Marlyse Baptista, Uriel Weinreich Collegiate Professor of Linguistics, will give a talk on "Convergence, Divergence and Innovation in Language Contact: A View from Creole Genesis."

ABSTRACT
From the early years of Contact Linguistics (Schuchardt, 1882), linguists have noted that when two or more languages come into contact, whether it is in the context of L2 acquisition (Ellis & Sagarra, 2011;Tolentino, L. C., & N. Tokowicz, 2014), bilingualism (Silva-Corvalán, 1994; Toribio, 2004), trilingualism (Rothman, 2010, 2015; Rothman & Cabrelli Amaro, 2010; Rothman et al., 2015) or multilingualism leading to language creation (Rougé, 1986; Kihm, 1990; Corne, 1999), it is often (but not always!) the case that the features that the languages in contact have in common promote acquisition or language creation. More precisely, the phonemes, morphemes, lexemes or syntactic structures that speakers perceive as being similar in the languages in contact, what we will call here, congruent features or domains, are likely to be acquired more easily in L2 (or L3/L4...) or are more likely to contribute to the grammatical make-up (and lexicon) of the emerging language in the case of creole genesis.

This paper represents a first step in a long-term research program exploring how new languages emerge in a multilingual setting. It examines the role of convergence in Creole formation and development, using a competition and selection framework. Specifically, it illustrates how morphosyntactic and semantic features are more likely to be selected into the grammatical makeup of a given Creole when they preexist and are shared by some of the source languages present in its linguistic ecology. This is empirically supported in this paper by numerous case studies and a survey of congruent features in 20 contact languages across 19 grammatical and lexical domains. In order to show how convergence operates, I propose an algorithm and a model of matter and pattern mapping, adapted to the multilingual setting in which Creole languages emerge. In addition to a set of variables, the model includes both the linguistic ecology (linguistic factors) and speakers' attitudes (non-linguistic factors) (Thomason, 2001) to predict (in a non-deterministic fashion) the features that are more likely to win within a competition and selection framework (Mufwene, 2001). It shows that even when a given feature is traceable to two or more sources, it readily diverges from the original sources and is innovative. The paper also explores cases where convergence does not take place and examines the conditions underlying such outcome.