Lewis, David. "Mad Pain and Martian Pain." The Nature of Mind. Ed. David M. Rosenthal. New York: Oxford UP, 1991. 229-234.
I. There is some sense in which pain is alike for all. However:
A. Beings may differ in the sense that their pain may have different sources.
B. Beings may differ in the sense that their pain may be experienced differently--e.g., a Martian's, a madman's.
II. Theory of mind that assumes the possibility of pain must account for causal and experiential differences in experience of pain--two different kinds of contingency.
A. Identity theory: accounts for common man and madman, ignoring (common) Martian; i.e., source the same but not experienced as such.
B. Functionalism/behaviorism: accounts for common man and Martian, ignoring madman; i.e., experience the same but source of pain different.
C. Embrace ambiguity: pain in sense A explained by identity, and pain in sense B explained by functionalism.
1. Too ambiguous to be useful; however,
2. Ambiguity theory properly reasoned may resolve dual contingencies.
III. Materialist theory of mind joining identity and functionalist theory
A. Pain (or other mental state) = concept that fills a certain causal role with certain typical causes and effects.
2. System of mutual referents to explain states
B. Whatever state fills that certain causal role = pain.
1. Does not mean nonpain = pain; however,
2. Does mean that causes of pain (or other mental state) might not have the effect of pain (or other mental state); therefore,
3. Nonrigid concept of pain (mental states)
C. Objection: that nonrigid concepts must be false; however,
1. If pain = a certain neural state, that identity is contingent; i.e., identity of pain (or other mental state) = a certain neural state that is not any other neural state.
2. Contingency derives from application of a mental state to that (and no other) neural condition, or to no neural condition.
...