Generative theories of diachronic syntax have always acknowledged the role of acquisition in language change (Lightfoot 1979, 1991, Roberts 1993, see for recent discussion Cournane 2019 and the commentary papers). One recent trend in diachronic syntax is the use of quantitatively precise models of language acquisition to explain variation and change (such as the variational learning model (Yang 2000), applied in Heycock & Wallenberg 2013 and the Tolerance Principle (Yang 2016), applied in Kodner 2020, 2022, Trips & Rainford 2022, among others).
In addition to the general sessions, one session will be dedicated to this new line of research. We encourage submissions that in some way combine quantitative or computational cognitive modeling with historical corpora or investigate questions of syntactic variation and change with these methods.
Invited speaker: Jordan Kodner, Stonybrook University
Who is responsible for language change? Researchers working in diachronic theoretical linguistics have typically implicated children who acquire different grammars from their caregivers. On the other hand, sociolinguists have marshaled evidence that young adults, not children, lead language change. I argue that this conflict is for the most part illusory. Child language acquisition is the primary locus of innovation and actuation of categorical changes to the grammar. But the observation that young adults "lead" change is a comment on population-level propagation, not on innovation (cf. Milroy & Milroy, 1985). Combining insights from competing grammar accounts (Kroch, 1994), the sociolinguistics of peer-oriented early childhood interaction (e.g., Roberts & Labov, 1995; Loukatou et al., 2021), and experimental studies of learning in the face of variable input (e.g., Hudson Kam and Newport, 2005; Austin et al., 2022), I provide insights into the relationship between language acquisition, actuation, and propagation of change, answering how and why some innovations may progress through actuation and gain a foothold in a population while others may not. This in turn provides a means for distinguishing instances of child-driven from adult-driven change in cases where direct observation is no longer possible.
Among the processes of language acquisition, generalization learning subsumes a diverse range of diachronic phenomena in phonology, morphology, and syntax can be subsumed under the process of generalization learning during child language acquisition. These include, among others, a secondary split in 20th century Menominee (Richter, 2021), the analogical extension of minority inflectional patterns at the expense of statistically predominant patterns in Late Latin past participles (Kodner, 2022), `Dative Sickness' ongoing in Icelandic morphosyntax (Nowenstein, 2021), and argument structure change for psych-verbs (Trips & Rainsford, 2022) in Middle English. This has broad implications for how we conceptualize language change: an ontology of effects in language change will not line up with an ontology of processes. An approach to the study of change which clearly delimits the individual- and population-level components of change and focuses on processes or mechanisms (including but not limited to generalization learning) rather than outcomes and effects stands to bring clarity to a confusing tangle of descriptive phenomena. It reconceptualizes the problem space in a way that cross-cuts and reduces traditional taxonomies of effects (analogical leveling, extension, phonemicization, secondary splits, grammaticalization, bleaching, etc.), unites the strengths of two subfields that have often been cast in conflict, and opens the door for new insights into when, why, and how language change occurs.
Interested in submitting to the session? Find our call for papers and submission link here!