We often say things like “Jane is intelligent,” with no further clarification. But we don’t simply say “Jane is an expert”— in the latter case, we need to specify the area of her expertise (e.g., “Jane is an expert in art history” or “Jane is an expert statistician”). No one is an all-around expert (although some people may have expertise in many areas, of course) the way some people are simply intelligent. We may recognize that even intelligence has components (and some might even argue about just how independent those components are of one another, but that is a debate for another day). For today, I am simply using the ways we typically talk and think about intelligence and expertise as tools to help us understand how we think—and how we should think—about creativity.

So can we say “Jane is creative” without further clarification? People do all the time, but that’s a mistake. Jane may be creative, but it’s unlikely that she is generally creative in all or most of the things she does. She’s probably creative in some things and not in others. Does this seem like nit-picking? Let me show you why it’s not.

Putting aside arguments about what IQ tests actually measure, we know two things about IQ scores: (1) they are correlated with performance in many domains and (b) if you score high on one IQ test, you are extremely likely to score high on other IQ tests. This means that there’s something domain general about whatever it is that IQ tests measure. Speaking of a person as intelligent therefore makes sense: It means that a person so identified has abilities of the kind measured by IQ tests that are significantly above average and that, on a wide range of tasks from diverse domains, the person is likely to do comparatively well on most of those tasks. If someone is intelligent, then it is reasonable to assume that they will have abilities in quite a few unrelated areas. This is true whether one equates intelligence with IQ tests or not. It’s a fairly general skill or set of skills that are useful in many very different areas.

Expertise doesn’t work that way. Calling someone an expert doesn’t really make sense unless their domain (or domains) of expertise is specified in some way. Even when people have expertise in several domains, one still needs to identify those domains. Otherwise, it would be impossible to understand what it means to call someone an expert. I need people with different kinds of expertise to fix my car, teach me calculus, or set my broken bone. Expertise varies widely by domain. Knowing that someone is an expert in California wines doesn’t lead one to assume that person will also be an expert in accounting, ice hockey, or modern dance.

So which is creativity more like, expertise or intelligence? Thinking about creativity the way we think about expertise is more accurate because both are highly domain specific. Just as someone may be an expert in more than one field, someone may be creative in more than one area, but neither creativity nor expertise generalize broadly across domains. This has been born out in numerous studies looking at the actual creative products produced by people in different domains. The creativity of the stories participants write is unrelated to the creativity of the collages, math puzzles, and other artifacts they create—average correlations are just a little higher than zero—and most of what little shared variance there may be is attributable to differences in general intelligence and access to educational opportunities.

Does this matter? Yes, because it influences both how we assess creativity and how we try to nurture it.

Efforts to assess creativity have been plagued by supposedly domain-general divergent-thinking tests like the Torrance Tests of Creative Thinking, although even Torrance knew they were measuring domain-specific skills. (He create two different versions of the test, one that used verbal tasks and another that used visual tasks. He found that scores on the two tests were unrelated —they had a correlation of just .06—so they could not be measuring a single skill or set of skills. They were—and still are—measuring two entirely different things.) Because these tests have been used in so many psychological studies of creativity, much of what we think we know about creativity may be based on invalid data. These tests have also been widely used in selection for gifted/talented programs — programs that have, in turn, often suffered because by assuming creativity was domain general, these programs often wasted students' time with supposedly content-free divergent thinking exercises (like brainstorming unusual uses for bricks) that really only develop divergent-thinking skill in limited domains.

Think for a minute how we build muscles. Does it matter the content of the exercises—whether they employ our biceps or our quadriceps or some other muscles? Of course it matters. To increase one's physical strength, it would be crazy to suggest doing only one kind of exercise. To increase overall strength one must do many different kinds of exercises, exercises that each strengthen different sets of muscles. Exercising one muscle will strengthen that muscle, not all of one's muscles. Because creativity is domain specific, the same is true of creativity training: one must do many different kinds of exercises if one wishes to strengthen many different creative-thinking "muscles."

The same is true when we think about the acquisition of expertise, of course. Because we know that expertise is domain-specific, we know it matters the kinds of education or training needed to develop expertise in a domain. Learning about Japanese history will help develop expertise in that area, but will do nothing for one’s expertise in calculus or auto repair.

If creativity were domain general, there would be essentially a single set of creative-thinking muscles that one could use with any problem, and many creativity trainers—seduced, perhaps, by the intelligence metaphor—operate in this way. If this were true it would make creativity training much easier (just as it would make muscle building easier if there were a single set of muscles), and this is perhaps one great attraction of domain-general theories of creativity: if they were true, they would allow training shortcuts. Because creativity is domain specific, however, training creativity must be more like building muscles or developing expertise. If one's goal is to enhance creativity in many domains, then creativity-training exercises need to come from a wide variety of domains—just as we must provide a broad general education if we want students to acquire modest levels of expertise in many areas.

But if one's goal is to increase creativity in just one domain, such as one might want to do in a gifted program focusing on one domain (such as a program in dance, poetry, math, etc)., then it would be appropriate for all of the creativity-training exercises to come from the particular domain of special interest. This also parallels what we would do if we wanted students to acquire expertise in a single domains—all (or at least most) of their studies should be in that area. If one wants to increase creative-thinking skills (or expertise) in many domains, however, one needs to use a wide range of domains in training.

Metaphors matter because they shape and direct our thinking and decision-making. Thinking about creativity the way we think of expertise will lead to a better understanding of creativity and better decisions about how to recognize and develop it.