In Defense of Semantic Externalism – Abstract Panu Raatikainen Helsinki Collegium for Advanced Studies P.O. Box 4 FIN-00014 University of Helsinki Finland E-mail: email@example.com Semantic externalism maintains, against the more traditional views on meaning, that the meaning of an expression may not always be determined by a description (or a cluster of descriptions) speaker associates with the expression, or by what is in speaker’s mind, but that the physical and social environment may also play some role in it – that some words refer in virtue of relations that are partly external to the mind. The view has now several adherents, but it has been also vigorously attacked by many distinguished philosophers. In this paper, I review critically what I take to be the most influential philosophical critiques of semantic externalism. One very popular move has been ‘the common concept strategy’; that is, one has repeatedly replied to Putnam’s Twin Earth thought-experiment by the suggestion that ‘water’ had, when nobody knew the chemical constitution of water, in its extension both H2O and qualitatively similar XYZ (e.g. Zemach, Mellor, Crane, Bach, Segal). But this strategy, if generalized, leads to implausibly strong empiricism – to the view that everything that looks the same really is the same (McCulloch). Also, one has often suggested (e.g. Searle, Bach, Crane, Stanley, Segal) that in the relevant cases (e.g. in the Twin Earth case), one does not have a full or complete understanding of the word (say, ‘water’). This is, however, both unnatural and irrelevant. It is unnatural because it entails that we do not really understand the basic part of our native language. It is not relevant because the real issue is whether people are able to successfully refer (‘mean’) by an expression even when they only ‘partially grasp’ the meaning of the expression - arguably one can. Also, this reply takes for grated that understanding itself must a ‘narrow’ concept. But one may question this. Many critical reactions are based on quite oversimplified and uncharitable interpretations of semantic externalism, and ignore the later developments and improvements in the theory. For example, it was never claimed (contrary to what e.g. Unger and Searle suggest) by Kripke, Putnam and others, that all expressions, or even all names refer along the lines of the causal- historical picture. It was always accepted that some expressions really do abbreviate descriptions (e.g. Kripke’s ‘Jack the Ripper’, Putnam’s ‘vixen’, Devitt’s and Sterelny’s ‘pediatrician’). But this already enables one to reply many criticisms. Perhaps the most important further refinement of the causal-historical theory of reference is Devitt’s idea of ‘multiple grounding’. That is, the early, sketchy formulations of the theory seem to imply that reference can never change. But that is, of course, implausible, as Evans clearly pointed out. Devitt, however, has emphasized that it is not only the initial dubbing which determines the reference, but a word typically gets multiply grounded in its bearer in other uses of the word relevantly similar to a dubbing (that is, they involve application of the word to the object in direct perceptual confrontation with it.). This idea enables one to reply, not only to Evans’ reference change worry, but also to e.g. Unger’s much-cited alleged counterexamples (Unger has devised ingenious variations of Putnam’s thought experiment that seem to support contrary intuitions). Critics typically ignore this modification. More recently, Crane (admitting now that ‘the common concept strategy’ cannot work) has suggested that one should rather give up the assumption that meaning (which he identifies with ‘what is grasped’) determines reference; he submits that it does so only relative to a context (also Fodor made at one point the same suggestion). But it is exactly externalism’s claim that meaning, in that sense, does not determine reference. So there is no disagreement here on the basic negative conclusion that the traditional picture of meaning, which assumes the contrary, must go. Anyway, refuting externalism requires further that there is no place for a wider notion of meaning which does determine reference. However, this would require that all seemingly ‘standing’ sentences containing such apparently non-indexical words as ‘tiger’, ‘gold’, or ‘water’, are in reality context-relative such that two sentences can have exactly the same meaning but they nevertheless differ in truth-value. This is extremely strong and counter-intuitive assumption. I argue that it has certain intolerable consequences. Finally, I consider briefly two recent approaches which aim to question semantic externalism, namely Causal Descriptivism (e.g. Lewis, Jackson, Kroon) and Two-Dimensional Semantics (e.g. Chalmers, Jackson). I argue that they fail to provide a convincing real alternative to semantic externalism.