The Study of Moral Revolutions as Naturalized Moral Epistemology

I argue for the merits of studying historical moral revolutions to inform moral and political philosophy. Such a research program is not merely of empirical, historical interest but has normative implications. To explain why, I situate the proposal in the tradition of naturalized epistemology. As Alison M. Jaggar and other scholars have argued, a naturalistic approach is characteristic of much feminist philosophy. Accordingly, I argue that the study of moral revolutions would be especially fruitful for feminist moral and political philosophers.

of argument changing hearts and minds in the real world requires a capacious conception of argument. Participants of social movements typically make a number of stock arguments; some have filtered down from those of philosophers and other intellectuals, while others originate bottom-up, emerging from everyday discourse. Moreover, arguments need not be explicit to be powerful; social movements are often inspired and energized by the arguments implicit in activist art, such as the key role that Uncle Tom's Cabin played in the abolitionist movement. Of course, argument of any kind never exists in a vacuum. Certain material conditions must obtain for argument to receive uptake, spark contentious social action, and eventually create institutional change. Each of these factors of social change has traditionally been the subject matter of disparate disciplines-social and intellectual history, literature, and of course sociology and psychology; a complete picture of a moral revolution thus requires a deeply interdisciplinary approach. For their part, philosophers can contribute a grasp of the structure and subtleties of argument-a crucial aspect of social movements that social scientists, characteristically focused on more quantifiable historical data, tend to treat superficially.
Although the study of moral revolutions can provide insight into many areas, I believe the most promising is moral epistemology. The task of moral epistemology is to provide an account of how value claims can be justified. But instead of trying to come up with such an account solely from the armchair, one can look at how value claims are actually justified in the real world to see how such justification in fact proceeds. Moral revolutions are especially interesting cases of such justification since they tend to involve changes in fairly fundamental moral beliefs such as one's conception of justice.
The prospect of studying moral revolutions is no doubt a surprising and even radical departure from the traditional methods employed by philosophers. Accordingly, it faces significant objections. The first is that the project conflates two importantly different issues: the descriptive, empirical issue about how moral justification does go, and the normative, philosophical issue about how moral justification should go. Indeed, the most prominent example of the project so far-Kwame Anthony Appiah's The Honor Code: How Moral Revolutions Happen (2010)has been read in just this way, as being of primarily historical, rather than philosophical, interest (Fischer 2011, 99). The second objection is that the project assumes that moral revolutions involve a genuine change in moral consciousness brought on by rational argument. Yet a purported moral revolution may not represent a change in moral consciousness but only superficial lip service to the moral ideals which we flatter ourselves that we hold. And even when members of a society truly change their minds, we cannot just assume that they do so on the basis of rational argument-a philosophical ideal not always reached, even by philosophers themselves.
The aim of this paper is to argue that, contrary to these worries, the study of moral revolutions is a legitimate and potentially quite fruitful method for moral and political philosophers. I situate the method within the tradition of naturalized epistemology, which provides the resources to respond to the two aforementioned objections. Moreover, the study of moral revolutions will be an especially appealing method for feminists since much feminist philosophy is characterized by an underlying naturalistic approach (Harding 1986;Anderson 1995;Solomon 2001;Antony 2008;Jaggar 2000).
A complete argument for such a method would include illuminating results which come from practicing the method-in methodology as in many other things, the proof of the pudding is in the eating. Of course, an empirically adequate study of even a single moral revolution is well beyond the scope of any single article. Nevertheless, a prolegomenon to that larger study is necessary to explain, at a general level, why such a method is philosophically legitimate and promising.

Naturalized Epistemology: Theoretical Ambiguities and Feminist Affinities
Naturalized epistemology has generated a massive amount of research, not only in epistemology but also feminist philosophy, moral philosophy, and the philosophy of science (e.g., Kuhn 2012). Accordingly, I sketch the view's fundamental commitments in only the broadest strokes, focusing on the simpler case of nonmoral knowledge before moving on to cases of moral knowledge.
Although naturalized epistemology has been arguably practiced since the early modern period (Kitcher 1992), it owes its explicit articulation as a method to W. V. O. Quine (1985). The traditional project of epistemology was to discover some criterion of justification to apply to beliefs, and the traditional strategy for carrying out this project was foundationalism. Quine (1951) famously argued that the failures of foundationalism were insuperable, necessitating that philosophers reconceive the very project of epistemology. According to this new vision, the project of epistemology is to study how beliefs are actually formed. In Quine's words, "The stimulation of his sensory receptors is all the evidence anybody has had to go on, ultimately, in arriving at his picture of the world. Why not just see how this construction really proceeds?" (1985,19).
One advantage of this conception of epistemology is that, in studying how beliefs are formed, we may avail ourselves of all of the resources provided by science. Traditional epistemology, which sought to vindicate science by way of the criterion of justification, could not use empirical data to do so, since such a move would be circular. But once we have given up the project of vindicating science using pure philosophy, we are free to use the results of science in philosophical research.
This philosophical use of science is why Quine's method is termed naturalized epistemology. This is a form of methodological naturalism, since it concerns the method by which epistemology is to proceed, to be distinguished from substantive naturalism, according to which all phenomena may be reduced to natural properties. (The two views are certainly compatible, however.) Naturalized epistemology is thus a method which advocates the empirically informed study of how beliefs are formed.
Quine's article, while groundbreaking, was ambiguous in one crucial way: he did not specify which beliefs we are supposed to study the formation of. The most obvious (one might even say "natural") way to take his claim is that we should study the formation of any and all beliefs. But this leads to an obvious problem, most famously articulated by Jaegwon Kim (1988): epistemology so conceived lacks all normativity. Studying how beliefs are formed is merely a descriptive enterprise. Kim argues that Quine "is asking us to set aside the entire framework of justificationcentered epistemology . . . [and] put in its place a purely descriptive, causalnomological science of human cognition" (388). Kim's objection is a general class of the earlier worry about the study of moral revolution-that it is of purely descriptive, and not normative, significance.
One consequence of epistemology so conceived is that it could make no normative distinction between the formation of our best-established scientific theories and, say, the formation of occult beliefs. As Laurence BonJour (1994, 289) points out, There is, after all, no reason to doubt that occult beliefs are caused in some way by the total sensory experience of the individual, and thus no reason to doubt that psychology can offer an empirical account of how they are produced. Such an account would no doubt differ in major ways from that which would be given for more properly scientific beliefs, but the differences would not, within psychology, have any justificatory significance.
The problem is not merely that naturalized epistemology lacks one theoretical desideratum, normativity. Rather, the problem is that the entire enterprise and point of epistemology has been thrown out the window. If epistemology is an inherently normative discipline, aimed at judging some beliefs to be better than others, then "naturalized epistemology" is not merely misguided, it is in fact impossible-a contradiction in terms.
However, because Quine's position is ambiguous, it permits of an alternative way of understanding naturalized epistemology. Instead of studying the formation of any and all beliefs, we might instead study the formation of a certain set of beliefs which we independently judge to deserve our epistemic confidence. On this view, the project of epistemology is the empirically informed study of how certain beliefs-specifically, beliefs which are clear cases of successful knowing-are actually formed. The advantage of such an interpretation is that naturalized epistemology retains normativity. When we judge a belief to a case of epistemic success, a descriptive understanding of how that belief is formed takes on a normative significance. Beliefs which are the product of similar practices in similar contexts would deserve our confidence, too, and beliefs which are the products of very different practices would look more doubtful. 2 As Louise Antony puts this interpretation of naturalized epistemology, "The decision to treat epistemology as the empirical study of the knower requires us to presume that we can, at least for a class of clear cases, distinguish epistemic success from epistemic failure" (2008,565).
At this point we can sharpen our understanding of naturalized epistemology to see it not merely as a claim about the proper subject matter of epistemology (the formation of beliefs) but as a method with distinct steps: Step 1. We begin by selecting some clear cases of successful knowing.
Step 2. We then use the relevant scientific disciplines to analyze these cases and generate a descriptive picture of how these beliefs are formed.
Step 3. We then use this this descriptive picture to generate standards about the justification of beliefs in relevantly similar contexts (Jaggar and Tobin 2013), standards which we can use to judge less clear cases of belief.
Obviously, this is quite schematic-for instance, we still need to determine what counts as successful knowing in the moral realm and how science could generate a descriptive picture of that knowing. But the point is simply to show where the normativity of naturalized epistemology is supposed to come from, while retaining what is distinctive about the naturalist approach. Perhaps all good philosophy involves starting from clear cases so that one may illuminate less clear cases; naturalized epistemology is distinguished by its commitment to doing this with the aid of the sciences. Why has this approach been so appealing to feminist philosophers? Alison M. Jaggar (2000) and others have argued that naturalism is implicit in a wide variety of major feminist arguments. Naturalism is thus characteristic of, though not definitive of, feminist philosophy; Jaggar in particular has called the feminist naturalist approach "feminism's contribution to moral epistemology" (2000,453). Consider as an example of this contribution Carol Gilligan's (1982) work on care ethics. Gilligan identified differences between men's and women's responses to ethical dilemmas, each rooted in a different type of moral thinking; in light of this, Gilligan argued that dominant conceptions of moral reasoning, which focused on impartial rules to the neglect of caring interpersonal relationships, were male-biased and impoverished. Gilligan's research illustrates all three steps of naturalized epistemology: the identification of clear cases of successful knowing (women's beliefs about the proper response to a moral dilemma; on this, see also Ruddick [1989]); the use of scientific inquiry to generate a descriptive picture of how the beliefs are formed (the account of care-thinking); and the use of the descriptive picture to generate standards for judging beliefs in similar contexts (the criticism of justice-thinking as impoverished). Of course, the specifics of Gilligan's account are contestable, and since her work was published it has been criticized on both empirical and conceptual grounds (Jaggar 1995). But her work nevertheless serves as an illustrative example of the naturalism implicit in many feminist arguments.
It may nevertheless be surprising that feminist philosophers are attracted to naturalism, with its emphasis on the study of the formation of beliefs by scientific investigation. After all, feminist philosophers of science have argued that many major scientific theories were and are male-biased (e.g., Lloyd 2008). And indeed, some feminist philosophers (Code 1996, 2-3) have criticized naturalized epistemology as advocating a narrow scientism, prioritizing "hard" sciences like chemistry, biology, and physics as the best resources for understanding beliefformation. However, naturalized epistemology need not be interpreted in such a narrow way. Quine himself clarified in a later work that the proper empirical study of belief-formation included disciplines like history and sociology (1995,49). Naturalism's primary commitment is to use our best science to understand beliefformation. Feminist philosophy of science does not contradict naturalism but complements it. As Jaggar (2000, 464) points out, feminist critiques of bias in the sciences serve to sharpen the naturalistic method by rejecting the assumption that our best science is identical with whatever scientific theory is socially dominant. The project of uncovering male bias thus helps the feminist naturalist to determine what our best science really is.

The Study of Moral Revolutions as Naturalized Moral Epistemology
I shall now argue that the study of moral revolutions is a promising version of naturalized epistemology, especially for feminists. Seeing why requires a deeper understanding of the nature of moral revolutions.
The paradigmatic examples of moral revolutions are instances in which largescale changes in moral consciousness resulted in the political recognition of those moral claims. I have already mentioned the women's suffrage movement, the civil rights movement, and the more recent embrace of marriage equality. Appiah (2010) argues that the banning of the transatlantic slave trade, the rejection of dueling, and the turn against Chinese foot-binding also constitute moral revolutions. The hallmark of a moral revolution is that there is a "new normal" afterwards. It is not that literally everyone has been persuaded; there are always some retrograde holdouts, sometimes quite a few. But a new moral norm is assumed within group discourse, and its rejection is treated with incredulity and even social censure (Anderson 2016). A moral revolution thus results in a new assumption within discourse that becomes part of the common ground of the communicative group (Haslanger 2012, 455-457;following Stalnaker 2002).
These are epistemic signs that a moral revolution has taken place. The same is true of institutional or structural changes. Although a moral revolution is often completed by such changes, they could occur in a top-down way without a corresponding change in moral consciousness; we might imagine a dictator who imposes a policy repellant to existing social norms. We would be reluctant to call such a shift, even if sweeping, a moral revolution. Constitutively, then, a moral revolution is not (except derivatively) a matter of institutional change but rather a change in a group's moral belief. More precisely: A moral revolution occurs just in case a substantial majority of a society, culture, or subculture comes to accept a general moral belief that had been previously rejected by a substantial majority of that society, culture, or subculture.
The analysis I offer is meant to capture the similarity amongst paradigm cases and be consistent with linguistic usage of the term "moral revolution." In each of the above examples, some movement took aim at a general and widespread moral belief: the morality of limiting the franchise by gender, the justice of segregation, and the propriety of restricting marriage to opposite-sex couples.
Perhaps the most striking feature of this analysis is that it claims that a widespread change in moral belief is a necessary condition of a moral revolution, but it does not specify whether such a change is for the better. This is perhaps obscured by the examples of moral revolutions I have offered, since all of them are widely accepted to have been changes for the better. Yet the aim of the definition is also to be consistent with linguistic usage, and felicitous use does not require one to agree with a moral revolution in order to diagnose it as having happened. Both the political right and the left may agree that the Reagan Revolution happened, understood as a rightward turn in beliefs about the role of government from the 1980s on, while having lively debate about whether it was a change for the better. Likewise, both right and left may agree that the sexual revolution happened, understood as looser, more secular norms around sexual behavior from the 1970s on, while disagreeing about whether it was a good thing or not.
The first step in the method of naturalized epistemology is to discern clear cases of successful knowing. Moral revolutions can typically be described as a change in propositional belief, as opposed to practical knowledge-the moral recognition spurred on by a moral revolution is typically knowing that, rather than knowing how. That is not to say that moral revolutions do not result in different moral practices. Morality is by its nature practical, and when a purported moral revolution involves no change in practice on the ground, we rightly become skeptical of whether the alleged change in moral consciousness has actually occurred. Nevertheless, we can describe moral revolutions as the acquisition of a general moral belief-that women have the same rights as men, that legal segregation by race is unjust, that there is no moral difference between same-sex couples and opposite-sex couples. These count as cases of successful moral knowing when the moral revolution is also a change for the better. How exactly one wants to characterize this epistemic success depends on one's metaethics-a realist will conceive of success as acquiring or moving closer to the truth; a constructivist will conceive of success as having principles more and more adequate to the rational choices or agreements of moral agents. The method of studying moral revolutions can be ecumenical on this point, inconsistent with only those metaethical positions (such as nihilism or noncognitivism) which deny the existence of epistemic success in the moral realm.
Although moral revolutions do not by definition include epistemic success, they constitute unusually good starting points for looking for clear cases of successful moral knowing. While there are some cases of controversial moral revolutions, the researcher who studies moral revolutions has many clear cases to choose from. Few would now seriously argue that women should not hold the franchise or that we should return to de jure racial segregation; accordingly, these are excellent starting points in naturalistic inquiry. Indeed, the advances of previous moral revolutions often constitute our most central and cherished moral beliefs. As Rawls pointed out, moral revolutions such as abolitionism can create "fixed points [in our moral reasoning]: ones we never expect to withdraw, as when Lincoln says: 'If slavery is not wrong, nothing is wrong '" (2001, 29). These are fixed points partly because they are important chapters in the story we tell ourselves about moral progress and our current moral identity. Moral revolutions are thus an unusually robust source of clear cases of successful moral knowing partly because we grow up in a world shaped by those very revolutions.
The second step in the method of naturalized epistemology is to use the relevant scientific disciplines to form a descriptive picture of how the selected beliefs were formed. In the case of moral revolutions, this involves an interdisciplinary analysis with a capacious conception of inquiry which rejects scientism and sees epistemic value in disciplines like literature, art history, and sociology. These various disciplines can help provide us with an understanding of both the material conditions which allow (and sometimes encourage) social change, along with the arguments which direct social innovation, as recorded in congressional debate, public writings by leading figures, court cases, but also activist art and fiction. It must be pointed out that one cannot limit oneself to arguments which drove the moral revolution; reactionary arguments which seek to preserve the status quo can be just as illuminating, showing what the reformers were up against and the dialectical structure of the debate as a whole. (As noted at the outset, at this point the historical record may cause us to lose confidence in our initial belief that hearts and minds had changed, or that it was argument that changed them; I shall consider this possibility in more detail shortly.) The third step in the method of naturalized epistemology is to use the descriptive picture to generate normative standards for evaluating arguments in relevantly similar contexts. This constraint means that one should always be circumspect in generalizing from individual cases. It would be unjustified to take even a richly developed account of a moral revolution to describe all moral discourse or even all moral revolutions. All results from naturalistic inquiry are limited in their generalizability to relevantly similar contexts. But this illustrates why a version of naturalism focused on moral revolutions is especially relevant to feminist philosophers. Because feminists are centrally concerned with the transformation of social structures regulating gender, moral revolutions which changed formal and informal social structures provide an unusually close basis of comparison (e.g., Medina 2013).
Furthermore, any adequate study of moral revolutions is apt to take into account how the agents and groups involved are socially situated, both in terms of how social position influences moral beliefs and how it influences the practice of moral argument itself. This avoids the atomic, hyper-idealized accounts of moral justification which feminist philosophers have argued so persuasively against (e.g., Walker 2007). Jaggar (2012,240) describes the alternative, feminist account of moral justification as consisting in the on-going evaluation of individual actions and social practices by people in actual communities of discourse who collectively construct historically specific ideals, norms and values. On this understanding, moral justifications are socially developed and contingently situated; 'the' moral point of view loses its transcendent status and becomes not single but multiple, rooted in the social world rather than floating above or outside it.
Deriving one's insights about moral justification from the study of moral revolutions helps meet Jaggar's criteria for an empirically adequate account; it is harder for one's theory to become abstracted from real-world concerns when it originates directly from actual struggles for justice and equality.
Having mapped the study of moral revolutions onto the three steps of naturalized epistemology, we are now in a position to answer the first objection mentioned at the outset of this paper. Doesn't recommending the study of moral revolutions in order to understand moral justification conflate the descriptive, empirical issue of how people justify beliefs with the normative, philosophical issue of how they should justify beliefs? The answer is that while the descriptive and the normative are of course different conceptual categories, it does not follow that they are epistemically unrelated. An examination of the origins of moral revolutions is an examination of how a certain set of widespread moral beliefs are formed. When those beliefs are clear cases of successful knowing, as they often are in moral revolutions, a descriptive picture of how that knowing occurs provides insight into what is epistemically valuable in such cases. And this in turn provides a normative basis for judging other, less clear cases. The study of moral revolutions turns out to be both descriptive and normative, empirical and philosophical.

Two Directions in Naturalism
Let us now return to the second objection: The method outlined above assumes that moral revolutions involve a genuine change in moral consciousness brought on by rational argument. It is important to distinguish two ways in which this assumption may turn out to be false.
First, historical investigation may reveal that a purported moral revolution involved no real widespread change in moral consciousness. (The qualifier "purported" is necessary since such a change is constitutive of a moral revolution.) For instance, before the Civil War, many Northern whites opposed the expansion of slavery into the western territories. This initially looks like principled, moral opposition to slavery; yet it may simply be that white settlers did not want to compete economically with slave labor. Moreover, even when people espouse an explicitly moral belief, what people say and what they actually believe can come apart. I noted above that a hallmark of a moral revolution is that a new moral norm is assumed within group discourse. Yet we have seen the way norms of discourse may look dominant, and yet many in the discourse community embrace values forbidden by the new norm, which they then regard disparagingly as political correctness imposed from the outside. The first step of the method of naturalized epistemology is to identify clear cases of successful knowing, yet historical investigation may reveal that they are not clear at all.
A related problem is that one can characterize the moral knowledge gained by a moral revolution in a wide variety of ways. For instance, one may describe the moral knowledge gained by the civil rights movement as knowledge of the fundamental equality of whites and blacks, the injustice of segregation simpliciter, the injustice of de jure segregation, the empirical falsity of the "separate but equal" doctrine of Plessy v. Ferguson, and so on. Yet some of these lessons (such as the injustice of de jure segregation) have clearly been internalized far more than others (such as the injustice of segregation simpliciter). 3 Historical analysis may cause us to revise, sometimes radically, our description of the moral knowledge gained by a moral revolution. Indeed, the arguments made by reformers are rarely accepted wholesale; the general trend of history is that the institutional changes reformers get fall quite short of the deep social changes they want.
The second way the assumption may turn out to be false is that a moral revolution may involve a genuine change in moral consciousness, and yet the change is brought about by factors other than moral argument. For example, Wyoming was the first territory to grant women's suffrage, not because the population had been persuaded by suffragist arguments, but because men outnumbered women so greatly in the new territory that some means of attracting women was needed. 4 Philosophers will have an understandable professional bias in focusing on the role of argument in bringing about changes in moral belief; it is all too easy to project onto the complicated history of moral revolutions a simpler story of reason triumphing against unreason. The second step of the method of naturalized epistemology is to generate a descriptive picture of how these beliefs are formed, yet the resulting picture may reveal only a small role for moral justification.
There is no a priori way to rule out these possibilities-they are real and must be grappled with. Finding out that a moral revolution does not actually involve a change in moral consciousness means having to rewrite our moral history. And finding out that a moral revolution was not brought about by reasoned argument means that we cannot in that case find the positive lessons about moral justification we set out for. However, whether the scenarios envisioned above actually obtainwhether the civil rights movement failed to change the nation's moral consciousness, whether women's suffrage was achieved by moral argument-is an empirical question. As such, they cannot be a reason to forgo the study of moral revolutions; the only way to actually determine whether this worry about the method is legitimate is to actually do research according to the method.
In fact, just these sorts of results have historically been derived from naturalistic inquiry. Nietzsche, Marx, and Freud, for instance, are all skeptical of the idea of moral progress and reason's role in it (albeit for very different reasons). Yet as Leiter (2004, 77) has argued, they should all be classified as methodological naturalists. Nietzsche (1998), for instance, is concerned with the causal origin of Christian moral beliefs, believing that once it is fully understood, it will be revealed as mere "slave morality," thus debunking the idea that the Christian virtues are worth striving for. Marx (1978) likewise sees the dominant ideology as the expression of the interests of the bourgeoisie, thus casting doubt on the idea that real liberation will be found through the enactment of policies that the dominant class recommends. Freud (1990) similarly argues that many taboos are the result of nonrational forces like subconscious impulses. Although their objects of theorization were immensely varied, these thinkers shared a common approach; as Leiter puts it, "When one understands conscious life naturalistically, in terms of its real causes, one contributes at the same time to a critique of the contents of consciousness" (2004, 77; emphasis in original). The possibility of empirical inquiry debunking our initial conceptions is in fact part of the naturalistic tradition.
The real distinction at issue is not between naturalistic and a priori methods of inquiry, but between naturalistic inquiry which gives rise to a negative, debunking result and naturalistic inquiry which gives rise to a positive, constructive result. I have already mentioned the positive contributions made by feminist naturalists. Within mainstream epistemology, reliabilism is an attempt at a naturalistic theory of epistemic justification, since it is concerned with the processes by which beliefs are formed in clear cases of successful knowing (e.g., Goldman 1979).
This widely varied group-Marx, Nietzsche, feminists, and mainstream epistemologists-illustrates that naturalized moral epistemology can pull in very different directions. It can pull in the positive direction of gaining new justified moral beliefs or the negative direction of undermining existing moral beliefs or practices. Both directions are genuine possibilities-neither should be considered the necessary aim or result of naturalized epistemology as such. In short, we cannot rule out the worry that naturalistic inquiry may end up undermining our impression of the moral revolution we sought to learn from. However, this is not a reason to forgo the inquiry-it serves instead as a reminder of the possibilities of that inquiry, so that we do not project onto history our preconceived notions about its course.

Conclusion
Philosophers are trained to work from the armchair. By contrast, I have proposed the study of moral revolutions as an empirically informed alternative method for moral epistemology. As I noted at the outset, the true test of any method is to be found in the results it produces, and to this end I hope to share in the future what I believe are significant findings. Nevertheless, a general defense of the method is necessary, since we should take seriously the objections that any results would be of purely descriptive importance, if they end up showing rational changes in moral belief at all. I hope it is now clear that moral and political philosophers have good reason to study moral revolutions and thereby avail themselves of the rich resources provided by history.