Is There a Link Between Creativity and Madness?

If you’re schizophrenic, depressed, alcoholic, or bipolar, are you likely to be more creative than the average normal person? Most people think the answer to this is an obvious “Yes.” We’ve all heard about famous writers who’ve been alcoholics or have committed suicide (Sylvia Plath, Ernest Hemingway) and musicians (Kurt Cobain) and even comedians (Robin Williams).

This is why so many people are surprised to learn that there’s no scientific evidence of a link between creativity and mental illness. In fact, there’s substantial evidence that creative people are more happy and mentally balanced than average; for example, check out this new study that found that writers “have more optimistic perspectives and improved quality of life”.

In the latest issue of The Chronicle of Higher Education, reporter Tom Bartlett has written an excellent story that starts with his visit to the 2014 Aspen Ideas Festival, where Professor Nancy Andreasen gave a standing-room-only lecture arguing that creative people are more likely to be mentally ill–specifically, to have mood disorders like depression and bipolar disorder. But after speaking with several other prominent creativity researchers who were also invited to speak at the event (including me), he was surprised to discover that most of us don’t believe there’s a link.

Of course, Professor Andreasen strongly defended her research when interviewed by Bartlett after her talk. So when Bartlett concludes his article, he throws up his hands and says he can’t figure out what the real truth is:

The discussion too often gets derailed by wildly varying definitions of creativity and mental illness, terms that are so hopelessly broad that simply asking if there is a link between the two is unlikely to ever lead to a satisfying answer. The research that appears most promising takes a narrow look at particular fields and distinguishes between everyday creativity and bleeding-edge genius. It remains a bewildering puzzle, one hampered by our still-evolving knowledge of neurological differences, the challenge of categorizing creativity, and the cultural biases that can’t help but influence our conclusions. It’s enough to drive anyone crazy.

It’s not Bartlett’s fault; after all, there are one or two professors who stand out against the broader scientific consensus, most notably Andreasen, so what is a journalist supposed to do? Journalists generally like to represent every side of a controversial issue, but in this case I think Bartlett worked too hard to seem to treat both sides equally. When he interviewed me, I told him about the four most definitive scientific studies of mental illness and creativity, all of which found that among creative people, there isn’t a higher incidence of any mental disorder (I summarize all four in my 2012 book Explaining Creativity). I told him about the many studies showing that exceptional creators are more likely to be mentally healthy. I told him that most of my creativity research colleagues are certain there isn’t a link:

  • Robert Weisberg (2006) says this is a myth
  • James Kaufman (2009), the author of Creativity 101, says that studies claiming there’s a link are flawed and that the link has never been proven
  • In his influential textbook, Mark Runco (2007) says “there are indications that creativity has benefits for health” and that the only reason people are still talking about this is “because it is newsworthy”

One seeming exception is Dean Keith Simonton, who’s quoted by Bartlett as saying there might be a correlation between very highly exceptional creativity and mild psychopathology (although not full-blown mental illness). But even Simonton doesn’t think there’s a link between regular, everyday creativity and mental illness. And more importantly, he doesn’t argue that the link is a causal link–the idea that being exceptionally creative makes you more likely to become mentally ill, or the idea that having a mental illness makes it more likely that you will be exceptionally creative.

The first thing you learn in an undergraduate psychology course is that correlation is not causation. Even if there were a correlation between being creative and having mild symptoms of mental illness–and this rather mild claim is the only one that gets any traction with any more than one or two of creativity researchers–that link would not necessarily be causal. There are lots of ways that it could be almost accidental:

  • Artistic professions don’t police their borders the same way other professions do. If you’re bipolar, it may be hard to get through law school or medical school, but no one can stop you from taking up a pen and paper and writing short stories.
  • Our society has lots of stereotypes about how creative people are “supposed” to behave, and these include quirky behaviors and eccentricity. If a painter is a bit unconventional, no one gets upset. But if your accountant starts to act a bit quirky, you’ll go in search of another accountant. In other words, if we believe there’s a link between artistic creativity and eccentricity, it becomes a self-fulfilling prophecy.
  • Finally, most artists and writers are not that successful, at least not right away. Their paintings don’t sell for a lot; their short stories and novels take years before they become famous (if ever). Engaging in work that you really care about, and experiencing rejection for days, months, and years, could stress out the most mentally stable individual.

All three of these explanations are perfectly good accounts of why there might be a correlation between artistic creativity and mild levels of mental unusual-ness. So it’s actually kind of surprising that all four of the major studies did not find any statistical evidence of a correlation between creativity and mental illness. (It’s probably because engaging in creative activities actually benefits your mental health.) And none of these explanations require us to posit a causal link. But of course it’s the causal link that gets people so interested–the belief that if you’re mentally ill, you can tap into some inner reserve of creative potential that is closed off to us normal people. For THAT claim, there is no evidence whatsoever.

Mental Illness and Creativity: Two New Swedish Studies

Is it true that creative people are more likely to be mentally ill than non-creative people? Are artists and writers are more likely to be alcoholics? Clinically depressed? Commit suicide? Most people can think of at least one famous artist or writer who committed suicide (Hemingway, Plath) or did something wildly eccentric (Van Gogh cutting off his ear). Hollywood seems to believe in a link; think of movies like A Beautiful Mind (2001) and Shine (1996).

But these are just isolated anecdotes. Is there any research? Voila: a new study has just been published by a Swedish research team, providing suggestive evidence that creativity may be connected to mental illness. This exciting new publication, along with an earlier 2011 publication by the same team, was reported in BBC News, CBS News, and here on the Huffington Post. The lead author of the study, Simon Kyaga, has presented his research at TEDx in Sweden. Kyaga’s research seems to provide confirmation of one of our deepest beliefs about creativity: that it is somehow linked with mental illness.

There’s a problem with these news stories, however: There is no link between creativity and mental illness. Creative people are not more likely to be diagnosed with mental illness, and mentally ill people are not more likely to be creative than normal people. Multiple studies, going back over a century, have consistently found the same proportion of mental illness in creative people as we find in the general population. In this post, I’ll review the research, I’ll discuss why the Kyaga findings came out differently, and I’ll suggest what the findings might really mean.

There are four major past studies that are similar to the new Kyaga study–they examine large numbers of creative and eminent individuals. None of these studies found an elevated degree of mental illness in creative people:

  • In 1904, British psychologist Havelock Ellis studied 1,030 eminent individuals and found that the incidence of mental illness was essentially the same as in the general population. He concluded that there was no relationship between genius and mental illness. In 1947, W. G. Bowerman replicated Ellis’s study and found only a 2% incidence of mental illness among geniuses.
  • Austrian psychiatrist Adele Juda spent 16 years analyzing 19,000 German artists and scientists who worked between 1650 and 1900. As with Ellis’s study, the incidence of mental illness was only marginally higher than in the general population. In her 1949 article about this research, Juda concluded “There is no definite relationship between highest mental capacity and psychic health or illness…Psychoses, especially schizophrenia, proved to be detrimental to creative ability” (p. 306)
  • In 1962, the married couple Mildred and Victor Goertzel published a study of 400 eminent people, and they found less than a 2% rate of mental illness–lower than the general population. Their son, Ted Goertzel, became a professor of sociology at Rutgers university, and republished their study in 2004. The study also found that the parents, brothers, and sisters of the 400 geniuses had a lower than average incidence of mental illness (pp. 237-238).
  • In 1994, Felix Post published a study of 291 biographies of famous men. None of these geniuses met the criteria for any DSM-III psychiatric diagnosis.

Despite these many studies that found no link between creativity and mental illness, in recent decades contemporary researchers have continued to search for evidence to support the Western cultural belief that mental illness is connected to creativity. (Most non-Western cultures do not share this belief.) These more recent studies, when well designed, consistently deliver the same finding: There is no link between creativity and mental illness. (Also see my 2009 blog post here.)

In my 2012 textbook overview of creativity research, Explaining Creativity, I conclude:

Despite almost a century of work attempting to connect creativity and mental illness, evidence in support of a connection has been remarkably difficult to find….The consensus of all major creativity researchers today is that there’s no link between mental illness and creativity (p. 171)

See the end of this post for a list of prominent creativity researchers who agree that there is no link.

If you search around on the web, you’ll see that people who claim there’s a link between creativity and mental illness often refer to studies by Nancy Andreasen, Kay Redfield Jamison, and Arnold Ludwig. But these are not well designed (using overly broad diagnostic criteria, or using sampling methods that might easily bias the results) and have been criticized by many scientists (see pages 168-171 of Explaining Creativity, the book The Insanity Hoax, and my review of it here).

So what’s going on with Kyaga’s findings? When a scholar publishes new research that contradicts a century of past scientific research, we have to pay close attention. Here’s what I think is going on.

The strength of Kyaga’s study is that he had access to huge databases of diagnoses, provided by the Swedish national health system–in their 2011 article, the team had data on anyone who had in-patient care for schizophrenia, bipolar disorder, and unipolar depression, between 1973 and 2003. But the weakness is that they didn’t have any way to measure these people’s creativity. So, as a proxy for creativity, the team used occupational information–also provided by the Swedish government, from self-reported information on the official census questionnaires in 1960, 1970, 1980, and 1990. In the 1990 census, the average age of the people Kyaga studied was over 45, so most of the people in this study are retired or deceased by now (22 years later).

Their definition of “creative people” was anyone who reported one of the following occupations on the Swedish government census: university teachers, visual artists, photographers, designers, display artists, performing artists, composers and musicians, and authors. For their control group of “uncreative” people, they used the occupation code for accountants and auditors.

Using these measures of mental illness and of creativity, Kyaga’s 2011 paper reported that there was no link between being in a creative occupation and being diagnosed with schizophrenia or unipolar depression. However, individuals with bipolar disorder were 1.35 times more likely to report being in a creative occupation than a noncreative occupation.

In Kyaga’s 2012 paper, again there was no link between having a creative occupation and schizophrenia or unipolar depression. In addition to the diagnoses examined in their 2011 paper (schizophrenia, bipolar disorder, unipolar depression), the 2012 paper also found no relationship between having a creative occupation and anxiety disorders, alcohol abuse, drug abuse, or committing suicide. The one exception, again, was bipolar disorder. But this 2012 study had a new finding: one specific occupational subgroup–authors–was more likely to be diagnosed with unipolar depression, anxiety disorders, alcohol abuse, drug abuse, and to commit suicide.

These two studies also analyzed the occupations of relatives of those individuals diagnosed with mental illness, and found that their relatives were more likely to be in creative professions (although the relatives were not diagnosed with mental illness: they were completely healthy). The Kyaga articles refer to a well-established genetic basis for mental illness, and suggest that relatives of people diagnosed with mental illness might have a latent propensity for mental illness that never manifested, and that this undiagnosed propensity is what caused them to choose creative occupations. There are several existing studies that provide evidence that an undiagnosed propensity to mental illness might lead to enhanced creativity, but these prior findings are controversial and these studies are difficult to interpret (see pages 172-174 of Explaining Creativity).

So the overall bottom line is that these studies did not find a link between creative professions and mental illness. But there were some sub-findings that seem to suggest a partial link–with bipolar disorder, with authors, and with healthy relatives. Why do the Kyaga studies seem to suggest a link between creativity and mental illness (although a fairly small link) when so many other large-scale studies found no link?

First, many of the subjects were diagnosed using the ICD-8 diagnostic criteria, which is no longer in use. ICD-8 used older and now-repudiated Freudian diagnostic criteria that American psychologists stopped using with the third edition of the DSM in 1980. (And that I suspect are no longer in use internationally in the current ICD-10.) Prior to the 1980 DSM-III, psychologists used overly broad conceptions of schizophrenia, and American diagnostic criteria have narrowed considerably; I suspect the same has happened with the ICD (See Explaining Creativity, page 167).

Second, regarding the 2012 finding that authors are more likely to commit suicide: Author occupation, as with all other occupations, was self-reported on the Swedish census. As Kyaga himself pointed out (2012), people with severe mental disorders are likely to have trouble finding a job, but they can still seek self-employment in artistic occupations. “Artistic occupations would then be the result of social drift rather than a creative propensity” (p. 7). In Explaining Creativity, I agree with this explanation (on page 171):

Artistic occupations don’t police their borders to keep people out, like other professions that require licensure…the arts can’t keep you out. And because our society has so many stereotypes about artistic professions being unconventional, it becomes a self-fulfilling prophecy: no one gets upset (and no one can fire you) if a painter is eccentric or unconventional. (p. 171)

Third, Kyaga’s measure of creativity, occupational choice, is a questionable measure of creativity:

  • Your occupational choice is the result of many things that are unrelated to creativity: Educational qualifications, opportunity, awareness, location of the job, salary, family values, and social commitments.
  • Although I’m a university teacher, I’d be the first to admit that a large percentage of university teachers are not creative.
  • Even within the artistic occupations, there’s a wide range of creativity. Take people who are photographers by occupation, for example. Some photographers are very creative, but most of them make their money shooting weddings and families, using highly constrained and repetitive formats.

It’s interesting that the Kyaga studies have received so much media attention, when they essentially conclude (2012 p. 6) that “individuals with overall creative professions are not more likely than controls to suffer from psychiatric disorders in general.” I think it’s because the very weak connections Kyaga has identified seem to affirm our widely held cultural belief that madness and creativity are related. Psychiatrist Albert Rothenberg, in his 1990 book Creativity and madness, observed “The need to believe in a connection between creativity and madness appears to be so strong that affirmations are welcomed and quoted rather uncritically” (p. 150).

But still, how else can we explain these results? I can think of a lot of alternative explanations that aren’t related to creativity at all. For example, it could simply be that accountants and auditors have a remarkably low level of bipolar disorder (lower than the general population) perhaps as a result of the steady and stable nature of their work. And likewise, they may have a lower level of suicide than unemployed authors, perhaps because they make a good salary and have a happy, stable life. Or, it could be that people who choose academic or artistic careers tend to come from higher social classes than people who choose accounting or auditing, and there’s something about class differences that accounts for differential diagnoses of mental illness (and this explanation could account for the different career choices of the relatives of the diagnosed patients). The two Kyaga papers have revealed some interesting patterns, that’s for certain, and I agree that they are worthy of further exploration.

If you’re a creative person, the good news is that there is lots of research showing that creativity is connected to normal mental functioning and elevated mental health. Much of creativity involves working with existing conventions and languages; you can’t make up your own separate universe. Creative success requires networking and interacting with support networks, and this requires social skill and political savvy. And creativity is mostly conscious hard work, not a sudden moment of insight; getting the work done takes a highly effective person. Many psychologists have demonstrated that when people engage in creative work, they attain a state of peak experience, sometimes called “flow,” that represents the pinnacle of effective human performance. Creativity is related to higher than average mental health–just the opposite of our belief in a connection between creativity and mental illness.

It’s interesting that those few scholars who still think there might be a creativity-madness link are psychiatrists (who treat mentally ill patients) rather than experimental psychologists. As Gordon Claridge wrote in 2009, “The purpose of researching the madness/creativity connection is not to rubbish the original but troubled mind; it is rather to give optimism to the mentally ill and their kin.” When mood disorder patients are told there’s a link between their disorder and creativity, it increases their morale and self-esteem. And they’re more likely to take their medication when they’re told “if you can control your disorder you have great creative potential” (Explaining Creativity, p. 176). As I conclude in my book:

Psychotherapists, in an understandable desire to help their suffering patients, could quite naturally be led to over-interpret what is rather weak evidence.

I empathize with these patients, and with their therapists, but I feel obliged to report the consensus that has emerged from decades of scientific studies: There is no link between creativity and mental illness. There may be a link between an undiagnosed tendency toward mental illness and elevated creativity; but we don’t really know, because that’s almost impossible to study (how do you study an undiagnosed tendency?). The jury is still out on that issue, and the Kyaga studies represent a contribution to that continuing debate. I’m impressed by the massive volume of data used in the Kyaga studies, and the patterns revealed are interesting, and worthy of further exploration.

This blog entry was republished on the Huffington Post on November 2, 2012 here.

The articles:

Kyaga et al., 2011. Creativity and mental disorder. British Journal of Psychiatry.

Kyaga et al., 2012. Mental illness, suicide, and creativity. Journal of Psychiatric Research.

The creativity research consensus:

All four textbook overviews of creativity research (mine and three others) reject any link: Weisberg (2006) says it’s a myth. Kaufman (2009) says the studies are flawed, and no link has been proven. Runco (2007) concluded that “there are indications that creativity has benefits for health” and that there are many flaws with the research claiming a link with madness, and concluded “this area receives so much study because it is newsworthy” (p. 152). Also see (Glazer, 2009; Rothenberg, 2001; Schlesinger, 2009; Waddell, 1998; Silvia & Kaufman, 2010; citations all from the bibliography of Explaining Creativity).

Weisberg, R. W. (2006). Creativity: Understanding innovation in problem solving, science, invention, and the arts. Hoboken, NJ: Wiley.

Kaufman, J. C. (2009). Creativity 101. New York: Springer.

Runco, M. A. (2007). Creativity: Theories and themes: Research, development, and practice. Burlington, MA: Elsevier Academic Press.