PhD grant in experimental grammar

The LabEx-EFL (Laboratory of Excellence Empirical Foundations in Linguistics) is offering a 3 year PhD grant in “Experimental grammar from a crosslinguistic perspective”, full time, about net 1700 euros/month, starting September 2023 at the University of Paris (Ecole doctorale Sciences du langage). The candidate must have a Master degree by September 2023, with a specialization in linguistics, computational linguistics or psycholinguistics.

He or she will be affiliated to the Laboratoire de Linguistique Formelle. Funds will be available for travelling expenses, equipment and experiments. He or she will be part of the Doctoral School Sciences du Langage, and of Paris Graduate school of Linguistics, and will attend doctoral seminars and may be able to teach courses in the Department of Linguistics.

Application

The application should be sent by June 1st 2023 (midnight MET) to Anne Abeillé and tavnik@gmail.com

It should comprise:

  • a CV with transcripts (Master) and diplomas
  • a motivation letter with explicit mention of the chosen workpackage and a short tentative project (max 2 pages)
  • the names and contact of two referees for reference letters

The candidate should contact their potential adviser.

The candidates selected for interviews will send their Master thesis or other written work supporting their qualification for the project.

They will be interviewed (remotely) early July 2023.

Projects/workpackages

Workpackage Lexeme

dir: Olivier Bonami codir Lucie Barque & Marie Candito

Profile: The thesis will investigate the relationship between formal similarity and lexical semantic similarity, using a combination of experimental and computational methods. Inspired by a growing body of research on colexification (Xu et al. 2020, Floyd et al. 2021), we will focus on pairs of concepts lexicalized by the same lexeme or two morphologically related lexemes, in order to assess the effect of formal similarity on meaning proximity. Of particular interest are 1) formally distinct vs unique word formation processes (e.g 'agent' and 'instrument' lexicalized by two distinct suffixes, like -ist in pianist and -or in elevator vs one, like -er in writer and hammer) and 2) word formation vs polysemy (e.g ‘container’ and ‘quantity’ lexicalized by suffixation -ful, or by a metonymic pattern, silver spoon/spoon of sugar). Empirical investigation will mainly concern English and French but experiments will also be conducted on more typologically diverse languages, drawing on recently developed multilingual resources (Vulić et al 2020).

References

  • Vulić, I., Baker, S., Ponti, E. M., Petti, U., Leviant, I., Wing, K., ... & Korhonen, A. (2020). Multi-simlex: A large-scale evaluation of multilingual and crosslingual lexical semantic similarity. Computational Linguistics, 46(4), 847-897.
  • Floyd, S., Dalawella, K., Goldberg, A., Lew-Williams, C., & Griffiths, T. (2021). Modeling rules and similarity in colexification. In Proceedings of the annual meeting of the cognitive science society (Vol. 43, No. 43).
  • Xu, Y., Duong, K., Malt, B. C., Jiang, S., & Srinivasan, M. (2020). Conceptual relations predict colexification across languages. Cognition, 201, 104280.

Workpackage Relative clauses and related constructions

dir: Anne Abeillé codir Barbara Hemforth

Profile: The thesis will investigate locality constraints on long distance dependencies in a crosslinguistic perspective, using experimental techniques and annotated corpora. It will test two hypotheses: that most locality constraints do not yield unacceptability; that they are sensitive to the kind of long-distance dependency, relative clauses and topicalization being more permissive than others.

References

  • Abeillé, A., Hemforth, B., Winckel, E., and Gibson, E. 2020. Extraction from subjects: Differences in acceptability depend on the discourse function of the construction. Cognition.
  • Chaves R, Putnam M. 2020, Unbounded dependencies, Oxford University Press.
  • Sprouse, J., Caponigro, I., Greco, C. & Cechetto, C. 2016. Experimental syntax and the variation of island effects in English and Italian. Natural Language and Linguistic Theory 34(1), 307-344.

Workpackage Pluralities, worlds and events (PLU)

dir: Lucia M. Tovena

Profile: The thesis will investigate issues at the interfaces of semantics with pragmatics and syntax. The research will be based on data from corpora and/or experiments and/or typological language comparisons.
Possible research questions:
1- Superlatives and definiteness (Romance languages do not have a dedicated morphological counterpart of the English -est (Heim 1999); instead, they convey superlative meanings by overtly using comparative markers associated with definiteness. Their seemingly morphological uniformity corresponds to quite different syntactic configurations (Loccioni 2018, Dobrovie-Sorin and Tovena 2022), in ordinary and modal superlatives. Semantically, different comparison classes seem to be needed.)
2- Argument structure modifications and verbal voice (Middle(-passive) voice markers are partly associated with valency change, and lexical semantics alterations (Inglese 2021). Crucial issues concern the external argument, whether the primary meaning appears to be one of constraining the agentivity of the subject, anticausativity, affectedness or heightened degree of autonomy.)
3- The thesis can focus on other themes including negation and the left periphery.

References

  • Heim, I. (1999). Notes on superlatives. unpublished ms. MIT.
  • Inglese, G. (2021). Towards a typology of middle voice systems. Linguistic Typology, 26(3), 489-531.
  • Loccioni, N. (2018). Getting ‘the most’ out of Romance. Ph. D. thesis, University of California Los Angeles USA

Experimental Pragmatics (XPrag)

dir: Ira Noveck

Profile: The thesis mines insights from the Experimental Pragmatic literature in order to test hypotheses about the way participants arrive at exact readings on classic number cognition tasks. We systematically compare ideographic representations (i.e. Hindu-Arabic numerical expressions) to glottographic ones (i.e. written — language dependent — expressions, such as onze) and across languages. Following up on Spychalska et al., (2019), who extended scalar implicature studies to number, Noveck et al. (2022) made two hypotheses: A) That glottographicnumbers, such as eleven, are understood with a lower-bounded (at least eleven) meaning which could then be pragmatically enriched at its upper bound to mean eleven but no more, i.e. exactly eleven; B) That ideographic representations of numbers convey exactly readings directly.

References

  • Noveck, I., Foegel, M., Von Voorhees, K., Turco, G. (2022). When eleven does not equal 11: Investigating exactness at a number’s upper bound. PLoS One. https://doi.org/10.1371/journal.pone.0266920
  • Spychalska M, Kontinen J, Noveck I, Reimer L, Werning M. (2019) When numbers are not exact: Ambiguity and prediction in the processing of sentences with bare numerals. JEP :LM&C, 45(7): 1177.
  • Van Tiel, Bob, Van Miltenburg, E. Zevakhina, N. & Geurts, B. Scalar diversity. Journal of Semantics 33, no. 1 (2016):137-175.

DIA : Dialog

dir: Jonathan Ginzburg

Profile: Language Acquisition: What comes after the one word stage?

Taking formal grammars for the one word stage developed by Moradlou and Ginzburg, the thesis (building on an extensive corpus study in M2) considers how such grammars evolve into the 2 and 3 word stage using a probabilistic semantics approach (see paper above by Cooper, Ginzburg and Larsson) and HPSG grammars for the grammar writing. It will develop a more grammar oriented developmental notion than MLU which is shown to be too coarse grained; it will develop computational methods (in partnership with Timothée Bernard) for classifying early utterances to scale up the corpus work to many subjects.

References

  • Robin Cooper, Jonathan Ginzburg, Staffan Larsson (2023):  `Learning Language Games Probabilistically: From Crying to Compositionality'  J-P. Bernardy, R. Blanck, S. Chatzikyriakidis, S. Lappin, A. Maskharashvili  (eds.) Probabilistic Approaches to Linguistic Theory, CSLI Publications, Stanford.
  • Sara Moradlou, Xiaobei Zheng, Ye Tian, and Jonathan Ginzburg, 2020 `Wh-Questions are understood before polar-questions: Evidence from English, German, and Chinese'.  Journal of Child Language  28:1-27.