Formation and utilization of the visual and verbal codes of pictures and words.

Abstract

Paivio's dual coding hypothesis proposes that the retention of pictures and concrete words can be mediated by both visual and verbal codes. The present study investigates the time course of the formation and utilization of the visual and verbal codes of pictures and words in a “same”-“different” RT-task involving physical (Picture-Picture; Word-Word) and semantic (Word-Picture; Picture-Word) matches. The ISI was varied between 50 and 2000 msec. The four matching conditions were either constant within a block of trials (Exp. I) or partially mixed (Exp. II). Exp. I showed that at long ISIs (>500 msec.) the comparisons in all matching conditions were based on visual codes; Ss either retained the visual code of the first stimulus (physical matches) or generated the visual code of the expected second stimulus (semantic matches). Under the conditions of Exp. II the visual code of the first stimulus was not actively retained. Except for very short ISIs (≤250 msec.) only a single code, the verbal code was used for comparisons of all stimulus combinations. Paivio's dual code hypothesis thus must be modified to account for code utilization in a matching task. The comparison between “same” RTs of the physical match conditions from both experiments suggests that two stages of visual storage were involved in the task, viz. the visual sensory store for ISIs ≤250 msec. and the visual short-term store for ISIs ≥ 500 msec.