CP tracking: sentence types and signers’ perspectives

Based on the diagnostic criteria established in section 3.3 for the assessment of the main structural properties of DGS associated with the CP level, including those dimensions that involve the syntax-discourse interface, we examined the samples for complex sentential constructions (including those involving referential shifts, that is, POVs) and interrogative clauses with a view to determining whether the structure available included the CP layer. Simultaneous constructions (two-handed) were considered as additional evidence for syntactic complexity and an advanced narrative level. All in all, the analysis reveals that while the CP is available to participants at the onset of this study not all of them fully exploit the sentential structure at the time. In later recordings, participants produce narratives characterised by the use of a broader range of complex structures serving diverse narrative functions, for example, the expression of characters’ thoughts and emotions.

Complex syntax: Interrogation. Because question formation involves mechanisms associated with the full CP structure, it is commonly used as a diagnostic criterion for the availability of the an expanded structure. Unfortunately, however, interrogative clauses not only appear seldom in the data collected in this study, they commonly consist of a single wh-word (cf. (341) an example from Muhammed’s file 1). Single wh-word interrogatives represent grammatical and appropriate utterances in DGS, a pro-drop language that knows no copula and allows for subject and object drop in certain contexts. Muhammed’s example in (341b) has a meaning that would correspond to the English question “where is the frog?”. Nevertheless, in our analysis we adopted a rather strict criterion and regarded the occasional production of this type of question as insufficient evi?dence. Interrogative clauses containing more than one element were considered as an indication of the availability of the target mechanisms for question formation. Only Maria produces this type of evidence from the onset of the study (cf. example (183) repeated in (342)), whereas other participants produce them in the third sample (compare Simon’s example in (343) and Fuad’s in (344)).

In the data collected there is no evidence of target-deviant question formation, with the exception of what was considered to represent a candidate for language mixing, namely, Christa’s interrogative clause with the LBG sign ist (cf. (345)). Because ist does not occur in other contexts, it seems Christa uses the interrogative clause in a formulaic manner in file 1, an assumption that is corroborated by her production of the same pattern in an embedded context (cf. (346)). By the time of file 3, the pattern has disappeared (Christa rather produces several single wh-word interrogatives, including the one in (347)).

Progress in the use of interrogates for narrative purposes can be observed in Muhammed’s narratives. This participant, who only used a single wh-word in file 1, makes a skilful use of interrogation for narrative purposes in file 3. In his lively narration of the frog story, he frequently addresses the audience, for example, by using yes-no questions (compare (122) repeated in (348)). Muhammed also makes use of single wh-word questions (cf. example (123) repeated here in (349) or (124) repeated in (350)). These often serve a rhetorical function (in the sense of an invitation to “guess what happened next” in (349) or an indication that what follows next represents additional background information).

Complex syntax: subordination and coordination. Participants vary concerning their production of complex sentential constructions at the onset of the study. Simon and Hamida, for example, only produce few other than those involving POVs serving the function of reported action. In general, the analysis reveals an increased range of complex constructions in the narratives of the third sample when compared with the complex sentential constructions produced in the first sample. Typically, the embedded clauses produced belong to the type of constituent clauses and they are selected by psychological verbs.³⁰ Commonly, they are placed after the matrix clause, as in Muhammed’s file 1 (cf. (351)) and file 3 (cf.(352)) examples. Occasionally, constituent clauses precede the matrix clause. This is the case in Maria’s file 1 (181) and file 3 (354) examples (example (181), repeated here in (353), is a construction that involves the determiner det_exist correctly appearing before the matrix verb know, a verb that also allows for a postposition of the embedded clause). The narratives of Simon and Hamida, too, now contain complex sentential constructions with embedded clauses selected by psychological or performative verbs. Simon, for example, produces the sequence in (355), with an embedded clause in which the object is expressed overtly prior to the reported action. The sequence in (267) from Hamida’s file 3, repeated here in (356), contains a relative clause. Christa’s file 3 narrative, too, includes a remarkable range of complex sentential clauses, such as the one in (294) repeated in (357) with the verb wish.

Studies on frog story productions in spoken language learners have revealed that 9-year olds make overt temporal reference to "next morning” (though not all); interpretative comments (about the emotions, intentions and states of mind) appear scattered in these narratives. To set off high-point events, German children use expressions like "suddenly” (also learners of Hebrew, but not so learners of English).

The participants’ command of syntactically complex structures also becomes apparent in their production of coordinated constructions. These occur frequently in Fuad’s file 3, for example. For further illustration consider the example provided in (358), in which Fuad recounts that the boy is looking at the frog and waits. Examples (234) and (235), repeated here in (359) and (360), illustrate Fuad’s use of the coordinating conjunction and to express the simultaneity of the activities of two different protagonists (the boy and the dog). Note that the expression of this type of simultaneity also occurs through the use of the adverb also (compare example (242), repeated in (361)).

Complex syntax: Referential shift. Turning to referential shift, the analysis reveals that participants vary regarding their use of shifted referential frameworks. Again, Muhammed and Maria stand out against the other participants regarding their skilful use of SRFs in file 1, which documents the mastery of non-manual linguistic devices to signal and mark referential shift (body orientation, eye gaze direction, facial expression). SRFs are chosen where they are grammatically required, that is, where POVs are lexically selected by verbs like see or regard or in constructions with direct quotation (where the performative verb might remain unexpressed). File 1 narratives of other participants also document the use of referential shifts and the signalling and marking of POVs. However, referential ambiguities make apparent that while these participants exploit the CP structure to shift reference at the onset of the study, failure to establish loci contrastively affects referential maintenance in constructions with SRFs. We will expand on this observation in the next section, where we will also learn about the progress achieved in this respect.