Wednesday, November 18. 2009

Verbal Behavior: 3 Key Strategies to Teaching Speech

Posted under: Research

Teaching children to talk (i.e., speech) is one of the primary aims of behavioral treatment in the first year. While focusing on functional communication through requesting (i.e., manding) is an important part of the process, there are other key strategies that often play a role in helping children who are primarily nonverbal become vocal speakers. Here they are:

  1. Pair ANY sound with reinforcement. I've worked with families who have carried a small fanny pack around with them, filled with their child's favorite food and toys. Whenever their child sporadically babbled throughout the day, they would rush over to the child and reinforce them with attention, praise, and even toys or food. This increased the number of times the child babbled throughout the day. For those children who vocalize infrequently, every sound they make is a learning opportunity, and providing reinforcement for every sound is one way to increase the frequency of those sounds.
  2. Pair reinforcement with a specific sound. Research has demonstrated that pairing a sound with reinforcement increases the likelihood the child will make the sound again later. With one three-year-old, we paired the sound "eee" with going down the slide. We would help the child go down the slide five to ten times in a row, each time saying "eee." Within a week, he was spontaneously saying "eee" throughout the day.
  3. Focus on imitation. First, make sure a child can imitate gross motor actions (e.g., jump, clap hands, push car). These are easier to prompt and easier for a child to discriminate. Then, focus on fine motor imitations that are less distinct (e.g., touch two fingers together, put thumbs up, etc.) This requires observation of more subtle cues as well as practice with fine motor control. Next, move to oral/facial motor imitation (e.g., open mouth, blink, tongue out, etc.). A child is now learning to focus primarily on your oral motor movements. Finally, make a sound while producing an oral motor movement the child has already learned to imitate (e.g., open your mouth and say "aahh," close your mouth and say "mmm," etc.). This gradual progression allows for steady advancement whereas requiring a child to say "ooo" when he wants juice rather than using sign language may be too difficult of a transition.

During the course of treatment, each child may demonstrate challenges in learning. For many children, one of the first big hurdles is learning to talk. A good behavioral treatment program does not focus on only one strategy to teach speech, but rather uses the wide range of strategies available in applied behavior analysis to create a comprehensive program that meets the individual child's needs. With that being said, are there any additional strategies you would add to this list?


This is related to manding and it dramatically increased my son's vocalizations: when manding for reinforcers, use up to a five second time delay to evoke a vocal response. So my son would sign for a specific item or activity, we would withold the reinforcer until we heard a vocal response, then we would differentially reinforce with greater magnitude. Shortly thereafter, he vocalized almost every time when manding.

The above suggestions are very helpful. I notice you mention that programs should use a variety of approaches to teach language. Why don't you mention mand training as one of the helpful keys as this has been shown to be very motivating and effective for most children. I don't know that there is research comparing the three methods listed above to mand training but I am sure all four are components you would want in a program. Could you elaborate on why mand training wasn't including in the three KEY strategies?

Megan DeLeon, M.S., BCBA

Another method to encourage language is by pairing vocalizations with a choice board when a child is manding. The child can begin by selecting a picture from the choice board to mand for preferred items. Once the child is successful in using the choice board, a vocalization paired with the picture can be required before reinforcement (the requested item) is provided. Approximations of the vocalizations can be accepted and differentially reinforced as the child begins to increase his/her repertoire of sounds and words.


Great question. I mention manding briefly in the first paragraph, but I chose 3 key strategies that I thought some individuals may not be familiar with. In my experience, the two strategies most people are familiar with are discrete trials to teach verbal imitation (echoic response) and incidental teaching or mand training to teach requesting (mands). The 3 key strategies I mentioned weren't meant to be an exclusive list, which is why I asked for additional strategies, and why I'm glad you responded...some people may NOT be familiar with the benefit of mand training to teach vocal language. Thanks to Suzy and Dagmar for flushing out some details on how mand training might work. There's still one other strategy that comes to mind when I'm trying to teach speech to a child? Anyone else have ideas? I'll post my other idea here after Thanksgiving.

Thank you for clarifying why mand training was not discussed as a key. Can you clarify what type of language (tacting, manding, intraverbals)you are referring to when you say you have one additional strategy in mind? If you are referring to language in general, I can think of a variety of strategies but the one that seems to most relate to what has been discussed on this blog so far would be the Automatic Reinforcement Procedure (Sundberg, Michael, Partington, & Sundberg, 1996)/Stimulus Stimulus Pairing which might be the same thing you are talking about in number 2. I have seen it done while doing something reinforcing (which I think would be more motivating) but the research also shows that it is effective in more artificial settings where the therapist says a sound and pairs it with a reinforcing item such as a piece of candy, toy, or other food item. The second way might be effective for children who do not find activities reinforcing but do have a preference for food, toys, videos, etc.

I can think of 2 other strategies that are used to promote verbal behavior.
Teach the child to expressively label his or her environment through tacting. This is when the child simply labels items (e.g., "Truck!") rather than requesting the said item. The team would promote the use of tacting and reinforce accordingly. This helps to develop spontaneous language.
Another strategy to promote verbal behavior is through the use of intraverbals. This is when you say, "I'm going to open the..." and the child finishes the sentence (e.g., "door"). Intraverbals can also come in the form of Wh- questions. In this situation you could ask the child, "What is something you play with?" without any toys present. The child would then answer with a specific toy (e.g., "cars"). Intraverbals allow the child to search for the appropriate answer within their repertoire of language. Imitation needs to be established prior to teaching both tacting and intraverbals so prompting is possible.

Thanks for everyone's insights. Keep them coming!
The last strategy that always comes to mind for me when teaching speech is the relevance of receptive language in the process. Let’s be clear, some children with autism come to treatment already able to produce some words or quickly begin to vocalize through mand training and verbal imitation. However, for those children who have a more difficult time learning speech, a behavioral treatment program can do more than simply continue with mand training and verbal imitation. Some other strategies have already been outlined above by both me and others who responded to this blog. Receptive language programs (e.g., SD: The instructor says, “ball” with 3 objects in front of the child. R: The child touches the ball.) can play a role in speech development by helping answer the question, “Does this child also demonstrate difficulty discriminating between the words I am making?” A child who demonstrates difficulty with speech and who also cannot receptively identify “ball” vs. “Mickey” may be different from a child who can receptively identify “ball” vs. “Mickey” but still demonstrates difficulty with speech. I know the second child at least “hears” the words I am saying. That is, they demonstrate by touching different objects that they are discriminating between the sounds I make.

Further, I have found that some children with emerging speech will begin spontaneously imitating the words delivered by the instructor during receptive programs. (For example, the instructor says, “ball.” The child repeats “ball” and touches the ball.) One benefit to this spontaneous speech is that it can develop without having to worry about tolerance issues associated with mand training (i.e., in mand training, one must be careful not to require too difficult a sound or series of sounds from a child when they demonstrate interest in an object or the child can begin to whine or tantrum). Because of the possibility that a child will begin to spontaneously imitate the instructor’s words, I think it is better in early receptive language programming to only name the object (e.g., “ball”) rather than say “touch ball” or worse yet, vary the words in the SD (e.g., “touch ball,” “where’s the ball,” “find the ball,” etc.).

In summary, emerging speech can be a complicated process. Much emphasis in recent years has been given to mand training as an important initial component in teaching verbal behavior. However, verbal behavior is different from vocal behavior. Strategies to teach verbal behavior will not always result in vocal behavior. There are a number of other key strategies that can help in the development of vocal behavior. However, this topic remains a prime area for further research.

Vince, thanks for that comment about the value of teaching selection/receptive language. My child had severe apraxia and it took literally years to develop vocal language, even with intensive speech therapy and motor imitation through the entire hierarchy including oral motor and vocal imitation. For reasons of rapid development of replacement adaptive behavior to compete with significant problem behavior we needed to use PECS after unsuccessfully trying to implement sign. To make a long story short, while vocal mand training had been somewhat unsuccessful, a breakthrough occurred when echoics began spontaneously occurring during receptive language drills, and which were differentially reinforced. In a fairly short time when those echoics became routine we were then able to cross transfer to tact instruction. It was after those echoics solidified and became reliable that vocal mand training was really able to take off and be worked on in a big way with better success. It's not what one would have anticipated considering what is known about motivation, but it's the way that it went.

Post Your Comment

Comments are moderated, and will not appear on this weblog until the author has approved them.

Enclosing asterisks marks text as bold (*word*), underscore are made via _word_.
Standard emoticons like :-) and ;-) are converted to images.
E-Mail addresses will not be displayed and will only be used for E-Mail notifications

Enclosing asterisks marks text as bold (*word*), underscore are made via _word_.
Standard emoticons like :-) and ;-) are converted to images.
E-Mail addresses will not be displayed and will only be used for E-Mail notifications