irls). Parents of infants in the Sign Training group (ST group) were individually instructed in ways to promote symbolic gesturing by modeling simple gestures themselves, always being sure to pair the gesture with the verbal word (e.g., "Birdie! [FLAP ARMS] See the birdie!"). They were told to use any physical motions that made sense to them and would be easy to remember in relation to the referent (e.g., clawing motion for cat, index finger wiggling for caterpillar). Videotaped examples of parents and infants using such gestures were used as illustration. In contrast, parents in the Non-intervention Control group (NC) were not involved in training of any sort, nor were they aware of our special interest in language. Inclusion of this particular control group allowed direct comparison of infants who had used.ResultsAcquisition of Symbolic GesturesTo determine if the independent variable had in fact been manipulated, initial analyses focused on whether the infants in the Sign Training group had acquired any symbolic gestures. The data relevant to this question came from the biweekly phone calls to the ST families during which parents described the contexts in McIntyre6which their infants used gestures to communicate concepts. The criteria used to determine when a gesture qualified as a "generalized symbol" were very similar to those used to assign symbolic status to infant gestures in Acredolo and Goodwyn (1988) and identical to those used for both gestures and words in Goodwyn and Acredolo (1993). Salient among these criteria were the following: (a) spontaneous usage by the child (i.e., not following direct modeling or elicitation), (b) occurrence in a stereotyped form, and (c) use in reference to multiple exemplars of the underlying concept beyond the specific context in which the item was initially taught. For example, use of a panting gesture for "dog" had to be extended beyond labelling just the family dog, and use of a gesture for ...
