Symbol Grounding and Transcendental Logic

Symbol Grounding tries to answer the question as to how it is possible for a computer program to use symbols which are not arbitrarily inter-pretable. Whereas the signs in conventional programs are just \parasitic on the meaning in our heads", grounded symbols should possess at least some \intrinsic meaning." This paper gives a brief overview of what Symbol Grounding is and summarizes some of today's connectionist Symbol Grounding models. Instead of concentrating on cognitive linguistics, we try to present an alternative view of Symbol Grounding. Our analysis reveals that Symbol Grounding is in fact the endeavour of automated model construction. Although it originated in a somewhat anti-formal spirit it is (necessarily) full of parallels to classical symbolic logic. We present our view that Symbol Grounding is in fact a con-nectionist version of transcendental logic, which is the basis for generating formal models of non-formal domains. Such formalizations are inherently logical, though not only based on formal but also on material truth conditions. 1 The Symbol Grounding Problem One of the best known criticisms of what is now often termed \the symbolic approach to AI" has been introduced by John Searle. His Chinese Room Argument Searle 80] is one of the most debated arguments in the philosophy of AI (and will therefore not be repeated here). Searle points us to the fact that all semantics in an AI program is but pure syntax. The central construct of such programs, the symbol, has no other meaning than that which the programer has given to it through connecting it to other, however again, symbolic descriptions. The reason for this lies in the very fact that generally symbols are arbitrary and that AI programming symbols are especially arbitrary since their concrete form implies no restriction on how they are processed. This diagnosis has been the starting point for Stevan Harnad, who in his 1990 paper \The Symbol Grounding Problem" de-nes a research problem for the AI community. This research problem is stated as the quest for a method which grounds symbols in something else but yet another symbolic description. Before we turn to Harnad's proposal of how to solve this problem it is worth pointing out some subtle diierences in Harnad's and Searle's accounts of the problem. In accordance with his previous work Searle 83] Searle is more concerned with the problem of intentionality than with meaning. His main argument is that symbol structures …