Error Theory in Retrospect
7
Moral Error Theory
Error theories can be defined by their allegiance to three central claims: (1) Our moral judgments aim to represent moral facts, (2) there are no moral facts, and so (3) none of our moral judgments are true. Historically there have not been many error theorists, probably because of the radical nature of the claim that, quite literally, nothing is ever either immoral (not even genocide or egregiously predatory behavior) or morally right (not even protecting the innocent, caring for the vulnerable, or the like).
Owing to the relative paucity of error theorists historically, most discussions of moral error theory begin with the Australian philosopher J. L. Mackie’s seminal Ethics: Inventing Right and Wrong.1 Jonas Olson fills in a bit more of the historical background of error theory by discussing some of the more or less influential twentieth-century figures like Axel Hägerström, Bertrand Russell, Ludwig Wittgenstein, and Richard Robinson.2
The Nature of Moral Error Theory
Moral error theorists affirm all four of the following theses:
1. We experience moral wrongness (e.g.) as an objective feature of the world.
2. This experience has its origin in some non-perceptual faculty; in particular, upon observing certain actions and characters (etc.) we have an affective attitude (e.g., the emotion of disapproval) that brings about the experience described in 1.
3. In fact, moral wrongness does not exist in the world.
4. When we utter sentences of the form “X is morally wrong,” we are misdescribing the world; we are in error.3
The first thesis concerns moral phenomenology, a matter of interior moral experience. Such experiences underdetermine whether what is experienced is mind-dependent or mind-independent, subjective or an experience of something objective, or whether moral properties actually exist. The second thesis is a matter of moral psychology offering, in Olson’s words, “a causal explanation of this moral phenomenology.”4 The third thesis pertains to moral ontology (the matter of what morally exists), and the fourth concerns moral semantics (how moral language is used). Mackie made clear, in his articulation of error theory, that his primary concern was ontological, the factual question of whether there are objective moral facts. Semantic questions of the meanings of moral terms are not at the core of the theory as he understood it.
R. M. Hare famously trivialized the ontological question: “Think of one world into whose fabric values are objectively built; and think of another in which those values have been annihilated. And remember that in both worlds the people in them go on being concerned about the same things—there is no difference in the ‘subjective’ concern which people have for things, only in their ‘objective’ value. Now I ask, ‘What is the difference in the states of affairs in these two worlds?’ Can any answer be given except ‘None whatever’?”5
Mackie demurred, arguing that there is indeed an important difference between the two states of affairs. After all, he wrote, “in the one there is something that backs up and validates some of the subjective concern which people have for things, in the other there is not.”6 Mackie likened Hare’s inability to recognize the important distinction to the positivistic denial that there is a difference between a real world and a Berkeleian mind-dependent one.
Noncognitivists, such as the classical expressivists discussed in the next chapter, readily affirm theses 1–3 but balk at 4 because, so they argue, moral claims express noncognitive attitudes and do not describe the world at all. Rather than 4, they claim that a moral judgment like “Murder is wrong” does not describe anything. Some expressivists might wish to reject 3 by appealing to a deflationary or disquotational analysis of truth. Again, by contrast, error theorists affirm all four theses.
Olson introduces a few additional wrinkles to this discussion and along the way describes error theory like this: “Error theory about some area of thought or discourse, (D), is commonly defined as the view that D involves systematically false beliefs and that, as a consequence, all D-judgments, or some significant subset thereof, are false.”7 So standard moral error theory is usually thought to entail the view that moral thinking, in particular, involves systematically false beliefs—typically about moral properties and facts—and that, as a consequence, all moral judgments, or some significant subset thereof, are false.8
Nonstandard variants of the theory also abound, but here we will direct our attention, as Mackie did, to standard moral error theory.
Mackie’s Relativity Argument
One quick preliminary before getting to Mackie pertains to the distinction between metaethical and normative claims, and this point will serve as an effective segue into Mackie’s exposition of error theory. This is the same distinction that Mackie wished to make in laying out error theory, insisting that it is a theory that applies only to metaethical or second-order moral claims.
The claim Mackie wished to defend is that there are no objective values, which he took to be a metaethical, or second-order, claim. “The claim that values are not objective, are not part of the fabric of the world, is meant to include not only moral goodness, which might be most naturally equated with moral value, but also other things that could be more loosely called moral values or disvalues—rightness and wrongness, duty, obligation, an action’s being rotten and contemptible, and so on.”9
What Mackie did not wish to do was extend moral skepticism to (what he called) first-order claims that he took as straightforward descriptive ones. That acts of cruelty and kindness objectively possess different qualities, he wrote, is undeniable. This is an objective fact able to be apprehended with clarity from attentive empirical observation. Mackie had no intention to deny this, even though he was not convinced that there is an objective disparity in moral value between them. “The present issue is with regard to the objectivity specifically of value, not with regard to the objectivity of those natural, factual, differences on the basis of which differing values are assigned.”10
In one sense, this seems eminently obvious, but on closer examination, when normative claims are taken as robust assertions about what morally ought or ought not to be the case, it raises a serious doubt about Mackie’s dichotomy. His claim that first- and second-order moral views “are not merely distinct but completely independent” strains credulity. At the least it does not seem obvious that if error theory is true as a second-order account, then it would have no implications whatsoever at the normative level of ethics.11 A standard interpretation of Mackie’s account is that a first-order moral claim like “Torture is morally wrong” is indeed false. With that preliminary done, we have said enough to set the stage for our discussion of Mackie.
With his characteristically liquid prose, Mackie offered five reasons to reject objective morality. The first is a relativity argument, then three considerations based on the odd features of morality, and the fifth reason is his account for why people stubbornly cling to notions of moral objectivity despite the fact that doing so is, to his thinking, mistaken.
His relativity argument is based on the widely fluctuating moral beliefs and practices throughout the world and history, which he thought moral skepticism accounts for better than moral realism (objective morality) does. One might be tempted to interpret his relativity argument along this line:
1. Moral beliefs and practices fluctuate a great deal.
2. If so, then there is no objective morality.
3. So, there is no objective morality.
Then one can subject the argument to critical scrutiny with the usual well-worn objections. For example, differences about moral beliefs do not entail differences in or denials of moral truth, any more than differences do in other fields of study. Besides, if relativity of belief about something means there are no truths about it, then the argument would be self-defeating, since neither 2 nor 3 is universally agreed on.
Mackie, though, did not lay claim to that sort of airtight connection. Rather, he was offering only “indirect evidence,” as he put it, arguing that his error theory could better explain moral disagreements and differences than moral realism can. So such criticisms of Mackie seem wide of the mark by treating his better explanation approach as an effort at providing the only explanation.
Besides, Mackie thought there was a relevant disanalogy between disagreements about history or science, on the one hand, and ethics, on the other. In the former, the disagreements do not vitiate the existence of objective truths to be known, at least in principle. The disagreements result from merely speculative inferences or explanatory hypotheses based on inadequate evidence, but he thought it implausible to think the same about ethics. “Disagreement about moral codes seems to reflect people’s adherence to and participation in different ways of life.”12
“In short,” he concluded, “the argument from relativity has some force simply because the actual variations in the moral codes are more readily explained by the hypothesis that they reflect ways of life than by the hypothesis that they express perceptions, most of them seriously inadequate and badly distorted, of objective values.”13
Mackie then anticipated an important objection along these lines: Perhaps what is not relative about morality are very general basic principles recognized at least implicitly to some extent in every society. For example, “the principle of universalizability, perhaps, or the rule that one ought to conform to the specific rules of any way of life in which one takes part, from which one profits, and on which one relies, or some utilitarian principle of doing what tends, or seems likely, to promote the general happiness.”14 Such general principles applied to specific and concrete circumstances and contexts would likely feature some latitude when it comes to the formulation of specific moral rules.
Mackie replied that such an appeal cannot well account for the confidence with which an assortment of specific moral guides to conduct are taken as authoritative. For many such principles are not thought derivative from more general axioms or ideals but are able to be recognized more directly and less inferentially. This makes sense, Mackie thought, of appeals to intuitions or moral sense rather than to the deliverances of reason itself. For such reasons Mackie was convinced that the argument from relativity remains in full force.
A later chapter will deal with relativistic models of morality at greater length, but for now just a word is in order. Among the moral data in need of explanation, as Mackie rightly discerned, is significant moral disagreement. Perhaps less prominent in his analysis is the felt need to address the significant swath of moral agreement that can be discovered both throughout human history and all around the world, including no small number of rather concrete moral axioms and principles.
This is the significance of the appendix in C. S. Lewis’s The Abolition of Man, after he had taken the trouble to compare the ethical codes from a wide range of times and places—from Egypt, India, China, Greece and Rome, Scandinavia, the Anglo-Saxons, Mosaic law, and the New Testament. Among the general headings of what he deemed these principles of universal morality—the “Tao,” he dubbed it—were the following: beneficence toward people in general and toward specific persons; duties to parents, elders, ancestors, spouse, children, posterity; justice; honesty and truthfulness; mercy; and magnanimity.15
By Mackie’s own admission, however, the relativity argument was merely preliminary. His main argument against moral realism was still to come: the so-called trifold “argument from queerness” (“queerness” is meant in the traditional sense of “oddness”).16 To this argument we will turn momentarily, but first one last word about moral disagreement that is germane to the project of this book. Persistent disagreement obtains not just with respect to first-order moral questions of various sorts but with regard to metaethics too. Moreover, the debates among and between adherents of these disparate theories can sometimes seem interminable or even practically incommensurable. Arguments and counterarguments and rejoinders abound, and rarely is there anything like a knockout blow to do away with a theory altogether. This can be discouraging, but from it we should glean both a warning and a reminder. The warning is to not just remain in our comfort zones, listening only to those with whom we already agree. The reminder is that disagreement usually does not mean that there is no truth to the matter. Moral theory is hard, and there are no shortcuts—and no single volume will clear everything up—but it is worth our time and effort to think it through, and prospects for progress are bright.
The Argument from Queerness and the Enduring Appeal of Moral Objectivity
Mackie’s “argument from queerness” has two parts, a metaphysical part and an epistemic part. The metaphysical component itself has two aspects, so all told this argument features three parts. The two metaphysical ways that Mackie thought morality to be odd are these: Morality would have to be binding and authoritative apart from one’s own desires and preferences, and morality would have to supervene on natural features of the world. The epistemic sense of moral oddness pertains to the challenge of accounting for knowledge of morality understood in that way. He thought realism likely to require a specific moral sense, which seemed to him unlikely and ontologically indulgent.
Mackie constructed a basic formulation of both the metaphysical and the epistemic concern like this: “If there were objective values, then they would be entities or qualities or relations of a very strange sort, utterly different from anything else in the universe. Correspondingly, if we were aware of them, it would have to be by some special faculty of moral perception or intuition, utterly different from our ordinary ways of knowing anything else.”17 Let us discuss Mackie’s metaphysical concerns, each in turn, starting with the authority of morality obtaining apart from one’s desires and preferences.
As the expressivists had done, Mackie found part of his inspiration in the iconic writings of none less than G. E. Moore. Moore thought that between the naturalistic fallacy and open question argument, prospects for a naturalistic account of ethics were unavoidably dim.18 So he gravitated to thinking of morality in unanalyzable nonnaturalistic terms. Such talk of nonnaturalism, however, can sound to some rather inexplicable, exuding a connotation of “unreal.” A strong commitment even just to methodological naturalism effectively confines one’s search for moral explanations to the naturalistic realm; any departure or deviation can seem otiose and obtuse, indulgent and infelicitous.
Nothing so ambitious as logical positivism or its verifiability criterion is needed to look askance at appeals to nonnaturalism. Mackie himself rejected such a criterion of meaning as well as its application to morality. He did not consider assertions that there are objective values or intrinsically prescriptive entities to be cognitively meaningless; rather, he regarded them false for being irremediably odd. He considered Plato’s Forms the archetype of objective values:
The Form of the Good is such that knowledge of it provides the knower with both a direction and an overriding motive; something’s being good both tells the person who knows this to pursue it and makes him pursue it. An objective good would be sought by anyone who was acquainted with it, not because of any contingent fact that this person, or every person, is so constituted that he desires this end, but just because the end has to-be-pursuedness somehow built into it. Similarly, if there were objective principles of right and wrong, any wrong (possible) course of action would have not-to-be-doneness somehow built into it.19
Mackie thought that Hume’s point that it is our desires and passions, rather than reason alone, that motivate behavior comports nicely with his own point. To the charge that values have a distinctive power to motivate, Mackie held, Hume would need to augment his analysis by appealing to something like the argument from queerness. For values would have to be, if they possessed such unique motivating power, something on a quite different order from anything else of which we are aware.
One way to encapsulate what is at stake at this juncture of the discussion is to speak of the authority of morality. This is what Mackie was getting at when he discussed morality as prescriptively binding. In the next chapter we will see how R. M. Hare tapped into the notion of Kantian universalizability: When we make a moral decision, we are essentially prescribing that everyone else in relevantly similar circumstances make the same decision. But Hare also denied the existence of moral facts, so what he ended up with was something like prescriptions without authority. Mackie did not adopt this watered-down version of ethics; he was convinced that too superficial an analysis, rather than explaining ethics, would only explain it away. Mackie thought that a distinctive feature of morality is that it is not only prescriptive but authoritative. It is prescriptively binding. This is also what makes it so odd. In essence, then, Mackie found dubious the notion that there are any true moral claims that are, as the right analysis of morality requires, genuinely authoritative.
The second aspect of metaphysical oddness Mackie accentuated, recall, was that morality would somehow need to be linked to natural features of the world. “What is the connection between the natural fact that an action is a piece of deliberate cruelty—say, causing pain just for fun—and the moral fact that it is wrong?” he asked. Mackie did not think it could be an entailment, yet it is also not merely that the two features happen to occur together. “The wrongness must somehow be ‘consequential’ or ‘supervenient’; it is wrong because it is a piece of deliberate cruelty. But just what in the world is signified by this ‘because’?”20 Among the moral data that Cuneo and Shafer-Landau identify as in need of explanation is this very matter of what connects moral properties with the natural properties on which they are thought to supervene. Mackie seemed wholly correct to raise this very question.
Incidentally, interesting to note is that Mackie anticipated the objection some might wish to raise that his invoking the specter of the more exotic products of philosophical fancy—Platonic Forms, nonnatural qualities, self-evident relations of fitness, faculties of intuition, and the like—stacks the deck against moral realism. Did he consider his argument equally effective against more garden-variety moral claims to objectivity captured in such everyday language as “You must do this,” “You cannot do that,” “obligation,” “unjust,” and “disgraceful”?
Mackie admitted it is not obvious that his argument is as forceful against these casual locutions, but he thought he could explain why. Such forms of speech, he argued, do not isolate the objective prescriptivity but are mingled with a variety of other things, like relations to desires and feelings, prevailing contextual standards, reasoning about ways to achieve certain ends, and so forth. There is nothing queer about such things, and “under cover of them the claim for moral authority may pass unnoticed. But if I am right in arguing that it is ordinarily there, and is therefore very likely to be incorporated almost automatically in philosophical accounts of ethics which systematize our ordinary thought even in such apparently innocent terms as these, it needs to be examined, and for this purpose it needs to be isolated and exposed as it is by the less cautious philosophical reconstructions.”21
Lastly, in terms of the epistemic oddness of moral realism, Mackie lumped Moore’s nonnaturalism in with what he deemed discredited appeals to mysterious and epistemically obscure faculties of moral intuition. Although Mackie thought that most could see the problems with intuitionism—primarily some sort of special faculty for apprehending moral truths—he thought it less commonly recognized that this central problem besets any and every objectivist view of values. Even if such views do not adopt a simplistic picture of moral epistemology according to which moral judgments are made or problems solved by just intuiting the answer, still the process of coming to discern the content of allegedly objective moral truths involves some such intuitive input. As Mackie put it, “When we ask the awkward question, how we can be aware of this authoritative prescriptivity, of the truth of these distinctively ethical premises or of the cogency of this distinctively ethical pattern of reasoning, none of our ordinary accounts of sensory perception or introspection or the framing and confirming of explanatory hypotheses or inference or logical construction or conceptual analysis, or any combination of these, will provide a satisfactory answer; ‘a special sort of intuition’ is a lame answer, but it is the one to which the clear-headed objectivist is compelled to resort.”22
Bearing in mind the supervenience concern, Mackie added that it is not enough to postulate a faculty that “sees” moral wrongness. “Something must be postulated which can see at once the natural features that constitute the cruelty, and the wrongness, and the mysterious consequential link between the two. . . . How much simpler and more comprehensible the situation would be if we could replace the moral quality with some sort of subjective response which could be causally related to the detection of the natural features on which the supposed quality is said to be consequential?”23
Before discussing this central matter of moral authority further, a summary of Mackie’s fifth and culminating consideration in favor of error theory is in order. On the basis of his previous arguments, especially the metaphysical and epistemic strangeness of objective moral authority, Mackie thought the last piece of the puzzle involved explaining how such authority has come to be so widely believed and is so resistant to criticism. Mackie offered several suggestive possibilities.
Whereas the language-centric expressivists tended to attribute enduring beliefs in moral objectivity to the human penchant for reification of misleading moral language, rendering something abstract as real, Mackie looked instead to moral psychology and phenomenology. He appealed to what Hume called the mind’s “propensity to spread itself on external objects.” Interpreting morality as realist naturally arises from the projection or objectification of moral attitudes. Mackie considered the process analogous to the “pathetic fallacy,” reading one’s feelings into their objects.24
In the moral case, though, there is also a social dynamic at work. Prevailing community standards of behavior exert pressure to conform and incentives to internalize them. Motives to objectify them include the ways in which certain behaviors are conducive to, and others detract from, harmonious social interaction. There are thus social reasons for wanting morality to be authoritative.25
Moreover, Mackie argued, even more important than the projection of feelings are wants and demands. He invoked Hobbes’s conjecture that the object of a man’s desire or appetite is what he dubs “good.” “We get the notion of something’s being objectively good, or having intrinsic value, by reversing the direction of dependence here, by making the desire depend upon the goodness, instead of the goodness on the desire.” We easily confuse a thing’s objective desirability with our sense of objective value.26
Mackie was also convinced that there is an important historical twist to the full explanation of the widespread convictions about moral realism. Notions of strong moral proscriptions and prescriptions, that some actions are not just wrong but wicked or forbidden, and the like, he thought, can almost certainly be traced to religious influences. With the diminution of religion, the vestigial imperatival force of moral language had resulted in a law without a lawgiver, retaining its supposed authority but having lost its foundations. Convinced by Elizabeth Anscombe’s famous analysis of modern moral theory, Mackie took to heart her concern that traditional moral language—like non-Aristotelian concepts of moral obligation—had merely “mesmeric” force with a “delusive appearance of content.”27
Again, though, Mackie was quick to add that persisting beliefs in moral objectivity even after belief in a divine legislator had faded is tenacious. The picture is a complicated one, in which several influences have converged to fortify ongoing convictions about objective and authoritative morality. Since such explanations are so readily on offer, though, this fact, when combined with the earlier arguments for error theory, makes plausible and justified the acceptance of error theory and rejection of moral realism.
Varieties of Alleged Moral Queerness
Olson lays out four versions of queerness objections, resisting the first three while taking the fourth to provide a good reason to reject moral realism and adopt error theory instead. Let us quickly consider the first three and then turn our attention to the fourth, which resides at the center of this important question of moral authority.
Olson begins by pointing out that, by the time Mackie arrived at discussing the queerness objection, he had set naturalistic ethics to the side. The target of the argument, as he conceived it, was nonnatural ethics. Mackie thought that the eventual result of Moore’s intuitionism was an undermining of good as nonnatural and unanalyzable—and thus mysterious if not inscrutable. The ontologically odd features of moral properties and relations did not make Mackie think assertions about them meaningless, but false. But note that he largely deferred to Moore’s semantic analysis (though not his presumed truth value). In the discussion to follow, then, what needs to be remembered is that the conflict should mainly be understood along the lines of error theory versus a nonnatural account of moral realism.
The first version of the queerness argument is what Olson calls the supervenience objection, predicated on two conceptual supervenience theses:
S1. Necessarily, for any item x that has a moral property, M, x ultimately has M in virtue of some natural property.
S2. Necessarily, for any two distinct items, x and y, if x and y are exactly alike in natural respects, x and y are alike in moral respects.
Olson does not want to beg the question against supernaturalist views according to which moral properties supervene on supernatural properties like being commanded by God. So “natural properties” are understood expansively as either naturalistic or supernaturalistic.
The challenge arises for nonnatural ethical accounts when the discussion moves away from concepts to properties, most specifically, the ontological implications of S1 and S2. How is the covariance of a natural property (like torturing a cat) and a moral property (like wrongness) to be explained? Mackie said nonnaturalists are committed to “synthetic necessary connections” between distinct properties. The reason such connections are thought to be metaphysically odd is something like “Hume’s Dictum,” which says there cannot be relations of necessary coextension between distinct properties because they would be irremediably odd.
Olson, interestingly enough, does not endorse this version of the queerness argument, mainly for this reason: The argument essentially depends on Hume’s Dictum, so its effectiveness is held hostage to whether the suggested counterexamples to it work.28 So he thinks nonnatural accounts of ethics can escape this objection unscathed; an example of such ethics that he uses for an effective counterexample to Hume’s Dictum is nonnatural realism predicated on hedonistic utilitarianism.
The issue that we would like to explore for a moment in this connection, though, is this: In The Miracle of Theism, Mackie considered a different sort of nonnaturalism—namely, theism.29 He broached the issue there of whether theistic assumptions might provide the needed explanation of the aforementioned synthetic necessary connections. It is not our intention here to make the case that theism is the best way to go; we have in mind the less ambitious goal of merely showing that such a possibility remains a viable, and not altogether remote, option.
Mackie frankly admitted that if we look at moral thinking “from the outside, as a phenomenon to be understood and explained,” there is potential for a moral argument for God:
If we adopted moral objectivism, we should have to regard the relations of supervenience which connect values and obligations with their natural grounds as synthetic; they would then be in principle something that a god might conceivably create; and since they would otherwise be a very odd sort of thing, the admitting of them would be an inductive ground for admitting also a god to create them. There would be something here in need of explanation, and a being with the power to create what lies outside the bounds of natural plausibility or even possibility might well be the explanation we require. Moral values, their objectivity and their supervenience, would be a continuing miracle . . . , a constant intrusion into the natural world.30
Of course, Mackie himself remained unconvinced, both because of his Humean-inspired skepticism about miracles and because of his error theory. It is telling, however, that he thought that, if moral realism held, it would provide some inductive reason to believe in God’s reality.31 Theism could provide just the sort of needed explanation for the relevant asymmetric dependence relation of moral properties on nonmoral ones. This is just the point C. Stephen Evans, echoing George Mavrodes, makes when he writes, “If the universe includes God, and it must if moral obligations are divine commands, and God is a moral being, then it hardly seems odd that moral qualities should appear to be ‘deep’ ones, and it will be unsurprising that humans have capacities to grasp such qualities.”32
Appropriately, then, the second variant of the queerness objection that Olson explores pertains to moral knowledge. The point of the argument is that we do not possess knowledge of moral facts and relations. Recall Mackie’s concern that the mental faculty by which we come to knowledge of wrongness and other moral properties would have to be metaphysically peculiar and prima facie implausible in excelsis. But as Olson points out, even Mackie himself wrote that, for the intuitionists, “there is no need to think of this ‘special faculty’ as a separate organ of the mind. As Richard Price says, all that is required is that the understanding, the aspect of our minds which can discover objective truths, has the ability to discover, among others, this particular sort of necessary truth of supervenience.”33
Early sentimentalists like Francis Hutcheson were surely susceptible to the charge of epistemic indulgence, but most eighteenth-century rationalists as well as most contemporary intuitionists like Michael Huemer chalk moral knowledge up to rather garden-variety operations of the mind: the ability to understand and to think.
Olson points out that if Mackie’s real concern was synthetic a priori know-ledge, this variant of the queerness argument (like the first) accomplishes too much. For the argument quickly extends beyond moral judgments. Much philosophical knowledge tends to be synthetic a priori, Olson points out. “Standard examples of synthetic truths thought to be knowable by a priori intuition include ‘Nothing can be red and green all over at the same time’ and the laws and inference rules of logic, such as the law of non-contradiction and modus ponens.”34
Besides, if skepticism about such knowledge per se is the culprit, it would show that the epistemic argument does not stand on its own. It is not at all clear that the way we come by moral knowledge is, in Mackie’s words, “utterly different” from our ordinary ways of knowing everything else. So because of its failure to isolate uniquely queer features of moral and normative properties and facts, this version of the argument falls flat.
The third formulation of the queerness argument pertains to the matter of moral motivation and applies uniquely to moral considerations. Recall how Mackie, using Platonic Forms as his model, insisted that moral realism entails that goodness has a “to-be-pursuedness” built into it. “Something’s being good both tells the person who knows this to pursue it and makes him pursue it.”35 The direction and motivation are not a function of the person’s predilections or proclivities but rather are furnished by the nature of the good itself.
Note that Mackie thought morality should be understood as providing both authoritative direction (a normative matter) and motivation (a psychological matter). Although Mackie did not distinguish sharply between normative reasons and motivating reasons, most philosophers today do; so the version of the argument from queerness about to be discussed will focus only on the motivational matter. The next section will take up normativity.
A natural and common assumption concerning Mackie’s motivation argument made by various expositors is that it derived from two commitments: a Humean theory of motivation and the thesis of motivation internalism. A Humean view of motivation is that what spurs one to action is never mere belief, but emotions, desires, preferences, and the like.36 Motivation internalism is the notion that there is a necessary connection between making a moral judgment and being to some extent motivated to act in accordance with it.
Mackie, however, actually rejected a Humean theory of motivation: “There might be some non-ordinary sort of what was still factual information, some special form of belief or even knowledge supplied by a moral sense as the objective interpretation understands it, which does make moral distinctions that can in and of themselves influence action.”37 “[The] psychological thesis [that] motivation always involves a desire as well as belief [is overstated]. The belief in objective moral requirements, made explicit by such writers as Price and Clarke and Reid, but implicit in much ordinary thinking, can in this curious way act as a motive on its own, even if . . . that belief is false.”38 Most tellingly, “it is evident that there can be sets of moral and factual beliefs which are, by themselves, motives to action.”39
For this reason, Olson offers this interpretation of Mackie’s motivation argument: Although some beliefs can give rise to motivation, moral beliefs are thought to guarantee it, and that would be strange indeed. Mackie seemed to think that the nonnaturalist moral realist was committed to insisting that anyone relevantly acquainted with moral facts would have motivation to comply with them, though Mackie himself arguably did not adopt motivation internalism. He thought no beliefs guaranteed motivation, and, as Olson puts it, “moral belief is not necessarily linked to motivation.”40 Mackie arguably made appeal instead to the Socratic doctrine that knowing the good entails motivation to do the good in order to drive home his point that moral facts possessing such motivating power would be queer.
Does this argument work? Likely not. A moral realist can handily accommodate the phenomenon of people becoming relevantly acquainted with moral facts and remaining unmoved by them. Not every moral infraction is committed out of ignorance; some might know the deliverances of morality and violate them anyway. Amoralists do it on a regular basis, but the phenomenon is wider than that. It is not unreasonable to think that all of us on occasion have exhibited weakness of the will (akrasia), knowing to do right but refraining, or knowing not to do wrong and doing it anyway. If a clear-eyed vision of the contents of morality were sufficient to ensure obedience to its dictates, then mere acquisition of the relevant knowledge would be enough to ensure complete moral maturation, which seems patently unrealistic.
Mackie assumed that nonnaturalist moral realists combine descriptive and prescriptive elements in their theory in such a way that conflates normativity and motivation. But as Olson puts it, “Non-naturalists who are also externalists about moral motivation maintain that just as we can fail to be motivated by true or false judgments about subjective and conventional reasons and requirements, we can fail to be motivated by true or false judgments about objective and non-conventional, or irreducibly normative, reasons and requirements.”41
Irreducible Normativity
The fourth and last way to interpret Mackie’s queerness argument is the one that Olson thinks has the most potential, and it arrives at the central feature of morality in need of explanation: its binding prescriptive authority, what Olson calls “irreducible normativity.” Reducible normativity pertains to standards that apply to specific domains, like etiquette or the rules of boxing. Irreducible normativity invokes standards that dictate behavior irrespective of an agent’s desires to comply or conform (perhaps some might wish to employ the aforementioned locution “stance-independent”).
Descriptions can vary, but they all try to narrow their focus to something like authority: binding prescriptivity, strong categorical imperatives, noninstitutional oughts and reasons that really bind agents, practical clout or “oomph” (as Richard Joyce puts it).42 Olson used to speak in terms of categorical reasons but now favors this language: “Moral facts are queer in that they are or entail facts that count in favor of or require certain courses of behavior, where the favoring relation is irreducibly normative.”43
Error theorists tend to agree with moral realists when it comes to the meaning of moral claims, and they agree, too, that morality is a cognitivist matter. Moral claims have a truth value, and they are what they appear to be: authoritative, binding prescriptions and proscriptions (among other things). As Olson puts it, nonnatural realists and error theorists agree on the “conceptual claim” that moral facts are or entail irreducibly normative reasons.
Where error theorists demur from the nonnaturalist moral realists, though, is on the ontology of morality. The former think that there is no good reason to think that such odd entities as moral facts—with their presumed authority that obtains irrespective of the agent’s desires—actually exist. There are no irreducibly normative reasons in reality. So as a result, moral claims are taken to be categorically false. Naturalist moral realists (like the Cornell realists) attempt to demystify morality, but Olson (rightly, we think) reiterates Mackie’s concern that such accounts leave out the “categorical quality” of moral judgments.
In terms of the debate between nonnaturalists and error theorists, their disagreement can easily be seen as intractable, despite the fact that their views on moral semantics largely coincide. Whereas the error theorists see normative reasons as unavoidably odd and metaphysically mysterious, the nonnaturalists assume as axiomatic the authority of morality. The realists insist that if such authority requires some rich ontology, so be it—one person’s modus ponens is another’s modus tollens. As Olson puts it, nonnaturalists “could maintain that it is a fundamental fact about reality that there are irreducibly normative reason relations, and they could refuse head-on to admit that there is anything queer about such relations.” (David Enoch denies such queerness, for example.) The standoff can easily seem to result in a stalemate where they “stare incredulously” at each other. 44
Error Theory in Retrospect
By way of some overall reflections on error theory, let us begin by paying it some compliments. To our thinking, the view has several strengths. Its moral semantics are solid; its advocates do not settle for watering down morality and, in the process, defanging and domesticating it. To the contrary, a recurring subtext of their analysis tends to be that, if moral realism were true, that would be an important fact revelatory of the nature of reality. It would tell us something vitally important. Mackie went so far as to suggest that moral realism would provide powerful inductive reason to take theism seriously.
Of course, this is half of the moral argument. It is the other half with which error theorists disagree: that there really are moral truths. By “moral truths,” remember, they mean something rather specific. They are not content to settle for prescriptivity without authority. Moral realism, as they construe it, requires both of these. It involves categorical imperatives, authoritative moral truths, and binding prescriptive values. Such imperatives and irreducible normativity, of course, are exactly what is thought to raise the specter of metaphysical oddness.
This point reminds us of perhaps the central challenge confronting moral apologists. On the one hand, they need to appeal to a quite robust sense of moral realism—of the type that error theorists describe—if it is to possess evidential weight for God’s existence. On the other hand, the moral facts in question should not stand in such obvious need of a theistic grounding that skeptics about supernaturalism would not be inclined to believe in them. As a result, most versions of naturalistic ethics, though dubbed “moral realist,” would not be adequate to undergird the moral argument for God’s existence.
Following Mackie’s way of laying out his argument, the present discussion mostly involves comparing and contrasting error theory and nonnaturalistic moral realism. They agree on moral semantics and on the nature of robust moral realism but part ways on the matter of moral metaphysics. The error theorists deny that there are prescriptively binding moral facts, while these moral realists affirm that there are. Let us quickly review the two main strategies that moral realists typically use against error theory.
One response is the “partners in crime” approach, which says that if moral facts are ontologically odd, they are not alone. Mathematical, logical, and epistemic facts are in the same boat. So if morality goes, it is not all that goes. Hilary Putnam is one among others who find this argument convincing, and later we will look at Terence Cuneo’s argument for moral realism that takes this approach.
The other approach is what has come to be called a “Moorean” approach, according to which we have better reason to believe in moral facts than to believe in the sorts of considerations that produce error theory. Such Moorean arguments essentially involve plausibility comparisons. For example, which claim do we have more confidence in: that we have a categorical moral obligation never to torture children for fun or that binding prescriptive moral facts are so mysterious and inexplicable that we rationally ought to refrain from believing in them?
We do not intend to ask that question in a way that stacks the deck in favor of moral realism, but the use of an example like child torture is obviously intentional. It is also important, we think, because it speaks to the aforementioned point about the need for morality to be readily accessible to those who do not already subscribe to supernaturalism. Most people, believers and unbelievers alike, do not hesitate to say that some conduct is morally beyond the pale.
This common occurrence gives us reason to think that, epistemologically speaking, morality is more bottom up than top down. Folks from a wide array of worldview perspectives can and do agree on basic moral principles. This is an important response to the factual premise of arguments from relativity, a topic we will consider in more detail later. Our point is not to try settling difficult metaethical debates by appealing to “folk beliefs” about morality but simply to point out that, empirically speaking, convictions about objective ethics are widespread. This does not settle the question about the right metaethical analysis. It leaves that question underdetermined one way or the other, but it is at least telling that error theory flies in the face of most people’s pretheoretical commitments of deep ingression. Error theory would dictate that we jettison and eschew them, and this requires quite a powerful argument.
That point, though, is just one of several that needs to be emphasized here. We do not pretend to have a single knockdown argument against error theory. At the crux of this debate, after all, are issues in at least close vicinity to “first principles,” fundamental beliefs that function primitively and primordially. It is notoriously hard to know how best, if at all, to argue for what practically amounts to one’s axiomatic starting points. If there is a good rational case to be made for moral realism, it likely will involve a constellation of considerations and will require the patience to think the issue through deliberately, carefully, and systematically.
Our basic metaphysical and epistemic convictions help shape what seems rational to us. A commitment to something like ontological parsimony, for example, may ensure never believing in something that does not exist, but it needs to be tempered with the courage, on occasion, to risk belief in order to avoid not believing in something that is real. William James was convinced that whereas avid proponents of parsimony feared superstition, prioritizing simple ontologies, he feared desiccation, too emaciated a worldview owing to an overly ambitious excising of anything remotely mysterious.45
At root, error theory is a veto of morality, and the price is prohibitive. It is a denial that torturing children or exterminating whole people groups is wrong, or that racism is perverse, or that integrity is morally better than duplicity and corruption. What it assumes is that our moral experience—the visceral aversion we have thinking about the Holocaust, our phenomenological antipathy to cruelty, our abhorrence of mindless barbarism or rapacious greed, the deep satisfactions of morality we are capable of feeling, our ravenous hunger to see justice done—is systematically misleading. There is nothing in the universe that is moral after all. Nor anything immoral. It is a world stripped bare of any moral features at all. In this sense it is a quintessential picture of the desiccation that James feared.
Such considerations will nevertheless be unconvincing to someone committed never to believe in anything morally authoritative. The appeal can thus be thought of as circular, preaching to the choir, bolstering the views of those already convinced. We do wonder, though, how practicable such an approach as error theory really is if it is so costly. We also find ourselves wondering if the impracticability of a view provides a certain amount of evidence against it. We suspect it may.46
Attentive readers may recognize the Moorean spirit behind several of our considerations so far, but this is hard to avoid. Remember the point that what shapes our view of rationality is what seems most basically true to us. We are inclined to think that the wrongness of child torture is an axiomatic truth that furnishes insight into the nature of reality. (And we wonder who, if they were forced to watch or endure such torture, would not think the same.) But this requires a willingness to listen to the voice of conscience and to heed the lessons of morality in order to give it a place at the table, as it were, in our efforts to figure out the answers to our largest philosophical questions. If instead we are determined to answer all of our metaphysical and epistemic questions first—bracketing considerations of goodness and justice and beauty—then morality will never be seen as even potentially significant evidentially. William James once said that a rule of thinking that precludes coming to see truth that is really there is an irrational rule.
What seems plausible to us is that reason needs to be construed along rather expansive lines. Too constrictive a view of rationality, or too narrow an empiricism, potentially leaves too much out of the picture of reality and the human condition. Morality, to our thinking, appeals to people in a number of respects—intellectually, affectively, relationally, aesthetically, perhaps even spiritually. But an a priori rejection of anything sacred or transcendent or even mysterious will invariably stop our ears from hearing what it has to say.
We admit to being heartened by the fact that robust moral convictions linger not just among most people but also among brilliant philosophers and ethicists who have spent their lifetime thinking hard about such issues. Sophisticated nonnaturalistic ethical analyses have been provided by the likes of Terence Cuneo and Russ Shafer-Landau, David Enoch and Erik Wielenberg. We may disagree with them about the adequacy of something like modern-day Platonism, but in terms of robust moral realism, and resistance to error theory, they are kindred spirits. Nonnaturalistic ethics, in fact, has experienced a major resurgence of interest among leading metaethicists in recent years.
An expansive rationality and empiricism remains open to potential deliverances from moral and aesthetic experiences and from all three elements of Aristotelian rhetoric: logos, ethos, and pathos. In a similar spirit, we think of moral realism as including not just deontic features like binding moral obligations but also value considerations like the intrinsic value of human beings and epistemic considerations like moral knowledge. An unpalatable implication of error theory is that human persons do not, ultimately, have any moral value at all: It is not the case that upholding dignity should be a moral premium or priority; fighting for equal rights is not the right thing to do; battling injustices is no morally better than perpetrating them. Justice is an empty shell of an idea reflecting nothing. This is not an ad hominem attack against error theorists themselves, who are usually eminently fine people, but an honest set of implications about the view itself, rendering it difficult to endorse without swallowing really hard. Again, might that very challenge be a relevant piece of evidence to consider?
This section of the book is not providing a moral argument for God’s existence but rather setting the stage for it in part 3. The last set of points we wish to emphasize will deal specifically with theism. In his book God and Moral Law, Mark Murphy distinguishes between explanans- and explanandum-driven analyses of God and ethics.47 Most moral arguments are explanandum-driven approaches that begin with the distinctive features of morality and then argue that God is powerfully or uniquely able to account for them. An explanans-driven approach, which Murphy favors, instead assumes God’s existence for purposes of discussion and then shows how such a supernaturalistic reality can account for morality, but he argues that such an approach has no apologetic value. We suspect, though, in agreement with Mike Almeida, that if God can offer explanations not otherwise on offer, such explanans-driven approaches may not be bereft of evidential significance after all.48
The point here is close to Robert Nozick’s strategy in philosophy, which is less about finding knockdown arguments to persuade others on pain of irrationality or stupidity, and more about finding insights to satisfy various puzzlements.49 Rather than seeking after compelling proofs, he aims to generate insightful explanations that could help solve problems, dispel confusions, and resolve dilemmas. Nozick prefers possible and plausible explanations to coercive proofs. He sees his own work in philosophy in this light. Various philosophical matters need to be explained, and he introduces theories to explain them, to render them coherent and better understood.
He is often content with postulating merely possible explanations; indeed, he thinks even impossible explanations can be illuminating and instructive, and here he exhibits some of his penchant for philosophical playfulness. Even false hypotheses can offer some understanding. To get at reality requires true hypotheses, but he did not take this to suggest that hypotheses offered as possible explanations of important phenomena have to be antecedently plausible. Such an insistence would preclude philosophy revealing something new and surprising.
This approach emphasizing possible and perhaps even surprising explanations of important truths contrasts, to Nozick’s thinking, with philosophical argument akin to a metaphorical sledgehammer. The latter, he wrote, is not a nice way to behave toward someone. Nor does a desire to produce uniformity of belief best cohere with the sense of puzzlement and curiosity that initially drives people to do philosophy. A successful proof transmits convictions from premises to conclusion, but an explanation may introduce hypotheses not already believed. And its explanatory power may finally lend support to and induce belief, previously absent, in the hypothesis.
In the light of Murphy’s and Nozick’s points, then, recall the three central queerness concerns that Mackie raised: the ontological strangeness of authoritative moral facts, the question of how such facts supervene on natural facts, and the challenge of accounting for our knowledge of such robust normative truths. We will consider each in turn.
What about the ontological strangeness of authoritative moral facts? As Mackie himself wrote, “Intrinsically prescriptive features, supervening upon natural ones, constitute so odd a cluster of qualities and relations that they are most unlikely to have arisen, without an all-powerful god to create them. If, then, there are such intrinsically prescriptive objective values, they make the existence of a god more probable than it would have been without them.”50 Set aside for now the explanandum-driven way Mackie put it; less ambitiously, he is acknowledging that if God were to exist, morality would not be odd after all.51
What about moral facts supervening on natural facts? Mackie again: “If we adopted moral objectivism, we should have to regard the relations of supervenience which connect values and obligations with their natural grounds as synthetic; they would then be in principle something that a god might conceivably create; and since they would otherwise be a very odd sort of thing, the admitting of them would be an inductive ground for admitting also a god to create them.”52 Again, the less ambitious point to stress is that if God exists, we have a ready explanation of the supervenience of moral facts on natural ones. Why is not this just the sort of “indirect evidence” that Mackie earlier seemed to value?
Finally, what about the issue of knowledge of robust moral truths? Mackie once more: “Men cannot by observation and experiment discover exactly what life is ultimately most satisfying for them; but given that God knows this, desires that they should so live, and has somehow revealed corresponding explicit instructions to them, men can reasonably resort to such revelations to infer this indirectly, so as to complete their determination of this required and ultimately satisfying life.”53 Once more, construed minimally: If God exists, knowledge of robust morality is more readily explained.
These three components were by Mackie’s own admission the heart of the matter, and in each case he saw that theism could potentially resolve the issue at stake. In explaining how, he painted a possible picture of God’s relevance to the foundations of morality that renders the Euthyphro Dilemma impotent.
As Nozick puts it, what seem initially like prima facie implausible explanations to some might surprise us, and we should remain open to what they have to teach us. Rather than insisting on treating nonnaturalism (or theism) as a dead end, perhaps there is wisdom in exploring whether they might have just the resources needed to demystify and bolster moral realism after all.
1 J. L. Mackie, Ethics: Inventing Right and Wrong (Penguin Books, 1977).
2 Jonas Olson, Moral Error Theory: History/Critique/Defence (Oxford University Press, 2017), 21–72.
3 Richard Joyce, “Is Moral Projectivism Empirically Tractable?,” Ethical Theory and Moral Practice (2009): 56.
4 Olson, Moral Error Theory, 5.
5 R. M. Hare, “Nothing Matters,” in Applications of Moral Philosophy (University of California Press, 1972), 47, quoted in Mackie, Ethics, 21.
6 Mackie, Ethics, 22.
7 Olson, Moral Error Theory, 8.
8 Olson, Moral Error Theory, 8.
9 Mackie, Ethics, 15. He then also extended his case to aesthetic values.
10 Mackie, Ethics, 17.
11 At least if this is what Mackie meant, but his use of the phrase “first order” appears to be, if not idiosyncratic, not typical.
12 Mackie, Ethics, 36.
13 Mackie, Ethics, 37.
14 Mackie, Ethics, 37.
15 C. S. Lewis, “Appendix—Illustrations of the Tao,” in The Abolition of Man: Or Reflections on Education with Special Reference to the Teaching of English in the Upper Forms of Schools (HarperOne, 2001), 83–102.
16 At least this is the position he took in his book, but not in his earlier article “A Refutation of Morals,” Australasian Journal of Philosophy 24 (1946): 77–90. Jonas Olson is skeptical that widespread moral differences are effective at undermining claims of moral realism because, among other reasons, deep and pervasive disagreements figure in nearly all branches of philosophy.
17 Mackie, Ethics, 38.
18 G. E. Moore, Principia Ethica (1903; repr., Dover, 2004).
19 Mackie, Ethics, 40.
20 Mackie, Ethics, 41. X is said to supervene on Y if and only if some difference in Y is necessary for any difference in X to be possible.
21 Mackie, Ethics, 42.
22 Mackie, Ethics, 38–39.
23 Mackie, Ethics, 41.
24 This is the point at which so-called projectivist theories come into play, psychological and phenomenological explanations for why we tend to think and speak as if there are moral facts even though there are not.
25 Mackie largely put ethical and aesthetic judgments in the same proverbial boat, thinking the same sort of analysis applies, but here he admits a disanalogy. We have, he argued, less social reasons to objectify aesthetics than we do ethics, which is why, he thought, an error theory of aesthetics is, of the two, the more readily accepted.
26 Mackie, Ethics, 43. Mackie also emphasized how we tend to suppress the conditional clause of hypothetical imperatives and start interpreting them as categorical instead.
27 Mackie, Ethics, 45–46. See Elizabeth Anscombe, “Modern Moral Philosophy,” Philosophy 33, no. 124 (January 1958): 1–19. Anscombe was herself a devout Roman Catholic who would have been pleasantly surprised to see, in the decades subsequent to her essay, a strong resurgence of interest in theistic ethics, including rigorous philosophical efforts to defend it against various objections and to articulate sophisticated versions.
28 Here’s just one: the property of being the number 2 and the positive square root of 4.
29 J. L. Mackie, The Miracle of Theism: Arguments For and Against the Existence of God (Clarendon, 1982).
30 We interpret Mackie charitably here as referring to epistemic possibility, since not even an omnipotent God can rightly be thought to do the metaphysically impossible. Mackie, Miracle of Theism, 118.
31 The literature critiquing the efficacy of Hume’s arguments against miracles is a growing one. See, for one example, John Earman, Hume’s Abject Failure: The Argument Against Miracles (Oxford University Press, 2000).
32 C. Stephen Evans, God and Moral Obligations (Oxford University Press, 2012), 180. See also George Mavrodes, “Religion and the Queerness of Morality,” in Rationality, Religious Belief, and Moral Commitment: New Essays in the Philosophy of Religion, ed. Robert Audi and William J. Wainwright (Cornell University Press, 1986), 223–24.
33 J. L. Mackie, Hume’s Moral Theory (Routledge, 1980), 147.
34 Olson, Moral Error Theory, 102.
35 Mackie, Ethics, 40.
36 Interestingly, Olson argues that David Hume himself did not accept this view. See his Moral Error Theory, chap. 2.
37 Mackie, Hume’s Moral Theory, 54.
38 Mackie, Hume’s Moral Theory, 141.
39 Mackie, Hume’s Moral Theory, 53.
40 Olson, Moral Error Theory, 110.
41 Olson, Moral Error Theory, 114.
42 Richard Joyce, The Evolution of Morality (MIT Press, 2007), 62.
43 Olson, Moral Error Theory, 117–18.
44 Olson, Moral Error Theory, 136.
45 James wrote, “Your bogey is superstition; my bogey is desiccation.” William James, in a letter to “an English critic” [probably L. H. Hobhouse] in 1904, reprinted as a footnote to his essay “The Will to Believe,” in William James: Selected Papers on Philosophy, ed. C. M. Bakewell (E. P. Dutton, 1917), 99.
46 Interestingly enough, Streumer argues that something in its close proximity—error theory’s inability to be believed—is evidence for its truth. See Bart Streumer, Unbelievable Errors: An Error Theory About All Normative Judgements (Oxford University Press, 2017), 150–52. For a rebuttal of Streumer, see Eric Sampson and Alexander Hyun, “On Believing the Error Theory,” Journal of Philosophy 111, no. 11 (2014): 631–40.
47 Mark Murphy, God and Moral Law: On the Theistic Explanation of Morality (Oxford University Press, 2011).
48 See Mike Almeida, “God and Moral Law: On the Theistic Explanation of Morality,” Notre Dame Philosophical Reviews, May 6, 2012, https://ndpr.nd.edu/reviews/god-and-moral-law-on-the-theistic-explanation-of-morality/.
49 Robert Nozick, introduction to Philosophical Explanations (Harvard University Press, 1981).
50 J. L. Mackie, The Miracle of Theism: Arguments For and Against the Existence of God (Clarendon, 1982), 115.
51 A suggestive forthcoming article in Religious Studies by Nick Hadsell, “Divine Authority as Divine Parenthood,” provides a distinctive fine-grained theistic account of where the authority of morality comes from.
52 Mackie, Miracle of Theism, 118.
53 Mackie, Ethics, 231.