Get access

Innate and Learned: Carey, Mad Dog Nativism, and the Poverty of Stimuli and Analogies (Yet Again)


  • This paper arose as a contribution to a symposium on Susan Carey's The Origins of Concepts at the joint conference of the European and American Society for Philosophy and Psychology in Montreal, July 2011, as well as from ideas sketched in Gross and Rey, 2012. I'm much indebted to both Susan and Steven for many immensely useful discussions of the topic, as well to the Maryland Discussion Group on Carey, which included Steven, Mark Engelbert, Andrew Knoll and Michael Zenz; and to Ned Block, Noam Chomsky, Carsten Hansen, Joseph Levine, Eric Margolis, Paul Pietroski and especially Jacob Beck and Mark Greenberg for extended exchanges. Versions have also been presented at CSMN, at the ENS in Paris, and at the universities of London, Oslo and Bielefeld. Undated, parenthetical page references are to Carey's book, unless the context indicates otherwise.


In her recent (2009) book, The Origins of Concepts, Susan Carey argues that what she calls ‘Quinean Bootstrapping’ and processes of analogy in children show that the expressive power of a mind can be increased in ways that refute Jerry Fodor's (1975, 2008) ‘Mad Dog’ view that all concepts are innate. I argue that it is doubtful any evidence about the manifestation of concepts in children will bear upon the logico-semantic issues of expressive power. Analogy and bootstrapping may be ways to bring about the former, but only by presupposing the very expressive powers Carey is claiming they explain. Analogies must be understood, and bootstrapping involves confirmation of hypotheses already expressible; otherwise they can't select among an infinitude of hypotheses compatible with the finite data the child has encountered, a fact rendered vivid by Goodman's ‘grue’ paradox and Chomsky's poverty of stimulus argument. The problems have special application to minds, since there is no reason to expect a child's concepts to be ‘projectible’ or to correspond to mind-independent natural kinds. I conclude with an ecumenical view that concepts are reasonably regarded as both innate and often learned, and that what is learned can in fact increase what really concerns Carey, the functioning psychological expressive power of the child, even if it leaves untouched what concerns Fodor, the semantic expressive power. Less ecumenically: maybe Fodor (2008) miscast the debate, and the real issue that bothers people concerns not nativism, but an issue on which Carey and Fodor surprisingly agree, his conceptual Atomism, or the view that all mono-morphemic concepts are primitive and unanalyzable. The issue deserves further discussion independently of Mad-doggery.