(C) Agents are committed to cognitivism.
The problem, of course, is that there is a gap between fact and value. There are no such things as values (according to Mackie).
(D) Moral values do not exist.
How can the moral anti-realist preserve the gap while explaining the commitment? Mackie calls this quandary the state of “believing in morals.” In other words:
(Q) What is the account of the consistency between C and D?
The consistency in question is not merely logical. Rampant self-deception or vulgar emotivism is logically possible. It is not, however, very interesting. Instead, Mackie seeks an account which harmonizes C and D in a sensible, compelling fashion. His method is to take C as a background assumption, gesture towards the plausibility of D, and provide an harmonious account of Q.
Mackie recognizes three attempts to support D. While none is sufficient on its own, the three together are meant to provide some cumulative force for D.
The first attempt to support D is an application of the property “queer” to morality. This is not meant to be pejorative—or, at any rate, merely pejorative. It is simply the observation that moral qualities, taken to be the things they are by those who are in a state of “believing in morals,” do not fit into a reductionistic account of what exists and/or can be known to exist. Moral properties appear “queer” in a Russellian universe. This very ambitious a priori methodism fails, since it immediately raises the need for prior justification for its very invocation. Mackie concedes that commonsense moral particularism seems prima facie to defeat this methodological constraint.
A second, much more plausible attempt to support D is an analysis of moral disagreement. An analysis of moral disagreement is an argument for D. It also forms part of the explanans for why C obtains—viz., why agents participating in the moral kind of life from the moral point of view are tempted to think that their evaluative beliefs express factual content. Moral disagreement should not be analyzed by reference to real disagreement about unchanging moral values. Moral disagreement should be analyzed by reference to amoral, contingent, and evolving social habits of valuation. In other words:
Historically, human social orders come to view some social practices as conducive to survival and others as destructive to survival. Survival is not a particularly moral phenomenon. It is biological and sociological. That is, survival is natural. For example, a stable social order arguably cannot be sustained in the face of unmitigated “ethnic cleansing” (i.e., genocide). Social demands of stability and survival require that behaviors qualifying as genocide be quashed. There is nothing intrinsically immoral about the mass elimination of particular humans of a particular ethnic category. This sort of elimination, however, reduces the ratio of social stability to social chaos. By dint of natural selection, modes of living that threaten survival appear to be endowed with definite sensations of pain or discomfort—a sort of “natural conscience.” The predicate “evil” is applied to such states of affairs. The appellation “evil” intensifies emotions. In conjunction with “natural conscience,” this influences the social habits and desires of citizens against ethnic cleansing. At some point, ethnic cleansing comes to be thought of as exemplifying real moral qualities of badness. On this model, moral kinds of habits, feelings, and valuations arise from social habituation.
A third relevant argument is the covariance of two factors. First, when an individual agent considers a morally charged state of affairs, there is a complex web of approbation, disapprobation, and desire for or against that state of affairs. Second, there is a definite moral belief about that state of affairs. Mackie suggests that the first factor is logically and genealogically prior to the second.
The cumulative force of these three considerations supports D. Combined with C as a background assumption, when an agent utters “This act is right,” the logical meaning must be something like, “I approve of this act.” Mackie recognizes the classic criticisms of this simplified emotivist view. Introspection reveals that agents never intend to mean what simplified emotivism requires them to mean. So then, perhaps agents are not stating, but expressing or exclaiming. Introspection reveals again, however, that agents do not think that is all they are doing. How does Mackie explain the fact that agents reason and argue, apparently rationally, with others about moral beliefs allegedly expressing moral qualities? When one tries to translate the language of morality into the language of hoots, grunts, and whistles, the implausibility is laid bare. This is simply to dramatize that Q surfaces because of the very view Mackie favors. He is well aware of this.
Mackie takes upon himself the burden of answering Q by giving an account of moral belief, under the assumption that the three cumulative considerations imply that moral qualities do not exist.
The explanans is the process of objectification and psychological transfer. First, objectification:
Propositions containing moral predicates are invented and instituted in social practice, or perhaps objective values or qualities corresponding to invented moral predicates are thought to exist, and this doxastic phenomenon becomes embedded in social practice.
Second, psychological transfer:
Psychological attachment to SV objects is transferred to OP propositions or values.
It is a three-step social etiology that combines the process above and the second and third attempts to support D. Abstractly, the social etiology follows some pattern like the following:
(1) Some state of affairs S is desired in an overpowering sort of way. It is, perhaps for survival-enhancing features coded through evolution, psychologically non-negotiable. Likewise, S may be generally perceived as a necessary condition for social perpetuation.
For example, that persons facilitate the survival of their offspring appears to be an important foundation for social order.
(2) Some proposition P is constructed in the social imagination. This constructed proposition, of the form “S is good” or “S is right,” becomes objectified—viz., thought by many to exist in earnest.
(2*) P is believed to be an irreducible moral fact.
(2**) “Goodness,” as an irreducible moral quality or value, is thought both to exist and be expressed by the predicate in P.
(3) The overpowering psychological attachment to S is transferred to the entities in (2).
The moral language becomes embedded in the social practices as part and parcel of the mechanisms which enhance the survival value of social order. Hence, this explains the sort of authority and influence attributed to invocations of moral language in social order.
The belief that moral value is thought to exist in earnest is quite consistent with its not existing at all. Similarly, the belief that moral propositions are constructed is consistent with their always being false, since there are no moral truth-makers for them. Moral predicates do not express moral properties. Hence, Mackie thinks he can consistently preserve C and D in a way that answers Q.
The result of Mackie’s social etiology in the moral domain is a moral fictionalism. Moral propositions and moral qualities are like literary propositions and qualities regarding Shakespeare’s Hamlet. There is a perfectly sensible manner in which literary agents can rationally talk about Hamlet, argue about whether or not he really loved Ophelia, whether or not he really went mad, whether or not he really was loyal to fair Denmark, or whether or not he really was admired by Fortinbras in the end. Propositions, qualities, and thoughts regarding Hamlet even draw tears from (some) eyes and incite passionate disagreements at professional conferences. The man Hamlet, however, does not exist. There are no real qualities predicable of the man Hamlet, because he does not exist. For example, he is not any particular height or weight, and to assert that he is over six feet tall is to assert a falsehood. Despite this puzzling state of affairs, no one needs to remind the literati about the non-existence of Hamlet. They, along with most educated persons, are quite aware that (many of) Shakespeare’s stories are make-believe. That Shakespeare’s work is make-believe detracts neither from propositional discourse about fiction nor from the enormous importance of fiction, as an embedded social institution, to human flourishing.
There is a striking similarity in the domain of morals. Morals do not exist. Taken as beliefs expressing real properties of the world, moral beliefs turn out always and everywhere to be false. This, however, does not, in the least, detract from the intelligibility of moral discourse any more than it does in the case of literary discourse. In fact, in the moral domain, since discourse leads often to conflicted action, it is all the more important to resolve moral disagreements in a rational manner. Moral rationality, however, does not require the existence of moral facts or moral properties any more than literary rationality requires the existence of objects in the world whose properties correspond to predicates in literary texts. The non-existence of morals, thus, does not and should not stop agents from passionate engagement in a moral kind of life or a moral point of view.
Thus, suggests Mackie, Q is answered through a social etiology that:
(1) explains C,
(2) supports D,
(3) exhibits the strong coherence of C and D, and
(4) preserves the logical gap between fact and value.