Exploring Creativity, Part III: How Not to Define Creativity

“There are forty kinds of lunacy but only one kind of common sense.”

-African proverb

Before I began this essay series, it had never occurred to me to consider creativity and rationality together. On the surface, they seemed like totally separate domains. Creativity belonged to positive psychology, a field that seemed scarcely more rigorous than humanist philosophy. And rationality belonged to a sphere of discourse that emphasized the proceduralization of one’s life in the service of eradicating bias. But I now recognize that creativity is central to rationality. One need not look further than the rationalist whose creative exercise spurred this essay series in the first place, Eliezer Yudkowsky, whose prodigious creativity is visible in all his pursuits, from fiction to scholarship to advocacy.

It is edifying to juxtapose Eliezer Yudkowsky, who privileges rationality over creativity (or at least views creativity as a tool of rational thought), with Mihaly Csikszentmihalyi (pronounced mee-HAI six-cent-mee-HAI), a positive psychologist who –as we shall see– allows his enthusiasm for creativity to deaden his critical thinking skills. Csikszentmihalyi is probably best known for his work on creativity, which he encapsulated in his book Creativity: Flow and the Psychology of Discovery and Invention (1997). In the book, Csikszentmihalyi outlines the characteristics of creative individuals that he had gleaned through extensive interviews:

Characteristics of the creative personality:

  1. Creative individuals have a great deal of energy, but they are also often quiet and at rest.
  2. Creative individuals tend to be smart, yet also naive at the same time.
  3. Creative individuals have a combination of playfulness and discipline, or responsibility and irresponsibility.
  4. Creative individuals alternate between imagination and fantasy at one end, and rooted sense of reality at the other.
  5. Creative people seem to harbor opposite tendencies on the continuum between extraversion and introversion.
  6. Creative individuals are also remarkable humble and proud at the same time.
  7. Creative individuals to a certain extent escape rigid gender role stereotyping and have a tendency toward androgyny.
  8. Generally, creative people are thought to be rebellious and independent.
  9. Most creative persons are very passionate about their work, yet they can be extremely objective about it as well.
  10. The openness and sensitivity of creative individuals often exposes them to suffering pain yet also a great deal of enjoyment. (Csikszentmihalyi 58-73)

If you’re like me, you find it easy to identify with these descriptions. Yes, That’s me! I do have these opposing tendencies! Personally, I felt a frisson of pseudo-insight when I read #5. My persistent amazement about how difficult it is to understand what it’s like to feel sociable when I am feeling introverted, and vice versa, is a perennial topic in my journal writing. These definitions also conform to our intuitive sense that creativity involves a kind of broad, exploratory, and boundary-defying mind. But we would be wise not to trust our intuitive sense that these descriptions point to a valid definition of creativity. Our minds are easily fooled by superficial resemblances, and are especially vulnerable to intuitive interpretations that seem psychologically comforting. Csikszentmihalyi’s descriptions of “the creative personality” are both superficially plausible and psychologically comforting. Truly, these traits don’t describe creative people. They describe people.

It is trivially easy to construct statements that apply universally, flatter everybody, and sound scholarly. Here are a few I came up with just now:

  • Creative individuals enjoy eating new and delicious foods, but also have the capacity to appreciate plain and familiar meals.
  • Creative people enjoy exploring the natural world, yet also delight in cozy indoor settings.
  • Creative individuals tend to link paper strips using tape, but sometimes they find it preferable to use a glue-stick.

Perhaps you think that the model’s generality is an advantage. After all, If creativity is a holistic concept with diverse manifestations, why shouldn’t the definition be just a general and multifaceted? The problem with this reasoning is that when your model is so elastic that it can explain everything, then it actually explains nothing. Such a model would be a Fake Explanation. As Yudkowsky explains, “the usefulness of a model is not what it can explain, but what it can’t.” A model of the world that is compatible with every possibility is as good as having no model at all.

But Fake Explanations are insidious. They trick us into believing that we have actually explained something when we haven’t. They the intellectual equivalent of junk food, satisfying our desire to be knowledgeable without actually increasing our knowledge. This kind of uncritical thinking is facilitated by imprecise language. In particular, the conflation of a word’s commonsense definition with its technical, scientific counterpart. In Csikszentmihalyi’s final synthesis, where he reveals the factor that unites the various “characteristics of the creative personality” that I enumerated above, we see an example of this irrational conflation:

“If I had to express in one word what makes their personalities different from others, it’s complexity. They show tendencies of thought and action that in most people are segregated. They contain contradictory extremes; instead of being an individual, each of them is a multitude.” (my bold)

This word, complexity, has a precise meaning in physics, mathematics, and computer science. Here, however, Csikszentmihalyi is not referring to that definition, but simply capitalizing on the association of the word complexity with scientific rigor. I do not mean to imply that Csikszentmihalyi is deliberately masking the vacuity of his argument with fuzzy language. That’s the most insidious part of a Fake Explanation: he probably thought that complexity was a suitable explanation for his observations. But it is It is extra disappointing because the scientific study of complexity is an fascinating topic. Did you know, for instance, that information, randomness, and complexity all have rigorous mathematical expressions? And did you know that they are equivalent, alternative formulations of the same basic principles? To say that creative individuals display “complexity” and contains “a multitude” is tantamount to throwing up your hands and admitting your ignorance. The most egregious sin against rationality is to believe yourself to be doing science when in fact you are not. As Richard Feynman wrote, “The first principle is that you must not fool yourself and you are the easiest person to fool.”

How, then, did Csikszentmihalyi fool himself so utterly? Csikszentmihalyi is famous for his notion of Flow, a kind of “effortless attention” that is widely described as a “peak experience.” The fact that psychology textbooks, rationality resources, mindfulness texts, and psychotherapy manuals all discuss Flow is a testament to its enduring utility as a concept. And Csikszentmihalyi is not a charlatan or a pseudoscientist; his work is repeatedly cited by the founder of behavioral economics, Daniel Kahneman, and other eminent scholars of the mind. Having researched the intellectual history of creativity, it seems plausible that Flow will someday be as much a fixture in the English lexicon as creativity currently is. My friend Mike Tizzano has written a cogent review of Csikszentmihalyi’s book on this subject.

To understand how Csikszentmihalyi could have made such patently irrational statements, we need to examine his evidence-gathering method. He interviewed 91 people who were publicly regarded as creative, and whose creative contributions had an impact on their respective fields. Full disclosure: I am a laboratory scientist, a consummate experimentalist who studies mouse behavior, physiology, and brain organization. I am just about the farthest thing from a positive psychologist, and I am eager to avoid the appearance that my critique is merely the chauvinism of an experimentalist. Truly, I am not impugning Csikszentmihalyi’s method, only the conclusions he draws from it. In psychology, there are two types of research methodologies. First, there are descriptive methods, such as naturalistic observation, case studies, or surveys. Then there are experimental methods, which involve manipulating variables to work out cause-and-effect. To an extent, the descriptive-experimental distinction is a false dichotomy. Surveys, for example, are usually classified as a descriptive method, but are much more quantifiable and generalizable than another descriptive methods, such as naturalistic observation or case studies. The experimental approach allows for more rigorous generalizations, while the descriptive approach provides evidence that is, at best, suggestive.

To be sure, Csikszentmihalyi’s approach was rigorous enough to cast doubt on certain popular misconceptions of creativity, such as the prevailing idea that highly creative individuals are “mad geniuses” who had distressing childhoods. In fact, Csikszentmihalyi found that most of his interviewees were not eccentric monomaniacs, but, rather, conscientious and agreeable. Moreover, Csikszentmihalyi’s sample generally reported happy or uneventful childhoods, with plenty of supportive adults to act as mentors. Csikszentmihalyi’s methodology was sufficient to demonstrate that these earlier theories of creativity were likely to be untrue. Although not maximally rigorous, this kind of study can serve as a stimulant for further research that is more rigorous.

Indeed, more rigorous survey methods have vindicated Csikszentmihalyi’s impression that the “mad genius” is mostly a myth. A apposite modern parallel is the work of Temple Grandin, an animal behavior expert who has written extensively on autism. Grandin, includes in her books the first-hand accounts of how people with autism experience the world, including her own. In her books, Thinking In Pictures and The Autistic Brain: Thinking Across the Spectrum, Grandin highlights these other perspectives not as a way of making an end-run around the peer review process, but as a way to highlight aspects of the autistic experience –such as the variety of ways-of-thinking and the role of sensory processing difficulties– that are currently being neglected by the scientific establishment.

The pursuit of a coherent vision of creativity begins with a recognition that intelligent, well-informed people will have slightly different intuitions about what constitutes creativity. This diversity is to be expected, since creativity is an abstract notion that cannot be observed directly. Nor does anyone have any privileged access to it. It is fundamentally unlike hard-to-observe physical phenomena like quantum tunneling, radioactive decay, or cell division. Understanding (and, eventually, enhancing) creativity requires more than building a better microscope or a more powerful supercollider. It will require a clear and explicit definition that tracks our intuitive sense of what creativity is. We begin with the observation that creativity is a broadly subscribed notion that exists in the minds of many individuals. In that regard, it is like a meme –a wildly popular meme, too. But the popularity of a meme is orthogonal to its truth value. False and pernicious memes –vaccines cause autism, feminism entails denigrating men, existing laws ensure humane treatment of factory-farmed livestock– propagate irrespective of their scientific merit. Memes spread because they are psychologically comforting and/or politically expedient.

Creativity is often framed as an unscientific concept, and unquestionably there are formulations of creativity that are not valid (i.e. that do not track our intuitions about creativity) or –as in the case of Csikszentmihalyi– define it in such a way that it is not possible to objectively distinguish creativity from its absence.The best way to classify creativity is as an intellectual construct. An intellectual construct is a phenomenon that cannot be attributed to a distinct cause or group of causes, but nonetheless proves a useful tool for categorizing and explaining other facts. The intellectual construct is a useful fiction, except the users of intellectual constructs are not trying to seed propaganda or hoodwink anybody. Nor are intellectual constructs resistant to objective study. There is, however, one sacrifice that you must make before embracing an intellectual construct. You must abandon the ineffable mystique that surrounds venerated notions like creativity, and deign to design a contrived test that is susceptible to measurement. And, finally, to accept such tests as proxies for creativity itself. This process of converting an abstract, holistic idea into a discrete, verbalizable form is called operationalization.

My first introduction to operationalization was in a Psychology 101 course I took when I was a junior in high school. Professor Tsiris explained that in addition to teaching, she also worked as a behavioral therapist at a special school for children with psychological problems. One day, Professor Tsiris asked us how we would make operational definitions to assess a child’s writing. The class volunteered that one should look at the legibility of the handwriting, and whether the child has written in complete sentences. No, she said. You had better make sure that the child is sitting down, has a piece of paper in front of them, is facing the paper, has a pencil in their hand, has the pencil tip pointing down toward the paper, and many other extremely basic considerations. The class was duly humbled by this explanation. “At the school where I work,” she said, “it’s my job to quantify the behavior of the children, so we can really assess the effectiveness of our interventions.” She explained that one of the children she monitored had a habit of banging his head on this desk. Matter-of-factly, Professor Tsiris explained that when he started doing this, she didn’t try to restrain or even reprimand him. Professor Tsiris lifted her hand and pantomimed holding up a clicker-counter. “Instead, in those situations, I count the number of times he hits his head against the desk.” The number of head-bangs was her operational definition of that child’s misbehavior. I remember being in awe of Professor Tsiris at that moment. Operationalization, it would seem, was a hardcore enterprise.

Like other intellectual constructs, such as intelligence, emotion, language, motivation, and attention, psychologists must define creativity in such a way that it is both valid and reliable. In the scientific context, valid and reliable have precise meanings. Validity refers to the fact that the operationalized descriptions actually measure what they purport to measure, in this case creativity. Intellectual construct like creativity actually correspond to our intuitions about who is and isn’t creative. Although there are several different aspects of validity that need to be assessed, as in the remark, “the idea was ridiculous on its face!” Unsurprisingly, face validity is sometimes referred to as a “sanity check.” Face validity is way of quickly checking a new idea, but it shouldn’t be used as a bludgeon for dismissing all nuanced or counterintuitive claims. The face validity metric doesn’t zero in on a specific aspect of a theory, but, rather, on the theory as a whole: its foundations, its proposed mechanisms, its falsifiability, and its overall relevance. In the case of Csikszentmihalyi, the glaring flaw is that his descriptions of creativity are unfalsifiable. His model is unable to distinguish creativity from its absence. To the extent that someone might claim to find the theory useful, it would be a case of mistakenly attributing their intuitive ability to recognize creativity to Csikszentmihalyi’s vague descriptions. Despite the challenges involved in fashioning a rigorous definition of creativity, we can all (to an extent) “know it when we see it.”

Moving a little beyond face validity, we might predict that people who a majority of impartial observers would describe as creative would perform better on the proposed tests of creativity than people designated as less creative. Put so starkly, this criterion sounds blindingly obvious. Why do researchers even bother with articulating such elementary principles? The short answer is that this component of face validity is not as blindingly obvious when encountered in the wild. And it is particularly hard to spot when it is your cherished hypothesis that stands to be eviscerated. Recall Csikszentmihalyi’s descriptions of “the creative personality” from earlier. Let’s consider #5: “Creative people seem to harbor opposite tendencies on the continuum between extraversion and introversion.” How would you design an experiment to find creative people according to this definition? Since no person is completely introverted or completely extroverted (what would such a person act like?), everyone must therefore “harbor opposite tendencies on the continuum between extraversion and introversion.” If everyone qualifies as creative according to this definition, then no one is creative.

Did you just bristle at my conclusion that “no one is creative?” I hope you did. I did,too –and I wrote it! That visceral sensation of doubt is a natural reaction. To be an effective rationalist doesn’t mean you have no such feelings, it means you have the skills to recognize these feelings for what they are, and deal with them constructively. Our incongruity-detecting mental modules are on the whole well-calibrated, but nevertheless occasionally misfire. Moreover, these misfires occur disproportionately in particular circumstances rather than at random. Part of your skill as a rationalist is to notice patterns in the circumstances that make these misfires more probable, and to then guard against them. If you can recognize the fallibility of your intuitive sense of doubt, then you are halfway to understanding why a scientific theory can propagate despite failing the test of face validity. You likely felt surprised and perhaps aggrieved at the conclusion of “no one is creative” because you interpreted the conclusion inductively (on the basis of your experience) even though the form of the sentence required deductive logic. The premise, I remind you, was “If everyone qualifies as creative according to this definition.” Your pre-existing knowledge of the instrumental and pragmatic value of the concept of creativity, as well as your emotional stake in conceiving of yourself as a creative individual, compeled you grant legitimacy to Csikszentmihalyi’s irrational formulation of creativity.

I have focused on face validity because, compared to other measures of validity, the most fundamental as well as the easiest to understand. But there are other forms of validity that probe deeper aspects of a theory’s applicability to its stated claims. These include internal, external, test, criterion, content, and construct validity. I will no doubt touch on these kinds of validity in future essays, though I probably won’t name them explicitly.

My purpose in this essay was to show through a case study some of the basic ways a definition of creativity could fail as a scientific theory. We saw that a definition must be falsifiable, operationalized, and stripped of any extraneous mysticism. As a rationalist, I see no principled reason for holding my personal beliefs to different, lower standard than my scientific beliefs. Put another way, there should be no firewall between the intellectual ethic of the laboratory and the rest of the world. In my next essay, I will look at the prototypical creative exercise, brainstorming, and evaluate its theoretical foundations, empirical support, as well as its intersection with the introversion-extraversion dimension of human personality.

References:

Csikszentmihalyi, M. (1997). Creativity: Flow and the psychology of discovery and invention. New York: Harper Perennial.

Tizzano, M. (2015, March 6). Book Review: Flow by Mihaly Csikszentmihalyi. Retrieved July 18, 2015, from http://www.miketizzano.com/ruminations/2015/3/6/book-review-flow-by-mihaly-csikszentmihalyi

Yudkowsky, E. (2007, August 20). Fake Explanations. Retrieved July 18, 2015, from http://lesswrong.com/lw/ip/fake_explanations/

Exploring Creativity, Part II: A Brief History of Creativity

In my previous post, I began my exploration of creativity with an intriguing proposal for circumventing bias via a creative exercise. My overarching goal in this series is to showcase the methods of rationality while also explaining the modern scientific understanding of creativity. I intended to follow up that post with a brief survey the history of the concept, from its origins in philosophy to its modern meaning. And then I planned to follow up with a discussion of some of the technical definitions of creativity used in modern psychology. Disappointingly, the intellectual history of creativity is not particularly relevant or interesting. I had expected creativity to have followed a similar trajectory to early sciences, like optics, which underwent a more or less smooth and unidirectional course of improvement from their ancient origins to their modern incarnations. The history of creativity, by contrast, is marked by a small number of theories going in and out of fashion throughout the ages, with the major conceptual breakthroughs all happening in very modern times.

Creativity as we understand it is an astonishingly recent invention. This is not to say that until recently no one thought or acted creatively; it is rather the case that they did not conceive of their thoughts and actions in those terms. More than anything, the history of creativity emphasizes that our current conception of creativity as a singular mental faculty is not the norm. Rather, creativity comprises a cluster of phenomena: it is a faculty of mind, a personality trait, a variety of subjective experience, and a skill that can be taught and exercised. This multifaceted thing we call creativity has also experienced a change in connotation over the course of its history, from slightly untrustworthy to uniformly positive. The mystical undercurrents of creative thought have similarly subsided. Whereas the term creativity originally referred to the divine ability to engineer life, it is now a thoroughly secular concept. Whereas creative thinking was once ascribed to the intervention of supernatural agents, both good (angels, geniuses, muses) and evil (daemons, djinns), it is now recognized as being instantiated in our physical brains. Despite its secularization, creativity retains some of its original mystique. This enduring mystique manifests itself in people who still argue that human creativity is qualitatively superior to the creativity of non-human animals, as opposed to both existing at different points along a spectrum of cognitive and behavioral flexibility.

The words creative and creativity have taken on a distinct meaning from their etymological origin in the word create. As I mentioned earlier, creativity used to refer exclusively to divine creation. The next step in the term’s evolution was toward something synonymous with “productivity,” in which there was no indication that the production was due to creative expression or not. We can illustrate the distinction between this sense and the modern sense using a simple thought experiment. We would not say that a person who produces a thousand origami cranes in a single day while working from an instruction manual is necessarily creative. However, we would say that the maker of the first ever origami crane was creative, even if it took him fifteen years to fold his first crane. The crane-maker in the first case is creative in the earlier sense, but not in the modern sense. The crane-maker in the second case is creative in the modern sense, but not in the earlier sense.

Before the twentieth century, it was commonly thought that creative expression was something field-specific: one used separate mental faculties for mathematical versus philosophical creativity, rather than employing the same basic faculty of mind but applying it to different subjects. The modern conception of creativity as a phenomenon that is fundamentally the same across disciplines is attributed to the philosopher Alfred North Whitehead, who, in 1927, coined the word in its modern sense. Whitehead’s references to creativity mostly track the current meaning, but, crucially, neglect the possibility that it could be an object of scientific inquiry. (Whitehead regarded creativity as a metaphysical concept.) The psychological establishment took little notice of creativity until 1950, when J.P. Guilford discussed creativity in his address to the American Psychological Association. Guilford’s message to the psychological community was that creativity was not merely a fuzzy metaphysical concept, but a something that could, in principle, be studied just as rigorously and quantitatively as topics like intelligence and emotion. Once creativity was reified as a legitimate scientific topic, the serious efforts to define it began. Precise definitions are crucial in all sciences, but particularly in the psychological sciences, where there most of the objects of inquiry are abstract concepts rather than physical phenomena.

When reviewing the early efforts to define creativity, I was delighted to discover another opportunity to bring in rationality. Next time, I will present another case study in the intersection of creativity and rationality. This will lead into a more technical overview of the different definitions of creativity that have been put forward by psychologists since 1950.