Agreement, assignment, and attraction in multiple antecedent contexts: A comparison of Latin and Ancient Greek
Time: 12:15pm - 01:30pm
Place: Seminar Room (Memorial Hall 317A)
Cindy Johnson (Ohio State University)The phenomenon of multiple antecedent agreement, where more than one noun controls agreement on a target, provides surprising and interesting data for existing theories of agreement, though, as Corbett (1991:261) points out, these data are often overlooked. In contrast with single antecedent contexts, which typically involve an exact match of the features of the controller and the target, multiple antecedent contexts always produce an agreement mismatch: the target’s features do not match those of at least one of the controllers. In Latin and Ancient Greek (among other languages), there are two main agreement strategies for multiple antecedent contexts: Resolution and Partial Agreement (cf. Corbett 1991, 2006; Wechsler & Zlatic 2003). Resolution is agreement with the collective noun phrase, where number and gender are assigned typically on a semantic basis (e.g. with reference to the animacy of the controllers). Partial Agreement occurs when the target assumes the features of only one of the antecedents; in most cases, the closest antecedent provides the features (“nearest antecedent agreement”), such that cases of Partial Agreement resemble what are typically referred to as attraction errors in languages like English (cf. Bock & Miller 1991). Corbett (1991, 2006) explains the distribution of the two strategies in terms of semantic and syntactic agreement: Resolution is semantic agreement, and Partial Agreement is syntactic agreement; their distribution can therefore by modeled by the Agreement Hierarchy (Corbett 1979) and the Predicate Hierarchy (Comrie 1975). In this presentation, I address whether such a distribution is actually observed in Latin and Ancient Greek and whether there are any notable differences between these two languages on the basis of two corpus studies in Latin and Ancient Greek. I also investigate whether the presence and influence of an additional agreement rule in Ancient Greek, which states that neuter plural subjects regularly agree with singular verbs, affects the application or output of the two existing agreement strategies. I claim that the types of strategies, their distribution, and the similarities and differences between Latin and Ancient Greek can be explained by appealing to more general principles within the language: gender assignment and avoidance (a term coined by Hock 2007 that encompasses several non-Resolution agreement strategies).
In category: Morphosyntax and semantics
Morphological structure in syntactic agreement: What we can learn from structural attraction 'errors'
Time: 02:30pm - 04:00pm
Place: Education 1230
Andrea Sims (Ohio State University)Structural attraction is when a target agrees with a local noun, rather than the proper but more distant controller. Attraction effects are rare in case-marking languages, but the likelihood of attraction increases in experimental conditions when a local noun is syncretic (homonymous) with nominative (Hartsuiker et al. 2001, Nicol and Wilson 2000). This paper explores the syncretism effect, based primarily on Croatian data (naturally-occurring examples and a forced production task). Specifically, it is unclear whether the syncretism effect, and structural attraction more generally, stems from shallow processing of the local noun's form or deeper processing of its inflectional feature structure. It has been taken as indicative of shallow processing, but other facts about attraction have led to the opposite conclusion (see, e.g., Bock & Miller 1991 for greater attraction to plural). I argue that neither assumption is well grounded, because both fail to consider the internal structuring of inflectional systems. There are indications that attraction in Croatian (and similar languages) is sensitive to the degree of systematicity with which two forms are syncretic. This suggests that structural attraction involves deep processing of feature structure, but differently than has been assumed -- processing of morphological structure, and not only syntactic features. Finally, structural attraction is widely considered to be an agreement 'error' that arises during syntactic processing, but it is strikingly parallel to some patterns of agreement that are deemed to be 'grammatically correct', some of which also exhibit a syncretism effect. I thus speculate about the relevance of processing effects for our understanding of morphosyntactic agreement more broadly.
Machine translation for under-resourced languages
Time: 12:00pm - 01:00pm
Place: Memorial Hall 401
Alex RudnickI'll discuss a few half-baked ideas for MT for under-resourced languages that Mike Gasser and I have been considering. First, I'll talk about cross-language word sense disambiguation, and our ideas for how to make lexical choice decisions jointly, and perhaps how to make lexical choice decisions jointly with morphological and syntactic disambiguation decisions. Second, I'll talk about our ideas for building (and using) translation memories with a system for children in schools to translate documents into their native language. Your input and pointed questions would be much appreciated; both of these projects are just getting started.
In category: Computational linguistics
Thirteenth Annual Conference of the Central Eurasian Studies Society
Time: 08:00am - 09:45am
Place: Indiana Memorial Union
The vernacular or the literary? Rural/urban divisions and the politics of Kazakh language purityGabriel McGuire (Nazarbayev University)IMU Walnut Roomhttp://centraleurasia.org/annual-conference-program/
In category: Sociolinguistics and pragmatics
Time: 10:00am - 11:45am
Institutionalizing educational equality: Bilingual education in the Tibet Autonomous RegionEveline Yang (Indiana University)IMU Redbud RoomProgress in 'bilingual' education in Xinjiang: A stock-taking and analysis of the present situationMuhaemmaetrehim Sayit (Minzu University)IMU Redbud Roomhttp://centraleurasia.org/annual-conference-program/
Time: 01:45pm - 03:30pm
Uyghur words, Chinese style: The impact of translation language and education on Modern Written UyghurNathan Montgomery (Indiana University)IMU Sassafrass RoomDevelopments in ethnic language publishing in XinjiangArslan Abdulla (Xinjiang University)IMU Sassafrass RoomHow much contact can change a language? The cases of Kyrgyz, Khalka, and UyghurJonathan Washington, Niko Kontovas, and Andrew Shimunek (Indiana University)IMU State Room Easthttp://centraleurasia.org/annual-conference-program/
The economic foundations of post-Soviet language reform: Central Asia in comparative perspectiveChao-yo Cheng (Academia Sinica)IMU State Room EastSocial transition of Mongolia and lexical evolution and change of the Mongolian languageOchirbat Sambuudorj (Mongolian Academy of Sciences)IMU Dogwood RoomThe features of advertising in Mongolian and English languagesShiirevdorj Serchmaa (University of California)IMU Dogwood Roomhttp://centraleurasia.org/annual-conference-program/
Code-switching in Georgian cinema: Diglossia, ambilingualism, contrapuntal sound, and political discourse Julie Christensen (George Mason University)IMU Walnut RoomEthnolinguistic identity and religion: The case of the Tat and Mountain JewsJohn Clifton (SIL International)IMU State Room Easthttp://centraleurasia.org/annual-conference-program/
JEvents v3.0.9 Stable Copyright © 2006-2013
Copyright © 2008 The Trustees of Indiana University | Copyright Complaints