Discovering the conceptual primitives

Discovering the Conceptual Primitives Lisa Aziz-Zadeh (lazizzad@usc.edu) University of Southern California, Department of Occupational Science Daniel Casasanto (casasan@stanford.edu) Stanford University, Department of Psychology Jerome Feldman (feldman@icsi.berkeley.edu) University of California at Berkeley & International Computer Science Institute Rebecca Saxe (saxe@mit.edu) MIT, Department of Brain and Cognitive Sciences Leonard Talmy (talmy@buffalo.edu) State University of New York at Buffalo, Department of Linguistics Keywords: conceptual primitive; development; linguistics; neuroscience; experiment; unified cognitive science Putative Conceptual Primitives Concepts that seem to be (nearly) linguistically universal are good candidates for the conceptual primitives. One immediate goal is to make an organized collection of such potential primitives and then try to design experiments to test which of these proposed primitives is always realized and therefore available as a basis for language (learning). There are many direct cases involving body parts, actions, desires, experiences, etc. There are several studies that indicate that the neural representation of words and concepts concerning direct bodily experiences are based (at least in part) on the circuits that carry out the underlying action, emotion, perception, etc. There remain open scientific questions about exactly how these primitives (e.g., emotions) are encoded, but that is not the current concern. However, there are also a large number of other potential conceptual primitives to explore and some fairly new experimental techniques that can help determine if a concept is primitive in our sense. Obviously enough, any concept that is learned is embodied somehow. The question is whether there is detectable neural encoding of mechanisms that help organize concepts and thus provide a basis for language and thought. Several developmentalists and linguists have suggested such possible conceptual primitives, but there does not seem to be any systematic collection of these suggestions. There should be. Summary We believe that Cognitive Science is now in a position to discover the neural basis for many of the conceptual primitives underlying language and thought. The main concern is conceptual mechanisms that have neural realization that does not depend on language and culture. These concepts (the primitives) are good candidates for a catalog of potential foundations of meaning. This multi- disciplinary, bi-coastal Symposium will explore the current prospects for postulating and then experimentally investigating such conceptual primitives. The five panelists are a developmentalist (R. Saxe), a linguist (L. Talmy), a neuroscientist (L. Aziz-Zadeh) and a psycholinguist (D. Casasanto), and an AI researcher (J. Feldman) all of whom work extensively on the subject. Introduction This is not about some question of innateness . We have known for decades that there is continual interplay between genetic and experiential (including cultural) factors starting from gestation and continuing throughout life. Strikingly, the basic neural wiring of prefrontal cortex is not complete until well past puberty. So it might seem to be a mistake to study fixed neural representations instead of developmental processes. But there do seem to be milestones of normal development and understanding these intermediate states appears to be essential. In particular, the development of language plays a crucial role in human experience and we may be on the threshold of determining the neural substrate of some of the pre-linguistic conceptual primitives involved. Of course, we should not expect every individual and culture to use exactly the same primitives (universality), but there should be enough commonality for an empirically grounded theory of language use and acquisition. Some Initial Examples As we mentioned, many conceptual primitives are linked directly to body parts, actions, experiences, mental states etc. There is already excellent work in several fields aimed at better understanding these. Much of the work on more abstract cognitive primitives has come from linguists and developmentalists. We list some examples of the kinds of postulated conceptual primitives that may be suitable for multi-disciplinary investigation.