I suspect that a crux in discussions and worries about (the absence of) “free will” is the issue of possibilities. I also think it is a key source of confusion. Different people are talking about possibilities in different senses without being clear about it, which leads them to talk past each other, and perhaps even to confuse and dispirit laypeople by making them feel they have no possibilities in any sense whatsoever.
Thinkers who take different positions on free will tend to emphasize different things. One camp tends to say “we don’t have free will, since all our actions are caused by prior causes that are ultimately beyond our own control”.
Another camp, so-called compatibilists, will tend to agree with the latter point about prior causes, but they choose to emphasize possibilities: “complex agents can act within a range of possibilities in a way crude objects like rocks cannot, and such agents truly do weigh and choose between these options”.
In essence, what I think the latter camp is emphasizing is the fact that we have ex-ante possibilities: a range of possibilities we can choose from in expectation. For example, in a game of chess, your ex-ante possibilities are comprised by the set of moves allowed by the rules of the game. And since this latter camp defines free will roughly as the ability to make choices among such ex-ante possibilities, they conclude that we indeed do have free will.
I doubt any philosopher arguing against the existence of free will would deny the claim that we have ex-ante possibilities. After all, we all conceive of various possibilities in our minds that we weigh and choose between, and we arguably cannot talk meaningfully about ethics, or choices in general, without such a framework of ex-ante possibilities. (Whether possibilities exist in any other sense than ex ante, and whether this is ethically relevant, are separate questions.)
Given the apparent agreement on these two core points — (1) our actions are caused by prior causes, and (2) we have ex-ante possibilities — the difference between the two camps mostly appears to lie in how they define the term “free will” and whether they prefer to mostly emphasize point (1) or (2).
The “Right” Definition of Free Will
People in these two camps will often insist that their definition of free will is the one that matches what most people mean by free will. It seems to me that both camps are right and wrong about this. I think it is misguided to think that most people have anything close to a clear definition of free will in their minds, as opposed to having a jumbled network of associations that relate to a wide range of notions, including notions of being unconstrained by prior causes and notions of ex-ante possibilities.
Indeed, experimental philosophy seems to paint a nuanced picture of people’s intuitions and conceptions of “free will”, and reveals these conceptions to be quite unclear and conflicting, as one would expect.
I believe that the two distinct emphases outlined above are both important yet insufficient on their own. The emphasis on prior causes is important for understanding the nature of our choices and actions. In particular, it helps us understand that our choices do not represent a break with physical mechanism, but that they are indeed the product of complex such mechanisms (which include the mechanisms of our knowledge and intentions, as well as the mechanism of weighing various ex-ante possibilities).
In turn, this emphasis may help free us from certain bad ideas about human choices, such as naive ideas about how anyone can always pull themselves up by their bootstraps. It may also help us construct better incentives and institutions based on an actual understanding of the mechanism of our choices rather than supernatural ideas about them. Lastly, it may help us become more understanding toward others, such as by reminding us that we cannot reasonably expect people to act on knowledge that they do not possess.
Likewise, emphasizing our ex-ante possibilities is important for our ability to make good decisions. If we mistakingly believe that we have no possibilities, ex ante, and if we in effect fail to think through all relevant (ex-ante) possibilities, it seems likely that we will create highly sub-optimal outcomes, whether it be in a game of chess or a major life decision. Aiming to choose the ex-ante possibility that seems best in expectation is crucial for us to make good choices. Indeed, this is what good decision-making is all about.
More than that, an emphasis on ex-ante possibilities may help instill in us the healthy and realistic versions of bootstrap-pulling attitudes, namely that hard work and dedication indeed are worthwhile and truly can lead us in better directions.
Both Emphases Have Pitfalls (in Isolation)
Our minds intuitively draw inferences and make associations based on the things we hear. When it comes to “free will”, I suspect most of us have quite leaky conceptual networks, in that the distinct clusters of sentiments that we intuitively tie to the term “free will” readily cross-pollute each other — a form of sentiment synesthesia.
So when someone says “we don’t have free will, everything is caused by prior causes”, many people may mistakenly interpret this as implying “we don’t have ex-ante possibilities, and so we cannot meaningfully think in terms of alternative possibilities”. This might in turn lead to bad decisions and feelings of disempowerment. It may also lead some people to think that it makes no sense to punish bad behavior, or that we cannot meaningfully say things like “you really should have made a better choice”. Yet these things do make sense. They serve to create incentives by making a promise for the future — “people who act like this will pay a price” — which in turn nudges people toward some of their ex-ante possibilities over others.
More than that, a naive emphasis on the causal origins of our actions may incline people to think that certain feelings — such as pride, regret, and hatred — are always unreasonable and should never be entertained. Yet this does not follow either. Indeed, these feelings likely have great utility in some circumstances, even if such circumstances may be rare.
A similar source of confusion is to say that our causal nature implies that everything is just a matter of luck. Although this is perhaps true in some ultimate sense, in another sense — the everyday sense that distinguishes between things won through hard effort versus dumb luck — everything is obviously not just a matter of luck. And I suspect that most people’s intuitive associations can also be leaky between these very different notions of “luck”. Consequently, unreserved claims about everything being a matter of luck also risk having unfortunate effects, such as leading us to underemphasize the importance of effort.
Such pitfalls likewise exist for the claim “you could not have done otherwise”. For what we often mean by this claim, when we talk about specific events in everyday conversations, is that “this event would have happened even if you had done things differently“. In other words: the environment constrained you, and your efforts were immaterial. This is very different from saying, for example, “you could not have done otherwise because your deepest values compelled you” — meaning: the environment may well have allowed alternative possibilities, but your values did not. The latter is often true of our actions, yet it is in many ways the very opposite of what we usually mean by “you could not have done otherwise”.
Hence, confusion is likely to emerge if someone simply declares “you could not have done otherwise” about all actions without qualification, since it risks obscuring the important distinction between whether we are constrained by our values or by our environment. And such confusion may well persist even in the face of explicit qualifications, since confusions that are found deep down at the intuitive level may not be readily undone by just a few cerebral remarks.
Conversely, there are also pitfalls of sentiment leakiness in the opposite direction. When someone says “ex-ante possibilities are real, and they play a crucial role in our decision-making”, some people may interpret this as implying “our actions are not caused by prior causes, and this is crucial for our decision-making”. And this may in turn lead to the above-mentioned mistakes that the prior-causes emphasis can help us avoid: misunderstanding our mechanistic nature and failing to act on such an understanding, as well as entertaining unreasonable ideas about how we can expect people to act.
This is why one has to be careful in one’s communication about “free will”, and to clearly flag these non sequiturs. “We are caused by prior causes” does not mean “we have no ex-ante possibilities”, and “we have ex-ante possibilities” does not imply “we are not caused by prior causes”.
Acknowledgments: Thanks to Mikkel Vinding for comments.