Leveraging Language into Learning

Abstract

I hypothesize that learning a vocabulary to communicate be-tween components of a system is equivalent to general learn-ing. Moreover, I assert that some problems of general learn-ing, such as eliminating bad hypotheses, deepening shallow representations, and generation of near-misses, will become simpler when refactored into communication learning prob-lems. I hypothesize that learning and reasoning can be a byprod-uct of translation between different perspectives. When two agents with different perspectives learn to communicate, some of the structure of the world relating their perspectives will be encoded into the communication system they gener-ate. If the two agents later communicate to bring their mod-els of a situation into alignment, then the process of transla-tion will effectively reason with the knowledge encoded in the communication system. The imperfect translation certain in such a system is an advantage rather than a limitation. Attempts at translation may serve to check for mistakes, refactor difcult problems, enable learning of deeper structure, and even generate orig-inal ideas. In my master’s thesis, I demonstrated two agents creat-ing a shared vocabulary and inections to communicate the-matic role frames.(Beal 2002a; 2002b) The two agents use very similar representations, so only identity relations are learned. With some generalization, other relations such as cause and effect or proximity should be nearly as easy to acquire. I propose to demonstrate that a group of agents, each responsible for one representation (e.g. vision, lan-guage, motor, social), can acquire a communication system that embodies knowledge about the world and act reasonably in complex situations as a byproduct of translating between representations.