As with bilingualism, it’s generally assumed that being an expert completely beneficial and has no downsides to performance. However we know that expertise tends to be domain specific, for example, chess grand masters can memorise chess boards far more quickly and easily that novices, but on standard cognitive tests tend to fare no better. In fact, if you arrange chess pieces to positions that would never be encountered in an actual game, again their recall is no better than chess novices, showing just how domain-specific expertise can be. But surely within a given domain, expertise can only be beneficial?

Castel, McCabe, Roediger and Heitman suggest not. They gave 40 students a memory test consisting of eleven animal names and eleven body parts. The twist here was that all the animal names were also NFL team names, like dolphins, colts, seahawks and bears. After the memory test, participants were given an NFL quiz, and the group was split into two, those scoring above and below the median on this test, to give high expertise and low expertise groups in the domain of NFL knowledge.

The results on the memory test for the two groups was then compared. Indeed, the NFL experts remembered more of the animal names than the non-experts, while there was no difference between groups on the body parts test. So far so good, however, the researchers also tested for incorrect answers — NFL animal team names and body parts that were not part of the original test. The results indicated that the experts were much more likely to make incorrect guesses than the non-experts. The authors suggest that this represents memory errors, the domain-relevant information of the experts got in the way of their accurate recall of the animal names. Since there was no difference between groups in body part experience, false answers were about even between groups on that test.

Is this really the case though? Or was it that the experts consciously noticed that the animal names belonged to the NFL teams and simply reeled off as many as they could remember during recall. Perhaps it was not a case of the existing schema interfering with memory, but a recognition that they already know these names, so why bother taking the extra effort to think back and recall? Why not just reel off my schema? I wonder if the results would be the same if participants were told that they would score 1 point for a correct guess, but minus 1 point for an incorrect guess, which might increase the incentive to actually recall. In other words, maybe this effect is a conscious strategy used in situations where there’s no cost to an incorrect answer.

However, there are other studies that support the authors’ conclusions, which I haven’t read so perhaps my question has been answered before or since. Either way, it’s an interesting thought that the knowledge base acquired by experts might be detrimental in certain tasks.

ref:
Castel AD, McCabe DP, Roediger HL 3rd, & Heitman JL (2007). The dark side of expertise: domain-specific memory errors. Psychological science, 18 (1), 3-5 PMID: 17362368