Motor Learning and Control

Watch a stutterer struggle to talk. You see overtense, overstimulated respiration, vocal folds, and articulation (lips, jaw, and tongue) muscles. Brain scans of adult stutterers have found overactivity in the left caudate nucleus speech motor (muscle) control area, during stuttering.

Fluency shaping therapy reduces speech motor activity. It trains stutterers to speak slowly by stretching vowels. It teaches speech with relaxed respiration, vocal folds, and articulation.

Motor Learning and Control

Motor learning and control is how brains execute complex muscle movements. Physical therapy and occupational therapy students study motor learning and control.

Sports coaches also study motor learning and control. The principles of motor learning and control are usually illustrated with examples from gymnastics, tennis, golf, or other sports.

You can study motor learning and control by reading a textbook:

Closed-Loop Motor Control

A muscle movement takes about 200 milliseconds (one-fifth of a second) to execute:

  1. Sensation, or neural transmission from sensory receptors in your eyes, ears, etc., to your brain, takes about 15 milliseconds.
  2. Perception, which retrieves long-term memories to organize, classify, and interpret your sensations, takes about 45 milliseconds. Perception changes sensation data into perceived information or meaning.
  3. Response selection takes about 75 milliseconds. You use current perception and past experiences to formulate a course or action. For example, in baseball, a batter watches the pitcher and decides whether to swing at a pitch, hit or bunt, hit to left field or right, etc. Psychologists differentiate conscious decisions from unconscious translations, or relating a particular stimulus to a particular response.
  4. Response execution of an action plan—a step-by-step sequence of events that make up the planned movement—takes about 15 milliseconds. In these events, motor neurons carry signals from the brain or spinal cord to muscles.

Under closed-loop motor control you use perception to consciously, continuously adjust muscle movements. For example, threading a needle. You look at the needle. You look at the thread. You move the thread towards the needle. You look at the needle again. You look at the thread again. You correct your movement. You do this many times until the thread is through the needle.

Each stimulus-response adjustment takes at least 200 milliseconds (one-fifth of a second). If you make ten adjustments, the task takes at least two seconds.

Closed-loop motor control has two advantages. It enables precise control, and it enables execution of novel movements (activities you’ve never done before). For example, threading a needle on the deck of a rolling ship.

Closed-loop motor control has two disadvantages. It’s slow, and it requires your full attention.

Closed-loop motor control is good for learning new skills, or for executing skills you rarely need. But you don’t want to use closed-loop motor control for fast-paced, frequently used skills.

Open-Loop Motor Control

200 milliseconds—a split second—may seem fast, but it’s too slow for many motor tasks. For example, a gymnast’s double-back somersault requires muscle movements lasting only tens of milliseconds.

How is it possible to execute a muscle movement in tens of milliseconds, when the sensation to execution cycle requires about 200 milliseconds? Simple—don’t do the sensation, perception, and response selection stages. Just do the response execution. This final stage of muscle movements can be performed in as little as 15 milliseconds. This is called open-loop motor control. Open-loop motor control is the execution of preprogrammed movements, called a motor program, without perceptual feedback.

The colloquial term for this is “muscle memory.” For example, gymnasts practice hours each day for years, until their muscles seem to know what to do without the mind getting involved.

After winning the gold medal in gymnastics at the 1984 Olympics, Mary Lou Retton said that coach

Bela [Karolyi] can really teach, I’ve learned so much from him. Many long hours were spent in the gymnasium …repetition, feedback, repetition, and experimentation. Somehow, after a lot of bumps and bruises, it got easier, as if I could float.

Karolyi added,

Someone should be able to sneak up and drag you out at midnight and push you out on some strange floor, and you should be able to do your entire routine sound asleep in your pajamas. Without a mistake. That’s the secret. It’s got to be a natural reaction.

Open-loop motor control has two advantages:

  1. It’s fast. You can execute muscle movements with split-second timing.
  2. It requires no attention. Movements under open-loop control are automatic and mentally effortless.

Open-loop motor control has three disadvantages:

  1. If your motor program contains errors, you’ll execute the errors. You can’t stop and adjust a mistake. You may not even be aware that you made a mistake.
  2. Developing open-loop control of a motor skill requires long practice—especially for adults. Children learn some motor skills easily, that adults struggle for years to master.
  3. Novel or new situations can’t be handled. For example, in the 2000 Olympics, officials set the gymnastic vault two inches too low. The officials didn’t correct the height until 18 of the 36 women had performed. These 18 athletes performed poorly, eliminating their hopes of winning medals. The American hopeful, Elise Ray, suffered a “devastating fall.” 1

Learning New Motor Skills

Use closed-loop motor control for learning a new motor skill. Then gradually increase your speed until you can perform the motor skill using open-loop motor control.

For example, a tennis or golf coach will have you start with swinging the club or racquet slowly, while she adjusts your knees, elbows, etc. When you’ve perfected your form, your coach will have you gradually increase the speed and force, while maintaining form. After extensive practice you’ll be executing perfect open-loop motor programs. You’ll smash the ball hard and fast and accurately without paying attention to your elbows or knees or anything other than the ball.

Speech Motor Control

Normal speech uses open-loop motor control:

  1. Speech is fast. Phonemes (speech sounds) are typically 20 to 40 milliseconds.
  2. Speech is complex, requiring coordination of hundreds of muscles to produce sounds.
  3. Speech is automatic and effortless. Speakers think about what they’re saying, not about the muscles they’re moving.

Fluency shaping stuttering therapy uses closed-loop speech motor control. You consciously relax your breathing. Then, as you exhale, you slowly increase your vocal fold tension, until your vocal folds hum. Then you slowly move your lips, jaw, and tongue to form the sounds of each word. Stuttering is impossible when using closed-loop speech motor control. Stuttering disfluencies are open-loop speech motor programs.

Making stuttering impossible might sound appealing, but

  1. Closed-loop speech motor control is slow. Closed-loop motor control takes about 200 milliseconds per muscle movement. Open-loop speech sounds are typically in the 20-40 millisecond range. Closed-loop speech motor control slows speech five to ten times, or one or two seconds per syllable.
  2. Closed-loop speech motor control demands your full attention. You must pay attention to your breathing, vocal folds, and lips, jaw, and tongue. This isn’t a problem when reading a list of words, but is difficult to use in conversations.
  3. Your speech loses prosody (emotional intonation). You sound like a robot with dying batteries.

A fourth possible problem may be that stutterers learn speech motor skills slower and retain them less than non-stutterers. 2

Prosody, Parameterization Schemata, Response Selection

Why closed-loop speech motor control loses prosody is an interesting question.

A study of television talk show guests found that 94% of what viewers remembered was prosody, or what actors call emoting, or what lawyers call affect. 3 Much—or almost all—meaning is communicated by prosody. Schemata theory suggests that you learn certain invariable characteristics of a motor skill, and you learn certain execution rules or parameterization schemata. You then combine the invariable elements with the rules to produce a motor plan.

For example, in a public speaking class I read algebra problems in an angry voice, in a sad voice, and then with the rhythm and emotional intonation of a stand-up comedian. The algebra problems were invariable—I read the same algebra problems each time. I changed the parameterization schemata to communicate different emotional states. Amazingly, the audience laughed at the “punchlines” when I did the stand-up comedy delivery. Even though the “punchlines” were just numbers, I made the audience think that a punchline was coming, and they laughed at the right times. 94% of the joke was the delivery.

Accents are another parameterization schema that conveys meaning. For example, a waitress from Oklahoma asked me if I wanted ah-iss. When I figured out that she was asking about ice, I affirmatively answered yay-iss. I knew the invariable characteristics of “yes,” and when I’d learned the rules of an Oklahoma accent—e.g., break monosyllabic words into two syllables—I was able to say a word I’d never heard.

In normal speech, we produce prosody through unconscious response selection of parameterization schemata. Different environmental cues cause us to select different responses. For example, you walk into a church and immediately lower your vocal volume. But if no one else is in the church, you could yell “I hate to wear pants!” while turning somersaults down the aisle. OK, that’s one of my eccentric hobbies, but most people wouldn’t do that.

Another example is a person who grew up spending summers in Vermont and winters in Georgia. When she’s in New England she speaks in a Yankee accent. When she’s in the South she switches to a southern accent. Different environmental cues cause her to unconsciously select different parameterization schema to produce each accent.

Like prosody and accents, stuttering is a parameterization schema. A stutterer responds to environmental cues to unconsciously select fluent speech parameters or stuttering speech parameters, which are then combined with invariable characteristics of words to produce fluent or stuttered speech. Thus you can treat stuttering by training stutterers to respond differently to environmental cues (“Responding to Stress,” page 154), or by training stutterers to use fluent speech parameterization schema (this chapter).

Training a stutterer to not feel fear or anxiety when answering the telephone is changing the response selection to an environmental cue (a ringing telephone). In contrast, training a stutterer to speak with relaxed vocal folds changes a speech parameter.

Snake Oil and Charlatans

Closed-loop speech motor control is the “wizard behind the curtain” of many stuttering therapy programs. Switch any stutterer to closed-loop speech motor control and he or she will speak fluently.

You can switch to closed-loop speech motor control by making any speech process conscious instead of unconscious. For example, focusing on relaxed, slow breathing will switch you into closed-loop speech motor control, with your vocal folds and articulators (lips, jaw, and tongue) following right along. Or you can focus on producing “gentle onsets” with your vocal folds. This will switch your breathing and articulators to closed-loop speech motor control. Or you can focus on “reduced articulatory pressure” and your breathing and vocal folds will follow.

Always these “wizards” claim that their therapies are 100% effective if the stutterer “really tries,” that is, if he devotes his full attention to closed-loop speech motor control. If he instead pays attention to a conversation, switches into open-loop speech motor control, and then stutters, then he wasn’t “really trying.”

And the closed-loop speech motor control effect has caused speech-language pathologists to hypothesize that stutterers have something wrong with their breathing, or with their vocal folds, or with their articulators, or even that stutterers’ brains are slow in some way. Everyone performs slowly when attentively learning a new motor skill, then their speed improves with practice. For stutterers in speech therapy, the new motor skill is fluent speech.

Some speech clinics tell stutterers that they’ll always have to speak slowly. That’s like teaching a student driver to drive 20 mph, then telling him never to go faster.

Slow Speech Is Not the Goal of Stuttering Therapy

If you learn tennis or golf, you’ll use closed-loop motor control when you’re learning to swing the club or racquet. As you practice, increasing your speed and force, you’ll gradually reinforce open-loop motor programs.

Similarly, you’ll use closed-loop speech motor control when working with your speech-language pathologist. She’ll train you to move your respiration muscles, vocal folds, and articulators correctly to produce fluent speech. When you’ve mastered this at a very slow speaking rate, she’ll help you to gradually increase your speaking rate, while staying fluent. The goal is fluent, automatic, effortless, normal-sounding and normal-rate speech. Slow speech is not the goal of stuttering therapy.

Severe stutterers usually don’t mind learning closed-loop speech motor control. If your stuttered speech is ten to twenty times slower than normal speech, then closed-loop speech motor control, which is typically five to ten times slower than normal speech, will double your speaking rate. Some severe stutterers are even willing to use closed-loop speech motor control outside of the speech clinic. Record conversations with and without using closed-loop speech motor control. Count your syllables per second. You may find that closed-loop speech motor control feels slower but is actually faster than your stuttered speech.

But mild stutterers don’t like closed-loop speech motor control. They can hide their stuttering by avoidance and substitution (of certain sounds, words, or speaking situations). They can sound fluent at a normal speaking rate. Closed-loop speech motor control would “advertise to the world” that they have a speech disorder. If they’re embarrassed to admit that they stutter, they won’t want to use closed-loop speech motor control.

Mild stutterers should consider that closed-loop speech motor control enables them to say anything they want. For example, a mild stutterer wants to buy a chess set. He’s afraid of s words, so he calls a toy store and asks if they have “one of those games with kings and knights and castles.”

The puzzled clerk responds that the store has many games with kings and castles and knights. After five minutes of conversation, the clerk asks, “Do you mean chess sets?” The stutterer says yes. The clerk never knows that the caller is a stutterer, but she thinks that the caller is an idiot. The stutterer wasted five minutes because he wasn’t willing to use ten seconds of slow speech.

Or the stutterer drives to the store and looks for a chess set, without calling first. If the store doesn’t have chess sets he wastes an hour, to save ten seconds. Saying what you want slowly is faster than saying something else, or not speaking at all.

Americans speak around 165 words per minute. Fast talkers who speak more than 190 words per minute get complaints from listeners unable to understand them. In contrast, Walter Cronkite trained himself to speak 124 words per minute in his newscasts. “Uncle Walter” may have earned his title as “the most trusted man in America” in part because he spoke slowly and clearly.

Analogy to Touchtyping

I’ve never taken a typing class. I type with two fingers, about 45 words per minute. (I may be the world’s fastest two-fingered typist!)

I tried to learn touchtyping. My speed dropped to less than ten words per minute. Touchtyping not only slowed me down, it required my full concentration. I couldn’t think about what I was writing, only about moving my fingers.

I gave up touchtyping within a week. If I’d kept at it, my speed would have increased and eventually surpassed my two-fingered typing speed. I might have been typing 80 words per minute now. The mental effort would have diminished, until touchtyping was automatic and effortless.

Coaches say they’d rather work with a novice who’s never played their sport, rather than with an experienced player who uses incorrect techniques. It’s easier to learn a new motor skill correctly than it is to correct an incorrect, deeply ingrained motor skill.

Stuttering is difficult to overcome because we learned to talk incorrectly. We have to learn new, fluent speech motor skills, and we have to not use our old, disfluent speech motor skills. We learned these disfluent speech motor skills in childhood, when our brains were growing. Now the disfluent speech motor skills are hardwired into our brains. Making fluent speech automatic and effortless, for a stutterer, demands more time and effort than learning a new sport or vocational skill.

Using DAF to Slow Speaking Rate

Many speech clinics use delayed auditory feedback (DAF) devices to establish fluency using closed-loop speech motor control. With only a little training a DAF device can help a stutterer maintain perfectly paced, steady, mentally effortless, slow closed-loop speech motor control.

The user’s speaking rate can be adjusted by turning a knob. A typical protocol is to train a stutterer to use closed-loop speech motor control with a 200-millisecond delay and one to two seconds per syllable. The stutterer practices this until he’s 100% fluent. That usually takes only a few therapy sessions. (A study found that without training a 195-millisecond delay reduced stuttering only 85%. 4)

When the stutterer can speak 100% fluently, the speech-language pathologist then has the stutterer use one- or two-second stretched syllables without the DAF device; in increasingly stressful situations (e.g., calling the speech-language pathologist’s answering machine); and then with the DAF device adjusted for faster speaking rates. The stutterer must stay on-target with 100% fluency, or go back to using the DAF device at 200 milliseconds and a one- to two-second speaking rate.

Typically, a 100-millisecond DAF delay is used with half-second per syllable stretched speech, a 75-millisecond delay is used with quarter-second per syllable “slow normal” speech, and a 50-millisecond delay is used with a normal speaking rate.

Three Stages of Motor Learning

We learn new muscle movements, or motor skills, in three stages:

  1. In the cognitive stage, an instructor demonstrates the motor skill to you.
  2. In the associative stage, you learn to perform and refine the motor skill. You perform the movements under closed-loop control.
  3. In the autonomous stage, the motor skill becomes automatic. You perform the muscle movements without mental effort, under open-loop control.

For example, imagine yourself learning golf or tennis. You watch the coach hit a few practice balls. Then the coach hands you the club or racket. The coach guides you through a swing, telling you to drop this shoulder or extend that forearm. Soon you can execute the swing perfectly, if you fully concentrate on each movement. You then practice the swing, and your game improves.

A few years later a novice admires your excellent swing and asks you to explain how you do it. “I don’t know,” you say, “I just do it without thinking about it.”

One summer I tried mountain bike racing. In four races I crashed four times. I then hired a coach. In twelve hours over three weeks, he taught me how to ride down hills, make tight turns, jump my bike over logs, climb hills, plus a few tricks such as picking up a water bottle off the ground.

Then I quit mountain bike racing. I’d completed the associative stage and learned how to do each skill. Now I would have to practice these skills hours a day, several times a week for months to make the skills automatic in the fast, high-stress environment of racing. In other words, I could do any of the skills if I thought about it, but my body didn’t automatically execute the moves without conscious mental effort. I decided that mountain bike racing isn’t important enough to me to spend so many hours practicing skills.

Stuttering therapy follows a similar course. A speech-language pathologist can show you the fluency skills—relaxed, diaphragmatic breathing; vocal fold relaxation (gentle onsets); and relaxed articulation muscles (lips, jaw, and tongue)—in ten minutes. Teaching you to execute these skills takes a few hours. You can then speak fluently in the speech clinic, when you mentally concentrate on each skill. Almost everyone successfully completes these cognitive and associative stages.

You then have to practice these skills thousands of hours to make them automatic and effortless, in high-stress situations. Many stutterers fail at this stage. But no one intentionally fails for the reasons I quit mountain bike racing. No one rationally weighs the alternatives and says, “Talking isn’t important to me. I’ll learn sign language instead, or write notes.”

Instead, stutterers fail at the autonomous stage because speech clinics don’t train this well. Speech clinics call this transfer. Perhaps your speech-language pathologist takes you to a shopping mall for an hour. But the autonomous stage requires thousands of hours of conversations, including high-stress conversations. Stutterers habitually avoid such conversations. You may find that the skills you learned in the low-stress speech clinic fail in high-stress conversations. Your therapy progress begins to fail. You revert to old habits and avoidances. Your stuttering returns.

This section has two more chapters. In the chapter Fluency Shaping Techniques we’ll do a deep dive into the cognitive and associative stages. Then in the chapter Beyond Fluency Shaping we’ll dive into the autonomous stage.

Submit your review

Create your own review

Casa Futura Technologies
Average rating:  
 0 reviews


  1. America Wins Olympics, October 2, 2000,
  2. Namasivayama, A.K.,  van Lieshout, P. “Investigating speech motor practice and learning in people who stutter,” Journal of Fluency Disorders, Volume 33, Issue 1, March 2008, Pages 32-51.
  3. Arielle Ford’s Complete Book Publicity Workshop.
  4. Stager, S., Denman, D., Ludlow, C. “Modifications in Aerodynamic Variables by Persons Who Stutter Under Fluency-Evoking Conditions.” Journal of Speech, Language, and Hearing Research, Volume 40, 832-847, August 1997.