There are some general, large-scale theories of creativity that attempt to describe all (or at least most) kinds of creative processes. The most widely known theory of this type goes by the name divergent thinking. It’s a descendant of Joy Guilford’s divergent production operations, which accounted for 20 of his original list of 120 intellectual abilities. (He later expanded this to 30 distinct abilities out of a total of 180.) Guilford believed that the many varieties of divergent production he had identified were components of many different kinds of creative thinking.

Divergent thinking is understood today to mean the process of thinking of many different possible solutions, answers, or responses to a problem, question, or prompt. Divergent thinking is no longer conceptualized as Guilford did, as a variety of distinct skills. It is instead thought of as a general, all-purpose kind of thinking that many people believe to be involved in creativity of all kinds and at all levels. (Divergent thinking should not be confused with brainstorming, by the way, although they are related. Brainstorming is a technique that encourages divergent thinking. Brainstorming is just one of many possible ways to produce divergent thinking, however.)

There is something intuitively appealing about the notion of divergent thinking, and I believe there really is something important in this concept. The problem is that divergent thinking may not refer to one type of thinking, but to many completely different types of thinking that only seem similar from the outside, as it were—to an observer—while on the inside (as cognitive processes) these various discrete and cognitively unrelated skills may have little or no connection or overlap. Just as the various things a painter and a poet and a pre-school problem-solver might do as they go about their work can all be called creative, even though the thought processes they use to produce poems and paintings and problem solutions may be completely dissimilar, so it is with divergent thinking.

If divergent thinking were a single process, then I could apply my divergent thinking skill in any area on any target or problem. But that’s not really how divergent thinking works. It’s not like a content-free computer subroutine that can simply be taken out of one program and dropped into another program doing a totally different task. Thinking of unusual uses for a brick, unusual recipes for a soufflé, and unusual ways to solve an equation are vastly different skills. If we didn’t give them all a shared label (such as creative or divergent thinking), there would be little reason to think of them as related. But calling things by the same name does not confer upon them shared properties. It does cause us to think of them as similar, however, if even they are not. (This might be why we tend to think more favorably of people who share our name, or even have the same first letter in their names. Completely arbitrary and random similarities, even when we know they are arbitrary and random, lead us to see connections that aren’t really there. But a similarity in name doesn’t make John from Baltimore any more like me than Michael from Philadelphia, even though I happen to share a name with John. Ditto for divergent thinking in poetry, divergent thinking in painting, and divergent thinking in physics.)

The leading developer of divergent thinking tests, Paul Torrance, demonstrated this himself. He created two different versions of the test—he called them figural and verbal divergent thinking—and gave the same tests to groups of people. There was almost no correlation between scores on these two tests of divergent thinking—0.06, which is essentially zero—which showed that the two sets of divergent thinking skill were completely unrelated. Divergent thinking is not one skill. It is many completely different sets of skills. Even the creator of the most widely used tests of divergent thinking knew that. Unfortunately, that message has gotten lost. Many school districts, and even some researchers who use Torrance’s tests, frequently forget this and act as if these tests were measuring some readily transferable, domain-general skill. The result is confusion about what divergent thinking is and the likelihood that many research findings that should have been rejected have been accepted uncritically. (Because scores on the two major divergent thinking tests are uncorrelated with each other, it’s likely—almost certain—that had the other test been used in any given study, the results would have been different.) Is it any wonder that creativity research seems to lurch this way and that and yet make disappointingly little progress overall?

So divergent thinking, the most widely known creative thinking skill (and the one that is the basis for most creativity testing and training), can’t really support a large-scale or general theory of creativity, and the same is true of many other supposedly domain-general theories of creativity. Here’s one more example. There is much evidence that intrinsic motivation (doing something that one finds interesting or personally meaningful) leads to higher levels of creativity than extrinsic motivation (doing something in order to earn a reward or in anticipation of having one’s work evaluated—to get a good grade). That seems to make sense, and there is evidence (disputed evidence—the truth, as usual, is both complicated and nuanced) that this is the case.

But is intrinsic motivation really a single thing, or is it a multitude of things that (once again) only seem related when looked at from the outside? Does one’s interest in cars have anything to do with one’s interest (or disinterest) in gardening? Can the same personal satisfaction that motivates a person to create a beautiful piece of music be used by that person to develop an interesting lesson plan or solve a challenging mechanical puzzle? Or is it not more likely that what we call intrinsic motivation describes not one but many completely different kinds of motivation? As with divergent thinking, intrinsic motivation might be an interesting idea relevant to different kinds of creative thinking, but it is not a unified theory, not a grand theory, not a single theory of creativity.

This doesn’t mean that divergent thinking and intrinsic motivation don’t matter. It does mean that even though having intrinsic motivation and divergent thinking in one domain may make a person more creative in that domain, this says nothing about their creativity in other domains—and this, in turn, has major implications for how we should nurture creative mindsets and teach creative-thinking skills.

When theorists lump together very domain-specific skills like being able to come up with lots of interesting and unusual ways to use a brick, many original ways to choreograph a dance, and many elaborate ways to link several discrete cosmological events— things that required completely different underlying skills—into an abstract concept and label it something (divergent thinking in this case), that can be very useful. It allows us to see connections that are not obvious, and this metaphorical connection might help us look at some phenomenon in new and possibly interesting and productive ways. (It might, for example, help us see that a technique originally designed to come up with new advertising ideas—group brainstorming—could also help administrators solve a scheduling problem, teachers find ways to reach challenging students, or architects discover new ways to construct living spaces, although as with all metaphors, we need to be crateful. Brainstorming might work well in some areas and situations and not others. The erratic results of studies of brainstorming—sometimes it produces creative ideas, but sometimes it appears to hinder creativity—suggests this might in fact be the case.)

The problem is that such metaphorical and tenuous connections can all too easily be reified as actual unified abilities, when in fact they may have no underlying cognitive similarities at all. Once this step has been taken, the slippery slope cascades into programs, tests, and theories that, although based initially on nothing more than a loose metaphorical connection, come to have realities (and constituencies) all their own—false realities and sometimes well-entrenched constituencies that can turn a potentially helpful metaphor into dogmatic thinking, thinking that impedes rather than facilitates progress in understanding complex phenomena like creativity.

I’ve focused mostly on divergent thinking in this post, but the same processes are true of many other skills, traits or attributes often linked to creativity. For example, intrinsic motivation, openness to experience, and content knowledge are all very domain-specific, and all have been shown in some domains and contexts to promote creativity. But one cannot use much of one’s knowledge of pre-Columbian pottery to solve a Rubik’s Cube, fix a bug in a computer program, or a write a haiku. Nor can a person transfer her interest in writing haiku to learning about pre-Columbian pottery, and one may not be equally open to new experiences in the areas of writing computer code and writing haiku. Like divergent thinking, intrinsic motivation, openness to experience, and content knowledge are all very domain-specific. The fact that we can find some interesting (and possibly useful) similarities in these ideas should not confuse us into thinking that these are three modules that operate in the same ways in all content domains and can be transferred freely among them.

I don’t doubt that the constructs of intrinsic motivation, openness to experience, content knowledge, and divergent thinking can be useful to creativity researchers, theorists, and trainers, but only if they are careful not to forget that these are all very domain-specific in their actual manifestations. When we make this mistake—when the metaphor is taken for a reality—then these constructs do more to cloud our understanding of how creativity works than to clarify its mystery.

As the saying goes, “Complex problems have simple, easy-to-understand, wrong answers.” Understanding creativity is certainly a complex problem, and the appeal of grand, over-arching, domain-general theories of creativity is that they tend to be simple and easy-to-understand. But they may all also be, necessarily (because of the complexity of the subject), wrong.