditions. Neither of these options work for pure proceduralists.

Possibility (i) is obviously unsuitable because, for pure proceduralists, ex hypothesi there is no procedure-independent moral reality out there for a form of moral consciousness to reflect. Retention across successive applications of the ideal procedure constitutes moral justification in the long run, not accurate reflection of moral reality. We cannot rule out in advance the possibility that the form of moral consciousness might be endorsed by the ideal procedure, without appealing to anything beyond the procedure itself. Because proceduralist moral progress is local and pragmatic – in the sense that it starts from the problems we currently face – and not teleological – in the sense of following an ideal trajectory from beginning to end – it makes not a bit of difference where our current form of moral consciousness came from; there is nothing in need of correction independent of the problems we currently experience.

Possibility (ii) is unsuitable for similar reasons. The fact that a form of consciousness arose from a morally reprehensible formative process does not imply anything at all about how that form of consciousness will fare when subjected to a corrective process under more ideal conditions. To condemn a form of consciousness on the basis that is morally ‘tainted’ or ‘contaminated’ by an ignoble formative process goes against the spirit of democratic contractualism: in the absence of any other problems, my squeamishness about a genealogy is quite unlikely to be endorsed as an urgent problem by a community of optimally informed, mutually sympathetic inquirers. Condemning the cause is not the same as condemning the effect.

For pure proceduralists, a non-ideal genealogy by itself shouldn't be fatal to a form of consciousness. At worst, a non-ideal genealogy might motivate us to ‘renew our license’ for subscribing to a moral norm. But a pure proceduralist might argue that the issue is not that a form of consciousness has its origin in a non-ideal formative process, but that adopting a particular form of consciousness makes future deliberations non-ideal. Maybe what matters isn't where false consciousness comes from, but where it might take us.

Something along these lines has been argued by Buchanan (2002; 2004). Buchanan argues that excessive ‘epistemic deference,’ of the sort exemplified by patriarchal moral-epistemic practice, puts us at moral and prudential risk. In the case of patriarchal false consciousness, the epistemic deference of women to men is both morally risky for men complicit in the unjustified and arbitrary frustration of women's aspirations, and prudentially risky for women who submit to male hegemony. Complacency about our form of moral consciousness risks going down a wrongful or harmful path. But given that we are ‘profoundly and unavoidably dependent’ (2004, 102) on our social context as a source of true information (or ‘true consciousness’), the optimal way to reduce our moral and prudential risk is not to eliminate social epistemic dependency altogether, but to design epistemically virtuous social institutions for moral inquiry.

Buchanan's view has some clear affinities with Kitcher's. Both recognise the importance of social institutions for making progress (Kitcher 2021, 96–98). Both recognise that the goal of these institutions is not truth per se but ‘significant’ truth (Kitcher 2001, 43–54, 63-82); we need institutions that are capable of ‘producing, preserving, and transmitting those true beliefs that are most important for acting prudently and morally’ (Buchanan 2004, 105). Both recognise the epistemic virtue of freedom of thought, speech, and association; following Mill, Kitcher has touted the moral-epistemic benefits of ‘tolerant’ (or ‘Deweyan’) over ‘dogmatic’ (or ‘conservative’) moral communities – those that permit individuals to try out alternative ways of living and to compare the lived results with an open mind (Kitcher 2017, 58; 2021, p. 97). Both believe that moral-epistemic authority should always be an achieved rather than an ascribed status; Kitcher hopes that ‘particular individuals – Deweyan philosophers – might emerge as particularly talented in identifying good topics for discussion, offering promising proposals for reform, and mediating the society-wide deliberations’ that lead to moral progress (2021, 98).

However, Buchanan's moral-epistemic argument for liberalism is an insufficient basis for pure proceduralist judgements of false consciousness, for two reasons. First, for pure proceduralists, the form of the ideal conversation is not instrumental to moral progress, as it is for Buchanan, but constitutive of moral progress. So long as the ideal procedure is followed, it is not clear how the form of moral consciousness with which we start can put us at moral risk. If a form of consciousness cannot be off-track, it cannot lead us off-track either. Second, methodological liberalism is not itself a form of moral consciousness, ‘true’ or ‘false.’ Liberal institutions may be characteristic features of the ‘Deweyan society’ (2021, 98) Kitcher envisions – part of the institutional framework within which democratic contractualism can most successfully be put into practice – but within this minimal framework, beyond strict practical compatibility with the democratic contractualist method itself, there are barely any limitations on the forms of moral consciousness that might take root, grow, and eventually become dominant.

The possibility that different forms of consciousness will lead us down different avenues of moral development is not so much a moral ‘risk’ for democratic contractualism but a consequence of the built-in pluralism of pragmatic naturalism. Kitcher recognises that different ‘lineages’ of ethical codes can exhibit ‘different varieties of ethical progress, and this points to a real incommensurability of practices. … Notions of truth and falsehood do not always apply in the ethical domain, for the core of ethical truth is surrounded by a periphery of pluralism’ (2011, 249). The fact that two forms of moral consciousness initiate divergent paths of moral development does not imply that one is ‘true’ and the other ‘false.’ The possibility of future divergence by itself is insufficient to undermine our confidence in one form of moral consciousness or the other.

To conclude: pure proceduralists must be careful to separate the inputs or outputs of a procedure from the procedure itself. Pure proceduralism cannot justifiably exclude an interlocutor on the basis of the content or genealogy of their moral consciousness. Nor can it preclude a possible resultant form of moral consciousness just because we, with our actual moral consciousness, find it distasteful; proceduralism is not determinism. Put differently, pure proceduralism precludes aberrant routes; it doesn't prescribe destinations.

Democratic contractualism cannot make sense of the ‘something wrong’ about failing to make moral progress in the absence of exigent claims. There is no non-question begging way to slide from an obligation to alleviate ‘a limitation felt as a confinement’ (Kitcher 2021, 25) to an obligation to avoid limitations tout court, without appealing to an extra-procedural criterion of the morally true or morally ideal. If other options for living lie beyond our collective imaginative horizons, and if there is no procedurally independent set of available and appropriate ideals of the self out there in advance waiting to be discovered, then democratic contractualists ought to deny, rather than affirm, that we have any reason to periodically reassess our options for living when they are perceived by none (or by what can legitimately be regarded as an epistemically unauthoritative few) to be problematic. Historical and hypothetical instances of extreme ‘false consciousness’ stretch our moral imagination, but our moral imagination is shaped by our actual moral-epistemic context and our own form of moral consciousness. It is only with hindsight that we diagnose the false consciousness of past wrongdoers. Judgements of extreme false consciousness appear anachronistic.

Of course, that's not to say that those women who did perceive their limited options as problematic were mistaken. The point is that the options they imagined for themselves were, tautologically, within their own imaginative horizons. Their problem, therefore, was one of undue conversational exclusion, not one of false consciousness. For the pragmatic naturalist, in the absence of challenges, there are no problems. And for the democratic contractualist, in the absence of perceived problems, there is nothing wrong – or at least, nothing blameworthy – with not making moral progress.

Democratic contractualists must either (i) accept that there can be no obligation to make, nor any sense to the notion of making, moral progress in the absence of exigent claims, or (ii) appeal to an extra-procedural criterion of the morally right or the morally ideal, collapsing their account into perfect or imperfect proceduralism. Of course, pure proceduralists about moral progress might take (i) on the chin; rather than giving up on pure proceduralism about moral progress, we should instead drop one of our peripheral intuitions about moral progress, namely, its obligatoriness in the absence of exigent claims.

I think there are compelling reasons for doing so. The idea that we can acquire moral knowledge by individual reflection – even if we are reflecting under a constraint of mutual sympathy – is arguably a relic of the unattainable Kantian ideal of enlightenment as ‘liberation from tutelage’ (Kant [1784] 1996, p.17f), an implausibly individualistic rationalism that fails to account for the actual social epistemology of morality.

On the other hand, Kitcher is aware that it would be silly to suggest that ‘whenever any of us faces a moral problem requiring deliberation, some group has to be assembled for the purposes of discussion’ (2021, 59, n. 29). Rather than backsliding into rationalism, he suggests that, when faced with a moral problem, our ‘individual deliberations should try to follow the contours of the conversation the pertinent group would have’ (ibid.). What I am suggesting here, though, is that it is incredibly difficult, if not absolutely impossible, for individuals to figure out on their own whether they face a moral problem or not – no reason to even initiate individual moral deliberation, whatever shape it takes. It is just as silly to suggest they ‘periodically’ initiate deliberation, just to be on the safe side. They must be told to do so by members of their moral community. Almost always, being told is the only way to know that something is morally amiss.

All of this is permitted by the reflexive naturalised epistemology underlying pure proceduralism. It is also possible that a form of pure proceduralism other than democratic contractualism might do better. However, as I will argue presently, accounting for some of our less peripheral intuitions about moral progress poses a deeper challenge.

Second challenge: making moral progress matter (in itself)

The very strongest intuition, the most inflexible semantic constraint on moral progress, is that moral progress consists of changes that matter – the idea of moral progress possesses a ‘special dignity’ (Foot 1972, 308), or, less charitably, a ‘mesmeric force’ (Anscombe 1958, p. 8). The exemplars or paradigms of moral progress are specifications of just those moral achievements that embody that special dignity beyond intelligible doubt. Alphabetising my bookshelf can't plausibly be viewed as moral progress because it isn't a change that matters very much. On the other hand, the abolition of the institution of chattel slavery cannot intelligibly be viewed as anything but moral progress, because it mattered so much. A more serious challenge for pure proceduralism about moral progress is to make moral progress matter in the appropriate sense.

This is not a challenge shared by perfect and imperfect procedural accounts, for which moral progress is made to matter by a criterion of the morally true or the morally ideal that is independent of the procedure intended to achieve it. Pure proceduralists, on the other hand, can appeal to nothing beyond the procedure itself as grounds for justifying the claim that moral progress matters – the procedure is constitutive of progress, and therefore progress matters only if the procedure itself can make it matter. The twin tasks not only of accommodating the paradigms but also of capturing the special dignity of moral progress fall entirely on one's constitutive procedure.⁹

Here's one plausible way a proceduralist might make moral progress matter. We start from an initial problem, a need or claim that is fundamental to the kinds of lives we live or the kinds of beings we are. We introduce a procedure that can serve and is justified by that fundamental need. We carry that justification forward through successive applications of that initially justified procedure, iteratively addressing new manifestations of that original need. The procedure and its outputs come to matter by virtue of the need it and they were originally intended to serve.

This approach can explain the paradigmatic status of certain progressive changes in our modern context. Previous applications of (approximations to) the constitutive procedure have, over time, shaped the intuitions we hold today about what can and cannot count as moral progress. The procedure licenses (or forces?) us to regard some things as settled issues, as the dialectical foundations of contemporary moral inquiry and of subsequent iterations of the procedure, placed (permanently or provisionally) beyond subsequent revision. As our moral consciousness is transformed (raised?), some claims become viable topics for moral discussion that were previously seen as inconsequential or discursively off-limits. In this way, what we now view as a paradigmatic instance of moral progress can come to follow necessarily, in retrospect, from the ideal procedure, without appealing to a teleology of moral progress. The more deeply embedded a particular instance of progress becomes – the more densely we build on top of it – the closer to a semantic constraint on the term ‘progress’ it seems in our modern context.¹⁰

In large part, the chosen need is what will define the flavour of pure proceduralism being offered. Pragmatic naturalists regard ethics as a ‘social technology,’ the function of which is determined by the aetiology of ethical practice (Kitcher 2011, particularly Ch. 1-4). The aetiological story at the heart of pragmatic naturalism goes something like this. Early humans possessed a capacity for psychological altruism that enabled them to live together in larger social groups. However, this capacity was limited. Thusly socially enmeshed, they found that they frequently ran into trouble in situations that stretched their limited altruism beyond its limits. Avoiding this trouble was the initial need – it was what Kitcher calls the ‘ur-problem.’ Our ancestors invented something: a system of norms and institutions for enforcing them, a practice of normative guidance that eventually became ethics. This practice helped them to ameliorate the ur-problem. But as with any invention, ethics had bugs that needed patching. Every solved problem churned up unanticipated new problems. They therefore needed a procedure for making moral progress, a way of building on the problem-solving successes of existing ethical practices to settle new claims as they arose. This procedure took the form of democratic contractualism.

According to pragmatic naturalism, we find ourselves in a situation very similar to our early human ancestors, the inventors of ethics and the very first engineers of moral progress. We are similarly socially enmeshed; we run into trouble with comparable frequency when our capacities for mutual sympathy are stretched too thin. However, this trouble manifests differently in our contemporary environment; modern ethics is tasked with treating symptoms that our ancestors could not possibly have anticipated.

Pragmatic naturalism construes the contemporary task of democratic contractualism as relative to our current, local problems and goals, recognising that those problems and goals are partly the result of prior progressive achievements:

Stepwise pragmatic progress is guided by local goals. … New problems emerge from the steps already taken to address older ones. How you go on depends on the decisions made in bringing you to your current place. (Kitcher 2021, 25)

This ‘stepwise’ conception of progress exemplifies the iterative approach described above. There is a need to update the ethical practices we inherited from our forebears in a way that builds constructively on their successes. We therefore inherit their need for a procedure for making moral progress. By virtue of facts about ourselves and our place in the world, so the story goes, the ideal conversation will serve a function that will always be pertinent to our lives, lest we and our place in the world change beyond recognition.

The ‘minimal conception’ of moral progress

We start with a need that will always matter, and we justify our procedure in terms of its continued suitedness to addressing that need. However, a great deal rides on exactly what kind of explanation is offered by the iterative approach, and how we think of the initial need our procedure was chosen to serve. If our very strongest intuitions about moral progress are explained by successive applications of a procedure designed to solve a non-moral problem, or to serve a non-moral desire, then it accommodates those intuitions at the cost of debunking them (cf. Cofnas 2020).

Democratic contractualism starts with the ur-problem, the issue of human beings’ limited capacity for altruism. But there is an ambiguity here: is the ur-problem (i) that of individuals seeking to avoid the prudential costs (subjective, evolutionary, or otherwise) associated with human beings’ limited capacity for altruism, or is it (ii) that of individuals seeking to expand our limited capacity for altruism for its own sake? Kitcher speaks in different places of (i) and (ii) as if they were the same. He writes summarily that the ethical ur-problem is ‘the problem of expanding human responsiveness’ (2021, 52), but this is misleading. Expanding human responsiveness through internalised ‘normative guidance’ is a solution to the problem of ‘socially disruptive behaviour, failure to share with others, the initiation of violence, and the like’ (53). How is this problematic? One possible interpretation is that social instability ‘leads to an evolutionary cul-de-sac’ (Kitcher 2021, 51) – social conflict is evolutionarily costly. This is the interpretation of Kitcher adopted by Buchanan & Powell in their critique of constitutive aetiological functionalism about moral progress (2018, 77–91).

My own reading of Kitcher's analytical history of ethical practice is that the pre-moral problem faced by our distant ancestors was the subjective cost of the limited human capacity for altruism: they ‘recognized the unpleasantness’ (Kitcher 2012, 316) of living with that limitation. By inventing moral practice, that limited capacity became a moral problem in its own right; overcoming it became one of the ‘internal goods’ (MacIntyre 1984, 187f) of that invented practice. This reading is supported by an important but understated footnote, in which Kitcher introduces a distinct mode of progress besides moral progress. He writes of the invention of morality not as morally progressive but as ‘socially progressive … and as introducing the conditions under which moral progress can occur’ (2021, 52 n. 23, his emphasis). For Kitcher, moral practice was – and remains – a means to ‘social progress.’ Whether or not democratic contractualism can make moral progress matter in the appropriate sense therefore depends on what social progress is and the way it matters for pragmatic naturalism.

Kitcher, reflecting on the idea of the good life, offers his own understanding of human flourishing. He identifies three criteria thereof (2017, 56):

1. One's life plan is autonomously chosen;

2. One's chosen life plan is intended in large part to promote the flourishing of others;

3. One's life plan meets with some success.

‘Social progress’ consists in the greater preponderance of human lives that meet (1)-(3). It is important to note that social progress is not on this interpretation an ultimate, free-floating goal for democratic contractualism.¹¹ If it were, democratic contractualism would be an account of perfect or imperfect procedural progress. Instead, I understand the criteria for social progress to be baked into the procedural maxims of democratic contractualism: social progress just is what happens when we model our discourse on the ideal conversation.

Democratic contractualism, interpreted as the vehicle of social progress, adopts of moral progress what Peter Strawson once called a ‘minimal conception’ of morality: as ‘of the first importance as a condition of everything that matters, but only as a condition of everything that matters, not as something that matters in itself’ (1961, 5). The only things that matter to us are the projects and plans we choose for ourselves. They matter – or appear to matter – insofar as they are capable of ‘capturing the ethical imagination’ (3). So construed, the interest an individual has in democratic contractualism is that their autonomously chosen projects for life will be more likely to succeed, and less likely to be obstructed, in a society that settles claims according to the methodological maxims of democratic contractualism. We may indeed share such a common subjective interest in the procedure of democratic contractualism and in social progress – but that is not enough to make it constitutive of moral progress. It must also be the case that the unobstructed fulfilment of our autonomously chosen projects for life is itself morally good – not merely in our individual interests, however benevolent they might be.

This metaethical connection, it seems, is something that must come from beyond the procedure of democratic contractualism. It is not enough to gesture at the (individually or socially) prudential benefits of imagining it to be so.

To his credit, Kitcher is aware of the challenge (2011, §41). He sums it up succinctly: ‘If [pragmatic] naturalism is to succeed, it must differentiate between felt authority from the genuine article and show why such rules are authoritative’ (267). Skipping over much of Kitcher's nuanced discussion, his reply to the challenge is roughly that, if genuine moral authority is to be anything at all, it must lie somewhere in the function of moral practice in human life (269).

I agree. But unfortunately, Kitcher's defence of the genuine authority of democratic contractualism equivocates. What exactly is meant by the ‘function’ of moral practice in our lives? If we mean the aetiological function played by moral practice in making our lives go the way we would like them to go, the minimal conception might plausibly do the job. But if we mean the role of moral practice in our everyday thought, the minimal conception fails.¹²

The role of moral inquiry in our thought is not to ameliorate the ur-problem (construed as above); I suspect this function would be met with bemusement by non-philosophers as a plausible justification for acting as morality requires.¹³ Rather, the role of moral inquiry in our everyday thought is the pursuit of moral propriety – the function of moral practice is to be morally good per se, and the function of moral progress is to become morally better per se. If there is no such way to be, if all we can do is change the subject from morality to prudence, what we have is not a vindication of moral authority, but an error theory of our moral thought and talk – and with it, an error theory of our thought and talk about moral progress. Our very strongest intuitions about moral progress are ‘accommodated’ at the expense of falsifying them.

Again, Kitcher is aware of the issue. He has objected to E.O. Wilson's attempts to ‘biologize’ metaethics along almost identical lines:

I suspect that Wilson … is genuinely torn between two positions. One hews a hard line on ethical objectivity, drawing the ‘profound consequence’ that there is no ‘extrasomatic’ source of ethical truth and accepting an emotivist metaethics. Unfortunately, this position makes nonsense of Wilson's project of using biological insights to fashion an improved moral code and also leads to the unpalatable conclusion that there are no grounds for judging those whom we see as morally perverse. The second position gives priority to certain desires, which are to be uncovered through sociobiological investigation and are to be the foundation of improved moral codes, but it fails to explain what normative standard gives these desires priority or how that standard is grounded in biology. In my judgment, much of the confusion in Wilson's writings comes from oscillating between these two positions. (Kitcher 2010, 383)

Kitcher appears guilty of the same ‘oscillation between two positions’ he attributes to Wilson, but at a higher theoretical level. One position draws an equally hard line on objective moral progress, rejecting the ‘Discovery View’; the other gives priority to a certain constitutive procedure, to be uncovered by historical investigation, and which is to be the foundation of ‘more systematic and more sure-footed’ future moral progress. But – crucially – the normative status of democratic contractualism consists in its well-suitedness to satisfying especially ‘deep’ human desires, namely, the desire to avoid the ‘unpleasantness’ of altruism failures. It must be shown how, without appeal to extra-procedural moral principles, these desires are taken to be privileged (paraphrased from 2010, p. 387). Anything less and the account collapses either into perfect or imperfect proceduralism – or into an error theory of moral progress.

Pure proceduralism and the possibility of error

One last gambit: a pure proceduralist about moral progress might swerve the error theoretical conclusion by going up yet another theoretical level and offering a theory of moral truth according to which their preferred form of proceduralism does not lead to an error theory. This is indeed what the pragmatic naturalist attempts to do, by offering an account of truth that is parasitic on proceduralism about progress.

On this account, ‘ethical progress is prior to ethical truth, and truth is what you get by making progressive steps’ (Kitcher 2011, 210). Along these lines, democratic contractualists think of moral progress as the process of a moral idea ‘verifying itself’ – of ‘truth happening to a moral idea’ (cf. James [1907] 2000, 88f). The role of method is not just to constitute moral progress, but also to constitute moral truth in the long run, and to license justified talk of moral truth in the meantime. Democratic contractualists award the badge of truth to ‘those judgements that prove stable in the long run, when the appropriate method of moral inquiry is followed’ (Kitcher 2021, 75, his emphasis).¹⁴ The suggestion that the method constitutive of moral truth leaves us with an error theory of moral talk seems contradictory.

But it is one thing to show that we can think of moral truth this way, and another thing to show that we ought to think of moral truth this way. For Kitcher, ‘[h]ailing a claim as true announces a resolve to rely on it in decision-making’ (Kitcher 2021, 77). We resolve to rely on it because we meet with greater success when we do. Kitcher therefore advocates a ‘Success-to-Truth’ rule for moral inquiry similar to that which he advocates for scientific inquiry (2012, 112f). The crucial question is whether the kind of success Kitcher has in mind can aptly be described as genuinely moral success.

Perhaps we can make a case in the affirmative by invoking a mode of progress that Kitcher has discussed elsewhere in relation to scientific progress (1993, 114–115): erotetic progress, progress in the form of asking better, more significant, questions. Erotetic progress and the attainment of significant truth can be viewed as two ends of the same process: we ask questions that penetrate more deeply into the essence of our problems; our resultant problem-solving success licenses an inference to the truth of our answers. Different domains of inquiry might warrant different notions of truth on the basis of the different kinds of questions they ask and the different kinds of problems they aim to solve; a question about how to live might call for a different kind of ‘true answer’ to a question about what the world is like.¹⁵ Perhaps moving away from the question of what matters ‘in itself’ to the question of what is and always will be in our subjective interests might be an example of erotetic progress in the moral domain. We are not merely changing the subject; we are asking a better, more significant, moral question (cf. Arvan's 2021 reply to Cofnas 2020). In other words, perhaps the question of what matters ‘in itself’ amounts to a misdiagnosis of the problem of living morally. I believe that this is Kitcher's view.

However, this gambit begs the question of what counts as a significant question and as a true answer in the moral domain. In Science, Truth, and Democracy, Kitcher proposes that significant questions in a particular domain of inquiry are those that would be chosen for investigation after a process of optimally informed deliberation amongst all affected individuals under conditions of mutual sympathy – that is, a question is significant if it is deemed so by something like the democratic contractualist method. But Kitcher is explicit that ‘[i]ndividual preferences should form the basis for our understanding of the personal good that inquiry … is intended to promote’ (2001, 116). If we apply this to moral inquiry, it is a tautology – not an insight – that reframing moral questions in terms of individual preferences is erotetic progress when the good of inquiry is framed in terms of satisfying individual preferences. On the other hand, reframing moral questions in terms of individual preferences is erotetic regress when progress in the moral domain is framed in terms of what is good ‘in itself,’ in a sense not grounded by individual (albeit somewhat altruistic) preferences. We have a choice between two incommensurable, self-vindicating perspectives on what constitutes a good moral question.

I think the only way to arbitrate a dispute between two competing views of moral truth, or of what makes a good moral question, is to consult our very best intuitions about it. Unfortunately, I think our very best intuitions about the kinds of moral questions that make sense tell against pragmatic naturalism and democratic contractualism. As Buchanan and Powell (2018, 86) observe, echoing Moore's open-question argument, ‘it always makes sense to ask whether coping with altruism failures is all that we should ultimately care about, from a moral point of view.’ Any form of proceduralism, starting from any particular need, must account for the ‘special dignity’ of the basis of its chosen procedure – it is unclear how this can be done without appealing to an extra-procedural criterion of the morally true or the morally ideal.

Conclusion: prospects for pure proceduralism about moral progress

The aim of this paper has been to use Kitcher's democratic contractualism as a test case, to assess the prospects of a more general family of metaethical views about moral progress that I have been calling pure proceduralism. I have been arguing that democratic contractualism struggles to do everything we expect from a theory of moral progress. We might drop some of our expectations in light of our investigations, but we cannot drop all of them, without abandoning the very investigation we set out to conduct in the first place. This difficulty is not shared to the same extent by perfect or imperfect procedural accounts. For such accounts, procedures for progress are instrumental to, as opposed to constitutive of, moral progress, and therefore the metaethical buck can be passed. Because pure proceduralists entangle their preferred procedure with their metaethics, they enjoy no such luxury.

Nevertheless, I remain upbeat about the prospects of pure proceduralism. Democratic contractualism is a good start but the potential viability of pure proceduralism as an adequate metaethics of moral progress remains an exciting open question. One potentially fruitful avenue for future exploration is hinted at in another of Kitcher's important but understated footnotes. He recognises that his ‘approach now seems too narrow’ in that it fails to include ‘joint projects’ alongside the autonomously chosen projects of individuals (2021, 52, n. 23). Unlike Kitcher, I do not view the inclusion of joint projects as a simple extension of democratic contractualism. I think joint projects are crucial to the prospects of pure proceduralism about moral progress. This is because some joint projects have the power to create epistemically objective truths – facts about the existence and properties of money, marriages, and even meetings (including ‘meetings of the tribe’) can be made true by our participation in joint projects (along the lines suggested by Searle 1995).

Our collective participation in some joint projects might license a legitimate success-to-truth inference for ‘institutional facts’ such as those listed above. We might think of moral facts similarly. While the aetiological function of our ethical joint project is very likely to mitigate the costs of altruism failures, this is not what makes the project matter; it matters because the joint moral project has come to be an ineliminable part of what makes human life intelligible – without it we cannot make sense of ourselves and our modern form of life. Perhaps morality's ‘special dignity’ can be located here.

Moreover, our ethical joint project is not static. Democratic contractualism, or something like it, might plausibly operate as a constitutive method for generating collective acceptance of new moral facts within our joint ethical project, thereby bringing facts about moral progress into being and making ‘new sense’ of ourselves and our form of life. Such a form of pure proceduralism would chime nicely with alternative Deweyan views of moral progress, such as that of Richard Rorty, for whom much of moral progress consists in modifying our practices ‘so as to take account of new descriptions of what has been going on’ (2001, 191). At least part of the achievement of the ongoing feminist revolution is not, on such a view, that of overcoming a ‘false’ form of moral consciousness, but that of redescribing the situation in which women found and continue to find themselves. The feminist movement is thereby seen ‘as helping to create women rather than attempting to describe them more accurately’ (197). Their success in doing so has depended and will depend on their methodology.

Viewed this way, by reframing questions of moral methodology in terms of the constitution as opposed to the discovery of moral facts, perhaps we can make erotetic progress in our inquiries into moral progress. Kitcher has started the conversation; let's see where we can take it.

Disclosure statement

No potential conflict of interest was reported by the author(s). This study did not generate any new data.

Correction Statement

This article has been corrected with minor changes. These changes do not impact the academic content of the article.

Additional information

Funding

This work was supported by the Arts and Humanities Research Council-AHRC [grant no 2406794].

Notes

1 This view of the jurisdiction of justice is articulated by Rawls (1971, pp. 54-55). The view is not uncontroversial; Cohen (2000, pp. 117-133), for example, questions why the principles of justice shouldn't also apply to individual choices within a just basic structure. For present purposes I’ll accept the Rawlsian demarcation. Kitcher notes these differences from Rawls (Kitcher 2011, p. 105, n. 1) but does not explore their implications.

2 In this regard, Kitcher's approach bears more than a striking resemblance to that of Jürgen Habermas. Compare especially Habermas’ ‘principle of universalisation’ (1990, pp. 62-68) with Kitcher's important seventh maxim for conducting moral inquiry (Kitcher 2021, pp. 33-37). One difference worth emphasising is that for Habermas, but not for Kitcher, the constraints on discourse are grounded in transcendental features of the ideal speech situation (Habermas 1990, pp. 76-98). For Kitcher, our conception of the form of the ideal conversation itself undergoes progressive evolution without ultimate endpoint (Kitcher 2021, p. 37, n. 24; pp. 76-77). Thanks to Philip for insisting I make this difference clearer. Transcendental or endlessly evolving, discursive constraints partly constitute moral progress.

3 How one should draw the boundary of the set of ‘all potentially affected individuals’ is a tricky question for Kitcher. He addresses it, very briefly, in The Ethical Project (pp. 306-311). Sadly, discussing the issue adequately here is impossible due to lack of space.

4 Here I understand certainty along the lines of the later Wittgenstein (1969): as that which cannot be intelligibly doubted in a particular context. I understand semantic constraints similarly: as criteria for intelligible use of a term, in a particular context. Intelligibility in both cases is understood contextually and interpersonally. I clarify my understanding of certainty to avoid accusations of foundationalism. However, if Wittgenstein isn't your bag, the arguments below also suit foundationalist semantic and epistemological theories, as well as ‘holistic’ theories favouring, ceteris paribus, conservative adjustments to our networks of commitment.

5 As noted, it is possible that historical inquiry into some instances of moral history might undermine, rather than vindicate, some of our strongest intuitions about what can and cannot count as a possible or actual instance of moral progress. Naturally, we want pure proceduralism to accommodate only those intuitions that are vindicated in the long run. However, it is also a desideratum for pure proceduralism to explain our undermined intuitions, in terms of deviations from or distortions of the constitutive procedure. Explaining failure is as philosophically valuable as accommodating success.

6 See (Gelfand et al. 2011; 2017) for succinct overviews of tightness-looseness theory. Many thanks to Charlie Blunden for introducing me to tightness-looseness theory and its bearing on the issue of false consciousness.

7 See (Knowles 2021) for recent critical discussion of the idea of adaptive preference.

8 I use the phrase ‘kinds of people’ to refer to the human-made pigeon-holes into which people are categorised, in a very similar sense to Ian Hacking's – see his (1986) and (1996). I emphasise the similarity of my use to Hacking's to disavow any ‘natural kind’ differences between human beings.

9 This challenge is a well-known one in constitutivist metaethics. Harkening David Enoch, we need to have an answer to someone who hears our proceduralist definition of moral progress and responds with ‘Progress, Shmogress!’ (Enoch 2006).

10 This process of moral ‘canalisation’ might also explain the intuition, discussed in previous sections, that had historical agents failed to make progress when they did, they would have done something wrong – in a sense independent of the exigent claims made at the time. Because those changes have since achieved the status of exemplars, by analogy with which we perceive the obligatoriness of certain contemporary moral practices, we cannot help but perceive them as having been obligatory back then as well – even in the absence of any perceived moral demand for them.

11 An attentive reader will note that Kitcher does describe social progress as ‘a well-grounded immediate goal’ in (2017, p. 59). However, Kitcher is explicit that social progress does not describe a perfect and ultimate end goal, in two ways. First, what we regard as a legitimate problem and what we regard as an adequate solution will vary as our quest for social progress runs its course, e.g., the right balance to strike between criteria (1) and (2); the ‘goal’ of social progress itself will evolve over time. Second, the evolving goal of social progress is intended only as a metric for assessing transitions from the current state to a temporally adjacent one, not transitions toward any ultimate end state. Understood properly, Kitcher's talk of ‘goals’ is provisional, pragmatic, and local. In both ways, Kitcher's view of moral progress resembles Kuhn's view of scientific progress (1962, ch. 12).

12 See also (Smyth 2017) for excellent critical discussion of the idea that morality's function can be inferred from its genealogy. My objection in the text is phenomenological: it doesn't feel like what we are trying to do is alleviate altruism failures when we strive to act morally. Smyth's objection is epistemological: he objects to the inference from genealogy to current function.

13 Kitcher's historical claim that, insofar as pioneering abolitionist John Woolman ‘had a moving insight, it lay in the possibility of regarding the selling of human beings as a form of altruism failure’ (2012, p. 318) stretches credulity. I am inclined to agree that the moral sensitivities exercised by Woolman are a product of kin and group selection against altruism failures – but, biographically, it is vastly more plausible that Woolman simply came to see the selling of human beings as morally wrong per se.

14 To reiterate: Kitcher also views the methodology itself as progressively evolving without an ultimate endpoint. On occasions in which ‘methodological adjustment allows for greater systematic success’ (2021, p. 76), we reflexively advance upon our conception of the constitutive method. See notes 2 and 11.

15 See (Price 1994) for illuminating discussion of the various functions truth can serve in different domains of discourse.