LiLa at UiB

Genuskongruens og transfer i norsk innlærerspråk

Fredag 10. januar, 2014, 14:15 – 16:00 (HF-bygget 217)

Her presenteres resultater fra pilotstudien til mitt pågående ph.d.-prosjekt om genuskongruens og morsmålspåvirkning (transfer) i norsk innlærerspråk. For å avdekke transfer, sammenligner jeg språkbruken til innlærere med morsmål (S1) som er ulike når det gjelder grammatisk genus og kongruens. Hovedhypotesen er at innlærere som har et S1 uten kongruens (vietnamesisk) vil ha større problemer med genuskongruens enn innlærere som haret S1 med kongruens (tysk, spansk og engelsk). Denne hypotesen springer ut fra en tidligere et S1 med kongruens (tysk, spansk og engelsk). Denne hypotesen springer ut fra en tidligere studie om genustildeling og transfer, der jeg fant at vietnamesiske innlærere hadde signifikant mer korrekt genustildeling enn både tyske, nederlandske, spanske og engelske innlærere. Man kan derfor spørre seg om Sabourin m.fl. har rett når de sier at:

It is possible that gender assignment can be done based on more general cognitive skills (simply learning or memorizing what gender goes with what item) and thus can, with enough experience, be learned by any L2 learner. In contrast, gender agreement, relying on more linguistic strategies, can only be learned in the L2 if the same strategies are present in the L1. (Sabourin, Stowe, and Haan 2006: 27)

Undersøkelsen min er korpusbasert. I ASK – Norsk andrespråkskorpus undersøkte jeg genuskongruensen i 100 tekster fra hver S1-gruppe. Pilotstudien var avgrenset til undersøkelse genuskongruens i én type substantivfrase: ubestemt artikkel + adjektiv + substantiv som i eksempel (1) og (2):

(1) en.MASK fin.MASK bil(MASK)

(2) et.NØYT fint.NØYT hus(NØYT)

Kongruensen i eksempel (2) kan i prinsippet manifestere seg på fire ulike måter, (A), (B), (C) og (D), i innlærerspråket:

(A) et.NØYT fint.NØYT hus(NØYT)

(B) et.NØYT fin.MASK hus(NØYT)

(D) en.MASK fin.MASK hus(NØYT)

Det første spørsmålet er om innlærerne har kongruens mellom den ubestemte artikkelen og adjektivet, som i (A) og (D), eller om de har ”inkongruens”, som i (B) og (C). Det andre spørsmålet er om det er forskjell mellom morsmålsgruppene. Signifikanstesting viser at det ikke er det.

Sabourin, Laura, Laurie A. Stowe, and Ger J. de Haan. 2006. “Transfer effects in learning a second language grammatical gender.” Second Language Research no. 22 (1):1-29

Silje Ragnhildstveit, stipendiat, Institutt for lingvistiske, litterære og estetiske studier (Forskergruppen ASKeladden), UiB.

A psycholinguistically plausible alternative to the standard pipeline in Natural Language Processing

Monday December 16, 2013 at 12:15 – 13:00 (HF – bygget: 301)

Research in Natural Language Processing has assumed that high level tasks such as Natural
Language Understanding can be solved by special purpose processes operating in a pipeline,
called the standard pipeline. Work is then done on building the individual components. However,
there appears to be no way of guaranteeing that such components will actually solve high
level NLP tasks when combined as suggested by the standard pipeline.

In my alternative model, syntax- and meaning representations are incrementally constructed
on a word-by-word basis. Processing is serial in that the model commits to a unique syntactic
and semantic increment for each word. Incrementally constructed structures are stored in states
that both result from- and influence processing. Meaning representations do not influence further
syntactic increments, although meaning representations can cause the process to throw an
error. I propose a reanalysis algorithm by which the process can recover from such errors.

Magnus Bakken, Master student at LLE, presents his MA thesis.

The Old Norse Gerundive

Friday December 6, 2013 at 14:15 – 16:00 (HF: 372)

“… þá er þat skýrandi.”

Especially in literature translated from Latin, Old Norse employs a form similar to the present participle in order to translate the Latin gerundive (Ekki er kristnum manni meir flýjanda en ofmetnaðr from Nihil magis Christiano vitandum est quam superbia – “Nothing should be avoided by a Christian more than pride”). While this gerundive is morphologically identical to the present participle, the semantics of the two differ drastically: the latter is active and factual while the gerundive is passive and deontic, i.e. it does not say what something is doing, but what must or can be done to something. In a typical construction the gerundive is used as a predicative with the copula (e.g. sjá kynning […] er nemandi – “this knowledge must be acquired”).

The syntax and inflectional behavior of this gerundive construction is not always as straight forward as in this case, and different constructions occur. They cannot always be easily told apart and appear to be influenced by other deontic, passive and impersonal constructions. I will present those different constructions and try to shed some light on their usage and distribution.

Robert K. Paulsen, PhD scholar in Old Norse (norrøn filologi) at LLE. He has his master’s degree from the University of Freiburg, and took up his PhD position at the University of Bergen 1 August 2013.

The elusive agent. A corpus-based study of Spanish passive and impersonal constructions.

Friday November 29, 2013 at 14:15 – 16:00 (HF: 217)

This project is a postdoctoral project for which I am currently applying for funding. It aims to reevaluate the passive/active status of the Spanish constructions “pasiva refleja” (PR), the periphrastic passive (PERP) and the “se impersonal” (IMP), using the LFG-framework and corpus linguistics, to ascertain the consequences of using linguistic constructions which conceal the agent when sensitive topics such as poverty are broached. No corpus-based study exists for all three structures and a suitable corpus will be collected by the NFR-funded POLAME project,(a comparative case study of how agenda-setting media in Argentina, Brazil, Colombia and Mexico convey and construct the notions of poverty). LFG is a well-attested suitable tool for explicating impersonals and the passive/active distinction in any language, and the leaf nodes of its constituent structures are based only on what is directly observable in an utterance, so it is well suited for corpus-based studies. The evaluation of the three constructions as passive or not will involve an examination of the inclusion of certain semantic properties for the agent-role in the argument-structure of the LFG-representation. It will also involve a revision of LFG’s Subject Condition for Spanish. A hybrid structure between the PR and the IMP, which has not previously been properly accounted for, is expected to be found in the Argentinian corpus, and will also be scrutinized. Finally, a comparison (frequency, semantics) will be made of the three structures in corpora treating poverty with those in corpora treating non-sensitive subjects. My hypothesis is that the use of passive and impersonal constructions to speak about poverty has a tendency to “depersonalize” it and to make it easier to evade uncomfortable notions such as culpability and cause.

Margrete Dyvik Cardona, Lecturer, LLE, University of Bergen.

Sharing research data and results

Gyri Smørdal Losnegaard (LLE) held vitskapsteoretisk innlegg tysdag 19. november kl. 10:15 i seminarrom 301 i HF-bygget.

Victory for Google Books and n-grams

Judge Chin ruled that Google’s scanning of more than 20 million books for an electronic database, and making materials available for online searches constituted fair use. Also the use of Google Books for linguistic research on ngrams played a role in the decision.

Live streams from #TromSlang 6-8 November

The Tromsø International Conference on Language Diversity will be opened Wednesday 6 November at 09:30 by President of the Norwegian Saami Parliament, Aili Keskital. This session as well as most of the other plenary sessions at the conference will be live streamed over the internet.

Please consult the conference web site for further information about the conference and the various sessions