Séminaire de Recherche en Linguistique

This talk is about the emergence of (reduced and full-fledged) clefts in French L1, in comparison with the emergence of Clitic Left Dislocation, in spontaneous speech production. I will show that these data have consequences for our understanding of the syntactic structure of clefts, the acquisition of syntactic structure, and the interaction between syntax and information structure (IS) in language acquisition (L1).

Our data have been retrieved in two corpora of spontaneous speech production: (i) the transversal corpus TCOF (ATILF 2018, fortnightly recordings of 15 children from ages 2-4) and (ii) the longitudinal corpus Lyon corpus (Demuth & Tremblay, 2008, three children: Marie, Anaïs and Nathan, recorded from age 1 until age 3).

My main conclusions will be the following:

On the syntax of clefts

(i)  The developmental path of the syntax of clefts is the following (with only some months between the first occurrence of each type):

a.    reduced clefts: “c’est + X” (the most frequent type at age 2)
          in answers to questions (1a) (see Belletti 2007) or not (1b)

b.    “cleft trials”
           - with premice of a CRC: “c’est + X + verb (phrase)” (2)
           - with juxtaposed clause: “c’est + X + subject + verb (phrase)”
           - with the subject of the clause coreferential (3a) or not (3b) with the pivot of the cleft

c.    cleft with adult-like cleft relative clause: “c’est + X + qui/que + verb (phrase)” (4)

Hence, between reduced clefts and clefts with adult-like cleft relative clauses, two other types occur, which have (as far as I can tell) not been mentioned in linguistic literature on the acquisition of clefts.

(1)         a.           Mother : qui c'est qu'a accroché l(e) ballon d(e)ssus ?
                            Child : C’est Chloé.
                            (age 2;07)

b.           [context : they’re playing a game]
             Mother : non laisse la avancer .
             Child : non, c’est moi.
             (age 2;9)

(2)         C’est Marie fait.(age 2;0.8)

(3)         a.           c'est Maéva elle a apporté ça. (2;4)

              b.           c'est à la crèche aussi il a régurgité. (2;10)

(4)         C’est moi qu’est fatiguée toute seule(2;6)

(ii)   Very young children already perform syntactic operations on the pivot of the cleft, such as wh-movement and CLLD of the pivot (see Haegeman et al. 2014). This seems to indicate that they acquire (or activate) the left periphery at the same time as the rest of sentence structure.

(iii)   Object clefts appear quite early (but later than subject clefts and clefts with an adverbial pivot), and in all syntactic syntactic types of clefts identified in (i). This seems to indicate that, given the right discourse context, children are able to produce object clefts.

On the discourse use (and semantics) of clefts:

(iv)   Whereas the first c’est (reduced and full) clefts are used as a focusing device (with corrective or contrastive focus), cohesive clefts (with a topical pivot) already emerge at age 4.

(v)    Even before their syntax is fully adult-like, the youngest children use their reduced clefts and “cleft trials” in adult-like discourse conditions. Hence, their acquisition of clefts starts with an adult-like Information Structure, even before the syntax is fully acquired.

(vi)   Some utterances of clefts with a non-adult-like discourse use involve additive particles (aussi ‘also’), which shows that children acquire the syntax and IS of clefts very early, but not the exhaustiveness of the construction.

About the comparison between clefts and CLLD

(vii)   The data of our corpora show that CLLD emerges shortly before clefts. Hence, in French L1, the clausal left periphery (where CLLD is situated, see Rizzi 1997 and subsequent work) and the “low position” hosting the pivot of the cleft (see Belletti 2012, Haegeman et al. 2014) are activated at the same time.