Imperativism works well for sprained ankles. Pains are a diverse bunch, though, and pain science presents a number of interesting cases. Much of my book is taken up with defusing potential counterexamples. These fall into three classes, which I’ll take in order of seriousness.
First, there are strange or maladaptive pains. These are often the first things that people turn to when they want to object. What about cancer pain? What about headaches? What about menstrual cramps? These seems weird. They don’t help — protecting your body doesn’t do anything, and often makes things worse.
Here I think the response is pretty straightforward. First, there’s only an objection to imperativism if there’s no obvious imperative content to these pains. And that doesn’t seem right. A headache tells you to protect your head. That may not do you any good, but there’s no confusion about what the pain wants you to do. (I go through more cases in the book, and show that there’s always a plausible content nearby.)
If there’s a problem, then, it’s for the background theory about homeostatic sensations. Here I give essentially the same response that Descartes’ gave in Meditation Six. The most adaptive pain system overall will be still be subject to all kinds of problems — referred pains, overblown pain, phantom pains, and so on. The pain system is messy in the way that any evolved system is messy. The fact that it goes wrong, then, is no evidence against it its overall usefulness.
The second, more serious sort of objection was first raised by Maura Tumulty. Consider phantom pain. Pain in a phantom hand tells you to protect your hand. But in an important sense you can’t do that, because you don’t have a hand to protect. (This was an even worse problem in my earlier published versions of imperativism, which focused on limiting movement. That’s one of the reasons I moved towards protection-imperatives.)
Here, I think imperativism has a unique and fascinating response available. The phantom pain says “Don’t do that with your hand!”. That’s not a problem with the content, though. It is still a legitimate, intelligible imperative. You just can’t do anything to satisfy it. We can be given commands that turn out to be impossible, even that we know to be impossible, and yet they motivate us to try regardless. (Friends of mine who’ve been in the Army say that this is one of the especially unpleasant features of boot camp.)
Of course, if imperatives reduced to something like deontic modals — if the phantom pain really said something like “You shouldn’t do that!” or “Doing that is bad!” or the like, then there really would be an issue. Assuming that “ought” implies “can,” such contents really would be bizarre. Imperatives don’t work like that, though. Indeed, I think that considering some pains as unsatisfiable imperatives can shed light on the unique harms of chronic pain.
The third, and most serious, sort of objection comes from pains which don’t motivate at all. While these have to be teased out with care, there do appear to be some cases where people feel a pain, say that they feel a pain, have no overriding motivation, and yet do nothing at all about it. That really does look problematic for pure imperativism. The whole point of pains is to motivate you; if you can feel pain without feeling motivated by it, imperativism just looks false.
The clearest such case is so-called pain asymbolia, popularised by Nikolai Grahek in his 2007 book. Asymbolia is a rare consequence of insular damage. Asymbolics claim that they feel pain, but don’t do anything to avoid it. They readily submit to injurious tests, and they are at actual risk of injury as they move around.
My response to asymbolia is complex. Much depends on the details of case reports about asymbolics. They don’t care about pain, but they also don’t seem care about any other threats to their bodies — reports paint them as indifferent to loud noises, being threatened with a hammer, having a truck bearing down on them, and so on. That suggests that the deficit is actually much broader than just a problem with pain.
Rather, I argue, asymbolics have ceased to care about the integrity of their body. That’s a crucial difference for imperativists. On the first day, I mentioned that we are motivated by bodily commands because we accept them as legitimate, which in turn means we accept the body as having the authority to issue commands. Once we accept the body as having authority, we are motivated by its commands, even when (as with phantom limbs) we know they’re silly. There is a good reason why, under ordinary circumstances, we can’t just give up on that acceptance. However, appropriate damage or drugs, might eliminate that acceptance. If so, pains would still command, but we would no longer be moved by those commands–they would be akin to the edicts of a deposed kind.
That’s how the imperativist can model pains which don’t motivate. Nor is this merely a dodge — it has straightforward empirical consequences. It says that you shouldn’t find non-motivating pains without some general breakdown of bodily authority. I suggest that the breakdown of bodily authority manifests as a kind of dissociative experience. I review some of the other standard cases (especially morphine pain and schizophrenia) and show how the dissociative model fits well.
I’ll end on a personal note. I got interested in pain by reflecting on my own experiences after a badly fractured ankle at age 13. I was given morphine in the E.R., and I had the experience described above — the pain was there, but I no longer cared. It was also a profoundly dissociative experience: the pain was off down there in my ankle, and I could no longer see what all that had to do with me. But (as first-person literature on opiate abuse often emphasises), the dissociation was more profound than that. I didn’t much care about much of anything after the shot — not my current predicament, not the surgery ahead, not the long rehabilitation to follow. Pain matters to us precisely because we care about our bodies in a deep and (usually) unshakeable way. That is the source of both the value and the disvalue of pain. Pure imperativism, I believe, is the best way to capture this complex relationship between our bodies and what we care about, and the vital role of pain in mediating between the two.
Hi Colin,
Fascinating stuff!
Here is a worry. I’m sure you’ve heard it before, but I nevertheless find it really important, and would be really interested in your response.
You claim that pains are commands, and commands require authority on the part of the thing that issues them, namely the body. The way you speak about this suggests that you take the analogy to ordinary commands (as in the army, or by parents to children) seriously. You do not merely claim that pains *resemble* commands, but that they *really are* commands.
But here is a disanalogy: a soldier can use reasoning in order to conclude that her commander does not have the authority to issue a certain command at a certain time (perhaps the command is illegal, or perhaps it is clear that the commander is not thinking clearly right now, or perhaps an officer outranking the commander took charge of the battle, etc.). Once the soldier concludes that the commander does not have the authority to issue the command in question (at that time), the soldier is not motivated to obey it. The same appears to apply to all cases of ordinary (i.e., clear-cut cases of) commands.
The case of pain seems different. We cannot use reasoning in order to undermine the authority of out body to issue commands. Even if we think the body is clearly “wrong”, or that, in the present context, we should override the authority of the body (say I’m running away from a wolf and my ankle hurts, making me limp, thereby slowing me down), still we are motivated to obey the command. Even if we are completely sure that we shouldn’t care about a certain part of the body, in a certain life-threatening context, this has zero influence on the motivating power of pain. No matter what I think and believe (via reasoning) about the authority of the body in this specific context, I will (e.g.) continue limping (because of my aching ankle), and consequently the wolf chasing me will catch me.
In short, the disanalogy is this: in the case of ordinary commands, we can undermine the authority of the commander *via reasoning*. In the case of pain, we cannot. Do you agree?
If so, I guess you’d say that the disanalogy, while real, is unimportant. I believe it is important, however. For, if pain motivates us to protect an injured ankle, and it does so regardless of what we think (via reasoning) about the authority of the body at the relevant moment, then the pain begins to resemble a *brute force*. The pain in the ankle appears to simply interfere with our walking, forcing us somehow to limp, regardless of our thoughts and reasons regarding the body’s authority.
Consider again the army analogy. A commander can order that I jump from a plane (with a parachute) or she can instead simply physically push me out of the plane with her hands. The first way will work only if I acknowledge the commander’s authority, and reasoning can undermine this. The second way will work regardless of my reasoning about authority. In light of the above, pains appear to resemble the second way, which suggests they are not commands.
What do you think?
Assaf,
Sorry for the delay — this is a fantastic question! So a bunch of things to say:
1) I discuss in passing the possibility that people might be able to talk themselves out of the authority of the body (you get things that sound like this in discussions of hardcore meditation, people managing long-term chronic pain, etc). I’m not quite sure what to make of these, empirically speaking. I think that if they are real phenomena then you should probably treat them as cases of talking oneself out of the authority of the body. (but see below).
2) I’ve adopted what I take to be the standard Raz/Hart line on authority, on which acceptance is kind of a background condition and then you simply take commands from an authoritative source to provide a first-order reason to act, along with second-order reasons to keep taking the first-order reason into account (that’s a simplification; there’s some modification and to-and-fro in the book).
It is, from what I understand, a known difficulty in the Raz literature on how to deal with cases like your thoughtful soldier. (I don’t know that landscape well at all, so what follows is my own view). I think the easiest thing to say is just that (a) this view is right, and so (b) once you start deliberating about commands, you’ve ceased to take them as authoritative in the right sense (though you still might be moved by them for other reasons: e.g. the fact that the command was uttered by this person might give you prudential reason to act), and (c) insofar as there’s deliberation involved about whether to accept a command, that must be ultimately cashed out as deliberation involved about whether to accept a source as authoritative, not about the particular command.
(I take it that this is complicated in part because most authorities have a limited scope, so you may have to deliberate at least about whether the present case falls under it. Presumably this is what you should say about generals giving orders to commit war crimes. But then just saying that you don’t deliberate is too quick.)
3) You might accept a source as a practical authority for a lot of reasons. Some of those might involve previous deliberations; if so, you could presumably undermine authority by further deliberations down the line. But that doesn’t seem to me to be a necessary condition, and certainly the imperativist shouldnt’ say that we once had to deliberate about whether we were going to listen to our bodies. So there, I think the the thing to say (and this is what comes out more fully in the Asymbolia chapter) is that our acceptance of the body as an authority is ultimately grounded in some pre-reflective sense of care we have for our bodies. You can’t give up on your body unless you stop caring for it, and there are good biological reasons why it’s hard to think yourself into lack of care.
That’s the quick answer; does that address some of your worries?
Nice post Colin, and I look forward to reading your book. Speaking to the second case made against the imperative theory of pain and your response to the case, I find it quite satisfying. If we are to regard pain as some sort of output to maintain homeostasis, such as hunger or thirst, it would probably be elucidating to look at scenarios analogous to phantom pains, in which these other outputs generate imperatives that seem illogical and are not possible to satisfy. For example, ceaseless and insatiable hunger in an individual afflicted with a mutation in the gene/receptor, which leads to leptin resistance, which leads to dysregulation of of appetite, excessive feeding, and obesity. The signal that one is full and should stop eating is there (elevated serum leptin produced by adipose tissue), but is not being received. This example does not mean that the hunger is not legitimate and is thus not a valid imperative, it is merely a case of a dysregulated feedback system, which as you state can happen with the messiness of biology.
Edward,
Thanks! That’s a great idea. I talk in passing about pathological itches, but that’s mostly just taken from an Atul Gawande article in the New Yorker and so didn’t have much about mechanisms. This seems like a much better case — I’ll check it out!
-c