Vygotsky argued that talking to yourself shapes thought. Thinking begins as a linguistic collaboration with other individuals, which gradually becomes internalized into a private dialogue with the self. When you see a child talking to herself during an episode of play or puzzle-solving, you are witnessing just such a private dialogue. As development progresses, these private dialogues become steadily less noticeable to an external observer.
Vygotsky's theory thus has big implications for our understanding of inner speech. In a separate post I noted that the study of children's private speech gives us a window onto some of the transformations that language undergoes when it becomes internalized. It also sets up the possibility that any particular child's use of private speech will be connected to their use of inner speech (the two are developmentally related, after all). The problem is, inner speech is covert. No one knows when you are doing it, or what you're saying to yourself when you are doing it. How are we to have a science of this unobservable phenomenon?
One option is to make an educated guess about when inner speech might be happening, and then try to block it. One venerable method for blocking inner speech is known as articulatory suppression, and it involves participants repeating an innocuous word (such as Monday or see-saw) out loud for the duration of the trial. In cognitive psychology, articulatory suppression is thought to block the systems that process phonological information (particularly the phonological loop component of the working memory system), systems which need to be functioning in order for internal and external speech to be produced. Asking participants to do articulatory suppression while they are performing a separate cognitive task, and assessing whether this affects their performance on that primary task, is a useful way of examining how much people rely on inner speech in certain contexts.
My doctoral student Jane Lidstone was keen to find out whether blocking children's self-directed speech through articulatory suppression affected their performance on a planning task. Planning is of interest to developmental psychologists for many reasons, but not least because it draws on neural resources in the prefrontal cortex, generally not thought to be fully developed until around age 5. Vygotsky thought that planning, as one aspect of the self-control of behavior, was a particularly important function for private speech.
Earlier findings from our lab  had shown that children who used more self-regulatory private speech were more susceptible to a pattern of memory errors (the phonological similarity effect) which indicates the use of verbal strategies in remembering information. Jane wanted to establish whether children who produced more self-regulatory private speech would perform at a lower level on a primary task when they were not able to use such speech (either covert or overt), as Vygotsky's theory would predict. So she picked a planning task, the Tower of London, which is known to stimulate private speech in this age group [2, 3]. In the version that she used, children had to move colored disks around on a frame made of three sticks of different lengths, to match a pattern she had given them. Using the standard articulatory suppression paradigm, Jane asked kids to repeat a word out loud to themselves as they thought about the task.
The findings of her two experiments were published recently in the Journal of Experimental Child Psychology . In the first experiment, 30 children aged 7 to 10 did either articulatory suppression or a control task (foot tapping) while solving Tower of London puzzles. Jane measured the time it took children to solve a certain number of puzzles, and how many moves it took them to solve each trial. All the sessions were videotaped, allowing us to get measures of private speech on the trials where children were not doing articulatory suppression.
Against our hypotheses, articulatory suppression led to no decrement in performance, compared to a condition where children had to perform a control task (tapping their foot). This suggested that inner speech was not functionally involved in this planning task. Crucially, though, we observed that children typically started moving the disks around on the sticks as soon as they were free to do so, which might have minimized any planning they might otherwise have done.
Before we gave up on the Vygotskian hypothesis, Jane ran a second experiment. This time, she tried to ensure that children thought about the task trial before they tried to solve it. The idea was that this would encourage the children to plan, instead of simply launching in and moving the disks around haphazardly. Another 30 kids of the same age took part, but this time they were asked to imagine moving the disks around in their heads (the planning phase), and to tell the experimenter how many moves they thought they could solve the puzzle in. They were then asked to demonstrate their solution by moving the disks around for real. In this experiment, the secondary tasks were only performed during the planning phase, and the same applied to the private speech measures.
This time, the results were more supportive of Vygotsky's theory. On the articulatory suppression conditions, children's performance was reduced relative to the control (foot-tapping) condition. We interpreted this as evidence that private and inner speech were typically used in the planning component, and that blocking both kinds of speech had a corresponding effect on planning.
There were also interesting findings relating to individual differences. Children who used more self-regulatory private speech on the control condition were more susceptible to articulatory suppression. Following previous theory, we predicted that this relation would only hold for trials that were neither too easy nor too difficult, and this turned out to be the case. We concluded that at least some children are heavily reliant on verbal mediation for solving this kind of task, and are thus more hampered by the blocking of self-directed speech.
Jane's findings fit with a growing body of work which has supported Vygotsky's view of the developmental function of private speech. It also raises some questions for future research. It is essential to bear in mind that no one has been able to observe inner speech directly. When we draw out inferences about the use of inner speech, they are just that: inferences. It is possible that future researchers will be able to back up these kinds of findings with subjective reports from children about the strategies they used, verbal or otherwise, although it is always going to be a challenge to get kids so young to report reliably on their own inner experience.
It's also interesting to note that the Tower of London, so often considered to be a classic 'executive' task with a heavy loading on prefrontal regions, has such a strong verbal component. Another recent study  showed a similar suppression-related decrement in Tower of London performance in adult participants. From a Vygotskian perspective, however, this involvement of verbal mediation makes perfect sense. We do our planning at least partly in language, whether it is overt (social speech), semi-internalized (private speech), or the kind that only we ourselves can hear.
1 Vygotsky, L. S. (1987). Thinking and speech. In The Collected Works of L. S. Vygotsky, Vol. 1. New York: Plenum. (Original publication 1934).
2 Al-Namlah, A. S., Fernyhough, C., & Meins, E. (2006). Sociocultural influences on the development of verbal mediation: Private speech and phonological recoding in Saudi Arabian and British samples. Developmental Psychology, 42, 117-131.
3 Fernyhough, C., & Fradley, E. (2005). Private speech on an executive task: Relations with task difficulty and task performance. Cognitive Development, 20, 103-120.
4 Lidstone, J. S. M., Meins, E., & Fernyhough, C. (2010). The roles of private speech and inner speech in planning in middle childhood: Evidence from a dual task paradigm. Journal of Experimental Child Psychology, 107, 438-451.
5 Williams, D., Bowler, D., & Jarrold, C. (in press). Inner speech is used to mediate short-term memory, but not planning, among intellectually high-functioning adults with autism spectrum disorder. Development and Psychopathology.