This paper analyses and discusses the speech of Vincent, a hearing child of deaf parents who acquired sign language as his first language. When exposed to spoken language, his progress was slow and abnormal. He utilised a number of grammatical devices which may have come from sign language, such as reduplication, copying around, doubling, and afterthought. These were imposed on a fixed kernel sentence which attributed more importance to form than to semantic relations. The possible reasons for this abnormal development are put forward, and their implications discussed.

Victor, a biologically normal child of normal hearing and good intelligence, had almost no exposure to spoken language until he was three years old; his only language until that time was the sign language which he leamed from his deaf-mute parents. Three structural features of sign language are described, and evidence is presented that Victor structured his sentences in those ways when his speech development began. It is argued that the presence of structural interference from sign language in Victor's speech suggests that the manner of functioning of our innate capacity to acquire speech may differ depending on the nature of prior linguistic experience.

No abstract

Subjects segmented patterns composed of dichotomous elements by grouping elements that seemed to go together. Three response conditions were used: (a) Subjects tapped the pattern on a bongo drum, (b) subjects placed slash marks between elements of a spatial-visual array and, (c) subjects vocally identified the pattern after auditory presentation. Subjects' organization of the patterns was found to follow principles of hierarchy and similarity. First, the segmentations given for each pattern fit into a tight lattice. Second, segmentations for all of the patterns were constructed so that the subgroups were similar in containing equal numbers of elements and either the elements within each subgroup were the same (e.g., x/oo/x/oooo) or the composition of adjacent subgroups was the same (e.g., oxo/oxo/oo). The segmentations found for each response mode were in an inclusion relation. In going from vocal to manual drumming to spatialvisual responding, new types of organizations emerged that resulted from different applications of similarity. There are general principles-hierarchy and similarity-that apply to all cognitive domains. However, the way these principles operate depends on the specific context.The authors would like to thank Delmar Yoder for his assistance in the experimentation.

Sign linguists routinely parse ASL sentences using the category 'topic' , by which is meant a constituent on the left edge of the main clause, structurally separate from it, and marked by a discrete formal symbolic event, more fully brow raise + backward head tilt + pause, although brow raise is sometimes considered sufficient. This paper provides evidence confirming suspicions that these leftdetached constituents need not be marked by brow raise, and suggests that brow raise is better regarded as signaling a type of momentary focus -thus explaining why sign languages tend to employ it as they do -and that it belongs to a larger set of 'topic-marking' devices whose iconicity remains active in day-to-day signing.