What may motivate?

Some draft material* for thought, since 'tis the season for New Year's Resolutions.

Prichard’s Fork divides the justification of moral theories from their normative power. The argument may be summarized this way. Say that ΓN is the set of statements (call them axioms) which justifies a moral theory N; these are statements within the bounds of normative discourse. Let us use for an example the deontological theory of Kant’s Groundwork , abbreviated name ‘G’, where the categorical imperative of that theory implies (in conjunction with those statements which give all relevant aspects of the situation S:{xh, …, xj} in which the rational agent required by the theory finds herself) the actions recommended by G: A₁, A₂, …, &c. Then presupposing the argument of the Groundwork ΓK, for an action An, since ΓK⊢G and {G ∪ S}⊢An, we may say that the axioms plus the situation (which is really superfluous for us since all ethical theories rely on a specification of the situation before recommendation of an action) imply the correct action to be taken: {Γk ∪ S}⊢An. Since the situation’s specification is, for our present purposes, superfluous, we may abbreviate this by saying the axioms of the theory imply the actions to be taken, as Γk⊢Ai.

Transcendental argument or no, it is not the case for any known ΓN that ⊢ΓN. That is, none are generally agreed to be true “in virtue of their form,” as it were: to be true on the basis of their logical structure. So there is no logical necessity attached to the supposed truth of any known argument for any ethical theory. Prichard’s point may be put as saying that if we do not presuppose a theory, as we did with G above, then there is a strong sense in which none of the recommendations of the theory are justified. But say that we take the arguments for each of the axioms: the axioms are the basic statements that justify a theory we are concerned with (and here I do not mean the example), and these, as basic, are not justified within normative discourse. That is to say, there are no moral reasons which show these premises to be true, whether they are in fact true or not—they are themselves unjustified. And on Prichard’s account they are unjustifiable, since any justification of them (or argument for their truth) must be given in terms that are not themselves within “normative discourse”: they must be justified by reasons that are themselves not within the realm of moral discussion. So justifications of Γ are not justifications on morally significant grounds, and the contention is that this then is no justification at all; for the sense of justification is wanted is the kind of justification that will motivate an agent to follow the prescriptions of the ethical theory Γ grounds. But justifications from without the normative (moral, ethical) realm of thought cannot so motivate actions. They could motivate only in conjuction with normative (within the realm of the ethical) beliefs such as, for a particular agent, “I ought to act upon the dictates of pure rationality insofar as it can be applied to rational actors.”

Here we have, however, reached the crux of the problem, and I want to move beyond Prichard’s specific challenge (which is itself aimed at deontological theories of morality). The problem is, either the ultimate reasons (the axioms, or whatever presumptions are made and taken as assumptions) for acting according to a moral theory are themselves unjustified, indeed unjustifiable, and therefore no case can be made that one ought to do one thing rather than another (follow one theory rather than another), or they are not justified on normatively significant grounds, in which case the “justification” can be dismissed as having no normative force (e.g. “Yes, but why would I want to be perfectly rational—I’m certainly not built that way?”). Suppose ΛN is the set of claims that justifies acceptance of ΓN, then whatever relation it is such that ΛN⊰ΓN, it (⊰) is not a relation that grounds acting upon what follows from ΓN. That is, {ΛN∪ S}⊬Ai for any action Ai. If it remains mysterious why this might occur, it may only be suggested that because A-ing follows from B, and we accept B on the basis of C, we may not have ground for A-ing on the basis of C alone. That kind of justification, logical justification, may be transitive, but what is wanted is normative justification, which does not appear in C (so to speak). Analogously, I may accept that the definitions of ‘2’, ‘4’, ‘+’, and ‘=’ justifies the claim that 2+2=4, but it does not follow that I must, in any sense, add two and two. Whereas, if I accept that addition is what I am doing, then if two and two arise, I shall (must) produce four.

I assume that this division between normative (specifically ethical) justification and other kinds of justification, and that the general problem extrapolated from Prichard is intelligible. That is, that the objection is not a phantom. Another way of putting the problem is to ask the following question, one fundamental to meta-ethics: even given that an agent is aware of some sort of necessary (or, one might say, objective) constraint on her behavior qua moral, what can be said or shown that will motivate that agent to act within those constraints? Why, in other words, ought we to be ethical rather than not?

A virtue-ethical framework for justification of acting morally (rather than not) may effectively answer this question. The aim is to produce a general schema or set of necessary (adequacy) constraints on what an effective answer is, and show how modern virtue theories may meet those constraints. A specific answer is not to be proposed herein...

[That is to say, we have only the introduction of a problem here, not a whit of solution.]

In the previous characterization of the transition from logical justification of the axioms that give rise to an ethical theory to the actions of an agent aware of that theory (if the theory is true, an agent acting in accordance with the right), practical reason was characterized as though it followed sentential reasoning in the predicate logic. That formulation ought not to be misleading, for this is merely a convenient way of characterizing the chain of justifications, by showing the disconnect between logical justification (implication from known truths) and normative justification (what happens when that which when given as input to the faculty of practical reason results in an action). It is not as though a practical syllogism is in play, though Aristotle’s mechanism is a useful metaphor....

An ethical theory must not only be true (and, likely, its truth must also be epistemically accessible) but it must actually result in morally good behavior. That is, it is a condition of adequacy on an ethical theory that one who understands the theory is moved by it to act in accordance with what it dictates. This condition makes a characterization of the motivation of “purely rational” agents from the statements that compose an ethical theory so perspicuous in the example case used in the first section. According to such a theory, motivation to action results from reasoning, which can be modeled as logical implication in the sentential calculus, thus it literally follows from the categorical imperative that agents like that act in accordance with it. But this psychologically unrealistic condition on morally good action was untenable even for Kant. Another failed attempt at changing human psychology so as to guarantee morally good behavior is psychological egoism, which despite its other problems is worth mentioning here for the fact that it has no explanation how anyone could fail to act well. An ethical theory needs an accurate philosophical psychology, in order to provide a proper explanation why an agent does anything, let alone act well or badly.

Here is a motivation test, then:
(M) Will a normal adult human who understands the ethical theory be
moved to act in accordance with it subsequent to the time
understanding begins?

The test does not imply that only one ethical theory can be true, but it does limit the field. The question that embodies the test, however, is full of philosophically loaded terms (‘normal’, ‘understands’, etc.), and it is part of the aim of this section of the essay to explain the conditions under which the answer to that question is “Yes.” If that is the answer with regard to a specific theory, it satisfies (M), which is merely one necessary condition upon its being a moral theory that applies to (has normative force over) humans.

*Those for whom the symbols here are not found ought to install Code2000 in their browsers.