Research – Marlyse Baptista


Broadly defined, my research straddles the subfields of Theoretical Syntax and Language Contact, and focuses on speakers’ cognitive (mental) processes in situations of language contact.

More specifically, I study the morphosyntax interface of Pidgin and Creole languages (and their source languages) and examine theories of language contact and language emergence, as they relate to cognition, language creation and change. 

My current work investigates the cognitive processes involved in contact situations and focuses on convergence.  I build on the linguistic evolutionary model of feature competition and selection (Mufwene 2001, 2008) and contact-induced change models (i.e., Thomason, 2001; Matras & Sakel, 2007) to account for the emergence of specific features in a Creole grammar.  Given a multilingual setting where languages A, B, C and D come into contact, my research seeks to answer the following questions:

  1. In a linguistic ecology where attested features from source languages A, B, C and D compete in a feature pool, which features get selected over others to contribute to the emergence of a new language?
  2. What are the linguistic variables and social factors that promote the selection of a feature in the new language over others?
  3. Why and how do speakers converge, diverge and innovate in a contact setting and what are the factors and cognitive processes underlying convergence, divergence and innovation?
  4. What are the grammatical properties of the emerging contact language that are not traceable to any of its source languages?

As such, my current research focuses not only on language convergence, but also on language divergence and innovation (as all three are inextricably connected), examining similarities and differences between the languages in contact and how the overlapping features affect the grammatical make-up of the new language.

I recently published an article where I investigate the role of convergence in 20 Pidgins and Creoles across 19 grammatical domains and developed an algorithm and a model accounting for how convergence may operate (Language, March 2020).

I direct the Cognition,Convergence and Language Emergence (CCLE) Research Group in the Linguistics Department at the University of Michigan.

I am currently involved in several collaborative projects:

  1. A psycholinguistic experiment testing the role of convergence in adult multilingual acquisition, using artificial language learning (with Susan Gelman, Rawan Bonais, Danielle Labotka and Emily Sabo).  In this experiment, we use two artificial languages that vary in their morphological congruence with one another and with English to determine whether overlapping features (form/function mapping) in the languages in contact facilitate acquisition and, by extension, play a role in creole formation.  This project builds on a previous paper using artificial language learning in second language acquisition: “Testing the role of convergence in language acquisition, with implications for creole genesis” (Baptista, Gelman and Beck, 2015, International Journal of Bilingualism 20 (3), 269-296).
  2. A project using field data and DNA samples to reconstruct the ancestry of Cabo Verde founding populations (with geneticists Paul Verdu, Noah Rosenberg, Trevor Pemberton, Ethan Jewett and linguists Sérgio Costa and Valentin Thouzeau).  Between 2010 and 2018, we collected DNA samples and speech data from individuals across all nine inhabited islands.   The first results of this long-term collaboration appeared in Current Biology in 2018.  “Parallel Trajectories of Genetic and Linguistic Admixture in a Genetically Admixed Creole Population.”  Co-authored with Paul Verdu [first author], Ethan Jewett [first author], Trevor Pemberton and Noah Rosenberg, Current Biology, Volume 27 (Issue 16), pp. 2529–2535.
  3. A project developing an agent-based model simulating language emergence, using 18th century Haitian Creole texts (with computational scientists Ken Kollman, Alton Worthington and mathematician Jinho Baik). The linguistic analyses reveal that some of the grammatical features of Haitian Creole grammar were already highly stable in the 18th-century text whereas others were highly unstable. The stability of some features may be due to direct L1 transfer into the developing Creole or to a convergence of features that were present in both sources, whereas the unstable features may be due to contradictory cues the original speakers may have received from the languages in contact. While we focus on Haitian Creole, this model is flexible and can be manipulated to accommodate various assumptions and parameters to fit diverse situations of language contact.
  4. A syntactic paper with Miki Obata [Obata as first author] entitled “Asymmetrical agreement: Evidence from Focus-Agreement in Cabo Verdean Creole.”
lsa logoum logoU-M Privacy StatementAccessibility at U-M