作者: Ina Bornkessel-Schlesewsky , Matthias Schlesewsky
DOI: 10.1016/B978-0-12-407794-2.00030-4
关键词:
摘要: This chapter summarizes the architecture of extended Argument Dependency Model (eADM), a model language comprehension that aspires toward neurobiological plausibility. It combines design principles from neurobiology with insights on cross-linguistic diversity. Like other current models, eADM posits auditory processing proceeds along two distinct streams in brain emanating cortex: antero-ventral and postero-dorsal streams. Both are organized hierarchically information takes place cascaded fashion. Each stream has functionally unified computational properties congruent its role primate audition. While dorsal performs sequence (i.e., noncommutative combinatorics) temporal receptive windows increasing size, ventral dependency formation (commutative comprising recognition increasingly complex feature-rich) objects their combinations. The discusses how these assumptions can derive range neuroanatomical neurophysiological findings.