improvement due to accumulated experience outside the laboratory. Ericsson
and Polson (1988b) found continual improvements in their expert waiter’s per-
formance of their standard task during about two years of weekly testing. It is
likely that part of the observed speedup resulted from the particular constraints
of the dinner orders studied. A more important determinant of the speedup,
however, was the fact that the real-life task of memorizing dinner orders was
not constrained by speed, because the customers required more time to decide
on their dinner selections than the waiter needed to memorize them. Only in
the laboratory situation with preselected dinner orders did the time required
for memorization become critical.
In sum, differences between real-life situations and analogous laboratory
tasks with respect to demands for maximum speed and the presented percep-
tual information are likely to lead to practice effects, even for experts, during
extended testing. But as long as the practice effects for the experts remain
comparatively small and the performance of the experts remains reliably supe-
rior to those for novices even after extended practice, we would claim that such
a collection of tasks can successfully capture the superior expert performance.
The effects of extended practice for novices will provide a major source of
empiricalevidenceaswenowturntoareviewoftheoreticalaccountsofhow
the superior performance of experts can be acquired through extensive training.
Accounting for Superior Performance by Experts: The Third Step
In all the studies discussed earlier, the assessed mechanisms mediating supe-
rior performance implicated cognitive structures that were specific to the rele-
vant task domains. The nature of the mediating cognitions allows us to infer
that they reflect acquired knowledge and previous experiences in the domain.
In order to account for those aspects of superior performance that are acquired,
it is critical to understand the role of knowledge acquisition and the important
effects of practice and training for their acquisition.
When we restrict ourselves to those task domains in which superior perfor-
mance has been adequately captured, the empirical findings can be summarized
relatively easily. The superior performance consists of faster response times for
the tasks in the domain, such as the superior speed of expert typists, pianists,
and Morse code operators. In addition, chess experts exhibit superior ability to
plan ahead while selecting a move (Charness, 1981). In a wide range of task
domains, experts have been found to exhibit superior memory performance.
What is acquired by experts? Superior performance in different domains
reflects processes and knowledge specific to the particular domain. The chal-
lenge is to account for the widest range of empirical phenomena with the small-
est of learning mechanisms and processes responsible for changes as a function
of long-term practice. Because it is not possible to observe subjects during a
decade of intensive practice, most of the empirical evidence is based on ex-
trapolationofchangesinperformancefoundasaresultofpracticeatlabora-
tory tasks over much shorter terms. Another important constraint is that the
proposed descriptions cannot posit performance capacities that would violate
the known limits of human information processing.
In this section we shall consider various accounts concerning the processes
and knowledge that experts have acquired. We shall first briefly describe the
Prospects and Limits of the Empirical Study of Expertise 539