Following NYU Shanghai’s debate on free will, Abiral Chitrakar Phnuyal follows up on compatibilism versus incompatibilism.
At the end of what was a pretty interesting debate on Apr. 24 about determinism and free will, clicker-poll responses to a couple of select questions suggested that the audience’s intuitions about the question of free will in a deterministic world had shifted from a less to a more incompatibilist point of view -– meaning that more people thought, at the end of the debate, that if the world is deterministic, then it cannot be that we are free in any relevant sense. This was, no doubt, the result of some excellent debating and sharp observations about the compatibilist view by Paula Velasquez and Professor Glimcher, for which they have my congratulations. However, I, at least, remain convinced about the correctness of the compatibilist point of view, and I cannot help but feel that at least some of the inferences about the compatibilist conception of free will relied on conceptual confusions and intuitions which may be common but that do not necessarily hold under scrutiny. So, in what follows, I’d like to pose some clarifications about the compatibilist position that Professor Weslake and I were defending, and mount something of a defense against some of the charges that were levied during the debate.
For the benefit of those who may be interested in the discussion but who may not have been able to attend the debate, I’m beginning with a broad overview of concepts and definitions. Section I is an introduction, where I lay out this background. In the second section, I consider the argument that free will is ‘unnecessary’ in a physicalist model of the universe, and can therefore be discarded. Finally, in the third section, I look at the argument that the definition of free will required to maintain a compatibilist position is deflated or meaningless.
If you were at the debate, it’s probably safe to skip the first section, unless you want to revisit certain concepts.
The question of free will – do we really have it? – is an ancient concern that has recurred in many different forms over the past couple thousand years; from stories and plays about fate, to theological disputes about divine will, to modern-day speculations about whether or not we live in a simulation. Since around the 17th century, the debate around free will has congealed around two central questions: the question of determinism, and the question of compatibility.
Determinism and Indeterminism
The question of determinism has to do with a view of how the world works: specifically, determinism refers to the view that all of the facts about the world at a given point in time in conjunction with all of the natural laws that act upon the world produce a set course of events after that point in time. It’s important to note that this is more than just a statement about general causation. If I say, ‘The air in Shanghai is bad this week because of a sandstorm in Mongolia’, this is not a statement about determinism – it is simply a description of the specific cause of a specific event. However, if I say, ‘Given all the facts about the universe and all of the natural laws that there are, the only way things could have played out this week was for the air in Shanghai to be this bad’ – this is a statement describing a deterministic attitude.
So the question is: is the world deterministic? If you think it is, you are a determinist. If you think it is not, then you are an indeterminist. There are a couple of different ways to be an indeterminist. You might think that some things really do just happen, without any kind of causal mechanism behind them (non-causal indeterminism). Or, you might think that everything is caused, but that there is generally more than one course of events that could take place – more than one way the universe can react to a given causal event (probabilistic indeterminism). Or, you might believe that most things are caused, but that there are ‘agents’ – like people, for instance – who can somehow act independent to chains of causation. This final position is known as libertarianism (unrelated to political libertarianism!).
Compatibilism and Incompatibilism
The question of compatibilism is one that follows on from the question of determinism: if determinism is true, then can it be that we have free will? In other words, are determinism and free will compatible? If you believe that they are, then you are a compatibilist. If you believe that they are not, you are an incompatibilist. So if you are an incompatibilist, you may believe that you have free will, and that therefore, it cannot be the case that the world is deterministic. This is a position described earlier as libertarianism. So libertarianism is that incompatibilist position which holds that (i) determinism and free will are incompatible, (ii) free will exists, and (iii) determinism is false. Conversely, you may believe that the world is deterministic, and that therefore it cannot be the case that you have free will. This is hard determinism, a position which holds that (i) determinism and free will are incompatible, (ii) determinism is true, and (iii) free will does not exist.
There is also a third incompatibilist position. Since around the 1920s, the natural sciences – and particularly physics – have taken an indeterministic turn. Especially following research in quantum physics, we no longer hold the Newtonian conception of the universe as a chain of causes and effects. While the precise implications of what quantum mechanics tells us are contested, it has led many to think that at some level, the world really is indeterministic. However, it is not clear that these kinds of findings seriously challenge deterministic theories about free will, even if they challenge determinism as a whole – for example, even if the truth of current theories about quantum physics would disprove determinism, it isn’t entirely clear what bearing this would have on the study of human actions and behaviour. Hence, the trends of indeterminism seen in physics are not necessarily seen in other sciences, or in fields that study human behavior.
Following from all of this, hard incompatibilism is the position which holds that (i) determinism and free will are incompatible, and (ii) free will does not exist. Someone who holds this position might agree with most of the major implications of hard determinism as far as human behavior goes, without having to say that all of the universe at all levels is deterministic. This allows them to make deterministic diagnoses of human action without committing to the truth of determinism. This way, the hard incompatibilist can produce an incompatibilist critique against free will without getting tangled in any controversies about how to interpret contemporary science.
A central term of contention in the free will debate is of course the meaning of free will itself. A standard way of defining it in philosophical discourse, which I used during last month’s debate, is as follows: free will is the capacity to act in such a way that the actor can be responsible for their action. (This definition was put into question, which I will address later.) This is quite a broad definition of a free will, so I’ll briefly outline two more specific ways of thinking about the concept.
One popular way of thinking about free will is the response to reasons approach. This approach holds that someone is acting with free will as long as they can reason about what action to take, and then respond to their reasoning by acting. In plain speech: if you can think about what you want to do, and then do it, then you’re acting with free will.
Some people reject the response to reasons approach by saying that it isn’t sufficient simply to be able to reason and act: it is also necessary to have had more than one possible course of action to choose from. This is known as the principle of alternate possibilities, and an approach from this principle holds that if only one course of action is possible – as determinism necessitates – then free will is impossible.
One more concept to clarify is that of responsibility. In ordinary speech, there are two slightly different ways of talking about responsibility. One might say that parents are responsible for making sure that their children go to school. This can make it sound as though the parents are obliged to send their children to school. This might be true, but the parents’ obligation is different to their responsibility. In the context of our debate, to say that the parents are responsible doesn’t mean that they are obliged, it just means that they can be held accountable for their act (of either sending or not sending their children to school).
There are two kinds of responsibility that are often spoken about, and it may be worth sketching them out. The first is legal responsibility; as is probably clear, this refers to the condition of being accountable from a legal perspective. The second is moral responsibility. Generally, someone is thought to be morally responsible if it is appropriate to react to their actions with attitudes like praise (when they do good) and blame (when they do wrong). So, for example, imagine that a person steals a cookie from a cookie jar (in a high-end confectionery shop). They might be morally responsible in the sense that we can blame them for stealing, and they may also be legally responsible, in that we might make them pay a fine. (You can also easily imagine cases where someone is either legally or morally responsible, but not both.)
On the day of the debate, Professor Weslake and I defended a compatibilist position. Although this is the position I think is the correct one, I recognize that it is not always the most intuitive one to come to. Hence, following the preliminary work of setting out definitions, I will sketch out one common argument as to why the compatibilist position is a plausible one to adopt.
Earlier, I sketched out the principle of alternate possibilities, which holds that for someone to be acting with free will, they need to have different possible courses of actions to choose from. In other words, if someone acts a certain way, and they could not have acted in any other way, then it cannot be that they were acting with free will. Intuitively, this is an easy position to agree with, and one reason why people might hold an incompatibilist view. But there are good reasons to think that it is incorrect. Consider the following example.
Mrs Caesar is the wife of the illustrious Julius, but she is also a powerful magician. Amongst her many fearsome powers is the ability to enter the heads of other people and control their actions. Being a prudent woman, she only uses this power when she needs to.
One night, she has a nightmare about something terrible happening to her husband. Being well-versed in the art of dream-interpretation, she knows that this means he should not leave the house that day, and she tells him so. But he, being pig-headed and self-important, insists that he will go out anyway, as soon as he finishes writing up some notes for a speech.
So finally, she decides to use her magic, and enters his head. She has no interest in stopping him from writing his notes, so for the time being, she lets him have control of his own actions. She decides that she will take control as soon as he finishes writing, and make him go to his room and take a long nap.
However, in addition to being pig-headed and selfish, Julius is also rather lazy. Once he has settled down to write some notes, he feels unwilling to go out. So he finishes up his writing, and of his own volition, goes to his room and takes a long nap.
Here are some facts about the above story: Julius decided to go to his room and take a long nap, and of his own free will, did exactly that. But here’s the other way the story could have gone: Julius could have decided to leave the house, or do something else, and his wife would have taken control to make him go to his room and take a long nap.
Here’s the interesting thing about this case: Julius made his own choices and acted of his own free will, but he could not have acted otherwise. The only course of action that could have taken place was him going to his room and taking a long nap. Hence, it seems that it is in fact possible for people to act with free will even when there is only one possible course of action.
This kind of case is called a Frankfurt-style case, named after the eminent Princeton philosopher Harry Frankfurt, and which follows the model of argument he set out in a 1969 essay. There are plenty of examples of Frankfurt-type cases (often involving brain-control devices), and it is easy to make up your own.
The important thing about Frankfurt-style cases is that they show how people can have free will, and therefore be morally responsible, even in a deterministic universe. Think about a case like the above, but instead of going to take a nap, the action in question is stealing a cookie from a cookie jar. Julius steals the cookie from the cookie jar, and he could not have acted otherwise; but also, he himself decided to act that way, and he performed the action. Here, even though he had no alternate possibilities, it seems that he is still morally responsible for the act of stealing, and we can justifiably blame him for it.
The Argument from Parsimony
The argument I’m concerned with here goes something like this: even if free will is plausible in a broadly deterministic world, it appears to be unnecessary to complete the picture. What would happen if we took the compatibilist version of what the world looks like and just removed free will? Nothing would break down, and everything would effectively remain the same. If free will is unnecessary to complete the picture, there is no reason to believe that it exists.
Before I engage with this objection directly, let me take a short detour.
If everything is physical, where is free will?
Here is one kind of argument that some people sometimes make to try and show that free will cannot exist:
If this world really is a physicalist world, then everything is composed of physical particles – atoms and subatomic particles. These particles do not have free will. If everything is just different arrangements of and interactions between physical particles, then where is the space for ‘free will’?
It’s a strange argument, and one that most people can immediately see through. Atoms, at a microscopic level, lack all sorts of qualities that we observe at a macroscopic level. For example, atoms have no color – but clearly, groups of atoms arranged in a certain way do. This also works for more abstract qualities. A painting might look like the ‘Mona Lisa’, but it’s clear that none of its atomic particles do. Or consider any behavioral trait: a person might be very humorous, but none of her particles or organs or constituent parts (apart from her funny bones) are necessarily the same. So it’s obvious that macroscopic objects often do not only have the qualities of their constituent particles. Another way of putting it is that there are emergent qualities of things that can be naturally produced when physical particles behave or are arranged in certain ways.
However, here’s one quite important thing about the above argument. The assumption being made is that for free will to exist, there needs to be something ‘extra’ to the physical world that allows for it – something like a soul, for instance. But for the compatibilist, this is a non-issue. There does not need to be anything ‘extra’ in the universe for free will to exist, because free will is not some mysterious, contra-causal force that has to act independently of natural arrangements and laws; free will is part of the natural world, is completely compatible with natural laws, and can depend on physical matter to exist in the same way any other quality depends on physical matter to exist.
Safely removing free will from the system
Let’s return to the objection described earlier. If free will is unnecessary to describing the world – that is to say, if you could remove free will from a picture of the world, and nothing would change – then why should we think it exists? This intuition draws some of its persuasive power from the principle known as Ockham’s Razor, which states that when you have two competing hypotheses, the simpler (more parsimonious) hypothesis is the better one. (It has been argued that Ockham’s Razor is not, in fact, a principle that applies equally in all conditions, and that it is not unconditionally justifiable. But I’ll ignore this possible response for now.)
The problem, however, is that this objection still seems to construe free will as an isolatable thing, something that you can point at and move around. But for the definition of free will that we’ve been working with, this just doesn’t make sense. Remember, free will is a capacity to act in a certain way (such that we have responsibility): to act with free will just requires that certain conditions are fulfilled when an action takes place. There doesn’t need to be any shadowy soul-substance for this kind of free will to exist, nor does there need to be a specific part of the brain or a particular event in the brain.
Let’s have a look at what is required, and what it would mean to ‘remove’ free will in this sense. Free will means that you can think about what you want to do, and then be able to do it. Hence, what’s required is (i) the ability to think, deliberate, and make decisions about what to do, and (ii) the ability to act. So to ‘remove free will from the picture’, you’d either have to remove people’s ability to think and reason, or you’d have to remove their ability to act (or both). In either case, it’s easy to see how the world would be very different.
The free will skeptic might push a little here and ask, what exactly is free will supposed to be explaining or describing here? The answer, just from the definition of free will, is that free will helps to explain why we are responsible for our actions in certain cases but not in others. Removing free will from a situation when it is already there, as I’ve tried to show above, is a bit of a weird concept; but a case in which someone is not acting with free will is perfectly coherent, and would be a case in which she is not responsible for her actions. So, for example, if you’ve just had surgery and are on medication that inhibits your ability to think clearly, and you therefore end up accidentally ordering fifteen cheese pizzas, you have not really acted with free will, and so you aren’t morally responsible for the mistake of over-buying (although in this case you probably will be held legally responsible if you refuse to pay for the pizzas). Here, free will explains how the question of responsibility is settled.
Hence, the argument from parsimony ends up being a conceptual confusion about what free will (in this sense of the term) means, and how it works.
On the Definition of Free Will
At a certain point during the debate, it began to look as though there wasn’t any real metaphysical dispute: both sides agreed that the world was physicalist, and that human behavior can be mapped onto physical (mostly neurological) processes and interactions. The disagreement from this point was about the definition of free will itself. Professor Glimcher made the claim that the kind of free will that people actually want is precisely the ‘libertarian contra-causal force that can change nature’ type of free will that both sides had accepted does not exist, and that defining free will in the way that we did made it deflationary and largely meaningless.
There are two closely connected assertions in this objection, so I’ll consider them individually. The first is that people generally require free will to be a kind of powerful force that will allow them to act against natural laws (and therefore make the world indeterministic through their actions). The second is that the kind of free will we proposed, which centers on people’s ability to reason and act, is a ‘deflated’ kind of free will that isn’t significant to us.
Free Will and Natural Limits
To expand a little, the first assertion is: if there are natural laws that place strong limits on the kinds of options available to us, then these are limits on our free will, and only being able to act against these limits grants us free will. The intuition here is similar to the principle of alternate possibilities: if all of your options are taken away from you, then it seems hard to say that you have free will. It’s a natural way of thinking, and in many cases it may be true. If you are forced to do something you do not want to – if you had no say or options in the matter – then you are probably justified in saying that you did not act with free will.
Two responses. First: notice that the case above makes perfect sense under the definition of free will we have been working with so far. If your reasoning makes you want to act in one way, but you do not have the freedom to respond to that reasoning, then certainly, you did not act with free will. But if you really did want to do what you were told, then it would seem that you did act with free will. This is just another variation of Frankfurt-type cases.
More importantly, the assertion being made is that people generally require that free will is the sort of thing that they can use to act against natural limits, and this seems obviously false. Think about one powerful natural force which places limits on your possible actions: gravity. Gravity dictates that all of us are bound to behave in certain ways – we cannot spontaneously float off the ground at will – and we cannot act against its laws. Now, it certainly may be true that many people think it would be cool to be able to float away at will, but this is a very different kind of attitude to what is being proposed. I take it that most people do not think that gravity, and other natural laws like it, somehow take away your free will: that because you do not have the option to float when you want, or the option to make objects materialize with a thought, or the ability to transform into a sponge who lives in a pineapple – that you therefore do not have free will. In almost any case, therefore, I think it’s fair to assume that when people talk about free will, they’re talking about free will within the laws of nature, and furthermore that they believe this to be a reasonable kind of free will to have.
What is free will good for?
The second claim is that when free will is being defined in the way we have – as the ability to act in a way that allows for responsibility – then by any ordinary understanding of what free will is supposed to achieve, we are deflating the meaning and value of free will. In other words, if free will is something that can comfortably live with determinism, then it ends up seeming meaningless. So this is an incompatibilist argument for why we should not define free will in the way we have.
I think the concern is misplaced, and that it turns the state of affairs on its head. To see why, perhaps rather than getting further into the reeds of technicality, it makes more sense to take a step back and look at the questions being asked at a larger scale.
I started my introduction by saying that the question of free will is a very old one; over time, it has come to be one of the most popular and discussed questions in philosophy, and not just amongst philosophers. It is clear that people see the stakes involved in the free will debate as being pretty high: it matters to a lot of people whether or not we are free to choose our actions. Why is this?
One reason is likely because the question of free will has to do with our self-image. We think of ourselves as agents in the world: deliberating, making decisions, acting, changing things. We think that our decision-making is somehow consequential: that it really is the making of a decision that leads to an action. But we also think that our actions are consequential in the sense that we can perform actions which are good, that we might be praised and rewarded for, and that there is the chance of us performing actions which are bad, which might lead to blame or punishment.
Further to this, this traditional view of a world in which we are agents regulates our relationships and interactions with other people. We like and praise our friends when they are kind to us, and we dislike and condemn the people who cause us harm. These kind of affective relationships are important to us, and they inform many of our values and principles. We believe it to be important to behave in certain kinds of ways, and for other people to behave in certain kinds of ways.
The worry is that without free will, this kind of world looks meaningless. If it is not really me who is acting, then it doesn’t mean anything to say that I am doing something good. If my friend has no free will and is not really making any decision to be kind to me, then there is no reason I should feel anything like gratitude or affection when she is kind. I might praise her externally just to extract more kindness (like feeding oil into a machine), but my world-view, with regard to anything I think or feel about myself or other people would have to be radically modified. In a very important sense, I would worry that my actions themselves would become meaningless when there is no free will.
These, I think, are the sorts of concerns that make the question of free will a popular one: this is what people worry about when they worry about free will.
If this is true, then to couple free will with responsibility is precisely the kind of definition that makes the most sense. If the concern is that taking away free will also takes away the significance of actions, then free will ought to be defined as the ability to act in a way that is significant: that is to say, in a way that makes possible all of these affective relationships of praise, blame, gratitude, resentment, and so on. And the name we give for this kind of quality is responsibility: you can be held responsible for an action, meaning that someone can praise or blame you, reward or punish you, because your action holds that much weight.
So it’s clear that the definition of free will that we started with is not deflationary. Quite the contrary, it engages with precisely the kinds of concerns that make the free will debate interesting to begin with. In fact, one might go so far as to argue that if anything is ‘deflated’ in this picture, it is the kind of deterministic picture that holds that nothing much needs to be changed about the way we live. If it is the case that the world is deterministic, but that we can still have all the kinds of beliefs and interactions that ordinary conceptions of responsibility grant us, then determinism (or these broadly deterministic ways of thinking) end up looking like mere technical details about how to describe the world. It does not give us a radical new metaphysics, or affect our lived experiences, or really engage with the concerns that ground the debate about free will.
This kind of view is an interpretation of arguments made by the British philosopher Peter Strawson in his famous essay ‘Freedom and Resentment’. Strawson is not concerned with any of the technical arguments whether for or against either compatibilism or incompatibilism; in fact, he claims in the essay that he simply does not know what determinism really means. He is concerned with evaluating what it would take for us to radically change our beliefs about and attitudes towards other people, and he concludes that the truth of determinism would not have that effect. I have not used exactly the same argument that Strawson did, and I am so far comfortable identifying as a compatibilist, but Strawson’s essay is perhaps my favorite piece in the free will debate. I think that his attitude towards the debate, in thinking that too much technical philosophizing distorts our understanding of what is important about the debate, is an important one.
This article was written by Abiral Chitrakar Phnuyal. Please send an email to email@example.com to get in touch.
Photo Credit: NYU Shanghai