The adaptive advantage of symbolic theft over sensorimotor toil

Using neural nets to simulate learning and the genetic algorithm to simulate in a toy world of mushrooms and mushroom-foragers, we place two of acquiring categories into direct competition with one another: In (1)"sensorimotor toil,” new categories are acquired through real-time, feedbackcorrected, trial and error experience in sorting them. In (2) "symbolic theft,” new are acquired by hearsay from propositions – boolean combinations of describing them. In competition, symbolic theft always beats toil. We hypothesize that this is the basis of the adaptive advantage language. Entry-level categories must still be learned by toil, however, to avoid infinite regress (the “symbol grounding problem”). Changes in the internal of categories must take place during the course of learning by toil. changes can be analyzed in terms of the compression of within-category and the expansion of between-category differences. These allow of similarity space to be separated, bounded and named, and then the can be combined and recombined to describe new categories, grounded in the old ones. Such compression/expansion effects, called"categorical perception" (CP), have previously been reported with categories by sensorimotor toil; we show that they can also arise from symbolic alone. The picture of natural language and its origins that emerges from this is that of a powerful hybrid symbolic/sensorimotor capacity, infinitely to its purely sensorimotor precursors, but still grounded in and dependent them. It can spare us from untold time and effort learning things the hard way, direct experience, but it remain anchored in and translatable into the of experience