in the beginning: word or deed?

Abstract

emphasizing that agents gain from culture-based patterns, i consider the etiology of meaning. since the simulations show that “shared categories” are not based in learning, i challenge steels & belpaeme's (s&b's) folk view of language. instead, i stress that meaning uses indexicals to set off a replicator process. finally, i suggest that memetic patterns – not words – are the grounding of language.