The Zodiac Killer's handwriting looks like the handwriting of a serial killer. His alphabet is sharp-looking and small. His words slant to the right, sometimes severely. Some of the loops of his letters don't close; he was writing in a hurry. His spelling is bad (intentionally, it was believed), his punctuation poor. His capitalization is irregular; he unfailingly granted an uppercase Z to his own name, but often kept others' in lowercase while treating improper nouns like proper ones. There is something about his penmanship that is distinctly male.

But here's the question: Does the Zodiac's handwriting look creepy to me because the writing itself is creepy? Or does it look creepy to me because I know it belongs to the Zodiac?

Copies of the Zodiac's letters are easily Googled, but I came across them first interspersed between the pages of the classic bestseller Zodiac, by Robert Graysmith. I witnessed them in very literal context; any judgments I might make about the handwriting are biased, informed by what I know about the writer. But if I received a cheerful postcard from my great aunt on vacation in Hawaii, written in the exact same handwriting as the Zodiac's, would I still find it creepy?

By and large, graphology has been shunted to the pseudoscience pile. The British Psychological Society puts the practice in the same camp as astrology.

GRAPHOLOGY, OR HANDWRITING ANALYSIS, is believed to have originated with a group of French clergymen in the 19th century. One of them, Jean-Hippolyte Michon, is today considered its father. Michon was first introduced to the idea that personality could be ascertained through handwriting by his philosophy teacher, Father Flandrin. It was Michon who came up with the term "graphology," and who later founded the Society of Graphology in Paris in 1871. Michon wrote two books on the topic (Système de graphologie in 1875 and Méthode pratique de graphologie in 1878) and traveled around Europe teaching their principles.

Interest in graphology soon spread across the ocean. June Downey, also the first woman to be named chair of her department in a state university, is among those credited with spearheading the study of handwriting in the United States. Her graduate thesis (submitted in 1908), "Control Processes in Modified Handwriting," examines the use of handwriting analysis in personality assessment.

As the graphology field grew, it also fractured. The American Grapho Analysis Society, founded in 1929 by Milton Bunker, established a subfield called "graphoanalysis," in which one's handwriting strokes are measured against a standard in order to assess one's psychological traits. Thus there grew to be two groups: those who adhered to Bunker's more stringent approach, and those who believed that graphology should be done holistically.

For a time, graphology was viewed with moderate enthusiasm—particularly as a potential hiring tool. If personality traits (morality and work ethic central among them) could be determined through prospective employees' handwriting samples, employers might be able to predict job performance, and avoid the duds that skated through other, more traditional screening tactics.

Graphology had cheerleaders elsewhere, too. In a paper called "Graphology: A New Marketing Research Technique," published in 1967 in the Journal of Marketing Research, professor James U. McNeal argued that marketing researchers would do well to incorporate graphoanalysis into their strategies for capturing a certain kind of consumer. McNeal suggested, for example, that graphoanalysis could be used by retail businesses to evaluate customers' credit applications or by companies to learn about the personalities of customers who write them letters of complaint.

But here, as with all potential uses of graphology, there is a major caveat. Although the quote is preceded by six positive paragraphs about graphology's potential uses, this comes straight from McNeal: "Another serious problem is the reliability and validity of graphology; there is little substantial evidence about these factors."

In fact, a 1982 meta-analysis that consulted over 200 graphology studies concluded that graphologists were largely unable to use subjects' handwriting samples to predict any particular personality trait or how they'd score on other known personality tests (like the Myers-Briggs test or the Eysenck Personality Questionnaire).

Neither has graphology, under scientific examination, proven effective in determining job performance; a 1989 study found that graphologists were just as bad at guessing how well someone would do in a new job as anyone else.

By and large, graphology has been shunted to the pseudoscience pile. The British Psychological Society puts the practice in the same camp as astrology. Psychologist Dr. Rowan Bayne, telling the BBC how he really feels, called it "useless" and "absolutely hopeless."

I HAD THIS IDEA that I'd print out copies of the Zodiac's letters, cut off the signatures and the dates, mail them to a few graphology experts, and compare the results. But then I thought of the kinds of questions that might raise, and I chickened out. Besides, it turns out my clever idea is not super original.

Analyzing the handwritten notes and letters of serial killers (and other forms of psychopath) is something of a hobby on true crime forums. Chris Yarbrough, self-proclaimed "computer geek" and owner of, a site dedicated to commentary on unsolved crimes, submitted one of the Zodiac's later letters to two different online graphology sites and listed the results he was sent from each.

The letter Yarbrough sent in for analysis is one of the Zodiac's most violent. In it, he claims responsibility for the murder of cab driver Paul Stine (which he proves by including a piece of blood-stained shirt), mocks the police for their inability to find him, and threatens to blow up a school bus full of children. As such, I expected the analyses to depict a cold-blooded killer.

Instead, the analyses manage to be both broad and contradictory: One (perhaps fairly) claims the Zodiac's unembellished "m's" and "n's" suggest an absent sense of humor, while the other calls him "sarcastic" and "funny." Studies suggest that part of what makes graphology "work" on people is the same mechanism that applies in astrology: the Forer effect, wherein people eagerly claim vague personality descriptions as highly descriptive of themselves, when in reality they describe almost everyone.

We don't know who the Zodiac Killer is (or was), so, I suppose, it's impossible to say with certainty whether the graphologists got any of it right. I'm no psychologist, but the analyses certainly seem counterintuitive. The first claims that because his words don't end with a curving stroke, he doesn't need to be noticed. This is a curious thing to claim about a letter sent to the police to boast about a murder.

Could there ever be a better example of missing the forest for the trees? The Zodiac is someone who "relates easily to others' problems," according to the second report. "He is a people person." Who knows, maybe he would have agreed.