Syntax-discourse interface

The narration of the frog story produced by Fuad in file 3 reveals a more skilful use of the linguistic devices necessary for the creation of cohesion and coherence.

Reference establishment and maintenance. In his file 3 narrative, Fuad’s demonstrates an advanced command of linguistic means used to establish and maintain reference. Consider, for example, the sequence in (237), in which det_exist establishes the locus for the group of frogs sitting behind a log. Notice that the locus associated with the frog family in (237a) is picked up in (238), in which we learn that the boy waves good-bye to the frog family. Note that Fuad signals the POV through a performative verb (say) and a change in body orientation and eye gaze direction to the right.

The use of determiners and pronouns in this narrative marks a difference to Fuad’s narration in file 1. For example, we can see in (239b) that Fuad uses a pronoun to address the dog in the context of a POV, in which he adopts the perspective of the boy. Notice that the locus associated with the pronoun corresponds with the end point of the sign stand-up produced in the utterance immediately preceding (239b) (that is, (239a)), in which the boy asks the dog to stand up.

In this context we must also note that, in some cases, the referential identity of the pronouns used is not unambiguous. For further illustration we might consider example (240), in which Fuad uses a pronoun in a reintroductory context, after his recount of the frog’s escape. This pronoun, produced with emphasis, as it would be the case in the production of a demonstrative pronoun in DGS, is associated with a location in front of the signer, slightly to his left. Now, reference in this case is ambiguous because Fuad has not associated the locus with a referent, so that only the story context might help to infer the identity of the referent associated with the pronoun (Fuad has just recounted that the frog wants to escape from the jar, which, as he explains in this sequence, is not perceived or heard by the boy or the dog because they are sleeping). Although the choice of this locus might seem uncommon at first sight, it must be noted that Fuad also associates a possessive determiner at the beginning of the narrative with this locus in front of him to refer to the room (“the boy’s room”) the story characters are sitting in. We can only speculate on the possibility that the choice is guided by the use of a contrastive criterion (opposite to the signer), although, as we remarked upon in section 3.1.4.2 signers usually choose a location right or left in the sign space for this purpose.

Further, the analysis reveals that det_loc is productively used to establish loci for locations referred to in descriptions of spatial relations. For example in (241) Fuad specifies the position of the owl (in the tree hole) by modulating det_loc toward the h2-classifier that backgrounds the information about the hole in a tree, introduced through a lexical antecedent in a previous utterance.

Taken on the whole, the analysis of the data reveals that the distribution of loci in the sign space is such that they contrast with respect to the vertical axis (top- down) and horizontal axis (left-centre), in narrative episodes that involve two characters. Consider, for example (242a) and (242b), the former involving the boy, the latter the dog as an agent. Fuad uses POVs to express the calling for the frog by either character, signalling the respective referential shift through a change in body orientation and eye gaze direction (the POV involving the boy is marked through body orientation to the right, eye gaze direction toward the top; whereas the dog’s perspective is marked through body orientation to the left, eye gaze directed toward the bottom of the sign space).

Non-manual means are also skilfully used in quotation environments introduced through a performative verb, as is illustrated in example (243), in which the shift to the perspective of the boy is signalled and marked by a change in body orientation and eye gaze direction: notice that object agreement (picking up the locus associated with the dog) in this case is marked non-manually via body lean forward to the left, eye gaze toward the same direction.

The only narrative sequence that continues to be difficult to interpret is the one in which we learn that the boy falls upon the deer’s back, the deer starts running and later throws the boy into the water. Referential shifts occur rapidly, and only the deer is referred to lexically via an NP.

Reference forms and functions. With respect to the reference forms used by Fuad to refer to the story characters, one remarkable aspect is that NPs are not only used for the introduction of referents, but also in nearly all instances of their reintroduction. As we can glean from Table 3.33, reintroduction of referents occurs primarily via NPs (77.8% out of a total of 20.7% of reference forms serving this function). If we consider this figure against the backdrop of the relative proportion obtained for file 1 (the relative proportion in that file amounted to 46.7%, cf. Table 3.31 above) the increase in the relative frequency of this form-function combination is certainly remarkable, as is the decrease of the relative frequency of subject drop in reintroductory contexts (from 33.3% to 16.7%). Repetitions of propositions, such as the one provided in example (244), involving the addition of an overt referential expression to unambiguously identify a character in a reintroductory context, might be taken as an indication of an advanced command of narrative constraints on referencing at this stage.

Clearly, the use of NPs in contexts other than introduction or reintroduction serves the purpose of making clear who is the agent of the activity described. This occurs particularly in the context of sequences involving referential shifts (see example (243) above).

Finally, it is interesting to note, if we compare the total distribution of reference forms in this file with that of file 1, that it is very similar (compare Table 3.31 above and Table 3.33). Yet considering the functions these reference forms serve, it becomes apparent that these change overt time.

Table 3.33: Reference forms and functions in Fuad’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-8.

Figure 3.8: Proportion of reference forms and functions in Fuad’s file 3.

Expression of spatial relations. We noted previously that Fuad adheres to the (S)XV format in constructions with locative complements. This is also the case in sequences involving a referential shift to describe the protagonist’s activity. In (245), for example, we learn that the dog looks into the jar. The ground (the jar) is expressed first, followed by the dog’s activity. Similarly, the ground is expressed first in the description of the dog sticking his head into the jar in (246) below. Finally, in (247) we can see that further details about the activities narrated appear in the context of repetitions, in which he provides additional information on the ground (compare (247c)).

Reference to the ground, as we can glean from Table 3.34, occurs via NPs or h2 classifiers. Also, lexical antecedents introduce information about the ground that is later backgrounded in complex classifier constructions, which contributes to the overall coherence of the narrative. For further illustration of the previous observations consider examples (248) and (249). In (248a) an h2-classifier is used to express that the boy looks into a boot, which is then described in more detail by shifting the referential framework and adopting the perspective of the boy. In (249) the h2-classifier backgrounds information on the location (the log) upon which the boy gets before he discovers the frogs.

Table 3.34: Expression of figure-ground relations in Fuad’s file 3.