It is widely known that writers -- but particularly poets -- have much higher incidences of mental ailments such as depression, alcoholism and bipolar disorders than the general human populace. Nobody has a mental image of poets as being perky and gregarious, especially since some fairly well-known poets have ended their lives with suicide:
Many more people who've been amateur or aspiring poets have also killed themselves, but at the same time, many poets (professional and otherwise) avoid or never even consider depression's final solution. While the reasons for poets' suicides are varied, a pair of researchers recently published a paper comparing the work of nine of the above poets (see footnote) with work from a closely-matched control group of nonsuicidal poets. The researchers found that the language choices of suicidal poets held some telltale signs of their downward spiral.
James Pennebaker (a psychologist at the University of Texas) and Shannon Wiltsey Stirman (a psychology grad student currently at the University of Pennsylvania) published their research in an article called "Word Use in the Poetry of Suicidal and Non-Suicidal Poets" in Psychomatic Medicine in 2001.
Instead of finding verses that overtly dwelled on doom and death, they found that suicidal writers' work displayed a sense of isolation and detachment from other people and extreme introversion.
"One of the most telling words of all is the word 'I'." Dr. Pennebaker told a reporter for a Reuters article. "People who are suicidal or depressed use 'I' at much, much higher rates, and there's also a corresponding drop in references to other people."
The closer the poets moved towards suicide, the less they used words like "listen", "share", or "talk"; nonsuicidal poets tended to use such references to human interaction more and more as they aged. And while the suicidal poets did use more sexual and death-related imagery, they didn't especially dwell on topics like hate or anger. There was no real difference in the emotional content (whether positive or negative) between the two groups of poets.
Pennebaker cautions that not everyone who writes self-preoccupied poetry is going to kill themselves; their research just indicates a higher risk, not a guarantee.
The poets listed here who were not included in the study were Thomas Chatterton and Marina Tsvetayeva. Please send me a message if you know of other poets who should be included here.
An abstract of the Pennebaker/Stirman article can be found at: http://www.psychosomaticmedicine.org/cgi/content/full/63/4/517