Supervisory authorities

CNRS UNS

Our partners

ILF MSHS-SE

Search





Home > News & Agenda > Le Cycle BCL > On some neglected garden paths in Italian & French

Cycle BCL - Carlo Cecchetto

On some neglected garden paths in Italian & French

jeudi 31 mai 2018 à 10h, salle 213

On some neglected garden paths in Italian & French

I will first report the main findings of an eye-tracking study (Staub, Foppolo, Donati and Cecchetto 2018, Journal of memory and language) which investigates the processing of syntactic ambiguity in strings such as “the information that the health department provided (a cure) reassured the tour operators”, where the that-clause can be either a relative clause involving movement of the head noun (the information) from the object position after “provided” (RC: “the information that the health department provided __”) or the clausal complement of the noun (CC: “the information that the health department provided a cure). Staub et al. showed that CC constructions were preferred and this preference was not modulated either by frequency of RCs with respect to CCs (RCs are exceedingly more frequent than CCs) or by frequency biases towards CC or RC continuation of the nouns followed by CC or RC. Staub et al. attribute the preference for RCs to the Minimal Chain Principle (De Vincenzi, 1991), which states that filler-gap dependencies should not be posited except when necessary.
However, a limitation of Staub et al.’s study is the fact that the two structures tested (CC vs. RCs) also differed in the attachment site of the critical clause (complement vs. adjunct in traditional terms). Thus, an explanation in terms of Minimal Attachment (Frazier 1978) is also plausible.
To further investigate this question, we report on-going work on two temporarily ambiguous structures in Italian and in French that only differ with respect to the presence/absence of the filler-gap dependency.
In some regional varieties of Italian there is an ambiguous che corresponding either to a that in a declarative complement clause (DC), or to a what in an indirect question (IQ), as shown in (a-b). The verb “capire” (to understand) is compatible with both continuations:

(1) (a) Ho capito che fare gli esami è difficile [Ambiguous V, DC]
(I) understood CHE to-do the exams is difficult
‘I understood that taking exams is difficult’
(b) Ho capito che fare agli esami difficili [Ambiguous V, IQ]
(I) understood CHE to-do at the exams difficult
‘I understood what to do during difficult exams’

Minimal Attachment does not distinguish (1a) and (1b) since they both contain a complement clause. The two sentences, however, differ in the role of che: in (b) it introduces an indirect question, thus postulating a gap is a necessary operation to understand the sentence. To make sure that any difficulty in (1b) is created by the availability of the continuation in (1a), we also compared sentences (1a-b) with sentence (1c):

(c) Ho chiesto che fare agli esami difficili [Unambiguous V, IQ]
(I) asked CHE to-do at the exams difficult
‘I asked what to do during difficult exams’

In (1c), the verb (in this case, “chiedere”, to ask) unambiguously selects an IQ: for this reason, a gap after the main verb is obligatory in (1c).
In French, we investigated a similar ambiguous construction (ce que) using the same methodology:

(2) (a) Yasmina s’habitue à ce que sa mère regarde la télé chaque matin. [Ambiguous V, DC]
Yasmina gets used to that that her mother watches TV every morning
‘Yasmina gets used to the fact that her mother watches TV every morning’

(b) Yasmina s’habitue à ce que sa mère regarde chaque matin. [Ambiguous V, FR]
Yasmina gets used to that that her mother watches every morning
‘Yasmina gets used to what her mother watches every morning’
Again, Minimal Attachment does not distinguish (2a) and (2b), as both contain a CC. The two sentences, however, differ in the role of ce que: in (2b) it introduces a free relative (FR), thus postulating a gap is required. In (2a), instead, ce que is just a variant of the complementizer que and the sentence contains no gap. As a control, we also compared sentences (2a-b) with sentence (2c), with a verb (here “voir”, to see) that unambiguously selects a free relative.

c) Yasmina voit ce que sa mère regarde chaque matin. [Unambiguous V, RF]
Yasmina overlook that that her mother watches every morning
‘Yasmina overlooks what her mother watches every morning’
We conducted two acceptability judgment studies with 68 adult Italian participants and 70 adult French participants. The results both in Italian and in French show a preference for sentences (a), that include a declarative complement clause with no gap, over sentences (b), which contain a gap (as an indirect question or a free relative). An explanation in terms of Minimal Chain Principle seems to be needed.

Carlo Cecchetto (based on joint work with Massimo Burattin, Caterina Donati, Francesca Foppolo, Ingrid Konrad and Adrian Staub)

published by Dijana Bojovic - updated on