카테고리 글 목록: 학술행사

2018년 제5차 언어학 콜로퀴엄(권수현)

2018년 제5차 언어학 콜로퀴엄

- 일시: 2018. 10. 29.(월) 오후 5시
- 장소: 신양인문관 국제회의실(4동 302호)
- 발표자: 권수현(서울대학교)
- 제목: Phonetic and phonological changes of Noam Chomsky: A case study of language change across lifespan and dialect shift
- 초록: In this talk, I present acoustic evidence of diachronic accent change of an adult speaker who relocated to a region characterized by dialect features different from those of his home region. Noam Chomsky was chosen as the subject because he exemplifies the effects of relocation on adult phonology. Chomsky was born and raised in Philadelphia, Pennsylvania, but at his age 27, he moved to Boston, Massachusetts. Three linguistic variables that have different systems in Philadelphia and Boston were examined in this study: (1) COT-CAUGHT distinction in Philadelphia and COT-CAUGHT merger in Boston, (2) the split short-a system in Philadelphia and the nasal system in Boston, and (3) strong fronting of /o/ in Philadelphia and moderate fronting in Boston. Acoustic analyses of Chomsky’s public speech in 1970 and 2009 show that his COT has significantly shifted along both F1 and F2 over the 40 years, while CAUGHT remained stable. Despite the substantial shift of COT, this change has not led to a merger of two vowel categories. Meanwhile, his short-a pattern has gone through a striking phonemic reorganization over the 40-year period; Chomsky exhibited the unsplit pattern in 1970 but by 2009 he appears to have adopted a (somewhat incomplete) nasal system. With regard to /o/, he did not show any significant change, exhibiting only moderate fronting both in 1970 and 2009. This study demonstrates remarkable intraspeaker changes over a lifetime, although the changes did not reach categoricity. Possible social and phonetic factors that play an important role in these changes are discussed.

2018년 제4차 언어학 콜로퀴엄(Sam Bowman)

2018년 제4차 언어학 콜로퀴엄

- 일시: 2018. 9. 10.(월) 오후 5시
- 장소: 신양인문관 국제회의실(4동 302호)
- 발표자: Sam Bowman (New York Univ.)
- 제목: Evaluating the Semantic and Syntactic Abilities of Neural Network Models
- 초록: Artificial neural network models for language understanding problems represent an increasingly large and increasingly successful thread of research within natural language processing. When developing these models in typical settings, though, it can be difficult to identify the degree to which they capture the structures or meanings of natural language sentences, and correspondingly difficult to identify research directions that are likely to yield progress on the underlying language understanding problem.
In this talk, I first introduce natural language inference, the task of judging whether one sentence is true or false given that some other sentence is true, and argue that that task is distinctly effective as a means of developing and evaluating sentence understanding models in NLP. I’ll cover the recent datasets SNLI and MultiNLI. I’ll then turn to syntax, and present our work on the Corpus of Linguistic Acceptability (CoLA), an NLP challenge dataset based on the classic linguistic task of acceptability judgment.

2018년 제1차 언어학 콜로퀴엄(James Whang)

2018년 제1차 언어학 콜로퀴엄

발표자: James Whang(MARCS Institute for Brain, Behaviour & Development)
일   시: 2018. 5. 16.(수) 오후 4시
장   소: 신양인문관 국제회의실(4-302)
제  목: Effects of phonotactic predictability on the production and perception of high vowels in Japanese
초   록: 첨부파일 참고

2017년 제7차 언어학 콜로퀴엄(강윤정)

2017년 제7차 언어학 콜로퀴엄

발표자: 강윤정, 윤수연 (Univ. of Toronto)
일   시: 2017. 11. 10.(금) 오후 4시
장   소: 신양인문관 국제회의실(4-302)
제   목: Phonetic drift in the speech of North Koreans in Seoul: time of arrival and word-specific effects
초   록: In this talk, we report on the results of our study on the speech of Northern Hamkyeoung speakers residing in Seoul. The goal of the study is to examine to what extent adult North Korean speakers’ speech changes in their subtle phonetic characteristics due to the exposure to Seoul Korean and if such phonetic drift, if any, is modulated by word-specific effects, namely, whether words that are typically used in North vs. South pattern differently. These effects are predicted under the view that mental representations of speech sounds include word-specific phonetic details linked to episodic memories, which are constantly updated with exposure to new speech experience.

Production data are collected from 58 Hamkyeoung speakers who are currently residing in Seoul, balanced for the time of arrival (recent: 3 years) and age (young: 40yrs). Comparison data are also collected from 20 Seoul speakers. This talk will report on the analysis of two phonetic variables that are known to have different realizations in North vs. Seoul: (i) the degree of aspiration of stops (VOT) and (ii) the place of articulation of lenis affricates (COG).

2017년 제6차 언어학 콜로퀴엄(강초롱)

2017년 제6차 언어학 콜로퀴엄

발표자: 강초롱(University of Southern California)
일   시: 2017. 10. 23.(월) 오후 5시
장   소: 신양인문관 국제회의실(4-302)
제   목: Selection of copy for pronunciation
초   록: 첨부파일 참고

2017년 제5차 언어학 콜로퀴엄(송지은)

2017년 제5차 언어학 콜로퀴엄

발표자: 송지은(UCL)
일   시: 2017. 9. 4.(월) 오후 5시
장   소: 신양인문관 309호(4-309)
제   목: Second language speech processing in realistic environments: An electrophysiological investigation
초   록: 첨부파일 참고

2017년 해외저명학자 특강(Elizabeth Hume)

2017년 해외저명학자 초청

발표자: Elizabeth Hume (Canterbury Univ.)
일   시: 2017. 7. 5.(수) 오후 4시
장   소: 신양인문관 309호(4-309)
제   목: Word Predictability Effects: Implications for Phonology
초   록: Word predictability has been shown to condition a broad range of phonological processes including segment deletion, reduction, epenthesis, nasal place assimilation and liaison. In English, for example, a word-final coronal nasal is more likely to undergo place assimilation if the word it occurs in is predictable from the following word (Turnbull et al., in prep). Such evidence has serious implications for phonological theory. One issue relates to the probabilistic nature of the conditioning factor and the fact that predictability is calculated by taking into account words that are related to the target word paradigmatically (e.g. word competitors, minimal pairs) as well as syntagmatically (i.e. the word preceding and/or following the target word). A further issue relates to the observation that one can predict the type of phonological process based on whether word predictability is high or low. One can also predict where in the word the change is likely to occur. These predictions do not, to my knowledge, follow from traditional approaches to phonology. However, drawing on Hall, Hume, Jaeger & Wedel (2016), I outline a new approach which not only predicts these patterns, but is also grounded in well-established principles about how communication systems work (Shannon 1948).

2017년 제4차 언어학 콜로퀴엄(윤지원)

2017년 제4차 언어학 콜로퀴엄

발표자: 윤지원(Stony Brook University)
일    시: 2017. 6. 12.(월) 오후 5시
장    소: 신양인문관 국제회의실(4-302)
제    목: Illusion of grammaticality in NPI processing
초    록: A growing number of studies report illusive licensing effects of negative polarity items (NPIs) in online processing, which provides important insights on the linguistic properties of NPIs as well as the human language processing mechanisms. However, most of the previous studies have been limited to certain languages (i.e., English and German). Moreover, the potential influence of prosody on NPI illusions has not yet been investigated. This study explores the case of Korean to see whether it exhibits NPI licensing illusions, and if so, whether the licensing pattern can be affected by prosody. The experimental results suggest the following: (i) Illusive licensing of NPIs does occur in Korean. (ii) The effect can manifest even in offline judgments. (iii) Adding prosodic information can help avoid NPI licensing illusions.