An excessively simple theory of pain

A physicalist research program is to identify physical state types that underlie mental state types.

Here is an overly naive physicalist-friendly theory of pain:

1. Pain is what occurs in the triggering of a damage-detector state that is linked to aversive behavior.

This theory is simple and elegant. Arguably, all actual instances of pain fits with the theory. So if there is an extensional problem with (1), it is that it classifies as pains states that aren’t pains. Plants feel pain on (1), and a program that monitors the health of a hard drive and relocates data away from damaged areas feels pain.

For the above reasons, I assume nobody will find (1) plausible.

The physicalist research program needs to be based on fitting physical stories to the data about mental states. This data has to be data about where a mental state type occurs and where a mental state type does not occurs. For as (1) shows, it is too easy to find physical stories that simply fit data about where mental states do occur. In fact, we can do even better than (1) if our only constraint is catching all cases of pain by giving this story:

2. To be in pain is to be physical.

We have a nice source of data about where pain does occur: our own experience, the reports of other persons, and the behavior of animals similar to us. But do we have data about where pain does not occur?

We could say this: I now know that I am not in pain. So (2) is refuted directly: I am a physical being, but I am not in pain. Slightly more subtly, I can refute (1) as follows. No doubt as I am writing this, some of my cells are being damaged by some factors in the environment, and my body is doing something aversive about it. But I am not in pain.

But I think this argument against (1) is not as evidentially strong as it seems. The leading physicalist theory is functionalism. If functionalism is true, then pain is some sort of a functional state. I exhibit this functional state in the brain. But my body could mutate so that my stomach would host that sort of functional state, without any connection between that state and my brain. When my stomach would host the state, then on the functionalist theory I would be in pain. But it would be a pain that I am incapable of reporting, because the state would not be connected to the speech centers in the brain. This would be a case where either there are two conscious things—I and my stomach—or a case where my consciousness is divided into a brain-based and a stomach-based consciousness, and only the brain-based consciousness is able to drive action. (Similar phenomena seem to happen with split-brain patients.)

Likewise, then, if some of my individual cells were currently being damaged, and my body detected that damage and engaged in something aversive, then it shouldn’t be expected that I would report pain, even if (1) were true. Rather, if (1) were true, then in a scenario like this, either there would be two conscious things, one located in and around the particular cells and the other in the brain, or else I would have a divided consciousness. In neither case would the absence of pain to the brain-based consciousness be a refutation of (1).

So it seems I cannot refute (1) by observing my lacks of pain, because the pains predicted by (1) could be occurring in a different conscious thing found in my body or in a consciousness divided from the one that is driving my paradigmatically human activity.

I think the best way to refute (1) is to rely on intuitions like that plants aren’t conscious. But if naturalism were true, I wonder if there would be any reason to think such intuitions are truth-conducive.