Author Topic: Developmental stuttering and models of speech production  (Read 5595 times)

Offline Torsten

  • New Linguist
  • *
  • Posts: 2
Developmental stuttering and models of speech production
« on: December 30, 2015, 01:23:18 PM »

let me tell you about my project: a causal theory of developmental stuttering. It is presented on my website where you can also read more about me and my motivation to do this work.

Developmental stuttering has been investigated for about 100 years, there are thousands of scientific publications, and we have a lot of empirical findings about the symptoms, about speech- and non-speech-related behavior and about the brain of people who stutter  – but the cause of stuttering is still unknown. So it is a fascinating problem.

Not a few children experience a period of stuttering, but most of them – about 80% – recover spontaneously, without treatment from the disorder. The prevalence of persistent developmental stuttering is about 1% of the overall population, with a male:female ration of 4:1.

Developmental stuttering is an extremely variable disorder. It is influenced by the speaker's expectations, by the communication situation, by stress, by alterations of auditory feedback (whether and how one's own speech is heard), and, not least, by linguistic factors like word category, sentence position, the information load of a word and its importance within an utterance. Additionally, stuttering is a disorder of connected speech: It onsets in childhood not before the child starts to form simple sentences.

Therefore, I don't believe that one will ever find a simple physical cause – a genetic mutation or a structural deficit in the brain. From my view, developmental stuttering can only be explained in the framework of a model of normal speech production. I chose Levelt's model as the basis of my theory, mainly because it might be the best-known and a widely accepted model, and because it includes assumptions about auditory feedback and self-monitoring. However, there are some problems with the Levelt model: Since formulation and articulation are strictly separated in this model, there seem to be no place for stuttering – see here.

So I made some changes to Levelt's model. First, I simplified it by putting together formulation and articulation: In spontaneous speech, sentences are formulated by speaking them, not before speaking – regardless the unconscious brain processes preceding speaking as well as all motor activities. Second, I included Engelkamp & Rummer's concept of speaking programs and acoustic word nodes into Levelt's model. The result is to be seen below; see also here for more explanation.

So far, a short presentation of my project and its relationship to linguistics. I would be happy to find some people here who are interested in a discussion about these topics. Greetings from Germany – and a Happy New Year to all of you who begin a new year.