Daily Archives: November 17, 2013

Agency and motive

Wherefore, the Lord God gave unto man that he should act for himself. Wherefore, man could not act for himself save it should be that he was enticed by the one or the other.

— Book of Mormon, 2 Nephi 2:16

I’m going to be approaching the problem of agency, alias free will, from as many different angles as possible, in the hope that eventually I can zero in on a meaningful understanding of the concept.

One way of conceptualizing the problem of agency is to consider the relation between agency and motive. The passage from the Book of Mormon quoted above summarizes the paradox neatly. (The entire second chapter of 2 Nephi, q.v., is essential reading for anyone interested in agency.) On the one hand, man has the capacity to “act for himself” (which entails, as the same writer explains a few verses later, the capacity “not to be acted upon”). Man is an uncaused cause, an unmoved mover, whose actions proceed from nothing other than his own free will. On the other hand, man cannot “act for himself” unless he is drawn in one direction or another by some external stimulus! This is a paradox. Man is something like an automatic door which won’t open unless you push it. If you have to push it, in what sense is it automatic? If man cannot act without the pushes and pulls of external stimuli, in what sense does he act for himself?


Lehi (the prophet in whose voice 2 Nephi 2 is written) makes a very important point: Motives — which come from “outside” the will, in the form either of external stimuli or of pre-existing dispositions — are necessary to free will. Man is pushed and pulled in various directions by his desires — desires which come to him unbidden, and which does not choose to experience. Yet it is not correct to say that he can act freely despite these desires; on the contrary, it is only because of these desires that he is able to act at all.

A moment’s introspection confirms that this is true. With no relevant motives, no reason whatsoever for preferring one course of action over another, a man can still “act” after a fashion — that is, he is capable of making a random choice; he won’t be paralyzed like Buridan’s ass — but such an act does not constitute a true instance of free will or of acting for oneself. The “liberty of indifference” (as Voltaire called it; see my earlier discussion here) is no liberty at all. What is not motivated does not come from the man; that is why a man can turn such decisions over to a random or pseudo-random algorithm such as a coin toss and feel that he has lost nothing.


So humans are not unmoved movers. We are — paradox alert! — moved movers. This is not merely to say that sometimes we are moved from without and sometimes we move ourselves; rather, every instance of acting for oneself is also an instance of being moved by external or pre-existing influences. The very same action is simultaneously self-caused and externally caused.

We express our intuitive understanding of this paradox every day when we ask such questions as “Why did you do that?” A true unmoved mover could only answer, “Because I chose to.” That would be the terminus a quo of all possible causal explanation. (“Yes, but why did you choose that?” “Sorry, does not compute.”) But we expect — and get — better answers than that.


When it comes to the motivation of free choices, no simple model of influence is adequate. We cannot, for instance, use conventional concepts of probability and say that a particular influence makes someone “more likely” to make a particular choice. Probabilistic reasoning applies only to rule-governed systems — which free agents, by definition, are not.  To say that I have an n-percent chance of choosing to perform a particular action is to imply that my choice is determined by some highly complex, random or pseudo-random process analogous to a roll of the dice. An act is not determined by any such process, though, but by the agent’s free choice.

Another model which is temptingly simple but inadequate is to think of various motives and influences as vectors, and the agent’s free will as one vector among others, with the final decision being determined by the sum of all relevant vectors. The vector model is inadequate for a couple of reasons. For one thing, it implies that the agent is not truly free — that sufficiently strong motives could force him to act in a particular way will he nill he. (If, on the other hand, his will is strong enough to counteract all other vectors, then those other vectors become irrelevant.) A more serious deficiency is that the vector model fails to explain how free acts are performed through motives rather than in spite of them. It treats motives and other influences as extraneous factors whose only role in the process of agency is to interfere with it. The vector model implies that someone subject to no influences at all — which means having no motives at all — would be a free agent in the fullest sense. In fact, though, we know that without motives, action is impossible — or, at least, meaningless and unintelligible. The Book of Mormon passage quoted at the beginning of this post sounds paradoxical, but it is undeniably correct: man cannot act for himself (that is, exercise agency) unless he is enticed (that is, subject to non-volitional influences). The vector model fails to explain how agency requires motives.


Another approach, which I consider more promising, is to recognize the absolute necessity of motivation*. There can be no action without a motive; motives are prior to actions and, in a sense, determine them. However, human action is not (or need not be) purely passive and mechanical, because each person is subject to multiple, often conflicting, motives — and can freely choose which to obey.

It is as if each human being is a slave and can do only what he is commanded to do — but he is a slave with multiple masters, and when the orders of one master conflict with those of another, he is free to hold to the one and despise the other as he sees fit. He has no power to do anything, though, unless at least one of his masters commands it.

Remember, though, that the “masters” in this analogy are a man’s own motives, so the “slave” in question is freer than the analogy makes him sound. In fact, he can do whatever he wants. His slavery consists in the fact that he can do only what he wants — and he doesn’t get to choose what he wants and what he doesn’t want, at least not directly.

This model is consistent with some aspects of agency as we actually experience it. It explains how our actions follow from our motives but are not determined by them — just as the “slave” in the analogy is actually free despite the fact that he does nothing but follow orders. One serious weakness, though, is that it fails to account for the “easiness” of some choices and the “difficulty” of others. Some motives are simply stronger than others — a consideration which seems to lead us back to the discarded “vector” model — and yet sometimes it is the weaker motive that prevails and realizes itself as action. I say “sometimes,” and yet (as discussed above) it is not a “sometimes” that can be quantified in the language of probability. All possible courses of action must be considered equally possible; none is more “probable” than the another — and yet some are “hard.” In what can this “hardness” consist? (My coincidental namesake, William James, addressed this question in his Principles of Psychology. While his analysis is ultimately inadequate, I do feel that it is a step in the right direction, and I will probably discuss it in a future post.)

The deepest weakness, though, of this and all other models, is its inability to explain how agency constitutes a third kind of thing, neither deterministic nor random — how mechanical motives (for the algorithms of the practical syllogism are no less mechanical than the laws of physics) and arbitrary free will can combine into something which is neither mechanical nor arbitrary, and which can therefore be meaningful. This is the problem of agency, the one that no one has ever come close to solving. I have some inkling of an approach which may prove profitable, but that, too, will have to wait for a future post.

In the meantime, I welcome comments of all kinds. These often prove to be very helpful.


* WordPress’s spell-checker helpfully explains that the word motivation should be avoided because it is a cliche.


Filed under Philosophy