Conversational skills of individuals with fragile-X syndrome: a comparison with autism and Down syndrome.

The conversational skills of 18 individuals with fragile-X syndrome (FXS) were compared with those of two other matched groups with autism and Down syndrome. The FXS group used more eliciting forms in conversation than those with Down syndrome, and also used partial self-repetition more often than the other two groups. The Down syndrome group had more speech dysfluencies than those with autism, but not more than those with FXS. The autistic group used more inappropriate phrases. Qualitative analysis of behavioral phenotype may reveal differences in communicative organization among subgroups whose retardation is based on different genotypes. In addition, analysis of verbal strategies during conversation suggests important differences between individuals with FXS and autism.

On the use of hidden Markov modelling for recognition of dysarthric speech.

Recognition of the speech of severely dysarthric individuals requires a technique which is robust to extraordinary conditions of high variability and very little training data. A hidden Markov model approach to isolated word recognition is used in an attempt to automatically model the enormous variability of the speech, while signal preprocessing measures and model modifications are employed to make better use of the existing data. Two findings are contrary to general experience with normal speech recognition. The first is that an ergodic model is found to outperform a standard left-to-right (Bakis) model structure. The second is that automated clipping of transitional acoustics in the speech is found to significantly enhance recognition. Experimental results using utterances of cerebral palsied persons with an array of articulatory abilities are presented.

"Can you give me that one?": The comprehension, production and judgment of directives in language-impaired children.

This research is an investigation of pragmatic abilities focusing on "requesting" in a group of 30 language-impaired children between the ages of 3 1/2 and 9 years. The subject's requesting abilities were examined in three situations: (a) operating in dyads in a role-playing situation; (b) production of requests in an experimental procedure involving handpuppets; and (c) perception of requests in that situation. Transcriptions were analyzed using a speech act model along the three dimensions of "purpose," "directness," and "surface form." The findings indicated there was a predominant usage of direct forms with only a slight increase of indirect ones in the older group. Regarding the experimental assessment, it was found that language-impaired children, although restricted in the range of linguistic devices at their disposal, appear to compensate by frequently using the structures that they have already acquired. In general, these subjects operated pragmatically at a level two years or more below chronological age (compared to the performance of the normally developing children studied by Bates, 1976) and it appeared that their ability to discriminate between requests on the basis of politeness did not reach an appreciable level until the age of 5 1/2 to 6 1/2 years.