ABSTRACT
There are two general approaches to linguistic meaning: symbolist theories and embodiment theories. The symbolist approach assumes that words and sentences activate mental symbols in our mind, which are abstract, arbitrary, and amodal. For symbolists, comprehension basically consists of a translation process from an external symbolic language (words) into an internal symbolic language (mental symbols). Thus, if you read or listen to the word “cat”, you would activate an abstract symbol in your mind, which is connected to other equally abstract symbols, none of which have any similarity to your visual, auditory, tactile, or motor experience with cats. The symbolist doctrine has been quite successful in cognitive sciences, because it allowed computational analyses of language meaning in terms of symbolic codes such as propositions, lists of features, semantic networks, semantic dimensions, statistical covariations, etc. One problem with symbolism, however, is that language meaning lacks grounding in the world (Harnad, 1990; Searle, 1980; Shapiro, 2008). Namely, words only refer to abstract symbols, associated with other symbols, which in turn are associated with other symbols, and so on; thus, language meaning remains in a sort of “Chinese room”, never interfacing with the world experience (Searle, 1980).
