Two kinds of psychophysical reduction, part 2: physical

by Dave Maier

Ted Chiang’s (very) short story “What’s Expected of Us” (collected in his recent Exhalation) tells of an unusual device called a Predictor:

Its only features are a button and a big green LED. The light flashes if you press the button. Specifically, the light flashes one second before you press the button.

Most people say that when they first try it, it feels like they are playing a strange game, one where the goal is to press the button after seeing the flash, and it’s easy to play. But when you try to break the rules, you find that you can’t. If you try to press the button without seeing a flash, the flash immediately appears, and no matter how fast you move, you never push the button until a second has elapsed. If you wait for the flash, intending to keep from pressing the button afterward, the flash never appears. No matter what you do, the light always precedes the button press. There’s no way to fool a Predictor.

This is because the device sends a signal back in time, flashing at time t if and only if you press the button at time t + 1 second. Your push causes the flash, even though the flash appears first. No guesswork necessary; that’s how the device works.

In the story, many people, including the narrator himself, take the Predictor to be a demonstration of the unfortunate fact that they have no free will, and that, as the narrator puts it, “their choices don’t matter.” I can see why they think this, but I’d like to try to undermine that intuition here if I can, since it underlies much of the non-science-fictional debate about free will and physical reduction that we started last time. (Note: we will not be solving the free will problem here today, nor do I claim originality for this line of thought; on the other hand, all infelicities in the exposition are my own.)

In this case, the particular intuition seems to be an expression not of determinism, free will’s traditional philosophical nemesis, but instead fatalism: the worry is not that one’s choices are determined by one’s past states, but instead that they have no effect on the future, which is “already fixed.” Yet of course they do have some effect: they determine whether or not the light flashes. It’s just that that causal effect is no longer in the future, but in the past. Let’s see whether this matters.

Discussions of fatalism often begin, reasonably enough, with an example made famous by Aristotle in chapter 9 of De Interpretatione. If I say “there will be a sea battle here tomorrow,” then if that statement is true at the time of utterance, then there must be a sea battle here tomorrow no matter what (and if false, then there can’t be one). That is, the future would be fixed by what is true now; as one commentator puts it, “the die is already cast.” Aristotle himself doesn’t like the sound of that, and his own solution is to deny that contingent statements about the future are either true or false. If this sounds funny – and it should – then think of it this way: their object, a certain state of affairs, doesn’t yet exist, so at the time of utterance there isn’t anything in reality to make it true or false.

I myself don’t think the truth of future-tense statements entails fatalism. It’s true that the sea battle hasn’t happened yet; but that’s just what our sentence says. We use the future tense to say how things will be in the future; that’s what the future tense means. “There will be a sea battle at [a certain time]” is true if and only if there will be a sea battle at that time; that is, if, when the time comes, the sea battle actually happens as predicted. If it does, the sentence was true at the time of utterance. Note that by the time the battle actually happens, it’s too late to use the future tense. The sentence is about the future. I’m just not seeing the threat here.

Putting the semantics of the future tense to one side, let’s try another tack. Naturally we expect – it looks weird even to write this – that when things happen, they happen just the one way and not also some other way. (Imagine a news story: “Airplane crashes in Morocco; also, it does not crash but instead lands safely.” That would be news!) 

Similarly, when we read an story (a normal one, not an experimental literary one or something), we don’t find it unrealistic when it ends only one way and not several different ways. We also don’t think it odd when if we read it again, the same thing happens as before. It doesn’t mean, that is, that the characters were fated to do as they did, or did not freely choose their actions, simply because we already know how things turn out (or even, for literary characters, because the author “determined” their choices for them: even in doing so, he represents them as having freely chosen and thus affected their futures accordingly, even if unsatisfactorily by their lights).

Now let’s say we read the pages out of order: we flip to the end to see how things turn out, and only then read the intervening pages. Here again we know what will happen even while we are reading about how it came to happen that way. Yet there is still no mystery about why it only happens that way and not some other way. It “had” to happen because … that’s what happens! What did you expect?

But wait, you will say, that’s different – the characters don’t know what happens at the end; they can only live the pages of their book in the proper order, when things are up in the air. Yet we all have plenty of beliefs about what will happen in the future, from the elaborate to the mundane, and when things turn out as we expect, which of course happens all the time (look, they delivered the paper again today, just like I thought they would!), we don’t think that things were fated to be that way simply because we knew (= believed truly) that that’s what was going to happen; and indeed the paper got delivered only because someone chose to do so. So who knows what when doesn’t seem to have much to do with it. People still have to act for anything to get done.

Interestingly, another of Chiang’s stories in Exhalation provides another perspective on the same subject. In “The Merchant and the Alchemist’s Gate,” various characters use the titular gate to travel both to the future and also to the past. Here, it turns out, past and future are equally unchangeable, because (as I have already put it) “that’s what happens.” If a character is speaking to his future self, they have different perspectives on what exactly their “future” holds, but at any one time, objectively speaking, only the one state of affairs is actual. Unlike “What’s Expected of Us,” this story relates not how people who used the gate suffered from shattered illusions of free will, but instead, poignantly, how they came to terms, in their various ways, with their inability to change what happens, whether in their pasts or futures.

In reality, of course, there are no such things as Predictors and time gates. But a famous real-world psychological experiment [Libet 1983] has seemed to many to amount to the same thing, and indeed we often see it cited as a disproof of the idea of free will all by itself. Most likely you’ve heard this one before (Daniel Dennett discusses this in a few places), so here is but a quick summary (from Daniel Wegner, The Illusion of Conscious Will (p. 53-4)):

The conscious willing of finger movement [as reported by the subjects] occurred at a significant interval after the onset of the RP [“readiness potential,” a characteristic brain event indicative of imminent action] but also at a significant interval before the actual finger movement. […] These findings suggest that the brain starts doing something first (we don’t know just what that is). Then the person becomes conscious of wanting to do the action. This would be where the conscious will kicks in, at least, in the sense that the person first becomes conscious of trying to act. Then, and still a bit prior to the movement, the person reports becoming aware of the finger actually moving. Finally, the finger moves. […] The conclusion suggested by this research is that the experience of conscious will kicks in at some point after the brain has already started preparing for the action [and thus, as Libet himself puts it] that “the initiation of the voluntary act appears to be an unconscious cerebral process. Clearly, free will or free choice of whether to act now could not be the initiating agent […]”.

You will have noticed that Wegner speaks here of “conscious” rather than “free” will; but subjectively, we’re talking about the same thing: a freely willed act, we feel, is one that I am conscious of doing, rather than one I am simply informed about later on, once my brain has already set the wheels in motion without me, so to speak. In any case these are striking results.

What exactly is our worry here, though? Contra those unhappy Predictor users we saw above, it’s pretty easy to convince yourself that your choices matter. Just as Hume easily left his skeptical worries “in the study,” so will only a madman, faced with the choice of going to a nearby but dodgy hospital or trying to make it to a more reliable hospital farther away, allow a philosophical thought experiment to leach his decision of the life-or-death meaning it actually possesses. Again, whatever conceptual weirdness there may be about the idea of choosing, whether you live or die will be the direct result of the choice you actually do make, and that should be enough.

Still, the felt experience of choosing can indeed seem “illusory” in the face of the uncomfortable thought that, as some commentators put it, our conscious acts of deliberation are not the “ultimate” cause of what actually happens after that. As Robert Kane for example says, we want to believe that with respect to our choices, “the buck stops here.” Libet’s experiment was unsettling enough, but there at least there was no question that our choices were at least somehow involved. Here we may wonder whether we have anything to do with our own choices at all. It can seem that not only free will, but also consciousness itself is, as we say, “epiphenomenal,” removed entirely from the causal chain. I won’t say anything more about this here except to ask that we really try to make sure that we know what we are asking for, and whether that makes sense. Can we really want to “change the future”? What would we be “changing” it from? Why can’t we just try to bring about the (single) future that we prefer?

More traditionally, in any case, the most potent threat to our sense of ultimacy has been the idea of physical determinism. Here, we don’t need to speak of choices or agents at all, free or otherwise: all that determines how things go is 1) the initial conditions of the physical universe, and 2) the deterministic laws of physics. We’re not going to get into that here, except to note that attempts to evade this result by appealing to quantum indeterminacy (I’m looking at you, Roger Penrose) are completely pointless. Unless we cause quantum events with our minds, in order to push the particles in our brains around just so, … never mind, it’s too gruesome.

Instead, let’s consider that regardless of whatever other, purely physical perspective we may take on “what happens,” we cannot do without the vocabulary of beliefs, desires, and choices. Without it we lack intentional explanations of people’s behavior (or indeed the capacity to see it as behavior at all, rather than mere movement or something), and that makes no sense. Q: Why did he get up and go into the kitchen? A: He wants the last piece of pie, which he thinks is in the refrigerator; but he doesn’t know that I already ate it. (Prediction: he will soon return and ask “Did you eat the last piece of pie??”) This explanation, and its associated prediction, is simply not available from the purely physical perspective; and it is only an all-too-common metaphysical prejudice that allows us to dismiss as “illusory” a set of conceptual tools that pays its explanatory way, simply because another, seemingly incompatible set is even more firmly established in its own context (physics is a science, you see, not just “folk psychology”).

In any case, when we consider the physical basis of behavior (which – don’t get me wrong here – is indeed undeniable), it is not the laws of physics which apply to the spot where the rubber meets the road. (They – the laws of physics – are off in the next county somewhere,  another reason why it’s pointless to zoom past the deterministic physical level of billiard-ball particles down farther away still, to the quantum level. You’re going the wrong way!)

So where is this rubber-road interface then? Even when playing with a Predictor (which of course I haven’t …) or considering Libet’s experiments or the free will problem generally, I find it hard to feel myself being pushed around by either the supposed fixity of the future or the laws of physics driving the particles which make up my body. I can still choose this or that without constraint (even while feeling pressure on my theoretical views about the choices I make). Without, I think, changing the subject, we can find a more salient conflict between our subjective experiences and the underlying physical substratum in the context we were looking at last time: that of the neurochemical phenomena underlying the experience of addiction. Here the physical story tells us not of abstract physical law but instead of brain structures communicating (or not) with each other in order to produce our experiences.

In that context, recall, the addict feels his choices constrained by the physical fact of “ego depletion,” an inability to resist urges which the subject recognizes to be destructive and tries to avoid. In neuroscientist Marc Lewis’s description, this happens due to the diminished chemical resources of the ventral striatum when the amygdala and nucleus accumbens are bombarding it with chemically manifested urges. Here the two ways of talking (intentional and physical) combine, allowing us to see the explanatory relevance of one to the other: we anthropomorphize the brain structures at the same time as unpacking our experiences into their physically manifested component parts. More immediately, we feel (that is, the addict does) the weakness which the neuroscientist has identified with a (“purely,” if you like) chemical phenomenon.

As I noted at the outset, this does not solve our theoretical problem – it is generally accepted that the felt experiences of “free choice” and of felt constraint on same do nothing to show that we have “free will” in the contentious sense, or that we do not for that matter. In the end, perhaps my advice is rather like that of the narrator of “What’s Expected of Us”: if you feel like your choices don’t matter, try not to think about that and just act to bring about the future you desire. What more can you do?