By Anne Smith, Ph.D.
Purdue University

From the Stuttering Foundation's 2008 summer newsletter

I am pleased to have this opportunity to update you on the progress over the past year from the Purdue Stuttering Project.  You may recall from our earlier articles that we are engaged in a longitudinal study of young children who stutter and their normally fluent controls.

With funding from the NIH's National Institute on Deafness and Other Communication Disorders, we have been recruiting 4 and 5-year-olds, and we will be following them yearly for a period of 5 years.  In fact, we have just received notice from the NIH that our project will be funded for an additional five years.

After the five years of testing, the children will be 9 or 10 years old, and approximately half of them will have recovered from stuttering, while half will persist. In each yearly visit to our laboratories, the children participate in an extensive set of experiments, including those that test basic motor timing ability, tests of the brain’s activity in processing language, and studies of speech production in which we record muscle activity and motion of the articulators.

The goal of the project is to determine if the results of these tests will help us to predict which 4 and 5-year-old children who are stuttering are likely to persist.

One of the speech production tasks that our kids do involves producing what we call “novel nonwords.” These are words like “mab” which sound OK in English, but which are meaningless. We designed a set of four nonwords of increasing difficulty, so that “mab” is the easiest one, and “mabshaytiedoib” is the most difficult.

When we tested adults who stutter, we found that they could produce the words correctly after hearing them, but they showed more variability in their oral movements, suggesting that hearing, encoding, planning, and producing a novel speech sequence was a much more difficult task for them compared to normally fluent adults.

Is this lower ability to encode and produce new sound sequences present at the onset of stuttering? Or is it something that develops after many years of disfluency? Our experiment is designed to answer this question. 

So far in the “mab” nonword experiment, we have analyzed data from 22, all 4 and 5-year-old children who stutter (CWS) and 14 children who are normally developing (CND). Interestingly, we found that all of the children could produce the simplest nonword, "mab" correctly. However, on the more difficult items, such as “mabshibe,” only about 25% of the CWS could accurately produce them, while about 88% of the CND could produce all four nonwords correctly.

Clearly in terms of these behavioral results, the children who are stuttering do not do as well when asked to produce novel speech sequences.

During the nonword experiment, we record the motions of the lips and jaw. We are able to look at oral coordination during production of the novel nonwords. We analyzed data for the nonword, "mab" since all the children could produce it. Given how simple and short this nonword is, we did not expect that CWS would be less able to generate a motor program and produce it. Surprisingly though, CWS were significantly more variable compared to CND in their coordinative patterns over 10 productions of “mab.”

Thus, we can see that already, in the earliest years of stuttering, even when the child's production is accurate and no disfluency is perceived, the underlying patterns of speech movement are different. Are the children who perform most poorly on this nonword repetition task the ones most likely to persist in stuttering? We will be able to answer that question in the later years of our project.

We are truly excited about the potential for this work to have major implications for treating young children who stutter.

If you know of someone who has a child who stutters and lives in our area (Lafayette, Indiana), please ask them to contact us  at