§I. Introduction
Terence Cuneo’s The Normative Web Footnote 1 has often been read as advancing a paradigmatic ‘companions in guilt’ argument for moral realism. Footnote 2 Such arguments aim to show that oft-cited reasons for going anti-realist about moral entities (i.e., moral facts, properties) should apply to other normative entities about which anti-realism is deeply problematic. The implication is that, given that both sorts of entities are sufficiently similar in ways relevant to realism, and given that we ought to be realists about the latter, we ought also to be realists about the former. To defend realism about the latter, then, many have also taken seriously Cuneo’s claim that ontological commitments to epistemic facts are in some sense ‘indispensable.’
Recently, however, Christopher Cowie has claimed that Cuneo has not provided a paradigm for companions in guilt arguments, and that such arguments must fail regardless. Footnote 3 Still, Cowie admits that, because Cuneo’s argument has a distinctive form, it ought to be evaluated on its own terms. Footnote 4 Cowie himself offers no such evaluation. In tandem, commentators on Cowie’s work have focused only on rescuing companions in guilt arguments from his criticisms. Footnote 5 In this paper, I will not be much concerned with the success or failure of those rescue efforts. Instead, my aim is to take Cowie’s admission seriously. That is, I set out to discern whether Cuneo’s argument—whatever its form—constitutes a viable defence of moral realism.
The paper proceeds as follows. In §II, I present Cuneo’s “core argument” for moral realism and explain why he views it as appealing to the guilty companionship of moral and epistemic facts. Footnote 6 I then discuss his argument for why we should be committed to the existence of epistemic facts, owing to their indispensability. I show why Cuneo takes all of this to justify realism about moral facts as well. In §III, I discuss an objection, anticipated by Cuneo, to the effect that his argument for the indispensability of epistemic facts actually undermines his argument for the guilty companionship of moral and epistemic facts, so that his defence of realism about epistemic facts cannot provide support for realism about moral facts. I then discuss his response, in which he makes a crucial appeal to the claim that there are ‘ontologically entangled’ moral and epistemic facts—facts that have both moral and epistemic properties inextricably. In §IV, I assess the cogency of this appeal. My conclusion is that it is illegitimate, and that Cuneo fails to offer a defence of moral realism as a result. In §V, I conclude.
§II. Companions in Guilt and Arguments from Indispensability
§II. I
The Normative Web is dedicated to defending the following “core argument” for moral realism:
-
1) If moral facts do not exist, then epistemic facts do not exist.
-
2) Epistemic facts exist.
-
3) Moral facts exist.
-
4) If moral facts exist, then moral realism is true.
-
5) Moral realism is true. Footnote 7
Cuneo does not avail himself of any particular theory of properties or facthood. Facts are often referred to interchangeably as ‘reasons’ or ‘norms.’ They are moral or epistemic inasmuch as they possess moral and epistemic properties. An example of a moral fact might be that <one ought not to slaughter innocent persons for fun>. An example of an epistemic fact might be that <one ought not to believe everything that one hears>.
Cuneo dubs (1) the “parity premise.” Footnote 8 To defend it is to show that moral and epistemic facts are sufficiently similar, such that arguments against realism about one type of fact should apply to the other. The relevant similarities are, by Cuneo’s lights, fourfold. First, there are both moral and epistemic facts that are categorical. Second, moral and epistemic facts are structurally isomorphic; facts of both types can be general, particular, deontic or evaluative. Footnote 9 Third, moral and epistemic predicates apply to the same sorts of entities: beliefs, intentions, and actions can each be subject to moral and epistemic evaluation. Footnote 10 Fourth, Cuneo claims that “some … [moral and epistemic] facts are not only necessarily coextensive, but also that, in some cases, there is no obvious way to disentangle (ontologically, at least) their moral and epistemic dimensions…” Footnote 11 I will refer to this claim as the ‘ontological entanglement thesis.’ Taken together, these four features of moral and epistemic facts suffice to show that they are sufficiently similar in ways relevant to realism—that is, a realist about moral or epistemic facts who accepts these features as characteristic of moral and epistemic entities, and who takes such entities to exist, is a “realist of a paradigmatic sort.” Footnote 12
Cuneo then examines those considerations that have typically been brought to bear against realism about moral facts. These are grouped together as concerns about their apparently “objectionable features.” Footnote 13 Moral facts are often problematized because their supervening on natural entities is mysterious, their ‘intrinsic normative force’ is metaphysically bizarre, Footnote 14 our epistemic access to them is difficult to explain, Footnote 15 and they are inessential to our best explanations of natural, social, and psychological phenomena. Footnote 16 Cuneo argues that, since epistemic facts possess these same features, the same problems should beset epistemic realists: that is, premise (1) should follow. If, however, one can justify ontological commitments to epistemic entities despite their objectionable features, then epistemic realism will nevertheless be justified. Then, given (1), the case against moral realism should falter when the case against epistemic realism does. This is the reasoning supporting the modus tollens inference from (1)-(3).
§II. II
How to argue for epistemic realism, then? The guiding thought is that, since moral and epistemic facts are sufficiently similar, anti-realist theories about them should also be isomorphic. For this reason, Cuneo examines various contemporary moral anti-realist theories in order to construct and evaluate their epistemic anti-realist analogues. If these analogues can be defeated, epistemic realism should stand alone. Here I focus only on Cuneo’s characterization of a certain expressivist form of anti-realism as articulated by philosophers like Simon Blackburn and Alan Gibbard. Footnote 17 My ultimate ambition is to engage with the positive argument for epistemic realism that emerges out of Cuneo’s evaluation of this sort of view.
In the moral domain, expressivists accept that there is an ‘internal’ perspective that “… captures what it is like to be an agent engaged in ordinary [moral] thought and discourse … giving and assembling [moral] reasons … uttering [moral truths] … and so forth.” Footnote 18 From this perspective, moral realist intuitions make perfect sense: moral disagreements are to be settled by appeal to the moral facts. The expressivist’s key claim, however, is that there is also an ‘external’ perspective from which we can make sense of the actual state of moral discourse and moral ontology. From this perspective, moral discourse merely consists in “… adjusting, improving, weighing, and rejecting … different [moral] sentiments or attitudes.” Footnote 19 Nothing in this picture appeals to the existence of, or any epistemic contact with, categorical moral facts, hence, the ‘non-cognitivist’ moniker often attributed to expressivists. Moral expressivists of this sort think that the external perspective properly represents moral discourse, such that the moral realist cannot avail herself of her desired ontology. Footnote 20
While this does not constitute a robust argument for so much as a characterization of moral expressivism of a certain sort, Cuneo sees enough in it to develop an epistemic analogue. He tentatively allows the epistemic expressivist to play off of the same sort of perspectival dualism as her moral expressivist counterpart. The internal perspective is to be characterized as that from which agents judge beliefs as strongly or poorly justified, assess the merits of various doxastic policies, and so on, where judgements about these matters are taken to track the existence of epistemic facts—categorical facts about what one ought to believe and how one ought to go about properly forming beliefs. From the external perspective, however, accepting or rejecting such claims merely amounts to expressing various attitudes toward epistemic judgements; our judgements do not, from this vantage point, track an objective epistemic reality.
But, Cuneo notices, the epistemic expressivist undercuts the possibility of justifying her own view. For “… it is very difficult to see how there could be a perspective in which a person at once engages in theoretical inquiry and does not believe (or take for granted) that there are epistemic reasons.” Footnote 21 The epistemic expressivist tries not to grant that there are such reasons, all while attempting to justify epistemic expressivism by way of positing reasons for the truth of her view. Her positing such reasons is inescapable if she is to be party to the debate in the first place. So it seems that “… anything we could recognizably call ‘theoretical inquiry’ … involves viewing ourselves as assembling reasons, epistemically evaluating claims, offering arguments, and so forth” which means that, “… anything we could recognizably call theoretical inquiry requires taking not the external, but the internal perspective.” Footnote 22 Even the epistemic expressivist must take herself as tracking reasons for the truth of her view, if she is to justify it. She cannot coherently adopt the external perspective, and so she cannot appeal to it in her argument against the epistemic realist. She cannot appeal to a perspective from which ontological commitments to epistemic facts are not required.
Cuneo later reiterates the point in terms of the ‘indispensability’ of epistemic facts. Footnote 23 Theoretical inquiry is, to use a phrase of David Enoch’s, an indispensable qua “rationally non-optional” project. Footnote 24 It is, for Cuneo, a project inescapably characterized by a “content platitude” of attempting to “represent reality aright.” Footnote 25 Partaking in this project requires quantifying over epistemic facts: we cannot even begin to theorize unless we accept that there are facts about how to do so in a minimally rational way. Footnote 26 Crucially, for Cuneo, we are justified in quantifying over such facts even though they may still display the objectionable features earlier discussed, precisely because they are indispensable. More than this, Cuneo believes that the indispensability of epistemic facts “… implies (or at least gives us powerful reason to believe) that epistemic reasons exist.” Footnote 27 Hence, we should be persuaded of the truth of epistemic realism.
Cuneo’s hope is that we should be persuaded of the truth of moral realism as well, so long as we are sympathetic to the parity premise of his core argument, as well as his argument for epistemic realism. In §III, however, I consider an objection to the effect that this is not so, since the argument actually undermines the parity premise.
§III. Ontological Entanglement and Indispensability
Although it may seem as though Cuneo has fruitfully combined the parity premise with an argument for the indispensability of epistemic facts in order to justify both epistemic and moral realism, one might worry that the two are actually at odds. The objection is that, while Cuneo’s argument from indispensability may justify ontological commitments to epistemic facts, there is no obvious sense in which moral facts are indispensable. If so, Cuneo’s argument from indispensability might unwittingly reveal a problematic disparity between moral and epistemic facts. Moral facts will not be protected from elimination since, unlike epistemic facts, there will be nothing about them (no indispensability) that suffices to trump concerns about their objectionable features.
Cuneo rejects this implication, however. At this point, it helps, so he thinks, to look at plausible examples of indispensable epistemic facts. He takes his inspiration from John Locke’s “rules for the conduct of the understanding” Footnote 28 in putting forward examples like <“If S [an agent] has excellent evidence for p [a proposition], then (all else being equal) S ought not willfully to ignore that evidence”>. Footnote 29 Epistemic facts like this are “bridge conditionals” in that they “link ordinary epistemic reasons with appropriate ways of governing our mental conduct.” Footnote 30 They are indispensable to agents, like us, who operate in “sub-optimal epistemic conditions” Footnote 31 rife with material and social impediments to easy knowledge acquisition. Footnote 32
The point of invoking these examples is to show they are also, so Cuneo thinks, instances of his fourth satisfier of parity—they are what he has called ‘ontologically entangled’ facts. These facts are entangled insofar as they possess both moral and epistemic properties. Cuneo makes the case for such facts having entangled normative properties by pointing out that failure to abide by them renders agents negligent or careless, where these constitute moral failings as well as epistemic ones. Footnote 33 This is especially obvious when we consider a potentially indispensable epistemic fact like <one ought to take the testimony of rational, trustworthy interlocutors seriously>. The idea is that abiding by it “consists in expressing respect of a kind” Footnote 34 where respect, however cashed out, is a moral attitude. A more general example might be that <one ought not to engage in habits of sloppy or biased thinking when discoursing with an interlocutor>. Footnote 35
It is not just that these facts are moral because abiding by them happens to amount to expressing respect of a kind, however. The entanglement of a fact’s moral and epistemic properties can also be explained, so Cuneo thinks, by noticing that “[moral content] platitudes tell us that … entities of certain kinds have one or another positive moral status in virtue of their being appropriately connected with the promoting, honoring, or sustaining of human flourishing,” and that “… cognitive states such as knowledge, insight, and wisdom” contribute to human flourishing. Footnote 36 Because all epistemic facts contribute to flourishing by contributing to cognitive states like knowledge and wisdom, and because this suffices to render them moral facts, Cuneo claims that “… epistemology is a branch of ethics.” Footnote 37 And now it seems that a response to the present objection is also at hand. For, given that there are indispensable, ontologically entangled facts, it follows that there are indispensable moral facts. So there are indispensable moral facts just as there are indispensable epistemic ones. So there is no problematic disparity between moral and epistemic facts.
Cowie has called Cuneo’s argument the “argument from hybrid facts.” Footnote 38 I adopt his moniker in what follows. For reference, I represent it in standard form:
-
1) Having to quantify over moral and epistemic facts is sufficient for justifying moral and epistemic realism.
-
2) We have to quantify over entities that are indispensable to likewise indispensable projects.
-
3) Theorizing is an indispensable project.
-
4) Some epistemic facts are indispensable to theorizing.
-
5) We have to quantify over some epistemic facts (1-4).
-
6) Epistemic realism is justified (1, 5).
-
7) Facts that are appropriately connected with human flourishing are moral.
-
8) Epistemic facts are appropriately connected to human flourishing.
-
9) Ontological entanglement: some epistemic facts are moral facts (7-8).
-
10) Some moral facts are indispensable to theorizing (4, 9).
-
11) We have to quantify over some moral facts (1, 2, 10).
-
12) Moral realism is justified (1, 11).
Recently, Cowie has wondered whether this argument really does exploit Cuneo’s parity premise. Footnote 39 Certainly, it does not occur as a premise in my reconstruction, though one of its satisfiers—the ontological entanglement thesis—does. In any event, Cowie believes that arguments exploiting the parity premise are subject to a dilemma. The first horn essentially restates the objection just addressed: that arguing for realism about one type of normative fact (here, epistemic) will tend to reveal that such facts have a special property (here, indispensability) not possessed by the other type, generating a problematic disparity. The second horn is that, if one can show that moral facts also have the special properties that epistemic facts have (as the argument from hybrid facts attempts to show), then it is pointless to argue for epistemic realism en route to arguing for moral realism. For, if there are indispensable moral facts, it directly follows that moral realism is true. Cuneo’s appeal to the existence of epistemic facts will be “dialectically redundant.” Footnote 40 The argument will not depend on an appeal to the guilty companionship of moral and epistemic facts even if some are, in fact, token identical.
This must mean that Cowie sees Cuneo as arguing for the existence of facts that possess indispensable moral properties just as they possess indispensable epistemic ones. For, if there are no such indispensable moral properties, the indispensability of moral facts will be entirely due to their indispensable epistemic features. If so, quantifying over moral facts will only be licensed insofar as they are identical with indispensable epistemic ones, which will require our recognizing their epistemic properties. But this will require appealing to epistemic realism en route to an argument for moral realism after all.
In what follows, I will not be concerned with whether Cuneo’s argument fails to require us to appeal to the guilty companionship of moral and epistemic facts. This is not my problem, since I do not care about the cogency of companions in guilt arguments in and of themselves: I am strictly concerned with whether the argument from hybrid facts suffices to justify moral realism. As such, the problem should not bother those sympathetic to moral realism so long as they can, as Cowie thinks they might, argue directly for moral realism by positing the existence of indispensable moral facts. My problem lies here: if the argument from hybrid facts is to justify moral realism, it must turn out that indispensable moral facts are indispensably moral, rather than merely indispensably epistemic. A fact will be indispensably moral if it possesses indispensable moral properties: if it is indispensable to the indispensable theorizing project in virtue of its moral rather than (or in addition to) its epistemic features. Unfortunately, as I will argue, there are no indispensable moral properties. So there is no reason to be forced to quantify over facts that have moral and epistemic properties.
Of course, settling the question of whether there are indispensable moral properties depends on figuring out what putative hybrid facts are indispensable in virtue of. Settling this question will depend on our ability to tell the difference between the moral and epistemic features, or contents, that hybrid facts purportedly have. One might think that this cannot be done, and that this bolsters the plausibility of Cuneo’s claim that moral and epistemic properties are inextricably entangled. Unfortunately, it really must be the case that we can make these separations of normative content, lest the argument from hybrid facts collapse into unintelligibility. For, if we cannot make these separations, we will be unable to justify our beliefs in facts that have moral and epistemic features. We will lose reason for thinking that we track entangled facts at all, where entanglement is a two-place relation of two normative properties. This is because our attributing normative properties to facts depends on our having the conceptual resources for doing so. But we deny ourselves the resources for attributing both moral and epistemic properties to facts if we deny ourselves the resources for distinguishing between their moral and epistemic contents. Footnote 41
In §IV, I will show how the moral and epistemic contents of putative hybrid facts ought to be separated. I will then show how, once we grant that putative hybrid facts admit of conceptually isolable moral and epistemic components, Cuneo’s argument from indispensability will only necessitate (if it necessitates anything) a commitment to epistemic properties and facts. My eventual conclusion will be that Cuneo’s argument from hybrid facts does not justify moral realism, since we will only be forced to allow (1)-(6) of the argument to go through. Footnote 42
§IV. Disentangling Ontological Entanglement
§IV. I
My concern about the differences between moral and epistemic sorts of indispensability can be brought to light by considering a dialectic that might play out between Cuneo and anti-realists of an error-theoretic stripe. The initial worry is simply that Cuneo’s argument helps itself to a flourishing-centric conception of the moral domain. Hybrid facts are, after all, taken to be epistemic facts that also have moral properties inasmuch as they contribute to human flourishing. Hence, if one rejects a flourishing-centric conception of the moral domain, there may be no reason to welcome such facts into one’s ontology.
Cuneo concedes that his flourishing-centric conception of the moral domain best suits his metaethical purposes. However, he acknowledges that even utilitarians and Kantians can conceive of epistemic facts as advantageous or obligatory along moral lines, such that they too have reason to quantify over hybrid facts. Perhaps abiding by certain epistemic facts ensures a good chance of contributing to the greatest utility for the greatest number of people, or that abiding by them is a matter of moral duty, owing to their universalizability. Footnote 43 Cuneo says no more about this, but let us concede that different conceptions of the moral domain can make a claim to tracking hybrid facts. In any event, it is clear that Cuneo simply thinks that it is plausible, as a matter of fact, that epistemic facts can be seen as moral-epistemic facts along any of these lines.
Moral error theorists, however, are free to admit that moral realists have an interest in tracking epistemic facts in moral discourse. They will only be required to deny that this point suffices to motivate positive ontological conclusions about moral entities. Footnote 44 Realists like Cuneo may respond by emphasizing that, since we have already admitted to the existence of epistemic facts on the ground that they are indispensable, and since these very facts are tracked in moral discourse, the moral error theorist is now advocating for a bizarrely gerrymandered error theory according to which she welcomes facts with epistemic properties yet refuses to do the same for their putative moral properties. She might be accused of a kind of selective blindness about the normative properties of certain normative facts.
In my view, however, the moral error theorist’s resistance is worth serious consideration in our dialectical context, since the question is whether epistemic facts must be tracked according to their putative moral features—whether such facts must be seen as indispensable from both moral and epistemic points of view. For, if an agent can satisfy the requirements of the indispensable theorizing project without committing to the existence of moral properties, then this should count as a strike against the indispensability of our attributing moral properties to epistemic facts. Again, the point has bite only if moral and epistemic points of view can come apart. I have already noted the problems that beset the argument from hybrid facts if they cannot. But I have yet to show how to separate them, given the need to do so.
As it turns out, the separation is quite easy to draw. One needs only to begin by imagining my encountering an agent engaged in theoretical inquiry of some kind. It follows, if one is sympathetic to Cuneo’s argument for the indispensability of certain epistemic facts to the indispensable theorizing project, that she must abide by these facts on pain of irrationality. If we take, again, an example to be that <one ought not to engage in habits of sloppy or biased thinking when discoursing with an interlocutor>, it follows that she ought to view this fact as epistemically categorical: it satisfies the epistemic content platitude of helping her to represent her reality aright and, in light of her sub-optimal epistemic conditions, she must abide by it.
Now, if I go on to suggest to her that her failure to properly abide this fact (to take it as a norm) would render her negligent, disrespectful toward her interlocutor, and so on, such that she is rightly subject to moral disapprobation (and such that she can rightly judge me to be of imperfect moral character if I fail to abide it), it is easy to imagine her as learning something new. She has learned that her epistemic conduct has implications for forming and maintaining bonds of respect with other agents, and that this ought to concern her. She has come to believe that there is a distinctively moral rationale undergirding her norm. It is not just that, in abiding it, she puts herself in a better position to properly justify her beliefs, track the truth, and so on. Now there are considerations about other agents that occur alongside her personal epistemic ones. She begins as an agent who affords her interlocutor a certain respect qua source of knowledge in order to attain to a minimally functional epistemic standing. She transitions into being an agent who abides the norm both for her own epistemic standing and to properly dignify her interlocutor.
The problem is that the intelligibility of her being able to transition from one to another conception of the normative status of her norm undermines the suggestion that she is necessarily committed to the existence of a hybrid fact from the outset. This is because, if she can successfully grasp an epistemic norm in virtue of its epistemic content alone, as it seems she can, she can thereby succeed in engaging the theorizing project without having to recognize the existence of moral entities—here, the moral properties that she might otherwise attribute to that very norm. Footnote 45 Granted, realists might prefer to imagine that she must have been committed to the existence of moral properties all along and has simply come to realize that her epistemic norms have them, but this alone cannot skew things in the realist’s favour. For, if I am right, this simply begs the question: if she could just as well deny (or ignore)—as the moral error theorist does—that my pronouncements about the moral properties of her norms are legitimate, then we have to ask what it is that forces us to accept moral realism in the way that we might be forced to accept epistemic realism. The realist, therefore, cannot respond by claiming that my argument conflates what we know about our ontological commitments with what they ‘actually are.’ Footnote 46 Insofar as we are trying to justify our ontological commitments to normative entities while taking concerns about their objectionable features seriously, and insofar as we are treating considerations of indispensability as sufficient to trump them, we are trying to determine what our commitments actually are by determining what they have to be.
The point can be pushed even further. For, one might just as well say that, in the above case, the agent who takes herself to learn that her epistemic norm has moral content, and goes on to explicitly endorse the existence of moral properties, does not discover something new about her epistemic norm but, instead, takes on a new norm. One can understand her as abandoning (or supplementing) her ontological commitment to an ‘individualistic’ fact which she successfully abides whenever she is sensitive to the testimony of others in order to represent reality aright, without necessarily having any view to what this does for her community’s general epistemic welfare, or for her interlocutor’s dignity. She might then be understood as coming to endorse a ‘social’ norm, superficial in appearance to the first, but which she successfully abides whenever she is sensitive to her interlocutor’s testimony in order to represent reality aright as well as dignify her interlocutor. If norms are dyadic in this way, as I think they are, we face the question of whether an agent needs to abide by both norms in order to successfully partake in the theorizing project. Footnote 47
In keeping with the above, I find it perfectly plausible to think that the first norm can do just as much as the second, at least where the goal is to satisfy the indispensable prerequisites for theorizing in sub-optimal epistemic conditions. It is perfectly well suited for this purpose even though it lacks moral content. But now the charge can no longer be, as it was at the outset of the discussion, that the moral error theorist who accepts the existence of epistemic facts is guilty of selective bias regarding the moral properties of her norms. She can now claim that she simply does not commit to the existence of putative moral-epistemic facts in the nearby woods: she does not commit to the existence of hybrid facts, even if they are conceivable. In claiming this, she is no longer potentially guilty of a kind of blindness about the normative properties of her very own ontological commitments. One must wonder anew what, if anything, could force her to acknowledge the existence of norms with moral properties.
Note, finally, that my argument is perfectly general, in the sense that it does not matter what conception of the moral domain with which one works. A fact may be conceived as that which produces the best consequences for the greatest number of people when abided, or as morally binding owing to its universalizability, or as contributing to human flourishing, but if the fact need not be seen in these ways in order for an agent to satisfy the necessary conditions for theorizing in sub-optimal epistemic contents, then it can hardly be cast as indispensably moral even if one grants that it is indispensably epistemic. Footnote 48 So there is a way to block premises (7)-(12) of the argument from hybrid facts even while granting premises (1)-(6). So the argument from hybrid facts fails.
§IV. II
Before concluding, I consider one important objection to my argument against the indispensability of hybrid facts. The objection is that I have helped myself to a particular account of fact (norm) individuation centreing on the individuation of normative contents which, in turn, involves separating out the various rationales (be they individualistic, social, or whatever) that serve as the sources of those norms. It may be complained, however, that my argument only works if there are no viable alternative accounts of norm individuation on offer. Footnote 49
If this objection is to get off the ground, it must be that there is an alternative account of norm individuation, the alternative must itself be plausible, and it must support the argument from hybrid facts. One such alternative might be called the ‘raw bearings on action’ account of norm individuation. The basic idea is that, if a moral norm and an epistemic norm both call for an agent to perform one and the same action, A, then those norms must be identical. Such a norm might be indispensable in virtue of its epistemic features alone. However, since that norm also satisfies certain criteria for being categorized as moral (since there are moral considerations that call for the action that the norm bears on), we must also accept that it is a moral norm.
There are two things to say in response. First, if the argument from hybrid facts ends up forcing moral realists to take up the raw bearings on action account of norm individuation, this is so much the worse for those who would hope to remain neutral on these matters.
The second point has to do with the viability of the raw bearings on action account of norm individuation itself. The worry is that, if norms are to be individuated strictly by the actions they call for and the categorizations they happen to satisfy, we will be forced to realism about too much. For instance, a norm may be categorized as moral and epistemic since it satisfies certain moral and epistemic desiderata and bears on action, A. But that same norm might end up counting as a schmepistemic just as it is epistemic. This is so even if the aims of epistemology and schmepistemology are at cross purposes—that is, even if the rationales supporting schmepistemic and epistemic actions are incompatible. For example, a norm may be categorized as schmepistemic insofar as schmepistemologists have an interesting in performing an action, A, because it helps one to form beliefs contrary to whatever judgements about the world one arrives at in performing it while one and the same norm bearing on that action, under its epistemic classification, rationalizes performing that action for wholly opposite reasons. Likewise, there might be a schmoral categorization of one and the same norm that advocates for avoiding habits of sloppy or biased thinking when discoursing with interlocutors because it satisfies the schmoral desiderata of helping one better understand and manipulate her interlocutor to one’s own will, while under its moral categorization it could never be endorsed for such a reason. Since we cannot avoid committing to the existence of a schmoral norm simply by saying that we reject the rationales of schmoral thinking, because norms are not to be individuated according to their rationales, and since the same is true for any normative categorization so long as one and the same action is tracked by it, the raw bearings on action account forces us to take one and the same norm as having schmepistemic, schmoral, moral, and epistemic properties. We will be forced to endorse the existence of all sorts of normative properties that are attributed to a single norm for reasons wholly incompatible with one another.
It is obvious to me that this result ought to be avoided: surely we want to allow ourselves to say that the ends of some normative systems are perverse or bizarre, and that we ought not necessarily to argue for the existence of the normative properties (be they moral, schmoral, or whatever) that these systems attribute to our norms. The only way to avoid these problematic ontological commitments is, by my lights, to say that schmoral, schmepistemic, moral, and epistemic norms (as well as any possible combinations thereof) are distinguishable despite the fact that they bear on the same actions, and to deny some of them into our ontology.
I do not claim that this is the last word on these matters, not least because there may be alternatives to norm individuation beyond the two that I have considered here. If there are, however, it is hard to see what they might look like, and it is hard to see why we should endorse them. Individuating norms by their rationales seems, to me, integral to understanding why norms should be endorsed, and for seeing what sorts of normative properties we really ought to take them to have. Just because norms can be variously categorized does not mean that we should be realists about all the normative properties such categorizations call for us to attribute to them. Therefore, I continue to hold that my account of norm individuation is superior, and that the argument from hybrid facts fails as a result.
§V. Conclusion
Even though the argument from hybrid facts fails, one might wonder whether Cuneo leaves us with the ingredients for a promising companions in guilt argument for moral realism. One such thought is that, if Cuneo’s argument for the indispensability of epistemic facts is successful, one might argue that a precedent of sorts is hereby set: a precedent of exceptions to the ‘rule’ that one ought not to quantify over entities with objectionable features. Perhaps, then, with Ramon Das, we should go on to think that this precedent in the epistemic domain legitimizes distinct considerations in favour of moral realism. Perhaps, in order to avoid odious moral conclusions like <“Hitler did not act wrongly in instigating the Holocaust”>, we ought to commit to moral realism. Footnote 50 One might even think that this line of reasoning supports quantifying over hybrid facts as well: that we should attribute moral properties to epistemic facts in order to avoid odious moral conclusions, even if those moral properties are not indispensable ontological posits.
In response, I have only two brief points to make. The first is that, having taken for granted the prima facie plausibility of arguments from indispensability, I have not thereby defended them. Debates about their tenability remain live. As such, the success of the line of reasoning just entertained on behalf of the moral realist is contingent on the results of these debates. But, more importantly, even if everything goes through as hoped, arguing for the existence of hybrid facts will play no special role in justifying moral realism. This should be clear since, if Das’s argument works, it also works to justify ontological commitments to putative moral facts like <“Hitler was wrong to instigate the Holocaust”> which do not have any epistemic properties. This means that Das’s argument does not require one to appeal to the existence of hybrid facts, since appeal to non-hybrid moral facts will also do. Appealing to the existence of hybrid facts will play no distinctive role in arguments for moral realism.
I have provided, then, reasons for rejecting the argument from hybrid facts as an argument for moral realism. The central strategy has been to pull apart the possible normative contents of putative hybrid facts by looking to the various sorts of (distinctly) moral and epistemic rationales that support them, and to argue that such facts are only indispensable in virtue of their epistemic rationales (if at all), such that it is only indispensable to attribute epistemic properties to such facts. More than this, I have argued that there may be epistemic facts wholly distinct from possible nearby moral-epistemic ones, and that the latter are not indispensable so long as the former can pick up the slack. I conclude that the ontological entanglement thesis cannot be made to prevent a problematic disparity between moral and epistemic facts from arising, and that the argument from hybrid facts fails as a result.
Acknowledgements:
The impetus for writing this paper came out of discussions of The Normative Web with Dr. Robert Myers, Samuel Steadman, David Rocheleau-Houle, Alexander Leferman, Bogdan Florea Alexandru, and Stéphane Savoie. I give special thanks to Dr. Travis Rieder and—once again—Dr. Robert Myers, David Rochelea-Houle, and Bogdan Florea Alexandru for reading early versions of this paper. Later versions of this paper were presented at the Louisiana State University Graduate Conference in Philosophy in 2015, as well as at a Meeting of the Canadian Philosophical Association in 2016. I am grateful to audiences in both venues for their valuable feedback. Finally, in addition to some of those already mentioned, I am grateful to Jamie Robertson, Olivia Schuman, Olivia Sultanescu, Andrew Molas, and Dan Barron for providing feedback on a more recent draft in a research seminar conducted at York University in the fall of 2016.
