Abstract
J. L. Mackie argued that moral thought and discourse involve commitment to an especially robust kind of normativity, which is too weird to exist. Thus, he concluded that moral thought and discourse involve systematic error. Much has been said about this argument in the last four decades or so. Nevertheless, at least one version of Mackie’s argument, specifically the one focusing on the intrinsic weirdness of the relevant kind of normativity, has not been fully unpacked. Thus, more needs to be said about the issue of how to interpret Mackie’s argument. Moreover, I argue that by looking closely at Mackie’s discussion, we can extract two distinct versions of the argument which together present a tougher, and also more precise, challenge for moral realism than extant versions. In this paper I thus revisit Mackie’s discussion with an eye to making progress on an important issue in contemporary metaethics.
1 Introduction
J. L. Mackie (1977) argued that moral thought and discourse involve commitment to an especially robust kind of normativity – he called it “authoritative prescriptivity” – which is too weird (or “queer”) to exist.1 Thus, he concluded that moral thought and discourse involve systematic error.
Roughly speaking, the argument goes as follows: The normativity in question would have to be authoritative in a way that does justice to the seriousness or gravity of moral issues. For example, it could not be understood in terms of individuals’ desires or societal convention, since that would render morality subjective and arbitrary. Also, it could not be a matter of straightforwardly objective natural or empirical fact, since that would lose the prescriptive directionality of moral requirements. Rather, moral normativity would have to be objective but somehow still prescriptive, and forcefully so. When we thus spell out what the requisite normative authority would have to consist in, we see that there could be no such thing.
Much has been said about this argument in the last four decades or so. It is a familiar point by now that the argument comes in several different versions. Not only would the relevant kind of normativity be weird in and of itself; it would also have to supervene in a weird way on natural or descriptive features; and in addition we would need weird epistemic abilities in order to detect it. These three versions, in turn, come in different versions, or have at least been interpreted or developed in different ways.2
Nevertheless, at least one version of Mackie’s argument, specifically the one focusing on the intrinsic weirdness of the relevant kind of normativity, has not yet been fully unpacked. Thus, more needs to be said about the issue of how to understand Mackie’s thinking in this regard. Moreover, I will argue that by looking closely at Mackie’s discussion, we can extract two distinct versions of the argument which together present a tougher, and also more precise, challenge for moral realism than extant versions. Or at least they promise to do so, since a full-fledged defense of the two arguments is beyond the scope of this paper. My aim is the more limited one of articulating the arguments and showcasing their virtues, rendering them philosophically interesting and worthy of serious attention. In this paper, I thus revisit Mackie’s discussion with an eye to making progress on an important issue in contemporary metaethics.
My project is similar to the one pursued by Jonas Olson in his 2014 book, Moral Error Theory: History, Critique, Defence, where he too looks to Mackie’s discussion in order to articulate the most forceful version of the argument from weirdness. Also, my conclusions are in some ways congenial with Olson’s, and if you like you can view this paper as a continuation of certain aspects of his project. However, unlike Olson I do not defend a moral error theory. First, I do not defend (nor do I share) the requisite views about moral thought and language. Second, since I do not fully defend the two arguments that I will extract from Mackie’s discussion, I do not assert their anti-realist metaphysical conclusions (even though I am inclined to accept them). Rather, my purpose is to articulate these arguments and point out their strengths, especially vis-à-vis other arguments in the vicinity with which they are easily conflated. Importantly, I will argue that Olson misses one of the two arguments completely, and I will make several distinctions and clarifications which help the arguments steer clear of influential objections.
The paper has two parts. The first part, comprising sections 2 and 3, is largely exegetical. I scrutinize how Mackie sees the relevant kind of normativity (section 2) and how he understands the crucial notion of weirdness (section 3). I will argue that Mackie’s notion of authoritative prescriptivity has two separate components, derived from two different aspects of moral phenomenology. The two components are objectivity and categoricity/overridingness, and these make for two very different arguments from weirdness. Both arguments make use of the same notion of weirdness, however, namely unintelligibility, which I contrast with non-naturalness (and later on also with unparsimoniousness). In the second part, consisting of sections 4 and 5, I argue that the distinctions made in the first part allow for a two-pronged attack on moral realism, where one might easily see just a single prong. I also argue that the distinctions from part 1 allow for responses to several important objections. In section 4 I extract the two versions of the argument, and in section 5 I explain how they are both more precise and less vulnerable than extant versions. I end the paper in section 6 with a brief summary of the main conclusions.
2 Authoritative Prescriptivity
As I mentioned above, Mackie uses the term “authoritative prescriptivity” for the special kind of normativity targeted by the argument from weirdness. My aim in this section is to unpack and clarify how Mackie sees this notion. I will proceed via a brief investigation of his moral semantics, as he views the matter through the lens of moral thought and talk.3
2.1 Externality and Prescriptive Force
Drawing on moral phenomenology, Mackie holds that morality purports to carry a special kind of practical authority. Moral requirements, if such there were, would thus have to be authoritative in this special kind of way. Evidently he thinks of this feature as having two separate components. To bring these out, it will be helpful to look at the way he argues against two rival metaethical views.
is a very natural reaction […] to protest that there is more to ethics than this, something more external to the maker of moral judgements, more authoritative over both him and those of or to whom he speaks […]. Ethics, we are inclined to believe, is more a matter of knowledge and less a matter of decision than any non-cognitive analysis allows.
mackie 1977, 32–33
Non-cognitivism thus fails to capture the phenomenology of engaging in moral thought and discourse, the sense of there being “external” answers to moral questions about which we might gain knowledge.
introduces a converse deficiency. On a naturalist analysis, moral judgements can be practical, but their practicality is wholly relative to desires or possible satisfactions of the person or persons whose actions are to be guided; but moral judgements seem to say more than this. This view leaves out the categorical quality of moral requirements.
mackie 1977, 33
Thus, naturalism also fails to capture a central component of moral phenomenology, namely the sense that the practicality of moral requirements is in some sense non-optional, and thus not held hostage to the contingent desires of those to whom they apply.
both naturalist and non-cognitive analyses leave out the apparent authority of ethics, the one by excluding the categorically imperative aspect, the other the claim to objective validity or truth.
mackie 1977, 33
As they stand, these arguments against non-cognitivism and naturalism are too sketchy to pose a serious threat to either of these views. My present concern is, however, not to evaluate these arguments, but to get a clearer picture of how Mackie sees morality’s supposed authority.
The critique of non-cognitivism brings out one important theme – what we might call the externality of moral requirements or prescriptions. Mackie repeatedly characterizes the perceived authority of morality as “external.” “We have some tendency to feel,” he says, “that the moral wrongness of a proposed act is an externally authoritative feature which tells us not to do this” (1980, 34). The common-sense belief in the “absolute authority” of moral values assigns to them an “external source” (1977, 43). Moral requirements “seem authoritative for us and external to us” (1980, 20).
This theme is also evident in some of Mackie’s metaphors. At one point he illustrates the notion of a demand for payment by using the metaphor of an “immaterial suction-pipe” or “invisible hook reaching out […] and fishing for the money.” Similarly, the concept of being obligated is likened to being “tied down [by] an invisible cord” (1977, 74). These metaphors may seem bizarre, but they reflect the vague, pre-theoretical sense that moral requirements are external forces, imposing themselves on us from the outside.
The critique of naturalism brings out another important theme: in addition to being external, the prescriptivity in question must also have some sort of oomph.4 This theme too is evident in Mackie’s metaphors. A hook grabs hold of us, and a cord binds us; they make themselves felt, whether we like it or not.5 This reflects the pre-theoretical notion that moral requirements are particularly weighty ones. They are supposed to be in some sense non-negotiable, non-optional, or inescapable. Mackie often expresses this thought by saying that moral requirements are supposed to be “absolute” or “categorical,” where these notions are explicated roughly in terms of independence of the agent’s desires (I will return to Mackie’s explication shortly). Let us call this theme the prescriptive force of moral requirements.
What I have done so far is to convey a sense of the pre-theoretical notion of moral authority underlying Mackie’s conception of authoritative prescriptivity. His critique of non-cognitivism and naturalism reveals the two themes of externality and prescriptive force. I now turn to Mackie’s explication of these notions.
2.2 Objectivity and Categoricity
Externality is explicated in terms of objectivity. Not only would moral prescriptivity not have its source in our mental states; it would not have its source in any mental states whatsoever, not even God’s. Moral requirements would be wholly mind-independent, or, as Mackie puts it, “part of the fabric of the world” (1977, 24). They would not, in other words, be requirements that we human beings impose upon ourselves, nor would they be requirements imposed on us by some other conscious being. Rather, they would “simply [be] there, in the nature of things, without being the requirements of any person or body of persons, even God” (1977, 59). They would be “commands for which there need be, and is, no commander” (1978, 352).6 Thus, the pre-theoretical sense that moral prescriptivity is external and not of our own making is cashed out in terms of the objectivity or mind-independence of moral requirements.
unconditional in the sense of not being contingent upon any present desire of the agent to whose satisfaction the [required] action would contribute as a means […]. The objective values which I am denying would be action-directing absolutely, not contingently […] upon the agent’s desires and inclinations.
mackie 1977, 29; cf. mackie 1980, 72; 1982, 115
The categoricity of moral requirements is here characterized as a matter of them not being contingent upon “any present desire of the agent,” and later “the agent’s desires and inclinations.” In other passages, Mackie formulates categoricity in terms of independence of wants, likes, purposes, ends, satisfactions, preferences, policies, and choices (1977, 27–30, 33). This inclusiveness is as it should be. As I explained above, the intuitive thought that Mackie wants to capture is that moral requirements are particularly weighty ones, and thus in some sense non-optional or inescapable. In order to capture this thought, categoricity should be understood as independence of anything that qualifies as one of the agent’s concerns, broadly construed.
Importantly, categoricity should not be understood as a matter of moral requirements existing independently of the agent’s concerns. Mackie has already made it clear that moral requirements, if such there were, would exist objectively, i.e., independently of everyone’s concerns (including those of possible non-human beings, such as God). Nothing would be added by saying that moral requirements exist independently of the agent’s concerns, and so this is presumably not what Mackie has in mind. Rather, categoricity is a matter of the requirements having enough prescriptive force to outweigh or override the agent’s own concerns, should a conflict arise. Moral requirements are thus independent of what the concerns of the agent are, in that they do not need to line up with those concerns to have their way, so to speak.
In Plato’s theory the Forms, and in particular the Form of the Good, are eternal, extra-mental, realities. They are a very central structural element in the fabric of the world. But it is held also that just knowing them or ‘seeing’ them will not merely tell men what to do but will ensure that they do it, overruling any contrary inclinations.
mackie 1977, 23
This notion of categoricity as overridingness with respect to subjective concerns is admittedly non-standard.7 Usually the term “categoricity” is rather taken to pertain to the existence of a requirement, the idea being that categorical requirements exist independently of the agent’s concerns, whereas hypothetical requirements do not. On this more common notion of categoricity, conventional requirements, such as those of etiquette, spelling, or grammar, are typically categorical in that they do not cease to exist just because they do not line up with the concerns of the agent (cf. Foot 1972, 308–09). But this is not to say that such requirements necessarily override the concerns of agents. Conventional requirements are not invested with that kind of authority.
As this illustrates, the more common notion of categoricity is not sufficient to capture the kind of prescriptive authority that Mackie is after. He also needs the notion of categoricity as overridingness with respect to subjective concerns. I will discuss this issue in more detail in sections 2.4 and 4.2 below. For clarity, in what follows I will use the term “overridingness” to refer to the latter notion.8
In order to do justice to moral phenomenology, then, moral judgments should be construed as descriptive of a special authoritative kind of prescriptivity, where this authoritativeness is understood in terms of externality and prescriptive force, explicated by Mackie in terms of objectivity and overridingness, respectively.
Each of these features is potentially weird. Keeping them separate is thus crucial to appreciating the full weirdness of authoritative prescriptivity, and I will return to this point later on. Before I proceed to further clarify the two features, however, I need to say more about what, in general, Mackie takes a prescription or requirement to be.
2.3 The Anatomy of a Prescription
When Mackie says that moral requirements would be “part of the fabric of the world” (1977, 24), what metaphysical structure is he imagining? What kind of a thing would such a requirement be, what kinds of things would it be related to, and how would it be related to them?
An answer is suggested by Mackie’s analysis of the general meaning of “ought” (1977, 73–77). According to Mackie, an ought-statement is analytically equivalent to a reason-statement: to say that a person ought to perform some action is to say that there is a (sufficient) reason for the person to perform it, where different kinds of reasons (such as moral, epistemic, or institutional) are invoked depending on the context of use. Thus, “oughtness” is identified with “reason-supportedness.”
[P] → φ
What this means is the fact that P counts in favor of, or normatively supports, performing the action φ. Let us call complex facts of this kind reason-facts. A reason-fact is thus a fact to the effect that some other fact – i.e., a reason – normatively supports some particular action.
As I interpret Mackie, he maintains that facts to the effect that actions ought to be performed are identical to facts about sufficient reason-support. And given the standard picture of reasons as favorers, such facts are ultimately grounded by reason-facts. The upshot is that prescriptions or requirements of all kinds (moral, prudential, epistemic, esthetic, institutional, or what have you) boil down to reason-facts. Thus, prescriptivity on this reasons-firstist picture is ultimately a relation of (pro tanto) support obtaining between certain facts and certain actions.
Note that prescriptivity is not here taken to be inherently forceful. The terms “reason” and “normative reason” are sometimes used for some special, intrinsically weighty kind of prescriptivity.10 What I have in mind is rather prescriptivity in its broadest sense. It is a further question whether a prescription or requirement, as I use these terms, has any oomph.
2.4 Minimal Realism and Ardent Realism
With a clearer interpretation of how Mackie sees prescriptivity on the table, I will now go on to further clarify the notion of authoritative prescriptivity. I said above that it has two components, objectivity and overridingness. I will discuss these in turn, beginning with objectivity.
[P] → φ,
constitutive mind-independence means that what it is for the reason-relation (represented by the arrow) to obtain is not for someone to have some disposition or feeling or attitude, or indeed any other kind of psychological property. Presumably this includes counterfactual psychological properties as well, such as attitudes that someone would have under certain conditions. Put another way, objective reason-relations would not be metaphysically mind-involving. This constitutive mind-independence of reason-relations will then spread, via reason-facts, to prescriptions or requirements in general.
[P] → φ,
might still be under the control of agents. If so, Gilbert Harman’s (1975, 5) archetypal bad guy – the contented employee of Murder Incorporated – would have no reason to refrain from killing his targeted victim.12 The idea that moral prescriptions might be held hostage in this way to the desires of villains does not square with the motivations usually underpinning moral realism.
There is another source of worry. Only agents can have reasons, or be required to do something. Although it is not obvious what the relevant agency-conditions are more precisely, and although this is presumably a messy, context-sensitive affair, it seems clear that sharks, for example, do not meet them. Thus, the behavior of a shark never enters as the right-hand relatum into any reason-fact.
A consequence of this agency-constraint is that any prescription will be escapable insofar as agents have the option of turning themselves into non-agents. This raises the question of how easy it is for agents to escape in this way. Many moral realists will have something to say here. At least some ways of construing the relevant agency will be unacceptable to them. For example, building in a concern for the well-being of others would make it too easy to escape. Indeed, Harman’s employee would have done so already. Of course, carrying out the assignment would no longer count as an action, but that may not be very comforting to moral realists (or to the victim). Put another way, while there may still be a moral reason against the employee committing murder, his lack of concern for others will guarantee that there is no moral reason against him committing “schmurder” (cf. Enoch 2006).
[P] → φ
We can ask (1) what the conditions of agency are (pertaining to the right-hand relatum), (2) what the facts that do the favoring are (pertaining to the left-hand relatum), and (3) what it is for actions to be favored by those facts (pertaining to the relation). As we have seen, issues concerning objectivity arise with respect to all three questions.13
Minimal realism addresses the third question, but most realists will also want to address the other two. In particular, the character that Matti Eklund (2017) dubs the “ardent realist” will insist that minimal realism by itself is too wimpy. On Eklund’s characterization, the ardent realist is motivated by “a desire to defend what may be called the objective authority of normativity,” and by the idea that people like the employee of Murder Incorporated are “somehow objectively out of sync with how to conduct one’s life” (2017, 1). Since these motivations align nicely with Mackie’s gloss of authoritative prescriptivity, and since “ardent realism” has the right kind of pre-theoretical “feel,” I will use Eklund’s term to refer to the view that there are authoritative prescriptions in Mackie’s sense (although I do not claim that this is precisely what Eklund has in mind). Ardent realism, thus understood, evidently requires more than reason-relations being constitutively mind-independent. It also requires at least some restrictions pertaining to the relata.
Note that, in order to accommodate the pre-theoretical motivations of ardent realists, these restrictions will have to rule out more than simple agent-dependence. Even if the relata are independent of the agent’s concerns, they might still depend on factors that are arbitrarily subjective by ardent realist lights, such as the conventions of the agent’s group or society, or perhaps the will of a supernatural being. Thus, the objectivity that ardent realists will want to impose on the relata of reason-facts goes beyond what I called “the more common notion of categoricity” in section 2.2 above. That notion will be satisfied as long as (1) the status as an agent, and thus the status of the right-hand relatum as an action, does not depend on the person having certain concerns, and as long as (2) the left-hand relatum is not some fact about, or relating to, the person’s concerns. But ardent realists will also want to rule out various other kinds of subjective dependence.14
As I also suggested in section 2.2, the ardent realist will want even more. Objectivity by itself does not rule out prescriptive impotence. Even if there are thoroughly objective prescriptions against the employee carrying out the assignment, this by itself does not preclude that carrying it out is the thing to do. Perhaps there are institutional requirements pertaining to professional conduct that outweigh the objective ones. Authoritative prescriptivity also requires overridingness.15
On my interpretation, the kind of overridingness that Mackie is after is a matter of objective prescriptions overriding the subjective concerns of agents. While overridingness in this sense is plausibly a necessary condition for authority, it is presumably not sufficient. This is because subjective concerns are not the only ones that need to be outweighed. The ardent realist will want a prescriptive hard currency such that any other kind of prescriptivity is overruled. This includes not just subjective concerns, but also various kinds of institutional prescriptions, such as those of etiquette or grammar. While institutional prescriptions are certainly subjective in a broad sense, they do not stand or fall with the individual concerns of agents.16
In this section I have clarified the notions of objectivity and overridingness, and I have relied on these clarifications in characterizing two kinds of normative realism: minimal realism and ardent realism. According to minimal realism, there are constitutively mind-independent prescriptions. The ardent realist agrees, but in addition holds that some prescriptions are authoritative. Authoritative prescriptivity involves not just constitutive mind-independence, but also some restrictions concerning the relevant kind of agency and concerning which facts qualify as reasons. It also involves full-fledged overridingness. Since ardent realism thus implies minimal realism, the weirdness of objective prescriptivity threatens both views. However, the weirdness of overriding prescriptivity threatens only ardent realism.17
Having identified and clarified the targets of Mackie’s argument, in the next section I go on to interpret the notion of weirdness.
3 Weirdness
According to Mackie, authoritative prescriptivity is too weird to exist. There are at least two different notions of weirdness in play in his discussion, however. In this section I distinguish between the notions of non-naturalness and unintelligibility.18
3.1 Non-naturalness
When we ask the awkward question, how we can be aware of this authoritative prescriptivity […], 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 clearheaded [moral realist] is compelled to resort.
mackie 1977, 38–39
If some supposed metaphysical necessities or essences resist [empiricist] treatment, then they too should be included, along with [authoritative prescriptivity], among the targets of the argument from queerness.
mackie 1977, 39
The idea here seems to be that if knowledge of something cannot be accounted for in empirical terms, then not only do we lack knowledge of that thing – the thing itself is non-existent. We can thus interpret Mackie as adhering to a naturalistic worldview, according to which everything that exists is natural in the sense of being empirically accessible, at least in principle. Weirdness is a matter of being empirically inaccessible, and thus ruled out by such a worldview (cf. Shepski 2008, 378–79).
The problem with this notion of weirdness, however, is that the requisite kind of metaphysical naturalism is a highly general and ambitious, and consequently vulnerable, philosophical position. After all, philosophical inquiry in general appears to be about facts and properties that are not empirically accessible in any interesting sense, and so weirdness as empirical inaccessibility would seem to generalize malignantly to most of philosophy. Indeed, the argument from weirdness would run the risk of self-defeat, since it is far from clear how this kind of empiricist metaphysical naturalism might survive its own scrutiny (cf. Shafer-Landau 2006). I certainly do not take these brief remarks to refute this type of metaphysical naturalism. The point is rather that there seems to be enough doubt about it that it cannot sustain a convincing argument from weirdness.
3.2 Unintelligibility
an objectively prescriptive element […] in a quite non-mysterious way: these would be literally commands issued by an identifiable authority.
mackie 1977, 231
Mackie certainly does not think that divine commands are naturalistically respectable. The point is rather that we can at least make sense of moral requirements if they are understood as God’s requirements.
This notion of weirdness as unintelligibility also seems to be what Mackie has in mind when he points to the “difficulty of seeing how values could be objective” (1977, 24), and when he says that “it is very hard to see how a possible action could have such a property as [objective] rightness or wrongness” (1980, 136). As Mackie (1977, 20–25) emphasizes, however, the point is not that we do not understand what it means to say that something is objectively or authoritatively prescribed.19 What we do not understand is rather how it could be true.
a state of affairs, property, or relation X is weakly unintelligible for a person A just in case A does not see or understand how X could obtain.
X is strongly unintelligible for A just in case A sees or understands that X could not obtain.
To illustrate, I do not see how the activity of a physical system, such as a brain, could give rise to subjective experience. This is weakly unintelligible for me. But it is not strongly unintelligible, since it is also not the case that I see that it could not. By contrast, not only do I not see how the circumference of a circle could be shorter than its diameter; I positively see that it could not be. Thus, the state of affairs that the circumference is shorter is not only weakly but also strongly unintelligible for me.
not seeing how something is possible,
whereas strong unintelligibility is a matter of
seeing how something is not possible.
It is not clear which of these notions that Mackie is employing.20 As I will explain later on, however, this ambiguity invites subtle equivocation when evaluating the argument from weirdness, and it is thus important to keep the two notions apart.
Having interpreted and clarified Mackie’s notions of authoritative prescriptivity and weirdness, in what follows I extract two distinct arguments from his discussion (section 4), and I argue that these together pose a greater, and also more precise, threat to moral realism than extant versions (section 5).
4 Extracting the Arguments
[P] → φ,
can have or lack objectivity. Each of the relata, as well as the reason-relation itself, might be mind-dependent or mind-independent to different degrees. Moreover, I understood the overriding force of a reason-fact in terms of it having more prescriptive force than any reason-fact built from a broadly speaking subjective reason-relation.
So, with this map on the table, what more precisely is weird? Let us begin with objectivity.
4.1 Objective Prescriptivity and Unintelligibility
Understanding weirdness as unintelligibility, what is so weird about objective prescriptivity? A useful device here is to look at some decidedly non-weird kind of prescription, and see which kinds of objectivity it allows for. I will focus on institutional prescriptivity.
First, there is nothing unintelligible about institutional prescriptions employing a thin notion of agency, which cannot be easily escaped. For example, one cannot escape the kind of agency relevant to etiquette just by not caring. Harman’s employee may be guilty of all kinds of rudeness, whether he cares about this or not.
We can escape the imperatives of etiquette by moving into a box on the street, but not even in a box on the street are we free of the imperatives of morality.
garner 1990, 138
Thus, unlike moral agency, the agency of etiquette can be escaped by simply extracting oneself from society, in which case the requirements of etiquette no longer apply.
This is not a promising way of locating the weirdness of moral requirements, however. It may be true that etiquette and many other familiar institutional prescriptions involve such constraints, but this does not seem essential to them. Indeed, if Mackie is right that there is such a thing as the institution of morality (1977, chs. 3–4), then it will presumably differ from other institutions precisely in not allowing its requirements to be easily escaped. This is simply due to the fact that we, the participants in the institution, do not accept the idea that, e.g., moving into a box on the street is sufficient to relinquish one’s moral agency. There is nothing unintelligible about this, and so the weird feature cannot be hard-to-escape agency.
Secondly, there is nothing weird about institutional prescriptions relating actions to agent-independent or even mind-independent facts. Institutional reasons are presumably just whatever facts the institution has in mind, so to speak, in constructing the prescription, and so virtually any kind of fact could be an institutional reason. (For example, the fact that a certain red light is shining might legally disfavor, or forbid, crossing the road.) Thus, the weird feature cannot be agent-independence or even mind-independence of the facts that are reasons either.
[P] → φ,
it makes no sense to suppose that the favoring-relation could be constitutively mind-independent. Thus, minimal realism, and by implication ardent realism, posit the instantiation of an unintelligible relation. Hence, minimal realism and ardent realism are both false.
It is hard to believe in objective prescriptivity because it is hard to make sense of a demand without a demander, and hard to find a place for demands or demanders apart from human interests and conventions. We know what it is for our friends, our job, and our projects to make demands on us, but we do not know what it is for reality to do so.
garner 1990, 143, emphasis in original
It seems clear that what Garner finds unintelligible is the constitutive mind-independence of objective prescriptivity. Demands without demanders – that is, demands that are constitutively independent of demanders – do not make sense.
Similarly, Olson (2014, 135–36) contrasts irreducible normativity with prescriptions of grammar and etiquette. While the latter are “reducible to facts about agents’ desires, roles, or engagement in rule-governed activities,” irreducibly normative favoring relations “appear metaphysically mysterious.” Although he does not say so explicitly, I take it that what Olson finds mysterious about them is precisely their constitutive mind-independence. This is suggested by their juxtaposition with constitutively mind-dependent prescriptions. Olson’s rhetorical question – “How can there be such relations?” – further suggests that he finds them unintelligible.
not seeing how something is possible,
whereas strong unintelligibility is a matter of
seeing how something is not possible.
Mackie’s talk of the difficulty of seeing how values could be objective is ambiguous here, and so is Garner’s talk of demands without demanders being hard to make sense of. Olson’s rhetorical question is similarly ambiguous.
The argument from weak unintelligibility
P1. Objective (constitutively mind-independent) prescriptivity is weakly
unintelligible.
P2. If objective prescriptivity is weakly unintelligible, then minimal
realism is false.
---------------
C. Hence, minimal realism is false.
The argument from strong unintelligibility
P3. Objective prescriptivity is strongly unintelligible.
P4. If objective prescriptivity is strongly unintelligible, then minimal
realism is false.
---------------
C. Hence, minimal realism is false.
(P3) is logically stronger than (P1), but on the other hand (P2) is logically stronger than (P4). There is certainly a risk of implicit equivocation here, and so it is important to keep these two arguments separate (I will return to this point in section 5.2).
4.2 Overriding Prescriptivity and Unintelligibility
What about overridingness? What might be weird about objective reason-facts having more prescriptive force than any subjective reason-fact? A potential source of weirdness here is that such a prescriptive force-property would have to be “interprescriptive.” That is, it would have to be able to settle conflicts between different kinds of prescriptions. But it is not clear that this idea makes any sense. After all, to claim that one prescription is weightier or more forceful than another is presumably just to make another prescriptive judgment, and so we can ask what kind of prescription is invoked. Whatever the answer is, we can raise the issue of the weight or force of that prescription. Thus, while objective reason-facts may be more objectively forceful than subjective reason-facts, subjective reason-facts may be more subjectively forceful than objective reason-facts, and any suggestion to the effect that objective force overrules subjective force will just move the bump in the carpet. Hence, overriding prescriptivity would have to involve some sort of “magic force,” as Philippa Foot once put it (1972, 315).
Let me spell this argument out in a bit more detail. The argument relies on two main ideas or premises, which we might call parity and untetheredness, respectively. I will explain them in turn.
Parity. Assuming that there are reason-facts built from constitutively mind-independent, or objective, reason-relations, it is no metaphysical stretch to think that there are also objective force-facts, which regulate the relative strengths of reason-facts. As I said, presumably these force-facts are simply further objective reason-facts, concerning what to take into account when deliberating, or some such. And it may well be that these second-level objective reason-facts are just the way ardent realists need them to be, siding with the first-level objective reason-facts in conflicts with subjective ones. But these second-level objective reason-facts may have to contend with their subjective mirror images. For example, an institution might deem its own requirements more important than any other kinds of requirements. Thus, consider the following potential conflict:
- (a)[P] →obj φ
- (b)[Q] →subj ¬φ
It may well be that (a) overridesobj (b). But then it may well be that (b) overridessubj (a). If so, there is overridingness parity. And invoking yet higher-level objective reason-facts to break this parity will merely invite a restatement of the issue. Somehow the objective reason-facts are supposed to be normatively gilded, but what this consists in cannot be explained, it seems, without simply helping oneself to the very thing to be made sense of.21
Untetheredness. The second idea is that subjective reason-facts are loose cannons, so to speak, at least from the point of view of the objective reason-facts. Wherever the objective reason-facts point, the subjective ones may (or may not) point in the opposite direction. Since intelligent creatures such as humans are in a position to create norms and prescriptions, there is no telling in advance which subjective reason-facts the objective ones will run up against. Likewise, there is no telling in advance which subjective force-facts the objective ones will have to contend with. The point is thus not that subjective reason-facts and force-facts will in fact come into conflict with objective ones (although they presumably will). The point is rather that there might be conflict.
Together, parity and untetheredness make it hard to see how the requisite overridingness makes sense. If there were some guarantee that the subjective force-facts would always side with the first-level objective reason-facts, then overridingness parity would be avoided. But there is no such guarantee, since our subjective concerns are untethered.
Plato’s Forms give a dramatic picture of what [authoritative prescriptivity] would have to be. 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.
mackie 1977, 40, emphases added
Authoritative prescriptivity is here ascribed two features. First, it would provide cognizers with a direction, thus telling them what to do. This is prescriptivity in its broadest sense, leaving the force of the prescription at hand an open question. In addition, authoritative prescriptivity would provide cognizers with an overriding motive, thus making them do as they are told. This, I take it, is Mackie’s way of spelling out overridingness, where a prescription having the requisite force means that agents who have gained knowledge of what they are authoritatively required to do would, as Mackie puts it, be “impelled” to do that thing, and “without any further motivation” than the cognition itself (1977, 24).
Overridingness thus construed would indeed be a magic force. But it does seem to render interprescriptive overridingness intelligible, since it tethers subjective reason-facts to objective ones. Awareness of objective reason-facts will simply rule out any subjective resistance, and so parity is avoided.
This way of spelling out overridingness is controversial, however, and it arguably conflates “right and might,” or at least prescriptive force and causal force. Overriding subjective concerns in the sense of causally overpowering them does not seem to do justice to our moral concepts. Indeed, one common moral realist response to Mackie’s argument is precisely to deny that our moral concepts commit us to overridingness thus understood (e.g., Shafer-Landau 2003, 120). Fair enough, but ardent realists cannot just leave it at that, since they do believe in the idea that objective prescriptions override subjective ones. And this has to be made sense of, on pain of leaving the present version of Mackie’s argument unanswered.22
5 Evaluating the Arguments
In this section I will explain the advantages of construing Mackie’s arguments the way that I have construed them. The advantages are twofold, having to do with both precision (section 5.1) and persuasiveness (section 5.2).
5.1 Why the Arguments Are More Precise
I have made several explicit distinctions which are at best implicit in Mackie’s discussion, and which provide a higher-resolution picture not only of Mackie’s metaethical thinking, but also of the challenges posed by the argument from weirdness. More specifically, I have distinguished between
- (i)objectivity and overridingness as separate, and independently weird, components of authoritative prescriptivity (section 2.2),
- (ii)between the different ways in which prescriptivity can be objective, pertaining to three different aspects of reason-facts (section 2.4),
- (iii)between weirdness as non-naturalness and weirdness as unintelligibility (section 3),
- (iv)and between weak and strong unintelligibility (section 3.2).
In what follows I explain why these distinctions matter.
5.2 Why the Arguments Are More Forceful
First, by distinguishing between objectivity and overridingness, and correspondingly between minimal and ardent realism, the arguments present two distinct challenges for the latter: (1) the issue of how there could be such a thing as constitutively mind-independent prescriptivity, when this seems just as unintelligible as there being constitutively mind-independent requests or promises (section 4.1); (2) the issue of how such objective prescriptivity could have more prescriptive force than any kind of subjective prescriptivity, such as that of institutional requirements, when this seems to require an unintelligible form of interprescriptivity (section 4.2).23 In other words, any kind of objective prescriptivity, forceful or not, seems weird, and this weirdness is compounded when the objective prescriptions are also supposed to somehow overrule our subjective concerns. And while Mackie’s way of spelling out the second concern may not adequately reflect our pre-theoretical notion of prescriptive force, this still leaves ardent realists with a difficult, additional challenge.
Second, by inspecting the anatomy of reason-facts, we can pinpoint the potential weirdness of objective prescriptivity: it lies specifically in the unintelligibility of constitutively mind-independent reason-relations. There is nothing weird, however, about reasons themselves (i.e., the facts that favor or disfavor actions) being mind-independent, and there is also nothing weird about hard-to-escape agency (section 4.1). This renders the version of Mackie’s argument that targets objective prescriptivity immune to responses that focus on the relata of the reason-relation rather than on the relation itself.24 Correspondingly, it shows that Mackie et alia will have to watch out if they want to countenance non-moral reason-facts, such as prudential ones, since such reason-facts might differ from the targeted ones merely with respect to the relata (cf. Bedke 2010, §iv).
Third, by distinguishing between weak and strong unintelligibility, we see that the argument from weirdness, as put forward by Mackie, Garner and Olson, invites subtle equivocation, relying on weak unintelligibility to establish the weirdness of the targeted entities, but switching to strong unintelligibility when inferring their non-existence. By making the distinction explicit, the arguments avoid the risk of being compromised by equivocation.25
Fourth, by appealing to weirdness as unintelligibility rather than non-naturalness, the arguments avoid relying on any kind of metaphysical naturalism. This is an advantage, since metaphysical naturalism is as vulnerable as it is ambitious, and since it would at the very least hold the arguments hostage to eternal debates in general metaphysics (cf. Olson 2014, 85–86).
Fifth, and relatedly, by appealing to weirdness as unintelligibility, the arguments avoid relying on considerations of ontological parsimony. This is important, since it defuses one influential “quietist” way of responding to the argument from weirdness. For example, Derek Parfit (2011, 465) maintains that Mackie’s argument fails with respect to his particular brand of non-naturalism. The reason is that on Parfit’s view, irreducible normativity exists only in a “non-ontological sense” (2011, §113). In short, the idea is that, along with abstract objects such as numbers, and unlike spatiotemporally located concrete objects such as atoms or rocks, instantiations of irreducibly normative reason-relations manage to avoid making the world ontologically heavier. Similar views have been proposed by Thomas Nagel (1986) and John Skorupski (2010), both of whom claim immunity from Mackie’s argument from weirdness (Nagel 1986, 144; Skorupski 2012, 215–16). Importantly, these views are not fictionalist. The idea is not that irreducible normativity is on a par with, e.g., the magic of Sauron’s Ring of Power. Neither are they disguised forms of nihilism about irreducible normativity. According to these views, there are irreducibly normative reasons. That is, there are facts which have the irreducibly normative property of favoring some course of action. It is just that positing such reasons requires no ontological commitment.
Thus, if irreducible normativity is held to be weird in the sense of being unparsimonious, then quietist non-naturalism may be in the clear with respect to that argument.26 But the unintelligibility versions of Mackie’s argument make no mention of parsimony. Instead, they appeal solely to the weak or strong unintelligibility of irreducible normativity. Hence, with respect to the unintelligibility versions of the argument from weirdness, quietist non-naturalism is just as vulnerable as ordinary non-naturalism.27
Sixth, and finally, by clarifying the potential role of motivational internalism (section 4.2), we see that at most the version of Mackie’s argument which targets overriding prescriptivity can be countered by disavowing motivational internalism. Two different versions of motivational internalism have been ascribed to Mackie in the literature: (1) the idea that there is a necessary connection between moral judgments and full/partial motivation to comply (e.g., Brink 1989, ch. 3); (2) the idea that there is a necessary connection between moral knowledge and full/partial motivation to comply (e.g., Stratton-Lake 2002, 11–13).28 And moral realists often respond to the argument from weirdness by disavowing one or both of these versions. But the argument targeting constitutively mind-independent prescriptivity does not, or at least need not, appeal to either of them. Neither version can be a constraint on prescriptivity as such, since truths about institutional prescriptivity, such as that of etiquette or grammar, can be believed and even known without any motivational impact whatsoever. And institutional prescriptivity is presumably a form of prescriptivity, even though it may not have any practical oomph. Thus, the claim that objective prescriptivity is weird is not informed by motivational internalism.
By distinguishing between the objectivity and overridingness versions of Mackie’s argument, then, it becomes clear that disavowing motivational internalism gets no purchase on the objectivity version. At best it answers Mackie’s intended understanding of the overridingness version. And even then, the challenge of rendering overridingness intelligible will remain, at least for ardent realists.29
Philosophers often dismiss [non-naturalist] moral realism […] as too ‘weird’, ‘spooky’, or ‘mysterious’ for them. This may well be the main source of resistance […]. No matter how forcefully I make the rest of the arguments in this book, [non-naturalist moral realism] will never be accepted as long as philosophers find it ‘weird’.
huemer 2005, 199
This observation is congenial with my own experience of talking to metaethicists of an anti-realist bent, whose qualms about robust normativity usually bottom out with some version of the argument from weirdness, especially with worries about intelligibility. I think that it is fair to say that weirdness-worries about robust normativity constitute the main stumbling block for ardent realists, and I have tried to articulate the most pressing worries of this kind in a more precise way. The most pressing worries are not about ontological parsimony or metaphysical naturalism. Nor are they mainly about motivational internalism. Rather, they have to do with the unintelligibility of constitutively mind-independent as well as interprescriptively overriding normativity. Moreover, this way of construing the core challenge is directly inspired by Mackie’s discussion.
[P]erhaps [the argument from weirdness] can be understood […] as merely insisting that there is just something pre-theoretically implausible, perhaps even unbelievable, about the thought that there are objective facts that, as it were, tell us what to do, that categorically demand behavior irrespective of people’s aims, desires, roles, and so on. My ability to respond to this challenge is somewhat burdened by the fact that I don’t share this intuition at all.
enoch 2011, 136, n. 2, emphasis added; see also kalf 2018, §4.2, esp. 123
Fair enough. But if I am right about the importance of the unintelligibility-worry to moral anti-realists, and if I am right about this being the most promising version of the weirdness-worry, then this tells us where we need to make further efforts. It may be that we have here reached “metaphysical bedrock,” as Olson (2014, 139) puts it, and that what remains is merely an exchange of incredulous stares. I doubt it, however. One interesting feature of the extracted arguments is that they start out with the notion of mind-dependent prescriptivity as a sort of paradigm, and then ask whether there could be any mind-independent variants and, if so, how they might override mind-dependent ones. This approach is very much in line with Mackie’s idea that our notion of authoritative prescriptivity stems from the projection and objectification of certain emotions and desires, often social in origin (Mackie 1977, 42–46). “[E]thics,” Mackie suggests, “is a system of law from which the legislator has been removed” (1977, 45). If this is one’s point of departure, then it is natural to think of putative objective prescriptivity as a sort of institutional-ish prescriptivity from which the mind-dependence has been (unintelligibly) removed. But ardent realists will presumably want to resist this whole approach. They will not see much family resemblance between mind-dependent and mind-independent prescriptivity. The question, then, is how much family resemblance can plausibly be denied, especially bearing in mind that ardent realists need to secure interprescriptive overridingness. Presumably they cannot say that mind-dependent and mind-independent prescriptivity are related purely metaphorically (like footsteps and logical steps), since that would seem to rule out the requisite interaction. Although this is sketchy, it seems to me to throw doubt on the bedrock-idea. At any rate, we can now at least stare at each other through somewhat higher-resolution lenses.30
6 Conclusion
In this paper I have revisited Mackie’s argument from weirdness, focusing on the intrinsic weirdness of robust normativity (“authoritative prescriptivity” in Mackie’s terms), and I have relied on this revisit to shed light on the contemporary debate about moral realism and the argument from weirdness. In short, by emphasizing the distinction between objectivity and overridingness as explications of two very different pre-theoretical aspects of moral authority (externality and prescriptive force), and by distinguishing between importantly different notions or aspects of objectivity and weirdness, we get not just a better understanding of Mackie’s metaethical thinking, but also a more precise and arguably more forceful weirdness-challenge to moral realism.
Acknowledgements
For useful comments and suggestions, I am grateful to Karin Enflo, Jonas Olson, Olle Risberg, and several anonymous reviewers. I am especially indebted to Jens Johansson. Work on this paper was supported by the Swedish Research Council (grant no. 2019–02828).
Biographical note
Victor Moberger is Associate Professor at the Department of Historical, Philosophical and Religious Studies, Umeå University, Sweden. He specializes in metaethics and has published several articles relating to the work of J. L. Mackie and the debate between moral realism and anti-realism.
References
Baker, D. 2018. “Skepticism about Ought Simpliciter.” In R. Shafer-Landau (ed.), Oxford Studies in Metaethics, Volume 13, 230–52. Oxford: Oxford University Press.
Bedke, M. S. 2010. “Might All Normativity be Queer?” Australasian Journal of Philosophy 88(1): 41–58.
Berker, S. 2020. “Mackie Was Not an Error Theorist.” Philosophical Perspectives 33(1): 5–25.
Brink, D. O. 1989. Moral Realism and the Foundations of Ethics. Cambridge: Cambridge University Press.
Dreier, J. 2010. “Mackie’s Realism: Queer Pigs and the Web of Belief.” In R. Joyce and S. Kirchin (eds.), A World without Values: Essays on John Mackie’s Moral Error Theory, 71–86. Dordrecht: Springer.
Eklund, M. 2017. Choosing Normative Concepts. Oxford: Oxford University Press.
Enoch, D. 2006. “Agency, Shmagency: Why Normativity Won’t Come from What Is Constitutive of Action.” The Philosophical Review 115(2): 169–98.
Enoch, D. 2011. Taking Morality Seriously: A Defense of Robust Realism. Oxford: Oxford University Press.
FitzPatrick, W. J. 2008. “Robust Ethical Realism, Non-Naturalism, and Normativity.” In R. Shafer-Landau (ed.), Oxford Studies in Metaethics, Volume 3, 159–205. Oxford: Oxford University Press.
Foot, P. 1972. “Morality as a System of Hypothetical Imperatives.” The Philosophical Review 81(3): 305–16.
Garner, R. T. 1990. “On the Genuine Queerness of Moral Properties and Facts.” Australasian Journal of Philosophy 68(2): 137–46.
Hare, R. M. 1972. “Nothing Matters.” In R. M. Hare, Applications of Moral Philosophy, 32–47. London: Macmillan.
Harman, G. 1975. “Moral Relativism Defended.” The Philosophical Review 84(1): 3–22.
Heathwood, C. 2015. “Irreducibly Normative Properties.” In R. Shafer-Landau (ed.), Oxford Studies in Metaethics, Volume 10, 216–44. Oxford: Oxford University Press.
Huemer, M. 2005. Ethical Intuitionism. Basingstoke: Palgrave Macmillan.
Joyce, R. 2001. The Myth of Morality. Cambridge: Cambridge University Press.
Joyce, R. 2006. The Evolution of Morality. Cambridge, MA: The MIT Press.
Joyce, R. 2008. “Replies.” Philosophy and Phenomenological Research 77(1): 245–67.
Joyce, R. 2022. “Moral Anti-Realism.” In E. N. Zalta and U. Nodelman (eds.), The Stanford Encyclopedia of Philosophy (Winter 2022 Edition). <https://plato.stanford.edu/archives/win2022/entries/moral-anti-realism/>.
Kalf, W. F. 2018. Moral Error Theory. Basingstoke: Palgrave Macmillan.
Kalf, W. F. 2019. “Mackie’s Conceptual Reform Moral Error Theory.” The Journal of Value Inquiry 53(2): 175–91.
Lossau, T. 2022. “Mackie and the Meaning of Moral Terms.” Journal for the History of Analytical Philosophy 10(1): 1–13.
Mackie, J. L. 1977. Ethics: Inventing Right and Wrong. Harmondsworth: Penguin Books.
Mackie, J. L. 1978. “Can There be a Right-Based Moral Theory?” Midwest Studies in Philosophy 3(1): 350–59.
Mackie, J. L. 1980. Hume’s Moral Theory. London: Routledge.
Mackie, J. L. 1982. The Miracle of Theism. Oxford: Clarendon Press.
Moberger, V. 2017. “Not Just Errors: A New Interpretation of Mackie’s Error Theory.” Journal for the History of Analytical Philosophy 5(3): 1–12.
Moberger, V. 2018. The Queerness of Objective Values: An Essay on Mackiean Metaethics and the Arguments from Queerness (Doctoral Dissertation). Uppsala: Department of Philosophy, Uppsala University.
Moberger, V. 2019. “The Mackiean Supervenience Challenge.” Ethical Theory and Moral Practice 22(1): 219–36.
Moberger, V. 2022. “Non-Naturalism and Reasons-Firstism: How to Solve the Discontinuity Problem by Reducing Two Queerness Worries to One.” The Journal of Ethics 26(1): 131–54.
Nagel, T. 1986. The View from Nowhere. New York: Oxford University Press.
Olson, J. 2009. “Reasons and the New Non-Naturalism.” In S. Robertson (ed.), Spheres of Reason, 164–82. Oxford: Oxford University Press.
Olson, J. 2014. Moral Error Theory: History, Critique, Defence. Oxford: Oxford University Press.
Olson, J. 2018. “The Metaphysics of Reasons.” In D. Star (ed.), The Oxford Handbook of Reasons and Normativity, 255–74. Oxford: Oxford University Press.
Parfit, D. 2011. On What Matters, Volume Two. Oxford: Oxford University Press.
Ridge, M. 2020. “Reinventing Ethics: Inventing Right and Wrong.” Journal for the History of Analytical Philosophy 8(4): 1–20.
Rosen, G. 1994. “Objectivity and Modern Idealism: What Is the Question?” In M. Michael and J. O’Leary-Hawthorne (eds.), Philosophy in Mind, 277–319. Dordrecht: Kluwer.
Schroeder, M. 2007. Slaves of the Passions. Oxford: Oxford University Press.
Shafer-Landau, R. 2003. Moral Realism: A Defence. Oxford: Clarendon Press.
Shafer-Landau, R. 2006. “Ethics as Philosophy: A Defense of Ethical Nonnaturalism.” In T. Horgan and M. Timmons (eds.), Metaethics after Moore, 209–32. Oxford: Clarendon Press.
Shepski, L. 2008. “The Vanishing Argument from Queerness.” Australasian Journal of Philosophy 86(3): 371–87.
Skorupski, J. 2010. The Domain of Reasons. Oxford: Oxford University Press.
Skorupski, J. 2012. “Reply to Cassam, Olson, and Railton.” Philosophy and Phenomenological Research 85(1): 210–21.
Smith, M. 1994. The Moral Problem. Oxford: Blackwell.
Stratton-Lake, P. 2002. “Introduction.” In P. Stratton-Lake (ed.), Ethical Intuitionism: Re-evaluations, 1–28. Oxford: Clarendon Press.
Timmons, M. 1999. Morality without Foundations. New York: Oxford University Press.
As Richard Joyce puts it, Mackie’s use of the term “queer” is becoming “increasingly anachronistic” (Joyce 2022, §3.2). Following Huemer (2005, §8.1), I use the term “weird” instead.
For an overview, see Moberger (2018). I discuss Mackie’s supervenience argument in Moberger (2019; 2022).
For recent and related interpretations of Mackie’s metaethical thinking, see Olson (2014), Moberger (2017), Kalf (2019), Berker (2020), Ridge (2020), and Lossau (2022).
The term “oomph” is from Joyce (2006, 62), who uses it to denote the kind of vague, pre-theoretical notion of practical authority that Mackie is after. For elaboration, see Joyce (2008, §ii, esp. 257–61).
I am not sure what a suction-pipe would do, but presumably it would grab our attention!
Mackie sometimes expresses this idea in terms of the supposed moral requirements being “intrinsic,” by which he means that “a situation would have a demand for such-and-such an action somehow built into it” (1977, 40). Hence, no extrinsic input in the form of someone’s attitudes toward the situation would be involved in generating the requirement.
As is Mackie’s motivational construal of overridingness. I will return to this issue in section 4.2.
I do not think that Mackie had the two notions clearly separated in his mind. While his comparisons with Plato’s Forms (1977, 23–24, 40) point toward the overridingness notion, some of what he says about categoricity points rather toward the more common notion (especially 1977, 27–29).
See, e.g., Schroeder (2007, ch. 1). Although there might also be reasons for non-actions (such as attitudes), for ease of exposition, I will speak only in terms of actions. For simplicity, I also construe the reason-relation as a two-place relation. One might want to include additional places, however, such as agent and circumstances. Presumably there can also be reasons against actions, in which case the reason-relation is rather a disfavoring-relation.
For example, Parfit (2011, 308–09) distinguishes between a “rule-implying” and a “reason-implying” sense of normativity, where the latter is taken to be inherently forceful (and the former to be inherently impotent).
To locate minimal realism in the metaethical landscape, we can note that moral non-naturalists are clearly minimal realists. Constitutive mind-independence is at least part of what non-naturalists have in mind with the notion of sui generis or irreducible normativity (see, e.g., Enoch 2011, §1.1). Naturalists may or may not be minimal realists, however, depending on the details of their views.
Although I have framed the discussion in terms of Mackie’s reasons-firstist picture, I take it that the three different aspects of objectivity can be distinguished also on alternative views. Any view of the metaphysics of normativity has to be able to distinguish between the relevant non-normative considerations, the normatively supported action, and the relation between the two.
To require independence of any mental states whatsoever would be going too far, however. Presumably the status as an agent cannot be mind-independent in this strong sense. Also, which facts enter into favoring-relations is a first-order normative issue, and ardent realists will not want to rule out views that attach normative significance to mind-dependent facts, such as facts having to do with the suffering of sentient creatures. (This is not to deny that certain forms of first-order ethical egoism, cultural relativism, and divine command theory do seem to clash with the motivations of ardent realists.)
There are related notions in the literature, such as judgment-independence, stance-independence, attitude-independence, and response-independence. See, e.g., Rosen (1994) and Shafer-Landau (2003, 13–18). Concerning the reason-relation, I think it important that we talk specifically in terms of mind-independence, since the idea that the reason-relation is constitutively mind-dependent (even if independent of judgments, stances, attitudes, and responses) does not square with the non-naturalist realist view that the reason-relation is irreducibly normative. And non-naturalist realism is, after all, the main manifestation of both minimal and ardent realism. Concerning the relata, specifying the right kind of independence will be a messy affair. As I have indicated, mind-independence is too strong, but the other notions might be too strong as well. For example, preference-utilitarians will want to attach normative relevance to desire-satisfaction, which would seem to make at least attitude- and response-independence too strong.
Often prescriptive force is assumed to be secured by talking in terms of reasons as opposed to mere rules (or merely “formal” prescriptivity). But this obscures the issue of what it is about reasons that allows them to trump mere rules. As I will explain in section 4.2 below, there is a real challenge here for ardent realists, and it is lost if they are merely taken to task for the objectivity of assumed-to-be-forceful reasons.
Overruling all possible subjective prescriptions may not amount to much, however, if such prescriptions have no force whatsoever (as many ardent realists appear to hold). Perhaps objective prescriptions would also have to be forceful in absolute terms, not just relative to subjective prescriptions. I will not pursue this point, however.
Locating ardent realism in the metaethical landscape is less straightforward than locating minimal realism, since the different components of ardent realism are not always distinguished clearly. However, I suspect that many, perhaps even most, of those who call themselves non-naturalists are ardent realists. A central motivation behind non-naturalism is precisely the idea that moral prescriptivity has some special force which no natural or descriptive property could supposedly accommodate, and the rhetoric of objectivity is usually salient (see, e.g., FitzPatrick (2008) and Enoch (2011)). I doubt that anyone else is an ardent realist, however. While naturalists may or may not accept minimal realism, the kind of rhetoric which would suggest adherence to ardent realism is usually found among naturalists who reject minimal realism (e.g., Smith (1994)).
An interesting question, which I will not pursue here, is whether these notions of weirdness capture what Mackie means by the term “queerness,” or whether they should rather be thought of as capturing different ways of being weird, or perhaps different grounds of weirdness. Presumably the latter interpretation is more plausible, since “weird” does not seem to mean the same thing as “non-natural” or “unintelligible” (let alone “unparsimonious”; see section 5.2 below). I do not think that this makes much philosophical difference, however, since the argumentative work is in any case being done by the more specific notions. Cf. Enoch (2011, §6.1).
As maintained by Hare (1972). Cf. Shepski (2008, 375–76).
Indeed, even the distinction between unintelligibility and non-naturalness is implicit in Mackie’s discussion, and the distinction between weak and strong unintelligibility is an addition.
Cf. Eklund (2017) on the parity resulting from the possibility of alternative normative concepts. See also Baker (2018). Note that the parity in question is symmetric, in that it would present an analogous problem for “ardent subjectivists,” who maintain that subjective reason-facts always override objective ones. Indeed, the same issue would arise also for mixed views, which assign overriding force to some objective and to some subjective reason-facts. The core of the problem is to make sense of interprescriptive overridingness as such.
I discuss other potential accounts of interprescriptive overridingness in Moberger (2018, ch. 4). For example, Chris Heathwood (2015, 241) suggests that the “essential practicality of [irreducibly] normative properties lies in the fact that they make those who attribute them knowingly to involve themselves in speech acts beyond mere attribution. [Their] “oomph” consists, at least in part, in the fact that we can’t help but commend or condemn, just by attributing them.” The idea here is that in knowingly attributing irreducibly normative properties to objects, we thereby perform the further speech acts of commending or condemning those objects, independently of whether the objects in fact have those properties, and also independently of background conditions, conventions, and the speaker’s intentions – the further speech acts, beyond the sincere attribution of the irreducibly normative properties, occur solely due to the nature of the attributed properties themselves (Heathwood 2015, §9.2). Adapting this idea to our purposes, ardent realists might claim that objectively prescriptive properties are overriding in that sincerely attributing them involves further speech acts of commendation or condemnation, whereas nothing analogous holds for subjectively prescriptive properties. Again, however, this seems to conflate prescriptive and causal force. As Heathwood himself puts it, the properties in question would “get us to do certain things” (2015, 241, emphasis in original).
These issues are often run together. For example, in his influential formulation of the argument from weirdness, Olson (2014, §6.1) appears to build both objectivity and overridingness into the target notion of irreducible normativity, and yet his critique focuses entirely on the weirdness of objective prescriptivity, thus potentially letting ardent realists off the hook too easily.
See Olson (2009, §3) for several examples of moral realists who have tried to counter worries about weirdness by in effect pointing to the non-weirdness of the left-hand relatum of reason-facts.
The distinction is also important in the context of responding to the argument from weirdness. For example, Huemer (2005, 201) responds to something like the argument from weak unintelligibility, but he does not address the argument from strong unintelligibility.
For parsimony-oriented interpretations of Mackie’s argument, see Shepski (2008, 375) and Enoch (2011, 135). Enoch responds to worries about parsimony by appealing to his argument from deliberative indispensability (2011, ch. 3), one implication of which is that the ardent realist’s posits do not violate any plausible requirement of parsimony. Even if successful in that respect, however, Enoch’s reply does not engage with the unintelligibility versions of the argument from weirdness.
See Olson (2018, §3) for a different critique of the idea that quietist non-naturalism manages to deflect the argument from weirdness.
I sided with the latter interpretation in section 4.2 above. See Olson (2014, 106–12) for a detailed defense of this way of reading Mackie. Cf. Dreier (2010, 81–82).
Note also that so-called Moorean objections, which appeal to first-order moral convictions, do not work against the extracted arguments, since they are not arguments for a moral error theory. One would need to supplement the arguments with controversial views about moral thought and language in order to get the implication, e.g., that it is not wrong to torture animals for fun. This point also speaks to the difference, mentioned in section 1, between my project and Olson’s.
How do Richard Joyce’s influential arguments in The Myth of Morality (2001) fit into the picture? Joyce does not target minimal or ardent realism, but rather the view developed by Michael Smith (1994), according to which moral normativity is constitutively desire-dependent but still objective in certain ways. Joyce argues (convincingly, it seems to me) that Smith’s account of moral normativity will in fact lead to a moral error theory, since it relies on overly optimistic expectations about our desires converging under certain counterfactual conditions. But I would not classify Joyce’s argument against Smith’s view as an argument from weirdness. Cf. Kalf (2018, 124).