Agenda
Enoch Aboh
Recombination, feature pool, population structure: Three factors underlying ‘grammaticalization’
In this talk, I propose that traditional theories of grammaticalization are misleading. Building on Chomsky (2005), I argue that the process involves three interrelated factors:
The first, recombination, is an innate human cognitive capacity which allows speaker/signer-learners (learners) to select specific linguistic features and recombine them into new syntactic variants.
The second involves the feature pool of such variants that learners are exposed to through contact, and which is subject to a process of competition and selection (Mufwene 2001).
The third, (commonly referred to as grammaticalization) relates to population factors (e.g., learner profile) which may favour or hinder the spread of specific variants across a speech community.
In this view, based on universal multilingualism and contact as cornerstones of acquisition and change (Aboh 2015, 2020), classical analyses of grammaticalization (e.g., from Latin lexical items to Romance conjugation affixes) are particularly misleading because they mix levels of analysis between I-language (whose extensions are idiolects) and E- language (a population behaviour). Such views of grammaticalization aggregate different populations of different learner profiles as if they represented homogeneous monolingual/monomodal communities living in the same ecologies (cf. Frasson 2022 for Romance heritage varieties).
In the proposed approach, language change is always the result of contact, whether at the level of the individual learner who recreates the grammar s/he is acquiring or at the population level. Commonly used notions such as language-internal vs. contact-induced change now become obsolete because they conceive of contact as the exceptional case.