Simulation Semantics: A computational framework for exploring the links between communication, cognition and computation.
The Berkeley/ICSI NTL project is an ongoing attempt to model language behavior in a way that is both cognitively plausible and computationally practical. Work within these projects coupled with a variety of converging evidence from cognitive linguistics, psychology and neuroscience suggests that language understanding involves embodied enactment which we call „simulation semantics“. Simulation semantics hypothesizes the mind as „simulating“ the external world while functioning in it. The „simulation“ takes the best-fitting model of the noisy linguistic input together with general knowledge and makes new inferences to figure out what the input means and to guide response. Monitoring the state of the external world, finding the best-fitting analysis, drawing inferences, and acting jointly constitute a dynamic ongoing interactive process. In terms of language processing, simulation semantics integrates probabilistic dynamic inference with deep semantic analysis based on construction grammar, frame semantics, and cognitive linguistics. This talk reports on a computational realization of the simulation semantics hypothesis and preliminary results on applying the model to vexing problems in language understanding. Based on these results and those from a variety of ongoing imaging and behavioral experiments, I will argue that simulation semantics provides a crucial bridge that ties the multi-disciplinary evidence together to produce new insights into the nature of language.