As far as I know, the term “externalism” is not used prominently by any of the philosophers on this panel to describe his own views.1 The popularity of the term has been, I think, inversely proportional to its clarity. Like the term “naturalism,” “externalism” warmly welcomes a running together of different doctrines, perhaps in the interests of a felt solidarity among those in need of asserting allegiance to a cause. Much of the problem lies in unclarity about what is external to what. Some answers to such a question use terms like “meaning,” “proposition,” and “content” whose uses vary widely among philosophers. I want to make a few remarks about important differences among views that have been labeled “externalist.” I will lay aside epistemic externalisms and focus entirely on externalisms relevant to language and mind. One family of views often termed “externalist” concerns language. It is

trivial that many entities that are in fact objects of linguistic reference are external to – independent for their nature and existence of – language. It is not trivial that some factors constitutively determining which objects linguistic terms refer to are irreducibly external to – independent of – idiolects, dialects, and communal languages. Such factors include causal chains, contextual parameters, and molecular structures, none of which individuals need not be able to specify. This point was established, I think, by Strawson for demonstrative reference (Strawson 1959), by Saul and Keith Donnellan for proper names (Donnellan 1970; Kripke 1972), and by Saul and Hilary for natural kind terms (Kripke 1972; Putnam 1975b, 1975c). I certainly endorse the point. It marks one sort of view often termed “externalism.” There are many importantly different and incompatible views about

meaning, as distinguished from reference, that are called “externalist views.” A hyper-Millian view holds that the meaning, or the specifically semantical contribution, of names and perhaps other terms is exhausted by its referent. No one on this panel is committed to this position. There is the less committal but still Millian view that elements of linguistic meaning, or linguistic propositional content – particularly, elements of the meaning of proper

names and natural kind terms – include objects of reference in the physical environment. Hilary accepted such a view in the 1970s and Saul has shown some sympathy for it. I am non-committal, shading toward the doubtful. I think that the notion of meaning has a root conceptual connection to the

notion of potential understanding. I think that a direct connection between understanding and physical objects has never been made clear or plausible. Understanding is certainly always perspectival. The idea of understanding, or indeed perceiving, a physical object (even as a component of meaning) neat is, I think, incoherent. So if physical objects in the environment are to be considered components of meaning, some account of meaning that loosens its relation to understanding must be developed. Clearly, some sort of understanding, some sort of psychological competence, is associated with proper names and natural kind terms. A serious development of the fundamental notions of meaning and understanding that underlie even this moderate Millian view has never, to my knowledge, been undertaken. I believe that if semantical discussion is to get beyond the impasses that have faced it in the last couple of decades, more attention needs to be devoted to the conceptions and motivations that underlie various notions of meaning and understanding. So I believe that the moderate Millian view is at best unclear. It does not help to add the negative point, originally made by Hilary, that

for “externalists” meaning is not in the head. Insofar as meaning is shareable, even those who are labeled “internalists” can and should say that meaning is not in the head. I believe that meaning is an abstraction, hence not located anywhere. So I am doubtful that there is an acceptable notion of externality according to which it is true, interesting, and distinctive, to claim that components of meaning are external to language, the individual, or what not. But such a claim is commonly associated with the term “semantic externalism.” I think that the term is so associated with unclarity and dispute on these issues that it is better to drop it. “Semantic,” “meaning,” and “understanding” are sufficiently in need of explication that adding another term “externalism” to the mix just piles up intellectual debts. In any case, I think that the power and interest of the arguments regarding

linguistic reference that were given by Donnellan, Saul, and Hilary transcend technical issues about exactly how to regard meaning or linguistically expressed propositions. To return to something I said earlier, the constitutive determiners of reference are partly independent of individuals’ idiolects and of communal languages. Perhaps here we have a position that the three of us share that is very close to what is popularly understood as “externalism” about language. Let us turn from language to mind. In my view, the positive results on

linguistic reference are ultimately founded on a broader set of phenomena regarding mind. And the fact that the determiners of reference are external to idiolects, dialects, and communal languages is ultimately to be understood in terms of the independence of such determiners from individuals’ psychologies.