Fall, 2013 Colloquium 6
※ 國立清華大學語言所102學年度第一學期專題演講六 ※
演講者：Joshua K. Hartshorne (Massachusetts Institute of Technology)
演講題目：How children learn verbs: New insights on old debates
時間：2013 年 11 月 14 日 (星期四) 中午 12 時 30 分
地點：清大人社院 B305 室
講題摘要：
How is it that children -- but not chimpanzees, pine trees, or computers -- learn language? There have long been two trains of thought, with researchers within the Empiricist tradition emphasizing domain-general representations and powerful learning mechanisms, while researchers within the Nativist tradition emphasizing language-specific representations and innate constraints. (There is an analogous divide in linguistic theory, with some theories positing narrow, language-specific representations and others positing abstract, cross-linguistically universal representations.)
Recent developments in our understanding of learning algorithms, information theory, and the nature of infant cognition have forced both Empiricist and Nativist researchers to adjust their theories. As a result, the two lines of research and theory have begun to converge. This is particularly clear in the study of verb acquisition, which for decades has been an important testing ground for linguistic and language acquisition theories.
In this talk, I describe the emerging consensus theory as to the nature of verbs and how they are acquired, with particular focus on how this relates to our understanding of the human mind. I describe experimental and computational modeling results that support this emerging theory.
演講者：Joshua K. Hartshorne (Massachusetts Institute of Technology)
演講題目：How children learn verbs: New insights on old debates
時間：2013 年 11 月 14 日 (星期四) 中午 12 時 30 分
地點：清大人社院 B305 室
講題摘要：
How is it that children -- but not chimpanzees, pine trees, or computers -- learn language? There have long been two trains of thought, with researchers within the Empiricist tradition emphasizing domain-general representations and powerful learning mechanisms, while researchers within the Nativist tradition emphasizing language-specific representations and innate constraints. (There is an analogous divide in linguistic theory, with some theories positing narrow, language-specific representations and others positing abstract, cross-linguistically universal representations.)
Recent developments in our understanding of learning algorithms, information theory, and the nature of infant cognition have forced both Empiricist and Nativist researchers to adjust their theories. As a result, the two lines of research and theory have begun to converge. This is particularly clear in the study of verb acquisition, which for decades has been an important testing ground for linguistic and language acquisition theories.
In this talk, I describe the emerging consensus theory as to the nature of verbs and how they are acquired, with particular focus on how this relates to our understanding of the human mind. I describe experimental and computational modeling results that support this emerging theory.