Andrew J. O. Whitehouse and Murray T. Maybery
University of Western Australia
Kevin Durkin
University of Strathclyde

Introduction

The functions of language are not restricted solely to external, interpersonal communications (Miyake, Emerson, Padilla, & Ahn, 2004). Covert, self-directed language plays an important role in cognitive processes and the regulation of behaviour (Carruthers, 2002). However, while difficulties with behavioural regulation (i.e., executive dysfunction) are well established in individuals with autism, very few studies have examined inner speech.

Task switching is one responsibility of executive function that is known to require inner speech. The current experiment employed a task-switching paradigm (Baddeley, et al., 2003), based on arithmetic, for which performance has been shown to be contingent upon inner-speech. If, as hypothesized, individuals with autism have deficits in their use of inner speech, then it would be expected that attempts to obstruct inner speech, through the addition of articulatory suppression (AS), would not significantly affect their performance during a task switching condition. font>

Method

Participants

Twenty three high functioning children with autism (M chronological age = 11;0) were compared with 23 typically developing children (M = 8;4) matched on verbal, non-verbal and reading ability.

Procedure

Participants were given slips of paper, containing 20 mathematical ‘equations’, that omitted the function and equal signs (i.e., 4 1 ____ ). The participants were asked to complete the mathematical problems as quickly and accurately as possible. There were three independent variables.

1. Task (block vs alternate) was a within groups variable, that related to how the participants were required to view the mathematical problems. Specifically, the block condition asked the participants to complete the list under the pretense that all of the problems were addition. The alternate condition asked the participants to complete the problems under the pretense that the first problem was addition, the second subtraction and so on (task switching).

2. Verbal Output (silent vs AS) served as the other within-groups variable. In the AS condition, participants were asked to repeatedly say ‘Monday’ in time with a metronome. In the silent condition, participants were given no instructions as to concurrent articulation.

3. Group (autism vs control) was a between-subjects variable.

Time taken (in seconds) to complete each list served as the primary dependent variable.

Results

As can be seen from Figure 1, task switching added a significant performance cost. More importantly, there was a significant Group x Verbal Output interaction, and the three-way interaction was also significant. Follow-up comparisons revealed that the control group took significantly longer to complete both the block and alternate conditions when AS was imposed. In contrast, the imposition of AS did not affect the performance of the group with autism in either the block or alternate conditions. Mean latencies for the two groups did not differ for any of the experimental conditions except the alternate-AS condition, for which, intriguingly, the control group took longer to complete the task (99.7 s) than did the autism group (76.2 s).

Figure 1. The mean time taken by the autism and control groups to complete each task (Error bars represent one SE of the mean).

Conclusion

Inner speech is known to be important in task switching performance, and hence disrupting inner speech during task switching should create a performance cost. While the addition of AS significantly affected the task-switching performance of the ability matched control group, it did not affect the performance of those with autism. This provides evidence for impaired use of inner speech in those with autism.

Future Directions

1. Ascertaining the exact nature of the inner speech deficit. For example do individuals with autism have a lack of inner speech in those with autism, a delay in the development of inner speech, or simply poor awareness of how to use inner speech.

2. How does the inner speech impairment relate to the prevailing cognitive theories of autism?

3. Can therapeutic benefit could be gained by teaching inner speech skills?

For more information please do not hesitate to contact Andrew Whitehouse at whitea08@student.uwa.edu.au