On the basis of such arguments, recent discussions of the VSSP have increasingl emphasized the analogies between the characteristics of the two slave systems (Baddeley, 1986, 1990; Logie, 1989).
The extent to which eye movements are implicated in such active rehearsal of material in the VSSP is questionable, and Baddeley himself admits that other types of rehearsal mechanism are conceivable and may indeed be more plausible (Baddeley, 1986).
One apparent problem for this two-component account of the VSSP is posed by the Baddeley & Lieberman finding that performance on the Brooks spatial matrix task was unaffected by a concurrent visual task.
The original argument put forward by Baddeley & Lieberman was that the VSSP employed amodal spatial representations.
So it is possible that the VSSP might be conceptualized as two separable subsystems: a 'spatial scratchpad' and a 'visual scratchpad'.
He argued that these results could be interpreted in terms of a couplin of the executive and VSSP during active processing, leaving the VSSP open to general disruption during encoding and retrieval of items, but not during more passive maintenance of material in the VSSP.
It should be noted that Morris' (1987) original arguments included the postulat that the maintenance of visuospatial material might not be immune to interference from secondary tasks directly utilizing the resources of the VSSP, rather than the more general resources of an executive.
However, bearing in mind the earlier caveats regarding the brightness judgement task, this would be consistent with Morris' earlier finding that the VSSP is not susceptible to general interferenc during maintenance rehearsal.
The present paper investigates whether a primarily spatial task is susceptible also to visual interference, and explores the characteristics of visual materia which are necessary for such material to gain obligatory access to the VSSP.
Should, however, the visual input interfere equally with performance of spatial and verbal tasks, then an explanation in terms of competition for the resources of the VSSP would not be tenable.
This extends the findings of Logie (1986) to another primary task that is also assumed to require the resources of the VSSP, and indicates that at least one locus of such interference is during encoding of visuospatial material.