Rationally Speaking encore: Does empathy negate physicalism?

[Originally published on November 1, 2005]

by Massimo Pigliucci

Tough question. It has been posed (and answered in the positive) by Michael Philips in a recent article in Philosophy Now. Let's see what this is about. Empathy, of course, is the ability that all normal human beings (there are some pathological exceptions, which are actually going to be very relevant in a minute) have of, in some sense, being in someone else's metaphorical shoes. Empathy, in other words, is that mental phenomenon that allows us to at least approximately feel the pain, or pleasure, being experienced by someone else, which in turn allows an understanding of other people's emotional situations.

Physicalism, on the other hand, is a philosophical term that indicates a family of theories about the mind-body relationship (for a rather technical summary see here). In particular, physicalism says that the mind in fact is a result of brain activity, excluding the possibility of any form of mind-body dualism. There are several versions of physicalism, but two major ones are the so-called "type identity" and "token identity" theories. Bear with me for a second, this is going to be interesting once we pass the technicalities.

A physicalist identity theory basically says that there is some correspondence between physical and mental states, i.e. that in order to have a given mental state (say, feeling pain) one has to be in a certain brain state, because the brain is the causal factor behind so-called mental events. If one subscribes to a token identity theory, then one is saying that any particular mental state corresponds to (it's identical with) a specific brain state. Only that brain state will cause that particular sensation or feeling. On the other hand, the more flexible type identity theory says that there is in fact a correspondence between brain states and feelings, but that this may be a many-to-one relationship, i.e. there may be several different configurations of a brain (or equivalent structure) that can generate a certain sensation in the subject. Keep this distinction in mind, it will be useful in a bit.

Philips, and other philosophers of mind, argues that physicalism is incompatible with the existence of empathy, because empathy implies the existence of qualia, and qualia cannot be accounted for by physicalism. Yup, we need to take care of another little bit of technical jargon. Qualia are so-called "secondary" properties of objects. Primary properties are independent of observers, for example shape. A box is a box regardless of who observes it, human or machine. Secondary qualities, however, are in some sense "in" the observer, for example in the case of colors. Yes, colors are elicited by the physical characteristics of light waves, but the experience of seeing a color (qualia are experiences) demands the subjective presence of a conscious being actually having the experience. (One can already object to this that, in fact, plenty of living beings -- for example insects -- experience colors in a physiological sense, and yet are not conscious in anything like the sense of the term when applied to human beings, but let that pass for now.)

Next, to the crux of the matter. Philips argues that empathy is made possible by qualia, because empathy is about feeling that we can experience something very much like what somebody else says she is experiencing (e.g. pain in response to a hammer hitting a finger). But how do we know what it's like to experience, say, pain? It's not because of a physicalist description of pain as a function of brain processes, but rather because we have the capability to experience qualia ourselves. In other words, the argument goes, physicalism may be able to tell us what sort of nerves and nerve impulses are involved in the feeling of pain, but that has nothing to do with the subjective experience of pain. So, physicalism cannot explain qualia; but since qualia are real (as demonstrated by the existence of empathy), then physicalism cannot account for a real (and important) mental phenomenon. Ergo, physicalism must be wrong, or at least grossly incomplete.

Philips' article goes into some detail into the possible responses open to a physicalist, and offers of course a series of counter-rebuttals by Philips. The problem is that one of the fundamental (and unspoken) premises of Philips' whole critique is highly questionable. It turns out that his arguments are pretty good against what I referred to above as "token identity" theory, i.e. the strictest variety of physicalism that claims that there is a one-to-one correspondence between brain and mental states. If that were the case, one could argue that a complete knowledge of brain circuitry would have to be sufficient to account for all mental phenomena, including qualia. But it turns out that subjective experiences are in fact difficult to pinpoint on a specific set of nerves and impulses. This isn't really surprising, because we already know that token identity theories must be wrong. It seems clear that different individuals, with different brains, can have apparently very similar qualitative experiences (such as perceiving colors).

But things get a lot more complicated when one moves to the more sophisticated type identity theory. In this case, the claim is simply that there are classes of brain structures and functions (e.g., nerves and nerve impulses) that can generate mental phenomena. But the same mental phenomena could be generated by different structures and functions, even by entirely different materials (which makes artificial intelligence possible, at least in theory), as long as certain properties are maintained by the system. Think of it as the idea that different types of hardware can run the same sort of software with relevantly similar (though not necessarily identical) results. While if token identity were correct there would be only one way to produce a word processor that looks and works like Microsoft Word, with type identity once can run different pieces of software (e.g., Word, OpenOffice, etc.) on different machines (PCs, Apples) and different operating systems (Windows, Linux), and pretty much get the same "qualia" (i.e., the same user interface) from all of them. If that's the case, type identity is compatible with the existence of empathy.

Finally, remember my initial reference to the fact that normal human beings can feel empathy? It turns out that some brain pathologies, such as the destruction of the amygdala, make it impossible for a human to feel empathy, because he himself has lost the ability to have emotions altogether. This and similar nightmarish conditions are described in a wonderful book on the human brain, Phantoms in the Brain, by neurobiologist V.S. Ramachandran. What these findings imply, however, is a pretty powerful blow to non-physicalist theories of emotions and feelings: if qualia aren't the result of the activity of certain brain regions (such as the amygdalas), why on earth would people with damage to those regions not be able to experience qualia? This objection is sometimes referred to in philosophy of mind as the "no ectoplasm" clause: we may not know exactly how the brain produces consciousness, but no brain = no consciousness, precisely as a physicalist theory would predict.

Something to ponder, the next time you'll look at the colors of a beautiful sunset...