and structure of recall of presented melodies between musicians and non-
musicians, which he attributes to shared extensive experience with music. Allard
and Starkes (1991) show that superior recall of briefly presented game situations
by elite players, as compared with intramural players, is not always found
in sports with speed stress, such as volleyball. Finally, Patel and Groen (1991)
demonstrate that levels of medical expertise have nonmonotonic relations to
the amounts of information recalled from presented medical cases, which they
attribute in part to the ability of experts to efficiently identify the information
relevant to the medical diagnosis. These findings show that superior memory
performance is not an inevitable consequence of attaining expertise.
It is thus questionable that a collection of tasks to measure the superior
memory of experts can be claimed to really capture the expertise in question.
With the exception of experts on memory tasks, superior performance by
experts in many domains does not include explicit tests of memory perfor-
mance. Moreover, there is no reason to believe that experts explicitly train with
the goal of increasing their memory performance. It is therefore unlikely that
their memory performance would have reached a stable maximum. We shall
later discuss in more detail the cognitive processes relating memory perfor-
mance and expertise.
An issue shared by studies of superior memory performance and studies of
superior performance in other realms is the problem of determining the stimu-
lus characteristics necessary to evoke performance in the laboratory analogous
to real-life expertise.
Finding the Appropriate Stimuli to Evoke Superior Performance.Incapturingex-
pert-level performance, one attempts to create a situation that is maximally
simple and yet sufficiently similar to the real-life situation to allow the repro-
duction of the expertise under laboratory conditions. The mere demonstration
that an expert-level performance can be reproduced under controlled labora-
tory conditions reveals something important about the mechanisms underlying
the corresponding expertise. It reduces the number of possible stimulus vari-
ables that are critical to performance, and it can also eliminate a number of
systematic covariations that would make the real-life performance much easier
than it would initially appear. Despite the critical importance of the process of
finding appropriate stimuli to evoke superior performance, that process has
rarely been documented. Ericsson and Polson (1988a, 1988b) investigated the
ability of expert waiters and waitresses to match meal orders to customers.
They reproduced under laboratory conditions the superior memory perfor-
mance related to dinner orders by simulating actual customers with photos of
faces. Similarly, Bennett (1983) reproduced the superior memory performance
related to drink orders by cocktail waitresses in a simulated situation with dolls
representing customers. Hence, highly schematic stimuli are sufficient to elicit
the perceptual and representational mechanisms that mediate superior memory
performance. Similarly, Chase and Simon (1973) found that the memory per-
formance of two chess experts did not differ for chessboards with real pieces
and schematic diagrams of chess positions, whereas a beginner at chess showed
poorer recall with schematic diagrams because of lack of familiarity with the
diagram notation. When they exposed a chess expert to an unfamiliar type of
Prospects and Limits of the Empirical Study of Expertise 531