Expression of spatial relations

The interaction of grammatical and discourse requirements also becomes apparent in the expression of spatial relations as choice of linguistic devices in this case is related also to the discourse status attributed to the information encoded (foreground vs. background).

We remarked previously that verbs in complex classifier constructions are correctly inflected, including the choice of classifier elements for figure and ground, and the spatial orientation of the two classifiers. However, participants differ regarding their use of these constructions at the onset of the study (whereby the lack of these constructions in Hamida’s file 1 might be related to her narrative style). We have not found whole body depictions to describe movements in our data as has been shown to be the case in the productions of young infants (before age 2). Neither do our data contain evidence for a sequential expression of meaning components as it has been observed in children aged 2;0-2;6 (Morgan et al. 2008) nor do our participants use of real-world substitutes as the subjects investigated by Tang et al. (2007) did.

As for the constructions produced in the first sample of the present study, it becomes apparent that some sequences are not clear, which, by assumption, reflects deficits at the syntax-discourse interface. Note that the deficits are not related to an incorrect selection of classifier elements as has been found to be the case in the production of the participants in Tang et al.’s (2007) study (see the discussion in section 3.2.3.2), which indicates that participants in the present study have a command of the constraints on the morphological composition of classifier constructions. Rather, the variation observed pertains to information encoded.

Foreground and background, as pointed out by Berman and Slobin (1994: 9) are not only determined by the logic of events in a narrative, but are also the result of the creative perspective taking used by the narrator to guide the listener through a subjective interpretation. To illustrate this point we might consider the range of variation observed in the participants’ narration of the episode involving

the frog’s escape from inside the jar. We will compare participants’ productions produced in the first sample of the data collected.

As we can see in (372), Muhammed does not recount the frog’s escape, but chooses to focus on the frog’s wish to leave and on his motivation to do so.

Fuad also recounts the frog’s intention but includes information on the motion involved in the activity intended. He uses a complex spatial predicate but chooses not to provide information about the nature of the location (example (216) repeated in (373)). The referent backgrounded via the h2-classifier remains generic.

Simon either provides specific information on the background. He focuses on the frog’s climbing out of the jar (example (374b)) and the manner in which he does so (silently).

Hamida chooses to describe the episode by shifting reference and adopting the perspective of the frog (cf. (375)). No previous information is provided on the location the frog escapes from.

Maria’s description of the frog’s escape in (376) includes a lexical antecedent providing information about the location the frog escapes from (= the ground); the information is backgrounded through the classifier on the non-dominant hand (the h2-classifier) in the complex spatial predicate used to describe the figure-ground configuration in the frog’s escape.

As we can see in (377a), Christa provides first a description of the frog’s escape without information on the location; this information is provided in a repetition of the propostion in (377b), in which a generic reference form (cl:form) used to designate the container is added in the target-like preverbal position.

As we can see, not all participants choose to provide details on the frog’s escape (some rather focus on the frog’s intention to leave); further, participants vary as to (a) whether they mention the location the frog is in before he escapes, and (b) whether they use generic or specific reference forms to refer to the ground (cl:form vs. glass^abowl). One participant uses an SRF to describe the frog’s activity. In sum, it becomes apparent that participants vary in their recount of a specific narrative episode, and that they do so also with respect to the background information provided. Not all of them use complex classifier predicates in their descriptions, and for those who use them, they do not use them in the same manner.

Our observations about variation regarding the provision of prior specification of the ground pattern with the observations made by Tang et al. (2007: 308; cf. also Morgan 2006; Slobin et al. 2003) regarding the productions of more advanced learners who had a tendency to omit lexical antecedents of the classifier referents. Regarding further development, our data reveal that participants that omit this information at the onset of the study provide it in file 3 (notably, Muhammed and Hamida). Hence, from a narrative perspective we are led to conelude that the appropriate use of complex classifier constructions is related to the expression of figure-ground relations. Their mastery represents a major developmental step in the attainment of the properties that involve the syntax-discourse interface.

Discourse buoys. Worthy of mention in this context is the participants’ skilful use of h2-classifiers for discourse regulatory purposes, that is, as discourse buoys. In several instances, h2-classifiers used to designate the background in complex classifier constructions describing a spatial configuration (particularly in the narrative episodes involving the beehive, the hole in the tree or the stone, behind which the deer is hiding) were retained during subsequent discourse stretches. By providing this additional information, the participants not only demonstrate their command of the classifier system; the appropriate use of discourse buoys is also indicative of their mastery of pragmatic constraints on the use of linguistic devices for the purpose of creating a cohesive account of the episode narrated.