LLS Colloquium: Phonological and syntactic approaches to tense omission in child English

On Wednesday 17th May 2023, Robyn Orfitelli (University of Sheffield) will be presenting on "phonological and syntactic approaches to tense omission in child English".

Talk: Phonological and syntactic approaches to tense omission in child English

One of the most well studied areas of child language acquisition is the telegraphic stage, during which children acquiring a variety of language, including but not limited to Danish (Hamann and Plunkett 1998), Dutch (Weverink 1989; Haegeman 1995), English (Orfitelli and Hyams 2008, and references therein), and French (Pierce 1989, 1992; Rasetti 2003) optionally omit certain grammatical morphemes including tense, determiners, plural marking, and the subject of the sentence.
The omissions that children make do not appear to be random: there are noted impacts of sentence length (Bloom 1970; Valian 1990; Valian, Hoeffner, and Aubry 1996) and prosody (Gerken 1991), although these predictive factors generalise less well outside of English (e.g. Hamann, Rizzi, and Frauenfelder 1996 for French). A predictive factor that does generalise cross-linguistically is grammatical contingencies: for example, children omit subjects in sentences more often in sentences which are also missing tense. Based on these grammatical contingencies, Rizzi (1993/4; 2000, 2005) proposed the ‘truncation’ hypothesis: until approximately 3;6, children optionally permit their syntactic representations to be truncated lower than the CP/left periphery. The optionality is key: children within the telegraphic stage may construct complete representations, and if they do, for example in wh-questions, they are predicted to not omit other material.
The theory of truncation gained a great deal of traction within generative linguistics, but one piece of data was noted to be especially tricky for the account: in English, children omit tense in wh-questions at a high rate (Roeper and Rohrbacher 1994, Bromberg and Wexler 1995). This anomaly led Song, Sundara, and Demuth 2009) to suggest that tense omission in English has a phonological cause (consonant cluster reduction), rather than a syntactic one (truncation).
This talk presents data from a novel large scale corpus study of English children in the telegraphic stage, which re-examines the predictions of phonological and syntactic approaches to children’s production of wh-questions. I conclude that the best way to satisfactorily capture the totality of data is a combination of the two causes: bare verbs in child wh-questions are caused by consonant cluster reduction, but bare verbs in other grammatical contests may also be caused by truncation of the sentence. I will conclude by linking this data to a larger model of language acquisition under which children generalise from their input extensively (Biberauer 2019), contra traditional models of the conservative language learner. 

The talk will take place at 3pm, and there will be an opportunity to ask questions at the end. There will also be an informal drinks reception afterwards in the Department of Language and Linguistic Science (Deborah Hines Room, 2nd floor). Everyone is welcome!