I am doing my MSc on this issue of "multimodalities", lingle.. specifically I am looking at gesture and language. My results so far (and those of others) seem to indicate that where gesture reinforces language e.g. it is more "sign like" and very specifically conveys a message, it seems to reinforce language understanding, but where it conveys something slightly different or supplementary e.g. you say "I'm going outside for a bit" while gesturing that you will smoke, it significantly slows down processing for everyone including people with typical language development. The effect is disproportionate for people with language and communication impairment, but it's not particularly clear cut..
Speech processing seems to be a major factor, unsurprisingly. The students who have performed most poorly (slowest reactions/most errors) in the simple, structured tasks we have been doing are those with speech disordered profiles, which surprised me. Many of the students with poor understanding responded faster to meaning conveyed across "two channels" than they did to speech only... the students with poor speech had far slower reaction times when they needed to listen for information in speech and also pay attention to gesture.
It is all very complex, really and there is so much to learn... we know so little really. Often difficulties are described as being about "social communication" or "pragmatics" when really they reflect difficulties with processing, understanding and using language form. In the same way, difficulties processing social information can be mistaken for difficulties with understanding or structuring language form. When you read transcripts of conversations with adults after stroke, you can often see this in action: the meaning is distorted. Yet in context with a supportive conversation partner, it's obvious that social understanding is there, but subtleties of the message have been misunderstood or the means to convey the social understanding isn't there. Yet socially and functionally, these students seem to "manage" this information better in everyday conversations than their peers with understanding difficulties....?
One of the reasons that some this stuff is so poorly targeted often is because of these subtleties. Simple sentences are easily broken down and "taught", and some basic complex sentences too (because.. when.. etc). However, the speed and flexibility of ongoing speech, combined with the multiple cues from context and nonverbal communication etc, is hard to break down into teachable targets. There is a desperate desire for "programmes" for "social skills" but teaching outward behaviours e.g. eye contact/volume/proximity etc just doesn't really match the level of demand when there are difficulties with language and/or social information processing.
I have had the most success with "supported conversation" type techniques... coaching students through their communication breakdowns in real, naturalistic situations... over time, I see real changes with this approach. In contrast, I have seen very little success comparatively with structured social interventions either in ABA or speech therapy.