서울대 언어연구소 해외 학자 초청 특강. 10. 31. 11. 2.

10월 말과 11월 초에 계획되어 있는 언어연구소의 외국인 저명학자 초청강연을 안내해 드립니다. 이번에 강연을 해주실 분은 언어유형론 및 호주 소수민족언어의 전문가인 Eva Schultze-Berndt 교수(University of Manchester, UK)와 국내에서 활동하고 있는 독일계 언어학자 Gerd Jendraschek 교수(상명대학교) 입니다. 구체적인 강연 일정은 아래와 같으니 참고하시고 많은 관심과 참여를 부탁드립니다. <해외저명학자 초청강연 : Eva Schultze-Berndt 교수(1차)> - 일 시 : 2017. 10. 31.(화) 오후 4시 - 장 소: 신양관(4동) 302호(국제회의실) - 제 목: Whoosh, off I go into another mode: The mimetic linguistic function and the word class of ideophones - 강연 개요 In Language users, when engaged in spontaneous interaction, do not restrict themselves to the kind of authorial description of a state of affairs that is the focus of most (Western) mainstream linguistic research. Rather, they frequently switch into what has been variously termed the “mimetic mode” (Güldemann 2008: 289), “demonstration” (Clark & Gerrig 1990; Clark 2004), “depiction” (Dingemanse 2011: 156) and “(verbalized) dramatization” (Kunene 1978: 3; Kilian-Hatz 2001: 155). The status of conventionalised mimetics (ideophones) as a cross-linguistic category is even more controversial than that of other parts of speech. It has variously been argued, for specific languages, that ideophones are a distinct part of speech, or a phonologically defined, sound-symbolic subset of e.g. adverbs or adjectives (Kulemeka 1995; Ameka 2001; Beck 2008). I propose that mimesis is a major syntactic-pragmatic function in language, on a par with reference, predication and attribution, and that giving it this status will enable us to both analyse more accurately the syntax of spoken language and to deal with the contested word class status of ideophones. ※ 강연 직후에 Manchester 대학교 언어학과에 대한 소개가 있습니다. <해외저명학자 초청강연 : Eva Schultze-Berndt 교수(2차)> - 일 시 : 2017. 11. 2.(목) 오후 3시 - 장 소: 신양관(4동) 302호(국제회의실) - 제 목: Word order, information structure and case marking in transitive clauses in Jaminjung: implications for constituent order typology and alignment typology - 강연 개요 This presentation reports on the patterns of clausal constituent order and information structure in a sample of transitive clauses with both subject (A) and object (O) NPs in Jaminjung/Ngaliwurru, a Mirndi language of Northern Australia. As is to be expected in a head-marking language with no obligatory phrasal expression‎ of arguments, clauses of this type are quite rare in discourse, making up only around 1% of all clauses in naturalistic speech. The dataset consists of all 123 such clauses found in a set of 112 texts totalling over 12,000 intonation units. Overall, the frequency of the six possible constituent orders confirm‎ the assessment of Jaminjung as a free word order language, since no pattern is at least twice as frequent as any other – the criterion used for dominant constituent order by Dryer (1997, 2013a). This study provides evidence for the often-repeated claim of a strong preference for topics to be agents (Givo?n 1983: 22; Du Bois 1985: 356; Comrie 1989 [1981]: 114; Aissen 1999). These claims are rarely underpinned with discourse data, and do not always distinguish between sentence topics and discourse topics. Indeed, the same principle has been invoked to explain the dispreference, in discourse, for overt, lexically expressed agents (e.g. Du Bois 2003; Everett 2009; Haig & Schnell 2016). The Jaminjung data clearly show that the preference for topics to be agents applies to overt sentence topics as well. The preference for agents to be topics is also one of the factors underlying variable agent marking (“optional ergativity”) in Jaminjung, in a pattern which bears a striking resemblance to the factors favouring case ellipsis in the nominative-accusative languages Japanese and Korean (Lee 2009, 2010, 2011): there is a strong tendency for ergative marking to be retained with focal agents (except for 1st and 2nd person pronouns), whereas it can be omitted on topics. The latter is a manifestation of the principle of Economy, since the agent status of topics is predictable. Information structure is shown thus to be a fundamental principle of grammar which not only underlies clause-level constituent order but which also interacts with the marking of grammatical roles. <외국인저명학자 특별강연 : Gerd Jendraschek 교수> - 일 시 : 2017. 11. 2.(목) 오후 4시 30분 - 장 소: 신양관(4동) 302호(국제회의실) - 제 목: Contact syntax in Basque and Romance: Grammatical convergence in genetically unrelated languages - 강연 개요 The Basque language has been surrounded by Romance for about two thousand years. The Latin variety spoken in an area adjacent to the Basque area evolved into Castilian, which after expanding southwards became known as ‘Spanish’. While Spanish is now spoken all over Spain (and, of course, its former colonies), in peripheral areas it coexists with other Romance languages—and Basque. Bilingualism and code-switching are frequent, making convergence likely, even though Basque is genetically unrelated to any other language, and has many grammatical features that differentiate it from Romance, such as lack of gender, verb-final constituent order, case suffixes, complex postpositions, and an overall high index of synthesis. Its typological profile is thus close to that of languages of the agglutinative type, such as Quechua, Turkish or Korean. In my talk, I will present a short Basque text to illustrate the most relevant grammatical constructions of Basque, and compare them with translations of that same text into Spanish and Korean. While Spanish syntax is overwhelmingly head-initial, the Korean corpus is one hundred percent head-final. Basque is located between these two poles, but its typological profile is in fact split, with intraclausal syntax preserving the Korean type, and interclausal relations following the Spanish—or more generally European—type. These findings corroborate the view that smaller constructions are stored as language-specific processing units, whereas larger constructions are less amenable to conventionalization and therefore prone to convergence in bilingual speakers.