1 2 3 4
讲座信息:M & T talks abstracts
来源: 外国语学院 更新时间: 2017-06-05 点击: 13

南京理工大学外国学院英品牌设:语言描述与言语加工系列讲座之一


The role of phonetic data in language description, documentation and typological studies


201765日星期一上午8:30-10:00

外国语学院309室学术报告厅


Joyce McDonough

Professor of Linguistics, Brain and Cognitive Sciences, University of Rochester

Language communities are speech communities or sign communities. Speech, or sign, is the primary vehicle for language. information about the speech and phonetic structure of a language is important to systematic documentation of a language, including information about the phonetic structure, its segments, suprasegments and prosody. This information is especially important for understudied languages where knowledge of linguistic structures is otherwise confined to orthographic representations. Phonetic studies serve as a basis for evaluating linguistic analyses of the phonology, prosody, morphology, pragmatics, syntax and can challenge ideas about typology. In this talk I will present several example of baseline phonetic studies and instrumental studies of under resourced languages and the challenges the data brings to theoretical and typological studies.

教授主页http://www.sas.rochester.edu/lin/people/faculty/mcdonough_joyce/index.html



南京理工大学外国学院英品牌设:语言描述与言语加工系列讲座之二


Studying spoken language using eye movements and targeted language games


201765日星期一上午10:15-11:45

外国语学院309室学术报告厅


Michael K. Tanenhaus

Bishop Professor of Brain and Cognitive Sciences, University of Rochester

Chair Professor, School of Psychology, Nanjing Normal University


I will present an overview of my research program that examines spoken language processing using eye movements coupled with standard factorial designs and targeted language games.  I’ll motivate the use of eye movements to study spoken language in the Visual World Paradigm (VWP).  I’ll discuss how the VWP, when combined with standard factorial designs, can be used to study use of fine-grained acoustic information in processing speech and spoken words.  I’ll then discuss how the VWP can be coupled with targeted language games to study real-time processing of unscripted interactive conversation in collaborative tasks.


教授主页http://www.sas.rochester.edu/bcs/people/faculty/tanenhaus_michael/index.html


南京理工大学外国学院英品牌设:语言描述与言语加工系列讲座之三


Evidence for a data explanation approach to speech and

language processing


201766日星期二上午8:45-10:00

外国语学院309室学术报告厅


Michael K. Tanenhaus

Bishop Professor of Brain and Cognitive Sciences, University of Rochester

Chair Professor, School of Psychology, Nanjing Normal University


I will present a series of visual world experiments conducted with Meredith Brown and Laura Dilley that examine how listeners’ real-time beliefs about speaker’s intended reference are constrained by the temporal unfolding of continuous information in ambiguous acoustic signals.   We use the ambiguities associated with phrases such as “saw (a) raccoon(s) swimming”.  By manipulating the speech rate of the surrounding context and the duration of the “a” and “s”, we shift interpretations between “saw a raccoon swimming” and “saw raccoons swimming”.  The results of our experiments suggest that listeners use gradient, temporally distributed information in speech to perform sophisticated, probabilistic inference about referential information in context, and that these inferences are subject to graded, dynamic updating based on subsequent acoustic information. In so doing we present crucial new information in support of emergent data-explanation approaches that have received support in perception and cognition generally and which have recently been extended to speech and language comprehension.


教授主页http://www.sas.rochester.edu/bcs/people/faculty/tanenhaus_michael/index.html


南京理工大学外国学院英品牌设:语言描述与言语加工系列讲座之四


Speech in the auditory system: Neural representations of speech sounds


201766日星期二上午10:15-11:45

外国语学院309室学术报告厅


Joyce McDonough

Professor of Linguistics, Brain and Cognitive Sciences, University of Rochester

Linguistic phonetics and speech research, including speech perception studies, depend heavily on acoustic analyses for the characterization of the properties of speech, to the extent that many researchers consider the acoustic signal to be part of the perception system, and consider by default the substitutability of the acoustic for the auditory signal.  However the acoustic signal, mechanical sound wave, is the output of the production system; it is not an accurate representation of the signals that are passed along through the auditory system. The signal is broken apart as soon as it reaches the structures of the inner ear in the peripheral auditory system. This fact was recognized since by Bjorn Lindblom and colleagues and their work on generalizations about linguistic sound systems in the early 70’s. In this talk, building on work by Lindblom, I will report on the collaboration of neuroscientist Laurel Carney and phonetician Joyce McDonough on the representation of vowels and vowel spaces in the auditory pathway of the midbrain and peripheral auditory system, the goal of which is to develop a realistic neural model of speech processing.

教授主页http://www.sas.rochester.edu/lin/people/faculty/mcdonough_joyce/index.html