The trouble with free will is not that there is no such thing, but rather that nobody has yet managed to say what sort of a thing it would hypothetically be, in a way that satisfies those that wish to believe in it.
One of the best-known definitions of free will is David Hume’s, which runs as follows (from ‘An enquiry concerning human understanding’. Section 8, part 1):
By liberty, then we can only mean a power of acting or not acting, according to the determinations of the will; that is, if we choose to remain at rest, we may; if we choose to move, we also may. Now this hypothetical liberty is universally allowed to belong to every one who is not a prisoner and in chains.
Ardent believers in free will (‘Libertarians’) reject this because it depicts us as purely at the mercy of the determinations of our will, rather than suggesting, as they would wish, that we are in some indefinable way in control of what we determine – that we control our will rather than our will controlling us.
A Libertarian typically wishes to say, in contradiction of Determinists, that the decisions we make are not even theoretically predictable based on a full description of the prior states of our brain, body and environment. But neither do they wish to say that those decisions are random, which would destroy any notion of responsibility for our actions. They wish the actions to be somehow caused, yet also uncaused. They wish it to be the case that the person ‘could have acted differently’.
This idea is hopelessly vague. First, it rests on the notion of cause, which is very difficult, yet tractable, as covered in my essay “What is a cause?”. But then we have the additional problems of deciding what is meant by ‘could have’ and ‘random’. ‘Random’ is a perennial difficulty. Nobody has yet come up with a satisfactory non-epistemological definition of randomness, as discussed in my essay “Some random thoughts on whether the world is random”.
The idea of ‘could have acted differently’ is also a major problem. Unless we are prepared to describe the action as one of a number of different possible random outcomes, the closest we can come to explaining this phrase is to say it means we can imagine the person having acted differently. But we can imagine many things that we know are impossible, so it is hard to see how imagination is helpful in this context.
I suspect that there is no possible definition of free will that would satisfy the Libertarians.
I further suspect that what determines our actions is the combination of our nature (genes) and our environment. There is no spooky metaphysical third factor that represents the act of ‘choice’. That opinion would make me a Determinist, except that Determinists have to deny that the world contains randomness and, since I deny that randomness is well-defined, I cannot say whether the world contains it.
But does such an opinion require me to believe that we don’t make choices in this world? That would seem an absurd conclusion, as we feel very strongly that we do make choices, all the time. Fortunately, it doesn’t require that.
Choosing is the process of discovering, through considering alternatives, what course of action you are going to follow. A Libertarian would say that the outcome of the process is neither predictable in advance, nor is it random. A Determinist would say that choosing is the process of discovering your destiny.
What happens when we choose an action is typically this. We think about all the alternative possible actions. We think about their consequences, and perhaps enumerate the good and bad consequences for each possible action. We consider how we feel about those consequences. Then we do the action that scores best against the values that are dominating our consciousness at that time. Hence, in short, we can say that:
Choosing is just the process of thinking about several alternative actions and then doing one of them.
The fact that the course of action may have been predetermined (as a Determinist would say it was) does not entail that the process of choosing – considering alternatives – was pointless. I considered the alternatives, and in the way that I did, because I am the type of person I am, because of my nature. And I chose the action I did because of my nature too. If I had not considered the alternatives I did then my nature would have been different, and my decision would likely have been different too.
Say I have a test at school that I very much want to pass. I will probably choose to study for the test. That choosing will involve considering the consequences of studying and of not studying, and comparing them. If I choose to study then I will probably pass. A Determinist would say that I was always going to pass, but that the reason I was always going to pass is that I was always going to study, and I was always going to do that because my brain is so constructed as to place a more positive assessment on the likely consequences of studying than on those of not studying. The ‘choosing’ to study, including the deliberation that precedes the choice, was an inevitable and necessary part of the process. So says the Determinist. But regardless of whether I agree with her on that, we can still both agree that I chose to study, in that I thought about the consequences of studying or not studying, and then studied.
Another example is being persuaded by an argument. Say I have been persuaded, by an ethical argument put to me by another, to become vegetarian. My choosing to become vegetarian is simply the process of my listening to the argument, considering it, especially the consequences of becoming or not becoming vegetarian, and then ceasing to eat meat.
While I was listening to the argument I was probably not aware which way I would choose. I may even have thought that I would remain omnivorous. A moment will come though, either while listening to the argument or in my subsequent reflection on it, when I realise that I will no longer eat meat. I realise that I have decided. That is the way it works for me (I can’t speak for others). My decisions are realisations. I just become aware that the decision is made. I do not decide to decide. If I did that I would have an infinite regress, because before I could do that I would need to decide to decide to decide, and before I could do that I would need to decide to decide to decide to decide, and so on.
The important feature of this example of a persuasive argument is that it counters the suggestion that there’s no point trying to persuade anybody of anything, because everything is predetermined. If it is predetermined (as Determinists would suggest) then every decision is a consequence of the genes and environment of the person deciding, and putting a persuasive argument to that person is a significant element of that environment. Perhaps it was predetermined that I would make that argument, and that I would persuade the person. But I can imagine an alternative world in which the argument was not made – perhaps because I was not there – and the decision was different.
So I conclude that, regardless of whether a satisfactory definition of free will is possible, regardless of whether our minds work according to that definition, and regardless of whether everything is predetermined, we still make choices.
Perhaps that is ‘Compatibilism’, or perhaps there is more to Compatibilism than that. I don’t place much stock in categorising ideas by ‘isms’ so I’m never sure whether any of my ideas fall into any of them.
A brief coda: Another satellite of the free will idea, to which many people are attached, like the notion of ‘choosing’, is that of reasons. What is the reason that I chose a certain action?
This is trickier than ‘choosing’, because we rarely seem to have a single reason for choosing anything. Often there is a network of interconnected reasons that are clearly visible in the choosing process. But even in the simplest decisions, where there seems to be only one reason, there are actually more. Think of any simple decision you’ve made, and your ostensible reason for choosing it. Now consider whether you would still have chosen it if it was guaranteed to make you suffer a prolonged, agonising death within one year. Probably not! So do you not need to include, as an additional reason for choosing it, the fact that you had no reason to believe that it would make you suffer a prolonged, agonising death within one year? We could find countless other such reasons for any choice, however simple.
A reason for an action is in many ways like a cause, and we need to be very precise about our definitions if we want to speak completely unambiguously about our reasons for choosing an action. I would say that the full reason for a voluntary action is the complete set of our expectations regarding the consequences of the action. Looked at another way, I choose an action because it accords more strongly than any of the alternatives, with the values that are uppermost in my mind at the moment of deciding.
Many of the considerations in my “What is a cause” essay will apply. In practice though, one reason may stand out as particularly noteworthy and we may just mention that (“the reason I am fining you is that you were driving at twice the speed limit; and I won’t mention the additional reason that you weren’t in a fire engine racing to put out a fire, as I think we all know that”).
Given that there is so much vagueness about the reason for any action, I think it is reasonable to adopt the following approach, which I like because it is broadly consistent with the way minds work (or maybe just my mind. I don’t know about yours), and because it reflects the idea that we realise we have decided, rather than deciding to decide. It’s vague, erratic and anomalous, but no more so than any other notion of a single reason for a decision:
The ‘reason’ why I did a particular action is the last thing I thought about before I first realised that I was going to do it.
Andrew Kirk. Bondi Junction, 24 August 2013