« Psalm 27 | Main | Taxation Federalism? »

May 02, 2007



In the last three paragraphs (including the update), Nathan appears to be clawing his way toward Dennett's position on free will, in which the person determines his actions by being a certain sort of person (such that Jesus or whomever could predict his future) but the same person can (and does, of course) make decisions at time T1 that change the sort of person he is at T2. This is a critical kind of free will worth wanting.

In fact, Dennett does not even care if we have radical free will - it's not a kind of free will worth wanting. We want to determine who we are and be impervious to certain kinds of temptation. It's character that transcends the vagaries of decision* that we want, but we don't want it to be static, totally immutable character immune to self-improvement. This is what makes us able to "own" our actions as opposed to being billiard balls impelled by fate.

*In the sense that, say, if asked to torture someone for a thousand dollars, we would consistently refuse irrespective of lighting, whether there's an apple sitting on the table, we've had a bad day, the man asking has a nice haircut, or all and sundry morally irrelevant differences in scenario.

Nathan Smith

Well, the last thing I expected is that my venturing into C.S. Lewis would lead to a convergence with Dennett, but it's not an unpleasant surprise. That said, I can only be bemused by statements like this:

"In fact, Dennett does not even care if we have radical free will - it's not a kind of free will worth wanting."

For the record: from my point of view, radical free will is not a "kind" of free will, it simply *is* free will. And wanting it is beside the point: it's simply a fact, regardless of whether we want it or not. And we would not be *we* without it; we would be robots. Dennett may have some sensible things to say about the phenomenology of free will, but it's an unfortunate distraction that he thinks he is offering an alternative to "radical" free will.

Lewis does *not* seem to be offering an alternative to ("radical") free will-- on the contrary, he emphasizes it-- but he does bring out the importance of character, which determines the choices that the will-- the voluble self-- is actually faced with, both (a) by filtering out certain choices-- I never *choose* not to rob a bank, or commit rape-- and (b) by imaginatively expanding the range of things that it will occur to us to do.

He also presents the interesting notion that the moment of choice is not the same as the moment of action.


"For the record: from my point of view, radical free will is not a "kind" of free will, it simply *is* free will."

So for Nathan, is the question "Did he do it of his own free will?" when used in common parlance, a variety of category mistake? It would seem we use "free will" in a variety of senses, of which will-based radical contingency is only one.

Nathan Smith

"Did he do it of his own free will?" generally refers to free-will-and. Free-will-and an absence of external compulsion. Free-will-and an absence of social pressure. Free-will-and reflection, not reflex. Actions that are strictly involuntary are rather rare cases, and usually the context would make it clear that that's not what's going on. For example, if we're speaking of giving $5,000 to a violent gang, that's not the sort of action one can do involuntarily, like breathing or jumping when startled. One would never ask whether a person voluntarily gave $5,000 to a gang in the strictest sense, it would be a "stupid question," because that action is always going to be voluntary. Other actions, such as going to sleep, are always involuntary. We can arrange the circumstances so that our falling asleep is likely to occur, but we cannot simply flip a mental switch and shut ourselves off. It would be foolish for a bank robber to order the tellers: "Now all of you fall asleep while I take this money, or else I'll shoot you." They can't do it just by wanting to.

If you want to ask whether any event of conscious choice was involved at all, you may have to define your terms and push your interlocutors to engage in a bit of introspection. We all know that paying money is always a voluntary action and falling asleep never is, but precisely because we know this so well, to say so would be as odd as to say that he was breathing using his mouth and lungs. But we reveal in our moral judgments that we know the distinction. We would never accept someone's explanation of paying $5,000 to a gang that they simply didn't mean to do it, that it just happened to them involuntarily. With falling asleep, we routinely accept that someone who falls asleep during our boring story might have done so without spite or even decision.


"But we reveal in our moral judgments that we know the distinction."

I *think* that was exactly the point I wanted to make. The vast majority of references to "free" will outside philosophy concern evaluating how a person's actions reflect on their character. We want to know about extenuating circumstances: Was it an involuntary bodily function? Were the alternatives morally worse? We're not generally interested in the metaphysics of an event's causality. "Do we have free will?" is a very odd question* that could really only be asked in philosophy, since a life spent evaluating differing prospects has little use for a concept of will with a singular prospect. (Note this is different from "fate" which is singular, but not a prospect, being unknown)

*In the case that free will is defined as only radical free will.

Nathan Smith

It's certainly true that the question of whether we have free will could only be asked in philosophy, since ordinary people take it for granted that we do. But ordinary people *are* interested in the metaphysics of an event's causality, even if that's not the language they would use. If the person chose to do Action X, he can be held accountable for it. If Action X was involuntary or compelled, whether by swords and guns or by deterministic physical laws, then the agent cannot be held accountable.

We are not interested only in "how a person's actions reflect on their character." Action X may be perfectly typical of Agent A, or it may be a bizarrely anomalous action. That really doesn't matter, or at any rate it's secondary. What matters is whether Agent A chose to do Action X or not-- free will.

Nathan Smith

By the way, it occurred to me that there may be another explanation for the casual use of the phrase "of his own free will." It has to do with the characterization of an action.

Suppose a woman's husband is on a trip, and she doesn't expect him back for three weeks. He comes home early, and she, taken by surprise, mistakes him for an intruder and shoots him. (The example is Hillary Putnam's, I think.) It would be true to say that the woman's shooting the presumed intruder was "of her own free will," but probably it would not be true to say that she shot her husband of her own free will.

A similar logic might apply to paying off the violent gang, even though Joe is not ignorant of the gang's identity. If you say "Joe gave the gang $5,000 of his own free will," this may suggest a misleading characterization of Joe's action, for it suggests that Joe simply desired to enrich the gang. Joe's real end, of course, was to save his life. Joe's choice was to save his life; paying the $5,000 was merely the means by which he realized his choice.


"If Action X was involuntary or compelled ... by deterministic physical laws, then the agent cannot be held accountable."

What about holding someone accountable for their actions is incompatible with the actions being deterministic?


Being held accountable, I should think means that the agent who did X will be altered from doing X again by way of the actions of some other agent. OR that people will in general be discouraged form doing X because of the added consequences imposed by system of law.

for example, I may steal a pack of cigarettes, and my doing so may be 100% a result of my uncontrollable addiction, that is nicotine pulls the strings and I can't help but steal.

However, If I get caught stealing, and the punishment is a week in jail and I have to pay a fine, etc. Then the fear of the punishment, especially the memory of previous punishment, may have more power over my actions that the addiction to nicotine.

Thus, the punishment was a deterministic force of cause and effect, that makes me to not steal again. Just as my original act of theft was not my "choice" it was the causal force of nicotine addiction.

Nathan Smith

Nato writes:

"What about holding someone accountable for their actions is incompatible with the actions being deterministic?"

Consider two scenarios.

(1) Suppose that I see a baby being crushed underneath a parked car. I go and try to lift the car to save the baby, but I'm not strong enough to do so. The baby dies. In this case, I cannot be held accountable for failing to save the baby, because of determinism: physical laws determined my failure to lift the car, and therefore choice was helpless, so I am not accountable.

(2) Now suppose I am strong enough to lift the car, or perhaps to roll it off the baby, but I just stand there and don't bother. Ordinarily, I would be held accountable in this case, because we ordinarily assume that my behavior in this case is *not* a deterministic given, but is subject to my own choice. However, if I am reducible to molecules and my action is as determined as a billiard ball, then I actually *could not* have saved the baby; my action was determined, Case 2 collapses into Case 1, and indeed in that case no one is responsible for anything, since no one could have done otherwise, except-- by one interpretation of Heisenberg's Uncertainty Principle-- by chance.

No doubt Nato will find this explanation inadequate, but I think it does correspond to the commonsense views of ordinary prephilosophical people, and that those views are based on a foundational intuition of free will which is epistemically deeper, as it were, than any of the inductive evidence by which we form a picture of the physical world. I don't think any prephilosophical person doubts it, and I don't think any philosopher can really succeed in eradicating this belief either; it is as practically ineradicable as the ontological distinction between persons and things, of which it is a part. And yet I don't like to make this kind of argument. Arguments about foundational intuitions and common sense and what prephilosophical people think inevitably have a sort of bullying tone, for what can one's interlocutor say?

Froclown suggests that accountability is all about social control: punishments affect people's behavior, that's why we have praise and blame. This is not true: we hold accountable, we praise and blame, strangers who we'll never meet again, and even the dead, whose incentives cannot be affected. The use of praise and blame for social control is a special case or even perhaps a corruption of our general habit of moral judgments.

Just because the existence of choice is a foundational intuition-- foundational, at any rate, compared to a highly *a posteriori* belief like the existence of a physical world with generally deterministic laws-- doesn't mean it's simple or unproblematic.

Nicotine is an interesting case, and can stand in for a whole class of other cases. If time is a continuum, then I can choose not to smoke at 6:00:00 p.m., and again at 6:00:01 p.m., and 6:00:02 p.m., and so on, 3,600 times, and I'll still only have passed a single hour. Can I choose *not* to smoke, not at this or that second, but *in general*? But even if at time t I make a resolution never to smoke again, I can reverse that resolution at time t+1. This is why temptation can be such an agony: one is forced to make the difficult choice not once, not a thousand times, but infinitely, and a single wrong choice will defeat a hundred thousand right ones. It's impossible-- or is there some way to make a permanent choice, to alter one's character so that the choice went away.

I think my wife got trapped in that situation. She could decide not to leave me this day, and the next, and the next, but she didn't know how to transcend the choice, to make a permanent decision.

If someone is trapped in a situation of temptation, and they don't know how to transcend it, is it realistic to expect them to make the choice of resisting the temptation an infinite number of times? Do they really have the option of resisting the temptation? As I said, it's not unproblematic.


our "habit" of passing moral judgments on the actions of others, has less to do with our desire to control them, and more to do with re-enforcing out own moral values, and codes of behavior.

If I denounce a murderer, I devalue him, his behavior, and what I do symbolically to the image of a man, is reflected in myself, I also degrade those behaviors in myself.

Also, social judgments help to encourage certain ethical codes and ideas and discourage others, in the community at large.

When we devalue some criminal on the news, it's the same process by which we devalue criminals in out midsts, and the alienation and shame associated with those behaviors serves as a detourant as well as coercion to stop those behaviors which are anti-social. These sorts of moral judgments that preserve the community are not intentionality set up however.

Religion comes on the scene when the laws that govern this socializing are studied and intentionally perverted to the aims of the clergy, politics is the same thing by natural rather than supernatural posturing.

True freedom, means freedom from all who would twist and manipulate images to control behavior.


Now *that* sounds fairly anarchist, froclown.


"However, if I am reducible to molecules and my action is as determined as a billiard ball, then I actually *could not* have saved the baby; my action was determined"

In what ways is a person reducible to molecules? Certainly we're composed of them, but it would seem that the properties of molecules do not exhaust the properties and behaviors of things they construct. It would be a classic fallacy of composition, but of course, Nathan intends to make no such move. After all, many entities have higher order properties (e.g. tensile strength of hemp rope) and behaviors (e.g. windmills swinging to face the wind and spinning) without having any moral currency like responsibility (in the agent sense) or will.

But that's just it - none of those things *act* like agents. They don't have agent properties and behaviors. It would be equally silly to evaluate the morality of an atom of boron, a cytosine nucleotide, and a windmill. What about a dog, though? We can and do somewhat hold dogs responsible for their behavior, praising good behavior with a belly rub and condemning the opposite with sharp words - something we wouldn't even contemplate with a windmill that was pointing the wrong direction or failing to spin.

Now, maybe dogs have radical free will - I'm unaware of Nathan explicitly denying the possibility - but we certainly do treat dogs as if they have an attenuated sort of free will. We assume that they can decide not to defecate on the rug but we do not assume they can decide to oppose taxes.

It would seem that we allocate the ability to be responsible for something based on prospective agent's perceived capacity to form contextually-sensitive policies. Perhaps this is because we "naturally assume" all agents fitting that description to have radically free will, but why would that be? Even if we wanted to say that dogs are morally irrelevant (because they are) automatons, we would still have to concede that treating them as moral animals capable of responsibility is productive. If we somehow discovered that a person who acts like a person in every way but featured no radical contingency, would this mean that the person is incapable of bearing any responsibility? That's a strange position in my eyes.

Val Larsen

A minor literary-critical point. I don't think the voluble self is the locus of choice in Lewis' account (as I think Nathaniel said). It is a distraction. It continues to babble after the choice has been made more deeply but its babblings are irrelevant.


How is being responcible for something not a causal event?

Is it not the rubbing of the dogs belly that causes the dog to repeat actions that lead to belly rubs, is it not the striking of the dog that causes the dog to avoid certain actions (at least to avoid being caught at them)

How is this not causal, and is not causal the same as determined.

If there is a reason of any kind why you would choose one thing over another, then that choice was determined.

Even a choice between two equal things, one is selected based on a need for something and the time limit imposed by circumstances, as well as the learned value of the type of thing in general.

A object does not move unless acted upon by a force.

Your thoughts are also caused, because thoughts are always about something, and it's only things you have experiences of, ie those events which determined what you have knowledge of.

Thought can not occur in a vacuum, It only happen in brains and brains only think when they are caused to do so by external events which interact with them causally, this every thought is determined.

tensile strength is the result of the atoms that make up the rope and the electromagnetic attraction between them. The strength and shape of that electromagnetic field is caused by the valence structure of the atoms and the types of atoms with various valance structures and the shape they connect into.

Graphite is slick because the carbon atoms form 1 layer thick sheets which slide over each other, where as Diamond is hard because the carbon atoms form dense geometric structures. Hardness is in the shape of the molecules and the energy level required to break the bonds between them. It is not an epiphenomena.

Val Larsen

In Lewis' account, a decision is made deeply, so deeply that it becomes like predestination. Nathaniel notes that we make many other decisions in the same way (not to smoke, not to slap others). And when we don't make that resolute choice, we are unlikely to maintain a true course to some end. The existence of settled choices (which probably constitute character or at least, without which, character probably lacks meaning) is an undeniable aspect of the phenomenology of choice. And it is an easy thing to explain if one has an account of free will like Nato's. I doubt that it can be adequately explained if one holds a radical concept of free will. For character to exist, past choices (or innate nature) must have causal purchase on future choices. The will must be woven into a matrix that gives it shape and direction, that causally impinges on it. But external causation (which is different from reality posing limited choice sets) is inimical, I think, to radical free will. Perhaps Nathaniel will have thoughts on how this problem (which he clearly recognizes) can be resolved. If not--given that character is as indubitable a reality as choice--then the radical account of choice must be held to be defective or at least inadequate.

The comments to this entry are closed.

My Photo

Only use a payday cash advance as a last resort.


Blog powered by Typepad