Syntax-discourse interface

Referential establishment and maintenance. Turning to referential establishment and maintenance the analysis reveals a progress concerning the use of linguistic devices to create cohesion. For example, the locus associated with the verb take in (297) to recount that the boy takes the frog back home coincides with the locus established previously for the frog found behind the log. (Incidentally, the sequence is also illustrative of the use of repetitions in file 3).

Referential shifts occur frequently in this narrative. In (298), Christa reports that the boy and the dog wake up, and that the boy sees the frog is gone. This latter proposition is expressed in the context of a shifted referential framework, marked through body shift and eye gaze to the left, toward the location associated with the frog at the beginning of the narration. Note that reference is maintained in (299), in which Christa narrates that the dog wakes up and also sees that the frog is gone.

Reference forms and functions. With respect to Christa’s use of reference forms and the functions they serve (compare Table 3.43) we can see that protagonists are predominantly reintroduced via a full NP. Indeed, the relative proportion of NPs out of the total frequency of the reference forms serving this function amounts to 77.3% (cf. also Figure 3.12), which marks a difference to file 1 (recall that in that file the rate of NPs was much lower [18.2%], with a clear predominance of subject drop (63.6%), cf. Table 3.41). As for the forms used to refer to a protagonist involved in a series of events subject drop clearly predominates (88%). Finally, we can see that at this stage the overall distribution of reference forms (in particular 27.9% NPs vs. 69.4% subject drop) is more in line with the distribution observed in the narratives of the other participants.

Table 3.43: Reference forms and functions in Christa’s file 3.*

* Expressed as a percentage of the total number of reference forms (proportions of forms used for respective function in brackets). Absolute numbers are provided in the Appendix Table C-12.

Figure 3.12: Proportion of reference forms and functions in Christa’s file 3.

Expression of spatial relations. Another major difference between Christa’s narrative in file 1 and her recount in file 3 concerns the expression of figure-ground relations, expressed in a more detailed manner in this file. For example, conventional signs, where they are available, designate the objects in question, backgrounded information is maintained via h2-classifiers, or classifier elements in SRFs. For further illustration consider example (300) which shows that Christa introduces the object (a jar) backgrounded subsequently via the h2-classifier used as a discourse buoy, as she goes on to recount that the dog looks into it before he eventually sticks his head into it. In a similar manner, the h2-classifier used to designate the object the boy is looking at in (301a) is retained as a discourse buoy in (301b), where Christa describes the thoughts of the protagonist.

Example (302) illustrates the complexity of the information expressed via twohanded complex classifier constructions and how the use of the h2-classifier as a discourse buoy throughout this discourse stretch including a description of the beehive (its shape) and its inhabitants contributes to the creation of cohesion.

Table 3.44: Expression of figure-ground relations in Christa’s file 3.