Abstract

The immediate serial recall of lipread material is disrupted by irrelevant speech (Campbell & Dodd, 1984; Jones, 1994) and by articulatory suppression (Campbell & Dodd, 1984). However, the interaction between these has not been directly examined. In Exp 1, participants recalled silently presented lipread digits in conditions of quiet, irrelevant speech, articulatory suppression and suppression/speech combined. Irrelevant speech disrupted recall, but not when articulation was suppressed. Experiment 2 demonstrated that participants were able to accurately lipread targets in all of the above experimental conditions. A third experiment contrasted predictions derived from the phonological loop model (PL; Baddeley, 1986) and the changing state hypothesis (CSH; Jones, 1993). The CSH predicts that tones and speech that vary in frequency to the same degree will disrupt recall equally (Jones & Macken, 1993), whereas the PL model implies that speech will always be more disruptive. The results support the CSH, and extend the findings of Jones and Macken (1993) to account for lipread stimuli. As with graphic presentation, the CSH provides a better account of the processes underpinning the irrelevant speech effect; however, it is argued that the recoding hypothesis from the PL model should be retained.