Modifying Phonated Intervals

We produce vowels by releasing air from our lungs, vibrating our vocal folds to produce phonation (a humming sound), and then shaping the humming sound by moving our lips, jaw, and tongue (articulators).

We produce voiced consonants, such as b and d, similarly to vowels. We produce voiceless consonants, such as p and k, by releasing air and moving our articulators without phonation.

Most words include both voiced and voiceless phonemes. We switch our vocal folds on and off many times per second. Stuttering is the inability to switch phonation on after a voiceless consonant.

If we measure the length of time that our vocal folds vibrate, the normal duration of phonated intervals is around 125 milliseconds (ms). Fluency shaping therapy trains us to prolong these phonated intervals to first 2000 ms (two-second stretch), then 1000 ms (one second), 500 ms (half a second), and finally 250 ms “slow normal” speech, which is a little slow but is within the range of normal speech.

By slowing phonated intervals and using other techniques such as gentle onsets, stutterers talk fluently.

Modifying Phonation Intervals (MPI) Therapy

Modifying Phonation Intervals (MPI) stuttering therapy takes a different approach. Instead of stretching all phonated intervals, MPI therapy eliminates phonated intervals that are less than 100 ms. In other words, the shortest phonated intervals are eliminated. The target phonated intervals are around 125 ms, or twice the speed of prolonged speech. The result is fluent speech at a normal speaking rate.

A biofeedback computer with a throat microphone is used to analyze the user’s vocal fold activity and display the duration of phonated intervals. The computer isn’t portable and can’t be used outside the speech clinic.

Currently MPI therapy is available in only a small number of speech clinics.

The Evidence

Seventeen adult stutterers were assigned to MPI therapy and ten were assigned to prolonged speech (fluency shaping) therapy. Of the 17 MPI participants, 11 (65%) were successfully treated. Of the 10 prolonged speech participants, three (30%) were successfully treated. In other words, MPI participants were more than twice as likely to be successfully treated. 1

Both groups received intensive therapy for three hours per day, six days a week. Participants progressed through five progressively demanding steps: oral reading, monologue, conversation, and two steps of telephone calls. To graduate from one step to the next required producing three consecutive trials of fluent speech at speaking durations of one, two, and three minutes. The duration of this Establishment phase ranged from three to twelve weeks per participant. Participants then completed a Transfer phase lasting 10 to 37 weeks. Graduation required three fluent conversations in six beyond-clinic settings. Finally the participants completed a Maintenance phase lasting 38 to 78 weeks. Follow-up data was then collected one year after completing the Maintenance phase. Average total time for the MPI participants was 155 weeks and for the prolonged speech participants was 168 weeks (approximately three years).

While success is important, examining how participants failed can be insightful. Of the eight participants who failed MPI treatment, two participants (25%) failed in the initial Establishment phase, i.e., they couldn’t speak fluently in the speech clinic. Two participants (25%) failed in the final Maintenance phase, i.e., they had initial success but relapsed within three years. Four (50%) participants failed for reasons unrelated to the treatment (two refused to undergo brain scans for a related study, one went to prison, and one dropped out due to domestic abuse). Of the five participants who failed prolonged speech treatment, two participants (40%) failed in the initial Establishment phase, i.e., they couldn’t speak fluently in the speech clinic; and three participants (60%) failed in the middle Transfer stage, i.e., they spoke fluently in the speech clinic but couldn’t transfer this fluency to “real world” conversations. This suggests that prolonged speech not only has a higher failure rate, participants also fail earlier in the program.

If we remove the four participants who were not successful for reasons unrelated to therapy, then 11 of 13 (85%) of MPI participants were successfully treated.

In a related study 2, brain imaging associated decreased activity in the putamen and claustrum areas with therapy success. The putamen area of the brain is associated with phonation and motor learning and control. The function of the nearby claustrum area is unknown.

Five other research papers are related to MPI therapy. 3 4 5 6 7

What the Experts Think

I’ve never met an expert who’s heard of MPI therapy. They don’t read the scientific journals and there’s no well-financed company paying them to provide MPI therapy.

My Experience With MPI Therapy

In 2012 MPI therapy was available only in one speech clinic, in California. I had throat microphones so I built my own iPhone app to produce the biofeedback. Here’s a picture of my app:

The MPiStutter iOS app displays your vocal fold activity as you talk.

The red, orange, and yellow bars indicate too-short phonated intervals. The blue and purple bars are too-long phonated intervals. The green bars are just right.

When I stuttered the red, orange, and yellow bars appeared.

When I did prolonged speech the blue and purple bars appeared.

The green bars were difficult to produce. I couldn’t intentionally make the green bars appear. I’d warm up with prolonged speech and then relax and just talk…and then the green bars appeared. It was a “zen” experience. My speech sounded fluent and relaxed.

The app ran on my iPhone so I intended to use it in conversations. But watching the screen while chatting in the supermarket checkout line wasn’t possible. So I made the app produce DAF in earphones when red bars appeared. But DAF switching on for 100 ms or less was annoying; making the DAF stay on longer made the feedback less effective as the DAF would stay on after the red bars were gone.

This was when it became clear that I needed a speech buddy. I needed someone that I could talk to while I watched my iPhone screen. The study above had stutterers talking with the MPI computer three hours a day, six days a week, for three to twelve weeks (up to three months).

My dog would be happy to listen to me talk to him three hours a day for a few months but I’m fluent with my dog. Stuttering therapy needs to be done in increasingly stressful conversations.

Here’s where the rubber hits the road. MPI therapy failed for me because I couldn’t talk enough hours, in stressful conversations, with a variety of people. What’s needed is a Zoom app where stutterers can talk to other stutterers practicing speech therapy.

Are Short Phonated Intervals Normal?

I did get to use my MPI therapy app for three days of talking to strangers ten hours a day. Every year at the American Speech-Language Hearing Association (ASHA) convention I’d get a booth at the trade show and talk to SLPs all day. I’d talk to between 100 and 300 SLPs in three days.

The SLPs loved the app. I’d show them the red intervals, the blue intervals, and then the green intervals. They could hear how my voice changed as they saw the colors changing on the bars scrolling across the screen.

But intentionally stuttering hundreds of times left me exhausted and unable to talk at the end of the convention. My stuttering was much worse than it had been in years. I’d proven that MPI therapy can make your stuttering worse, if you try to produce the red bars instead of the green bars.

I hired four SLP students to help me in the booth. All wanted to specialize in fluency disorders; none were stutterers. The three young women decided that intentionally stuttering wasn’t for them and let me do the demos while they handed out brochures and collected contact info from the SLPs.

The fourth student was a young man who was determined to do the demos with the app. He absolutely couldn’t produce the red intervals. Apparently red intervals aren’t produced by normal speakers; stutterers have a special gift to make red intervals. The third day is figured how to cough while speaking and make the red intervals appear. I have no doubt that in a few more days the app would have taught him to stutter. He wrecked his voice. Luckily the trade show ended at noon so he only did this for a few hours. He left the convention having glimpsed what it feels like to stutter.

Submit your review

Create your own review
Average rating:  
 0 reviews


  1. Ingham, R., Ingham, J., Bothe, A., Wang, Y., & Kilgo, M. (2015) Efficacy of the Modifying Phonation Intervals (MPI) Stuttering Treatment Program with Adults Who Stutter. American Journal of Speech-Language Pathology, January 28, 2015. doi:10.1044/2015_AJSLP-14-0076
  2. Ingham, J., Wang, Y., Ingham, J., Bothe, A., & Grafton, S. (2013) Regional brain activity change predicts responsiveness to treatment for stuttering in adults. Brain and Language, 127:3, December 2013, 510-519.
  3. Gow, M., & Ingham, R. (1992) Modifying Electroglottograph-Identified Intervals of Phonation: The Effect on Stuttering. Journal of Speech and Hearing Research, 35, June 1992, 495-511.
  4. Ingham, R., Kilgo, M., Ingham, J., Moglia, R., Belknap, H., & Sanchez, T. (2001). Evaluation of a stuttering treatment based on reduction of short phonation intervals. Journal of Speech, Language, and Hearing Research, 44, 1229–1244.
  5. Godinho, T., Ingham, R., Davidow, J., & Cotton, J. (2006) The Distribution of Phonated Intervals in the Speech of Individuals Who Stutter. Journal of Speech, Language, and Hearing Research, 49, February 2006, 161-171.
  6. Davidow, J., Bothe, A., Andreatta, R., & Ye, J. (2009) Measurement of Phonated Intervals During Four Fluency-Inducing Conditions. Journal of Speech, Language, and Hearing Research, 52, February 2009, 188-205.
  7. Ingham, R., Bothe, A., Jang, E., Yates, L., Cotton, J., & Seybold, I. (2009) Measurement of Speech Effort During Fluency-Inducing Conditions in Adults Who Do and Do Not Stutter. Journal of Speech, Language, and Hearing Research, 52, October 2009, 1286-1301.