Three experiments investigated the role of working memory in various a
spects of thinking in chess. Experiment 1 examined the immediate memor
y for briefly presented chess positions from master games in players f
rom a wide range of abilities, following the imposition of various sec
ondary tasks designed to block separate components of working memory.
Suppression of the articulatory loop (by preventing subvocal rehearsal
) had no effect on measures of recall, whereas blocking the visuospati
al sketchpad (by manipulation of a keypad) and blocking the central ex
ecutive (by random letter generation) had equivalent disruptive effect
s, in comparison with a control condition. Experiment 2 investigated t
he effects of similar secondary tasks on the solution (i.e., move sele
ction) of tactical chess positions, and a similar pattern was found, e
xcept that blocking the central executive was much more disruptive tha
n in Experiment 1. Experiment 3 compared performance on two types of p
rimary task, one concerned with solving chess positions as in Experime
nt 2, and the other a sentence-rearrangement task. The secondary tasks
in each case were both designed to block the central executive, but o
ne was verbal (vocal generation of random numbers), while the other wa
s spatial in nature (random generation of keypresses). Performance of
the spatial secondary task was affected to a greater extent by the che
ss primary task than by the verbal primary task, whereas there were no
differential effects on these secondary tasks by the verbal primary t
ask. In none of the three experiments were there any differential effe
cts between weak and strong players. These results are interpreted in
the context of the working-memory model and previous theories of the n
ature of cognition in chess.