What Is Expertise?

Everybody knows what an expert is: It’s somebody like Tiger Woods and Michael Jordan in sports, Albert Einstein and James Watson in science, Wolfgang Amadeus Mozart and Adele in music, or Bill Gates and Warren Buffett in business. However, it turns out surprisingly difficult to provide a formal definition that everybody can agree with. There are in fact many definitions, but most are unsatisfactory (see Gobet, 2015, for a detailed discussion).

One way to define expertise is to highlight experts’ know-how—their ability to carry out actions. This is undoubtedly an important aspect of expertise, for example in sports; however, in some fields, expertise is more about knowing that—that is, it is more about declarative knowledge than procedural knowledge. Such fields include history and philosophy.

A second way to define expertise is to use experience and the amount of time that one has spent in a particular domain. There is some correlation between the amount of experience and level of expertise, but it is rather weak. Just think of friends or colleagues who have been playing tennis or learning how to play the guitar for years if not decades, but are stuck at a low level of proficiency. Even when one limits experience to deliberate practice (Ericsson, Krampe, & Tesch-Römer, 1993)—that is, goal-directed and intensive practice informed by feedback—only part of expertise is accounted for. For example, deliberate practice accounts for only 29.9% of the variance in expertise in music (Hambrick et al., 2014).

Very often, expertise is characterized as fluid, automatic behavior without any conscious control (for example, Dreyfus & Dreyfus, 1988), and this is our third definition. While this is consistent with many domains of expertise, most notably sports and music, there are also important exceptions. Some authors actually define expertise in exactly the opposite way: fluid and routine behavior is the hallmark of non-expertise, and real experts address difficult problems, with the possibility of mistakes, expanding their knowledge along the way (Bereiter & Scardamalia, 1993).

A fourth common definition is to use official titles such as diplomas, PhDs, or professional certificates. Unfortunately, this approach has several weaknesses. In particular, diplomas often relate to declarative knowledge that is not necessary the same as the procedural knowledge actually used in a discipline; this is the case, for example, with the kind of theoretical knowledge acquired in medical schools, compared to the kind of clinical knowledge used in medical practice. In addition, some individuals perform at a high level without any qualifications. Epstein’s study (1996) of AIDS activism offers a classic example; it was found that AIDS activists without formal qualifications had acquired substantial knowledge in microbiology and statistics to the point that they could make contributions to research.

Some fields, in particular sociology, emphasize the social aspects linked to expertise, and in the extreme argue that expertise does not really exist but is a label given by society (or other groups) to some individuals. This damning judgment is true is some cases. For example, consider astrology, a highly popular field where “experts” allegedly divine human affairs and personality using the positions of the stars and planets. It has been scientifically established that astrologers do not predict the future or human personality better than chance (e.g., Carlson, 1985). However, there are domains where experts are not just the product of a society exercise in labeling. If you doubt this, just play a game of tennis against Roger Federer or a game of chess against Magnus Carlsen, and observe the ensuing destruction.

What is missing in the definitions discussed so far is a reliable measure of expertise. Some fields have such measures, for example the amount of wealth accumulated in business or the number of books sold in literature. These measures, while objective, are not ideal, because they are too sensitive to confounding factors, such as fashion in literature. Sports are in a better position, and numerous objective measures (e.g., time to run a marathon) can be used to reliably rank experts. Some domains (e.g., chess, table tennis and Scrabble) use a reliable measure of performance, known as the Elo rating (Elo, 1978), which updates one’s rating continuously as a function of the strength of the opponent and the result of the match.

One of us (Gobet, 2011; 2015, p. 5) has proposed the following definition: an expert in a given domain is “somebody who obtains results that are vastly superior to those obtained by the majority of the population.” This definition has several advantages. First, it can be used in domains where most people perform at a high level, such as breathing. Practitioners of hatha yoga can be considered as experts, as they can use breathing techniques that would be difficult and even dangerous for most people. Second, it can be used equally successfully to experts in domains were procedural skills dominate and in domains characterized by declarative knowledge. Third, it makes it possible to identify fields where genuine experts exist (e.g., tennis) and those where expertise does not exist (e.g., astrology). Finally, it can be applied to experts themselves: a super-expert is somebody whose performance is vastly superior to the majority of experts.