Harris and Free Will: Could versus Would.

P.F. Strawson's famous essay "Freedom and Resentment" influenced many philosophers.to be compatibilists.

I've been reading Sam Harris' Free Will and while I have a lot of issues with many of the things he says, I want to focus on one particular issue: his central idea that free will and moral responsibility are a simple function of whether or not we could have done otherwise at a particular point in time. In the beginning of his book Harris asserts:

Free will is an illusion. Our wills are simply not of our own making. Thoughts and intentions emerge from background causes of which we are unaware and over which we exert no conscious control. We do not have the freedom we think we have. Free will is actually more than an illusion (or less), in that it cannot be made conceptually coherent. Either our wills are determined by prior causes and we are not responsible for them, or they are the product of chance and we are not responsible for them. If a man’s choice to shoot the president is determined by a certain pattern of neural activity, which is in turn the product of prior causes—perhaps an unfortunate coincidence of bad genes, an unhappy childhood, lost sleep, and cosmic-ray bombardment—what can it possibly mean to say that his will is “free”? No one has ever described a way in which mental and physical processes could arise that would attest to the existence of such freedom. Most illusions are made of sterner stuff than this.

The popular conception of free will seems to rest on two assumptions: (1) that each of us could have behaved differently than we did in the past, and (2) that weare the conscious source of most of our thoughts and actions in the present. As we are about to see, however, both of these assumptions are false.

But this idea that free will is not 'conceptually coherent' seems to be a basic confusion between the logical form of an indicative conditional and a counterfactual conditional and I think this confusion is at the heart of Harris' and other writers' outright rejection of compatibilism. The metaphysical truth of determinism for the actual world has absolutely nothing to do with the fact that there are situations where had a person's intentions and attitudes and values and virtues been different it would have resulted in a different action. Our notion of moral responsibility for individual actions stems from whether or not such counterfactuals are true, not from whether or not our wills are determined by prior causes or if it were possible that the action could not have occurred. Harris is correct in recognizing that indeterminism is not a sufficient condition for moral responsibility, but the crucial mistake he makes is in not recognizing it is not a necessary condition either.

Suppose we have 2 sentences:

1. If Oswald did not kill Kennedy then somebody else killed Kennedy.

2. If Oswald had not killed Kennedy then somebody else would have killed Kennedy.

The first sentence is an example of an indicative conditional: A->B. It simply says if A is true then B must be true. Indicative conditionals are true by virtue of the truth values of the antecedent (A) and consequent (B) parts of the sentence. If Oswald did not fire the fatal shots at Kennedy then this fact seems to entail that somebody else must have, since it is an accepted fact of our actual world that Kennedy was killed by someone. It is not the case in the actual world that that the consequent of the first sentence can be false if the antecedent is true.

The 2nd sentence is an example of a counterfactual conditional A>B. Counterfactual conditionals are true by virtue of the existence of possible worlds where the protasis clause A is true and the apodosis clause B is true, that are closer to our actual world than possible worlds where A is true and B is false. The truth of the 2nd sentence actually depends on the entire state of our actual world at the time of the event and our reasoning about what would have been a likely turn of events had the first part of the sentence occurred. The first key thing to notice here is that while the first sentence is obviously true, then 2nd is not. And the second key thing is that how we evaluate the truth of indicative conditional sentences is completely separate to how we evaluate the truth of counterfactual conditional sentences. In the first instance we look for facts in the actual world that correspond to what is asserted by each part of the sentence. In the 2nd instance we look at the actual world as a whole and possible worlds where both parts of the counterfactual are true, that we think are nearer or more similar to our actual world than other possible worlds where the first part of the counterfactual is true but the second false.

Suppose I am trying to evaluate moral responsibility for some egregious action I took at time t. Maybe I tweeted something rude to Taylor Swift yesterday that caused her to block me. Now consider the following pair of sentences:

3. If all physical events prior to t remained the same then I could not have done otherwise.

4. If I had been a better person then I would have done otherwise.


If we accept Harris' account of the role prior causes and prior neural activity plays in our decision-making and choices, then the first sentence is true. But the second sentence is also true. The truth of the 2nd sentence does not depend in any way on the falsehood of the first. If I accept that my being a better person is simply a function of all the prior causes of our actual world i.e that I could not have been any other way than I was when I made the tweet and thus I could not have done otherwise, that does not change the truth of the counterfactual that a different set of attitudes and beliefs and intentions and values would have resulted in me being not so rude to Ms. Swift yesterday.

It's important to note that sentences like 4. are not always true. It might be the case than I was secretly subverted by a Beyonce music video that sent subliminal messages ordering me to send rude tweets to Ms. Swift. In this case 4. would not be true. And this is where the compatibilist account of free will begins. As long as we think that counterfactuals like 4. are true we believe individuals are morally responsible for their actions. If counterfactuals like these are not true, as in the case where a person's mental ability or judgement is impaired to the point that he would not be able to make a actual decision,  then we do not think individuals are responsible for their actions. The truth of counterfactuals like these is quite independent of determinism being false...indeed it is not hard to see that the truth of these counterfactuals might require determinism to be true. For how can it be the case that we know a person would have acted differently because of different intentions if we were not sure that a sequence of prior causes necessitates some action? This is where Harris and I think many contemporary writers on free will become entangled in conceptual confusion, but they do not see that there is an alternative way out.

Suppose I'm sitting in a train and a fellow passenger tries to get into the seat next to mine but ends up kicking me on my shin. My first reaction is to feel a certain resentment towards him, and his first reaction, assuming he has the same inclinations towards empathy or courtesy as the average civilized adult does, would likely be to apologize to me for his careless actions. We both feel that if he had been more careful then he would not have kicked me.

The use of would here is critical. It is not true to say if the passenger was more careful he could not have kicked me. We can imagine very easily the passenger being extremely careful and cautious getting into his seat, but due to the train hitting a particular bump he ends up kicking me anyway. Rather it is the truth of the counterfactual:

If the passenger had been more careful then he would not have kicked me.

that is the origin of both of our reactive attitudes to his action. 

I've been trying for the past day to figure out why it is that Harris in his book carries out essentially the same thought experiment P.F Strawson does in his famous essay Freedom and Resentment but arrives at essentially the opposite conclusion. Harris asks us to consider cases like the following:

4.A 25-year-old man who had been raised by wonderful parents and never abused intentionally shot and killed a young woman he had never met “just for the fun of it.”

5.A 25-year-old man who had been raised by wonderful parents and never abused intentionally shot and killed a young woman he had never met “just for the fun
of it.” An MRI of the man’s brain revealed a tumor the size of a golf ball in his medial prefrontal cortex (a region responsible for the control of emotion and
behavioral impulses).

and concludes:

Why does the brain tumor in case 5 change our view of the situation so dramatically? One reason is that its influence has been visited upon a person who (we must assume) would not otherwise behave in this way. Both the tumor and its effects seem adventitious, and this makes the perpetrator appear to be purely a victim of biology. Of course, if we couldn’t cure his condition, we would still need to lock him up to prevent him from committing further crimes, but we would not hate
him or condemn him as evil. Here is one front on which I believe our moral intuitions must change: *The more we understand the human mind in causal terms, the harder it becomes to draw a distinction between cases like 4 and 5.*

Harris thinks that the closer we come to establishing determinism as true, the resentment we feel towards the man for his actions in 4. will collapse into something like what we feel towards the man in 5. But where Harris goes wrong is that he assumes we condemn the man in 4. because we think the man must have had some kind of freedom the man in 5 did not. But the actual reason we condemn the man in case 4 is not simply because we think he could have done otherwise. This is trivially true: there's an infinite number of events that could have caused him not to shoot the young woman that have nothing to do with his intentions. Perhaps the gun could have jammed as he pulled the trigger.

The reason we condemn the young man without the brain tumor is because we know men just like him would have done otherwise because they had the right intentions or virtues or higher-order desires. That is, in a set of possible worlds close to our actual world a man just like the shooter would not shoot the woman because, for instance, he believes he has some duty to fulfill towards others. The truth of this counterfactual is the reason we hold the man in this case responsible for his actions. In the case of the man with the brain tumor this counterfactual is false. There are no possible worlds that resemble our own where the man would not have shot the woman because of different intentions, since his judgement was irreparably compromised by a physical condition. (Just as mine was because I was being mind-controlled by Beyonce).

We can see the truth of this analysis if we add the following 6th case:

6. The man in case 5 had been urged by his doctor to get a MRI as he had been experiencing mood swings and blackouts for months, but kept putting it off.

Why is it in case 6 our sense of moral responsibility switches back to holding the man responsible for his actions? It is not the case the man had any more control of his decision to shoot the girl than he did in 5. But we see that by considering the entire state of the actual world at the time of the shooting, it is certainly likely that he would have avoided shooting the girl had he listened to his doctor and been more responsible.

Harris' account of free will suffers from a lot of problems but I hope this analysis can eliminate the conceptual confusion he alleges that free will suffers from and show that compatibilism is a perfectly valid account of free will that Harris simply ignores.