The ideas in this post grew out of my reflections on a recent exchange with Agellius in the comments to this post. Agellius brought up the idea of the practical syllogism, which I dismissed as “an algorithm which is no less mechanical than the laws of physics” (since any mindless computer program can derive a conclusion from premises) and therefore of no use in constructing the non-deterministic model of causation which agency seems to require. Further thought has convinced me that I was wrong in this assessment. While any given practical syllogism is indeed a deterministic algorithm, the process of making choices via practical syllogisms is not deterministic and may indeed be relevant to the question of free will. * A syllogism is a deductive argument deriving a conclusion from two premises, for example:
- Major premise: All men are mortal.
- Minor premise: Samuel L. Jackson is a man.
- Conclusion: Samuel L. Jackson is mortal.
In this kind of syllogism — the theoretical syllogism, or syllogism properly so called — all three components are propositions, and any rational person who believes the first two proposition must believe the third also. We could make this explicit by writing our example as follows:
- I believe that all men are mortal.
- I believe that Samuel L. Jackson is a man.
- Therefore, I believe that Samuel L. Jackson is mortal.
Aristotle also introduced the idea of the practical syllogism — that is, a syllogism which concludes not in a belief but in an action. Unfortunately, he did not develop this idea very clearly, as can be seen in his example of a supposed practical syllogism:
- Major premise: [I believe that] everything sweet ought to be tasted.
- Minor premise: [I believe that] this particular thing is sweet.
- Conclusion: [Therefore, I believe that] this particular thing ought to be tasted.
This is a poor example, not only because the major premise is completely bizarre (the logical form of an argument is independent of the sanity of its premises), but because, as my bracketed additions make explicit, it is actually just another theoretical syllogism – a set of three propositions, belief in the first two of which necessitates belief in the third — and not “practical” at all; a person could assent to all three propositions in the syllogism without actually doing anything. So here I must part company with Aristotle and insist that in a true practical syllogism, the major premise should be the desire for a particular end; the minor premise, the belief that a particular course of action will effect that end; and the conclusion, the execution of that course of action. For example:
- Major premise: I am hungry.
- Minor premise: I believe that cheeseburgers satisfy hunger.
- Conclusion: Therefore, I eat a cheeseburger.
* Now there is a sense in which it is obviously true that something like a practical syllogism lies behind each of our conscious decisions (as opposed to “autopilot” decisions, which probably account for the majority of human behavior and which are matters of habit rather than of reason). If you wanted to explain why you chose to take a particular course of action, you would probably do so in terms corresponding to the major and minor premises of such a syllogism. However, it is also clear that the conclusion of a truly practical syllogism (like my cheeseburger example, as opposed to Aristotle’s pseudo-practical sweet tooth example) does not really follow from the premises the way it would in a theoretical syllogism. It would be manifestly irrational to affirm the two premises of my Samuel L. Jackson syllogism while at the same time denying that Mr. Jackson is mortal. However, there’s nothing at all irrational in affirming the premises of the cheeseburger syllogism while at the same time refraining from eating a cheeseburger. Why might a person be hungry, admit that cheeseburgers satisfy hunger, and yet choose not to eat a cheeseburger? Well, one reason is the existence of countless competing syllogisms, differing from the cheeseburger syllogism only in the identity of the minor term. (“Cheeseburger” is the minor term in the original.) Cheeseburgers do satisfy hunger, but so do schnitzels and burritos and apple pies and as many other things as you care to think of. To be sure, this is also true of our theoretical syllogism; Mr. Jackson is a man, but so are John Travolta and Jacquizz Rodgers and Takeru Kobayashi and a few billion other people. However, these other syllogisms are not in competition with the original Jackson syllogism because I’m free to believe as many things as I please. Realizing the conclusion of one of these syllogisms — say, that Mr. Travolta is mortal — doesn’t prevent me from realizing that Messrs. Jackson, Rodgers, Kobayashi, and any number of other individuals are mortal as well. In the case of the practical syllogism, though, the competition is real. Realizing the conclusion of one of the syllogisms can preclude the realization of its competitors. It’s physically impossible for me to eat all of the things that are capable of satisfying hunger. Even eating two of them is problematic; if I eat, say, a schnitzel, then I will no longer be hungry, and the cheeseburger syllogism and its other competitors will no longer be sound.
Even worse than these competing syllogisms which are practically incompatible with the original one, there may be syllogisms which are just as sound as the cheeseburger syllogism but which are logically incompatible with it — that is, which terminate in the conclusion “I don’t eat a cheeseburger.” For example:
- Major: I want to maintain my health.
- Minor: I believe that cheeseburgers are detrimental to health.
- Conclusion: Therefore, I don’t eat a cheeseburger.
Nothing like this exists in the world of the theoretical syllogism. If you have two valid syllogisms, the respective conclusions of which are “Samuel L. Jackson is mortal” and “Samuel L. Jackson is not mortal,” you can be sure that at least one of your premises is false. It is logically impossible for both syllogisms to be sound (that is, valid and with true premises). However, it is possible (and quite common, actually) for two perfectly sound practical syllogisms to have contradictory conclusions.
Theoretical syllogisms are therefore definitive and self-sufficient in a way that a practical syllogism can never be. So long as I know that all men are mortal and that Mr. Jackson is a man, I can safely ignore all other considerations and conclude with confidence that Mr. Jackson is mortal. I know that, so long as this one syllogism is indeed sound, no other sound syllogism can possibly contradict it. Once I am satisfied that the premises are true and the syllogism is valid, the case is closed.
In practical reason, though, the case is never closed — or, rather, logic will never dictate when the case ought to be considered closed. Be I never so convinced of a particular syllogism’s soundness, I can still never be sure that there isn’t some other equally valid syllogism out there which contradicts it. Nevertheless, at some point I do have to stop thinking and act – declaring the case closed by a free exercise of will. This is what makes the practical syllogism a possible vehicle of free will despite its superficially deterministic nature.
I can exercise my will in two main ways. The first, as mentioned above, is by deciding how long to think about a possible course of action. The longer I hold a contemplated action in my mind and dwell on it, the greater the number of relevant desires and beliefs that will appear and arrange themselves into syllogisms. At first I think only of hunger and the tastiness of cheeseburgers. Then health comes to mind. As I continue to think, any number of other relevant concerns may turn up: monetary cost, convenience, the morality of killing animals for meat, the social implications of eating prole food, the question of the extent to which I should maintain “American” eating habits as opposed to going native, and so on for as long as I care to think. Decide quickly, and I will tend to follow the path of least psychic resistance — little better than just going on autopilot without engaging consciousness at all. Dwell on the possible action for a long time, and complications will proliferate — giving me more options when I finally do make my choice, but also possibly leading to paralysis.
Assuming I have thought long enough to have come up with at least two syllogisms whose conclusions are mutually incompatible (either practically incompatible or logically incompatible, as discussed above), then I have a further opportunity to exercise my will be decided which syllogism (or which set of mutually compatible syllogisms) will “win” — that is, which of the mutually incompatible conclusions will actually be realized in action. How is this decided? Common sense has it that the “strongest” desire wins out — that if in the end I actually eat the cheeseburger, that goes to show that I wanted the pleasure of eating it “more than” I wanted the benefits of good health. The relative strength of the minor premises is also relevant, of course; perhaps my belief that cheeseburgers satisfy hunger is a near-certainty, while I am much less certain about their long-term effects on health. This is true enough of “autopilot” decisions in which consciousness and will do not play a part. Once consciousness is engaged, though, the relative strength of various desires and beliefs turns out to a very plastic thing, highly susceptible to the influence of attention.
Attention is the instrument of will, just as reason is the instrument of thought. To say that will just is attention wouldn’t be too far off the mark. Almost any belief or desire can be made stronger or weaker — or can be made to change its character in other ways — by the attention we choose to give it. In James Hogg’s novel The Private Memoirs and Confessions of a Justified Sinner, the narrator, after admitting to a “longing desire to kill my brother,” writes,
Should any man ever read this scroll, he will wonder at this confession, and deem it savage and unnatural. So it appeared to me at first, but a constant thinking of an event changes every one of its features.
Attention, in turn, just is imagination. (William James makes a compelling case for this identification in his Principles of Psychology.) The longer I dwell on the idea of eating a cheeseburger, calling up a mental image of the taste and the smell and the feeling of it in my mouth, the stronger those premises become, and the more likely their associated syllogisms are to prevail over the competition. (In fact, the process of writing this post required me to dwell on cheeseburger-eating at some length, and sure enough, I went out and ate a cheeseburger afterwards. My apologies if I should happen to induce a similar reaction in any of my readers.) If, on the other hand, I am persistent in my refusal to entertain such images, choosing instead to dwell on an image of myself in excellent health, the relative strength of the competing syllogisms will change, and the other decision will be made.
In my previous post on agency and motive, I proposed an analogy in which we are slaves with many masters — slaves who can do nothing but what we are commanded to do, but who can choose which of various competing commands to obey.
The “masters” can be identified with competing practical syllogisms — i.e., with imperatives derived from beliefs and desires. The slave’s freedom lies in his ability to choose who to listen to. Some of the masters have louder voices than the others, but it is still within his power to tune out the shouting of one premise and attend to the whispering of another — to “hearken and hear and obey.”