Hector-Neri Castafieda has recently stated a version of what I shall call the theoretical argument for the existence of other minds) I wish to discuss his argument, and offer what I think is a conclusive objection to it. The recent origin of this type of argument is Wilfrid Sellars' essay 'Empiricism and the Philosophy of Mind', 2 which Castafieda acknowledges led him to his own argument 3. Sellars' argument, whether successful or not, is important in at least two ways.
First, it substitutes the making of a theory for the simple induction of the classical analogical argument. Making a theory, whatever else it may be, at least involves the positing of unobservable states (or events, etc.) for the purpose of explaining observed regularities or (better 4) irregularities in phenomena. We can infer the presence of the theoretical (unobservable) state from the observed presence of the proper phenomena; we can explain these phenomena by appealing to the presence of the hidden theoretical state. Thus at one stage in Sellars' argument the state of being in pain 5 is a theoretical state whose presence in a person explains and can be inferred from the presence of familiar phenomena, such as that person's groaning, wincing, etc. 6 (At a later stage in Sellars' argument, pain and other mental phenomena become directly accessible, in that the person who has them learns to report them directly, and need not infer their presence.)
This theory, like other theories and unlike inductive generalizations (analogical arguments), must be assessed not only in terms of its ability to explain the limited range of phenomena that first suggest it, but in terms of its heuristic fruitfulness, its coherence with the favored theories for neighboring phenomena, and its simplicity and explanatory power relative to competing theories, criteria which are difficult precisely to specify but which for our purpose can be appealed to in a relatively crude way.