Objects, Motions, and Paths: Spatial Language in Children With Williams Syndrome
Barbara Landau
Johns Hopkins University
Andrea Zukowski
University of Maryland
The acquisition of spatial language is often assumed to be built upon an early-emerging system of nonlinguistic spatial knowledge. We tested this relationship by examining spatial language in children with Williams syndrome (WS), a rare genetic disorder that gives rise to severe nonlinguistic spatial deficits together with relatively spared language. Twelve children with WS, 12 normally developing mental-age matched children, and 12 normal adults described 80 videotaped motion events. Children with WS showed substantial control over key linguistic components of the motion event, including appropriate semantic and syntactic encoding of Figure and Ground objects, Manner of Motion, and Path. The expression of Path, although surprisingly spared, was more fragile among children with WS in contexts plausibly related to their nonlinguistic spatial deficit. The results show strong preservation of the formal aspects of spatial-linguistic knowledge and suggest that the nonlinguistic spatial deficits shown by children with WS have, at most, limited effects on their spatial language. These findings have implications for the relationship between spatial language and other aspects of spatial cognition.
One of our most fundamental capacities is the ability to talk about the objects and events around us. Even during early language learning, children take delight in commenting on objects and their motions through space. The privileged nature of the motion event in language learning stems, no doubt, from the salience and interest that such events hold for young children. However, equally important is the fact that languages are well designed to capitalize on this salience, with universal means of encoding objects, their motions, and the paths along which they move. The early emergence of such aspects of language has traditionally been viewed as evidence that language builds on nonlinguistic representations: Because children can and do perceive and represent objects, motions, and paths, they naturally tend to express these notions linguistically. In fact, the character of children's spatial language has often been taken as a clue to the nature of their underlying nonlinguistic representation of space (Brown, 1973; Clark, 1973; Bowerman, 1996).
From this perspective, severe impairment in the child's nonlinguistic representation of space would predict corresponding impairments in spatial language. Must this necessarily be true? In this article, we address this question by examining the linguistic expression of motion events among children with Williams syndrome (WS)āa rare genetic defect which results in severe spatial deficits together with relatively spared language. The striking difference in profile between the two systems of knowledgeāspatial cognition and languageāraises the question of the extent to which spatial deficits in WS have an impact on the acquisition of spatial language. At the extremes lie two possibilities: Spatial language in children with WS might show severe deficits, like other aspects of their spatial cognition. Alternatively, their spatial language might be relatively spared, like other aspects of their language. Between these extremes lies the intriguing possibility that spatial language may be selectively impaired in ways that closely reflect the nature of the nonlinguistic spatial deficit. In this case, examining spatial language in children with WS should shed light on how nonlinguistic representations of space support spatial language and what consequences for spatial language follow from a deficit in nonlinguistic spatial representations. The results of such an investigation can, more generally, be brought to bear on considerations of the relationships between spatial language and spatial cognition.
Background and General Issues
WS is a neurodevelopmental disorder caused by a hemizygous microdeletion on the long arm of chromosome 7 (7q11.32). The syndrome gives rise to a variety of characteristics, including pathology of the heart and other internal organs, a characteristic facial profile (often referred to as elfin facies), and mild to moderate mental retardation. Of particular interest to cognitive scientists, however, is the unusual cognitive profile typical of individuals with WS: They show severe spatial deficits accompanied by surprisingly spared language. The unique developmental profiles of space and language naturally suggest the possibility that the two systems may be developmentally modular, emerging independent of each other (Bellugi, Bihrle, Neville, Doherty, & Jernigan, 1992).
The spatial deficits of individuals with WS have been documented in a range of studies, but the most widely cited evidence comes from tasks which require that people observe a spatial configuration and then reconstruct it, either by drawing a copy of a pictured object or array or by putting together a set of blocks to duplicate an existing multiblock spatial design. The spatial performance of individuals with WS is typically far inferior to normally developing children of the same chronological age (Bellugi et al. 1992; Mervis, Morris, Bertrand, & Robinson, 1999) and it is even inferior to normally developing children of the same mental age (Hoffman, Landau, & Pagani, in press; Mervis et al., 1999). For example, Bellugi et al. (1992) reported that individuals with WS performed in the bottom percentile of their age group in a standardized block construction task, and Mervis et al. (1999) reported the same. Importantly, the performance of these individuals in these construction tasks is qualitatively different from that shown by individuals with comparable retardation of different etiology. For example, when copying a multiblock design, children with WS tend to duplicate the local elements at the expense of the global form, but children of the same mental age with Down syndrome tend to do the reverse (Bellugi et al., 1992; see also Birhle, Bellugi, Delis, & Marks, 1989). The pattern of spatial deficit appears early in development, and although there is some developmental growth in the ability to solve spatial construction tasks, the severe deficit persists into adulthood (Mervis et al., 1999).
In contrast to this impairment in the spatial domain, individuals with WS typically show language capabilities that meet or exceed expectations based on mental age. The relative advantage of linguistic over nonlinguistic abilities appears quite early, within the first several years (Mervis & Bertrand, 1997), and thereafter, children with WS have language that is surprisingly fluent, with rich modulation of tone and expression (Bellugi et al., 1992). Despite this strength, recent research indicates that language in individuals with WS may not be entirely intact. For example, mastery of morphological rules and aspects of lexical learning may be impaired (Clahsen & Almazan-Hamilton, 1999; Karmiloff-Smith et al., 1997) and the semantics underlying certain lexical items in individuals with WS may not be as in conceptually rich as in normally developing children (Johnson & Carey, 1998). Nevertheless, vocabulary measures show performance generally above overall mental age (Bellugi et al., 1992) and, to the casual ear, the children's spontaneous vocabulary and overall language skills are strong (Bellugi, Wang, & Jernigan, 1994). It is safe to say that the contrast between their profound spatial deficit and their relatively spared language is striking.
What do these profiles predict for the acquisition of spatial language? Because spatial language sits at the intersection of spatial cognition and language, it affords us an unusual opportunity to understand the degree to which the two systems of knowledge emerge independently and/or interac...