By Paul Newall (2005)
In this article we'll consider the problems associated with free will and determinism, starting by explaining the terms involved, the difficulty (if there is one), and then trying to understand the proposed solutions. The importance of the topic is plain enough: it comes up often, in many contexts, and is one that people can easily understand the relevance of; which is only to say that it isn't just for the philosophers.
What do philosophers and laymen mean when they start worrying about whether we have free will, or what the consequences of determinism must be? We'll begin by making sure we know what is at issue before we worry about the implications.
The idea of determinism is easy enough to explain in a simple fashion but considerably more complex if we want to use it to make an argument about free will, whatever that is. Initially, then, we'll define it in the common sense fashion: determinism states that the way things will be is a result of how things are and the work of natural laws. That is only to say that if we know exactly how things are at the present moment and the laws that govern how the world (or the universe) works, then we can derive how things will be at some future time. Bearing in mind that we've skirted over some problematic issues that we'll come back to, let's consider some examples.
Take a simplistic case first. We know that denizens of Auckland are keen on their rugby, and moreover that if their team wins then they are going to be happy. Conversely, if they lose then they are going to be morose in the special way that only rugby fans can be when the finest attacking side on the globe somehow manage to undo themselves yet again. Add to this the fact that Auckland have just won a game, and we can say that Aucklanders will be happy. We use the way things are (that Auckland have won again), together with the law (that Aucklanders are happy if their team wins), to determine that the Aucklanders will be happy again, at least until they throw another cross-field pass under their own sticks for the umpteenth time.
Now consider something that many people hope eventually to achieve, namely that all the laws of physics, or of nature, have been discovered and understood, along with the (hypothetical) situation wherein we know the position and other characteristics of all fundamental particles (whatever they may be) in the universe. We can then apply the laws to (again) determine how everything that follows for these particles will play out over time. (Whether this is possible in light of certain other theories in physics and elsewhere is not important for the purposes of our example.)
In general, then, we have an iterative process: we take the state of the world at some time t, a formula (or law) that tells us how to get from t to t+1, and hence we know how the world will be subsequently. In these simple terms, it is little different from figuring out how much money we'll have in our savings account at the end of the year by knowing how much we had and how to work out how it will change over the period of investment.
Now we'll look at tightening this definition. We said that determinism involved several aspects:
- The way things are;
- Natural laws;
- The way things will be.
Taking the first, what do we mean by "the way things are"? We can say that we are concerned with the state of the entire world, or universe, but why not take only a small portion instead? That would give us, say, determinism for a small region (or even particle) based on knowledge of how it will behave in future—a decidedly less ambitious endeavour. The problem here, though, is restricting the domain in this way: can we even do so at all? Although it may seem plausible initially, there are myriad factors that might influence the area we're looking at, especially if we're talking about laws of nature that are presumed to apply everywhere.
Let's look at our examples again. If we try to restrict our concerns to Auckland only, we fail to acknowledge that news from elsewhere is coming in, affecting the mood of Aucklanders. We also assumed—quite reasonably—that rugby is the only important thing to Aucklanders (indeed, the entire world), but there may yet be some people who have failed to heed the gospel and want to consider other matters, too, for their sins. It seems that our attempt to determine the mood of Aucklanders is doomed to fail because our restriction was impractical.
Now take our second example. If we want to consider only a portion of the universe—say the earth, or an area of it—we need to bear in mind that the laws of physics will apply everywhere. If we look at a law like gravity, for instance, we suppose that it applies everywhere, such that the other particles, planets, stars, and so on, in the universe will have an effect, however small, on the earth or the region we've taken. How can we determine the future behaviour of a restricted area like this if we explicitly discount the influence of the other regions of the universe?
In general, once again, we seem to be forced to take the whole of the universe in order not to miss the impact of whatever we leave out. That means that determinism will have to apply everywhere.
Looking now at natural laws, we can first say that there are some objections that can be made to the very possibility, which we'll come to later. For now, what do we require of these laws such that determinism makes sense? For one thing, it won't help to have natural laws that only hold for a certain length of time: if Auckland eventually becomes populated by people who don't appreciate their rugby, our attempts to determine their mood following games will be useless. To take a less horrifying prospect, if it so happens that the laws of physics only apply until, say, ten years in the future, at which point dropped items float in the air and tourists are no longer annoying, then our efforts to determine the future behaviour of the universe will be dashed. Plainly, then, we require these laws to hold at all times.
Secondly, suppose that our observations on Aucklanders only apply to those from Ponsonby; that would render a prediction (based on our determinism) for someone from Parnell useless (we simplify the actual rugby circumstances somewhat here). Similarly, if the laws of physics work well enough on earth and in regions nearby, but behave in a completely different fashion in some far away galaxy (perhaps a region of the universe in which all politicians are honest), then our determinism will fail. The moral of the story here is that we require the laws to hold everywhere.
To summarise, then, we want the natural laws under consideration to be universal (as befits the first part of our tale), applying everywhere and always. We also want the laws themselves to be deterministic: it wouldn't help much if they weren't, since a well-determined present plus indeterministic laws would make the future state indetermined, too.
Lastly, consider what we mean by "the ways things will be". One thing we could ask is why the determinism has to function only in one way? Why not, for instance, require that if we know the laws of nature and the way things are then we can also determine the way things were? Is the past not as fixed as the future under these circumstances? If so, it would seem that taking the "arrow of time" in the direction we're accustomed to is moot: if the past, as far back as we care to consider, is determined, then it seems trivial that what follows will be also. Quite often we find that talk of determinism ignores or minimises this notion.
This brings us to the question of how long the universe need be determined for: if we take it, say, that the way things are, together with natural laws, determines how the universe will be for perhaps the next five hundred years but no further, it seems that although the universe is not truly deterministic, it is for the time we'll be alive to think about it. It would appear, then, that we don't need to have a deterministic universe forever, although this is generally what we're thinking about when we worry about the implications for free will.
To conclude our study so far, we have the idea of determinism as being able to give the future state of the universe from its present state and the laws of nature governing it. Note, however, that although this has thus far been an epistemological issue, but determinism is very much metaphysical; determinists hold that the kind of thing we've been discussed about fixed laws determining the future state of the universe even if we don't know about it and perhaps can't know. If we knew the required information about the state of the universe and the laws governing it, then we could determine the future state, but that state is still determined in the same way even if we don't.
Let's now take a look at free will, so we can begin to appreciate the problem that will arise.
Much like determinism, what we mean by free will seems obvious enough (which is usually a good reason to suppose that it isn't). Take some examples:
- If I live in Auckland, I choose whether to support their rugby team or not
- If I'm interested in philosophy (or not), I choose whether I read this article or not
- It's up to me what my favourite colour is
- I choose what to pick from the menu (my finances notwithstanding)
We could expand such a list ad nauseum; in general, the decisions we make are ours to make in the first place: we choose according to our will. Obviously there are pressures put upon us (from our peers, our upbringing, our circumstances, and so on), but ultimately the choice is ours.
Alas, the matter is not so simple. To start with science fiction (or not, depending on our stage of paranoia), what if our decisions are influenced by others? We could, for example, be swayed by the wrath of Khan, some kind of government mind control designed to make us vote Republican (an implant of sorts) or else an alien influence. Are the choices we make then still examples of our free will? It hardly seems reasonable to say so. What, though, of the impact of the pressures we considered above? Some people are well able to ignore the advice of their parents and lose their teenage years in a fog of alcohol, as though they're the first and only ones to think of so doing, but various culture-specific (and other) influences are not so easy to escape. Take the following hypothetical list of circumstances:
- Everyone I know supports Auckland
- Everyone I know thinks that to support anyone else is tantamount to asking to be institutionalised
- My parents and grandparents support Auckland
- The few people I know who don't support Auckland are ridiculed endlessly
- I was brought up to support Auckland
If I then decide to support Auckland, was it really an expression of free will? (We could just as well ask the same question if I decided not to.)
Clearly we've been concerned here with the extent to which a freely willed decision is really ours. It does no good to say that free will is what we have when we choose one direction instead of the other possibilities when beyond the influence of such circumstances, because situations like that are few and far between (indeed, we could argue that they don't exist at all). What we want to say, then, seems to be that insofar as we can, free will is when we are able to choose more than one option and do so by our own volition.
Consider now, though, not just external influences but those from "within". If I have a sweet tooth and choose candy instead of an apple, did I exercise my free will? What if someone who suffers from kleptomania steals something? (To hint at what will come, was it my fault, or theirs, respectively?) The issue here is whether such decisions can be called examples of free will.
We can add to these the various desires we often have that are referred to (although not by everyone) as "base", such as opting to watch more television instead of studying for a test, or choosing to eat a forbidden food when on a diet. As we discussed in the ninth article, some thinkers want to discount such choices and consider free will to be what we would decide on if we were in full possession of the facts and, as it were, our own masters. The problem, as before, is to ask if we're ever in such a situation: when can I say my choice was freely made? We can approximate it, but when are we truly free of base desires and influences that would lead us to choose wrongly? More importantly, perhaps, how do we determine in the first place which is the "right" choice?
Although there are many more angles we could take, to draw these strands together we can look at the definition of free will that Aristotle advanced many years ago:
Based on what we've considered above, can it ever be the case that we're in a situation simplistic enough to satisfy this? If not, what does it mean for the notion of free will before we've even reached the problem?
Regardless of the difficulties we've considered above, we now come to the issue of compatibility, Here we have a fair idea of what's at issue: we talk of couples being incompatible if there's something about them that will make quarrels inevitable, say, or of supporting Auckland being incompatible with supporting North Harbour: that is, the one excludes the other, or the one circumstance must lead to the failure of the other. In the context of this discussion, we have two concepts: free will and determinism. Compatibilism suggests that the two can coexist; incompatibilism that they cannot.
Hopefully the problem should by now have become obvious, assuming it wasn't already known of beforehand: how can we have free will if everything is determined? Conversely, if everything is determined, how can we have free will? Still another version is to start with free will and ask how, then, can determinism be true?
Often this problem is framed in an ethical context: if the kleptomaniac we considered above steals something, we usually attribute blame and, if we catch him or her, we assign some form of punishment; at the very least, we generally take a dim view of theft. If the future state of the universe if determined by its past and the laws of nature, though, then the kleptomaniac was surely bound to steal; can we then blame him or her for so doing? It hardly seems fair to complain at something that had to happen, besides bemoaning that it did. After all, were we to fall from a tall building and find ourselves plummeting to the ground rather too quickly (assuming, of course, laws such as gravity), we may utter a good many things, mostly remarking on the unfortunate circumstances, but it would scarcely make any sense to declare "how strange that I am falling down instead of floating; who is to blame for this?"
Another aspect of the problem concerns accomplishment: suppose that someone has tried unsuccessfully to achieve some dream—like playing for Auckland, to take a realistic example. By dint of sheer effort, perpetual practice and—ironically enough—a great deal of determination, they eventually manage it and turn out at Eden Park. Similar stories abound and we hear about them every day in one context or another; usually we agree, readily, that the person is due our praise and congratulations. What, however, of determinism? If their achievement was bound to happen, on account (again) of the past and the inevitable laws of nature, then why pat them on the back at all? Why not, instead, say "so what? You could hardly have done otherwise?"
This problem is one that cuts against our common beliefs both that crimes should be punished (in some way or other) and that accomplishments (and many other things) should be praised. We can extend it to cover others, especially with a fertile imagination, but one area in which it has traditionally been of great importance is theology, where the question of how much freedom we can have if God is all-powerful and all knowing has caused much debate. To what extent are we free to sin or not to sin, say, if God already knows what we'll decide and has made it so in the first place? Rather than take this issue specifically, we'll cover it in the general case.
Arguments for incompatibilism
The basic form of argument for incompatibilism seems straightforward enough from the above examples, but it's important to note that neither compatibilists nor incompatibilists dispute the fact that we make decisions: the first issue lies in whether we cause our choices to come about in the right way; i.e. if we determine them for ourselves sufficiently (as we discussed above when looking at free will) or not. The second way to look at the matter is (as we also remarked on before) to wonder if determinism takes something away from us—namely, the power to choose one way or another, and hence the issues of moral responsibility that go with it.
One popular and easily understood argument is a form of the moral problem from before: suppose we've robbed a bank dressed in a particularly bad costume, leading to an embarrassing arrest; at the resulting trial, we're being defended by Lionel Hutz, the famously inept lawyer. Instead of mounting any kind of defence or providing a reasonable plea, he maintains instead that, as a result of determinism, we had no free will and hence no choice but to do otherwise. The syllogism here runs as follows:
- P1: We are only guilty of (or responsible for) robbing the bank if we chose to do so;
- P2: We could not have chosen otherwise, because of determinism;
- C: Therefore, we are not guilty (or responsible).
It's not clear that this is an argument for incompatibilism, since it rather suggests that there can be no such thing as responsibility at all—we could replace the bank robbery with any other action and get the same result—but that is hardly what most incompatibilists want to imply. Moreover, and even with strengthening the syllogism somewhat, we run again into the problem of defining free will: when, exactly, are we ultimately responsible for a choice, given all the influences and factors playing a role?
Another incompatibilist argument is to wonder about what we could have done in the past. We can rue our decisions where we now think we went wrong, but in the absence of a time machine we can seemingly do very little about it. If the past is fixed, though, and the future is determined from it on the basis of natural laws, then can the future be "open" at all? On the face of it, it would seem not.
The difficulty here is that causation—a very troublesome and mysterious subject in itself—appears to only run in one direction, from past to future. In that case, there is nothing that we can do to change the past, irrespective of determinism. The choices that we make may end up determining the future, but they can never have an influence on the past. What's going on here is very subtle: the past is closed because causation only works forwards, and for that reason the future is not.
This may not be convincing, so we can strengthen the incompatibilist position: take the decision to read through this article, and consider the propositions:
- I decided to read this article.
- I decided not to read this article.
Since we're reading, it seems that (1) is true. Now according to determinism, we have the past state of the universe and extant laws of nature to account for why (1) occurred and not (2)—we had no choice but to read. If we wanted to say that we had the possibility of choosing (2), then we would be suggesting that either we could alter the past state of the universe or the laws of nature, both of which seem rather beyond our meagre powers. Does this mean the incompatibilist is right?
Perhaps not; surely the compatibilist is not suggesting that the possibility of choosing otherwise requires a miracle every time we suppose we've done so? What the compatibilist is instead saying is that if we chose differently then either the past state of the universe or the laws of nature must have been different also. Once again, this is quite subtle, but all we're saying is that we might have done otherwise if the circumstances had not been as they were: we chose to read the article, but if the state of the universe had been slightly different at the time, we might have not. This is not merely a clever way to absolve your narrator of blame for boredom, of course.
A much easier way to argue for incompatibilism is to show that determinism is false, or that indeterminism must hold. Below we'll remark briefly on the prospects for both by considering the evidence and our physical theories to date. It won't help much, however, to find determinism on shaky ground if the notion of free will is in as much trouble.
Arguments for compatibilism
Taking now the other possibility, what of arguments for compatibilism? In our discussion immediately above, we've already seen some of the ways in which compatibilism appeals by answering the incompatibilist's ideas, as we'd expect: if it isn't obvious that the two concepts are incompatible, then we would invite some kind of justification to that effect. That said, the problem does seem to have something to it, even if we aren't yet convinced by the arguments.
One suggestion was advanced by Hume when he said that some kind of determinism is required if we want to have free will. After all, if we want to be free to decide for ourselves and hence make plans and choices, we expect the same action or cause to lead to the same result or effect each time—otherwise what use is free will if we never know what will come of our decisions?
Another way to think of compatibilism is to question the assumption of the past being "fixed" in some way, since some results in the sciences seem to cast it into doubt. Some physicists, as well as some so-called eastern philosophies, have suggested instead that determinism may be a relationship wherein every aspect of the universe has an influence on (or determines) every other part, with the links being more like a web than a chain.
Problems with the problem
We've already seen that there are a good many difficulties with defining free will in the first place. What other problems are there?
Several philosophers are skeptical of the existence of natural laws and have given some very powerful arguments against them, including particularly Bas van Fraassen, John Dupré, and Nancy Cartwright. How, for instance, would we prove that a proposed law was in fact what it claimed to be? Many such "laws" in the past have been found to be mistaken, so maybe the same will happen to what we suggest nowadays—does that follow? Not really (cf. our fifth article), but it should perhaps at least give us caution. If we want to assert that such laws exist, though, we might want to explain how we know this and why they should—a difficult task, to say the least.
Another aspect to this problem, which some might consider even more important, is saying what these laws are. An obvious candidate would be the laws of physics as we currently understand them, but that isn't entirely helpful. As we noted above, what if they're wrong? Once again, theories in the past that seemed virtually certain have proven to be mistakes, and the theory that many physicists take to be our best yet (the quantum theory) is said to be indeterministic (although that is open to severe critique in itself). If we could say that we're approaching some "final theory" that some people aim or hope for, then perhaps we could base an argument for or against determinism on the form it takes (for example, something similar to the quantum theory), but where does that leave us now?
There is a major epistemological issue with determinism. As we said above, some physicists (and others) interpret the quantum theory to be indeterministic, which, if it turns out to be the final theory or closely related to it, would cast much doubt on determinism. Advocates of determinism respond that this interpretation is doubtful (indeed, we could argue that such a reading is methodologically untenable), and that, even if it wasn't, the universe on a larger scale (i.e. not the microscopic quantum level) behaves deterministically enough. Indeterminists are not convinced. Other theories give support to determinism, but in general we can say that it is as yet far from clear what physics tells us about the deterministic nature of the universe or otherwise.
Back to the start
To conclude our discussion, then, the issues of free will and determinism, along with their relationship, are thorny ones. They have vexed philosophers for many thousands of years and involve considerations from other areas of philosophy and science alike. More important, perhaps, is the question of whether you decided to read this article yourself or not.
Dialogue the Tenth
The Scene: Trystyn and Steven have met for coffee and are discussing the events of the previous evening.
Steven: You know, all this has got me thinking.
Trystyn: How so? You seem to have calmed down a good deal anyway, whatever the reason.
Steven: Well, it occurred to me this morning that much of that stuff was pretty much bound to happen.
Trystyn: "Bound" how? Which parts of it?
Steven: (Takes a sip of coffee before replacing his cup carefully.) Think of it this way: you're a quiet guy, almost all the time. I should've guessed that you wouldn't just blurt anything out, and I shouldn't have expected it. Hell, I doubt you would've, even if you'd wanted to, right? (He doesn't stop for an answer.) Similarly, I always act like an idiot when I first meet someone—I don't stop to think, or to look from a different perspective. It never really occurs to me to wonder about what other people might want, I guess, so I blunder on regardless.
Trystyn: Even supposing that's all true, why does that mean it was bound to happen?
Steven: I'm sure I don't know or understand the philosophical ins and outs of it, but it seems to me that most of the decisions we make aren't really ours after all; at least a good part of them are already decided by other factors.
Trystyn: Like what?
Steven: Like your upbringing, or your social circle, or the way you behaved as a kid. If you didn't say much when you were little, you probably won't grow up to be the sort of guy who volunteers information; far more likely that you'll only talk when someone asks you something, at least until you get to know them better or become more confident. What's more, it doesn't seem fair to blame, say, you for not telling me something; after all, if I couldn't really have expected anything else, then it doesn't make much sense to get annoyed at you for acting just as I should've supposed you would—in fact, in the way that made you my friend in the first place.
Trystyn: Similarly, then, I can't really complain at you knocking books out of my hand?
Steven: Heh, not really. If I'm an idiot then you've just got to get used to me, I suppose. You see my point, though?
Trystyn: Sure, but how much of it is decided beforehand? Surely we still have some kind of responsibility?
Steven: Well, that's what I was thinking about this morning, instead of wondering what I'd have for breakfast. If it was all decided for you, I don't see how we could hold anyone accountable for whatever they did, however wrong it may seem. That's the end of the line for jails, or so it seems.
Trystyn: It'd be the same for whatever they did right, too.
Steven: Huh? How's that?
Trystyn: It's just the same thing in reverse: if I can't blame you for screwing up because you had to, neither can I praise you for doing well—you had to do that as well.
Steven: There goes my dissertation—don't tell my professor.
Trystyn: We'll keep it between the two of us—we'd have to.
Steven: Heh—fair enough. I don't see any way around it all, though.
Trystyn: There are several ways we could try.
Steven: It figures that you'd say that.
Trystyn: Of course. The first is to ask whether there really is a problem: if everything is determined, does it follow that we have no choice in any matter?
Steven: It seems clear enough that we don't.
Trystyn: All the more reason to suppose that we do, then. (He winks.) If everything is fixed in advance then things had to turn out the way they have; nevertheless, if they'd been fixed in a different way then they'd have turned out differently just the same. When we say that we chose differently, it just means that we could've done differently if the facts had been otherwise, but they weren't.
Steven: Um... what?
Trystyn: It took me some time to figure that out myself—just mull it over. If I wasn't so quiet, I might've spoken up; but I'm not, so I didn't. If the circumstances had been different, I'd have chosen differently.
Another way is to say that either the concept of free will is in trouble, or determinism itself, or both. There are some good arguments for either.
Steven: In trouble how?
Trystyn: Take free will: given that "no man is an island"—with there being so many different influences on us, our thoughts, feelings, ideas and behaviour, all the time—can any decision we make really be said to have been a "free" choice, of our own volition? On the other hand, if everything is determined by what came before, how is it that they're determined? You yourself probably know that scientific laws are not so clear-cut...
Steven: Alas, we're usually wrong.
Trystyn: Right (he winks again), and some people say the quantum theory is indeterministic while others insist to the contrary.
Steven: Don't let's start that off again. (He looks around the café...)
Trystyn: We could just wonder if this purported incompatibility really exists at all. It could be that every action influences and is influenced by every other, making things a whole lot more complicated—and beautiful—than all this talk allows.
Steven: I want to know if this means you're buying me lunch.
Trystyn: Only your horoscope knows...