The vast majority of handwriting analysts are self-taught from popular books or trained by self- accredited correspondence schools or unaccredited night school classes. "Watch one, do one, teach one" could be the motto of the field. Although I could not find a single reputable textbook in psychological testing that treated graphology with anything but disdain, graphologists still claim to be a misunderstood and unfairly maligned branch of psychology. Few graphologists, in my experience, have had anything close to an adequate background in psychological measurement or modern personnel methods. Though they claim persecution from a hostile establishment bent on preserving its turf, graphologists seem oblivious to the fact that if their techniques really worked, and the orthodox professionals were as venal as they claim, the licensed practitioners would long ago have stolen these powerful tools and muscled out the self-credentialed amateurs. Sensitive to their resemblance to fortune tellers, graphologists claim they do not foretell the future. But what conceivable value would there be in describing a stranger if it were not assumed that the description would predict how he or she would act in the future?

There are few areas of human nature and mental or physical health that graphologists do not feel competent to assess. That a single technique could apply in so many different areas flies in the face of almost everything in modern research on psychological measurement. Such grandiosity and ignorance of relevant research is almost grounds in itself for dismissing graphology.

Graphologists claim to discern temperament (e.g., self-confidence, optimism, profligacy, complacency, or an explosive temper). They also believe writing reveals mental qualities such as intelligence, reasoning ability, and intuitiveness, and social traits such as introversion, friendliness, and dominance. In the workplace, they claim to rate leadership, reliability, diligence, attention to detail, propensity to be a team-player, and far, far more. On the moral and ethical side, graphologists pass judgement on people's honesty, trustworthiness, generosity, piety, cruelty, jealousy, criminal tendencies, etc.

How to determine marital suitability occupies a large portion of almost every graphology text. Sexuality is also supposed to have a multitude of written signs. Although graphologists typically demand to know the gender of the writer in advance, they are happy to pronounce on his or her secret sexual orientation and/or deviance, as well as promiscuity and capacity for intimacy. Could it be that they want to know gender in advance because it is too simple to check the accuracy of such a guess? (Ironically, untrained novices can discern the gender of writers in an anonymous sample with approximately 70% accuracy.) Probably for similar reasons, handwriting analysts will not guess the writer's age, but are happy to rate slippery attributes like "maturity" that offer plenty of room to fudge if challenged.

The alleged ability to derive medical diagnoses from writing has been alluded to already. As I have explained elsewhere, certain medical problems do affect writing, but in not in the way the graphologists assume . In the psychological sphere, graphologists claim everything from neuroticism and general stability to psychoses, phobias, depression, psychopathy, and a host of other clinical symptoms are all there for the asking.

Many of the aforementioned categories are combined when graphologists approach the criminal justice system. They claim to expose actual or potential criminal behavior as well as deceitfulness, lack of self-control, violence proneness, and sociopathic tendencies. Graphologists say they can help the police apprehend suspects and aid the courts in selecting juries and determining both guilt and appropriate punishment. They also say they can determine likelihood of recidivism and suitability for parole. McNichol's highly touted 1991 textbook, for example, provides exercises on how to spot a murderer, a babysitter who might use drugs, and a shop owner who cheats his customers . Marne's Sex and Crime in Handwriting offers numerous ways of exposing different kinds of criminals . Unfortunately, the betraying signs are all recognized after-the-fact in the writing of previously convicted felons. Marne, as usual, offers no evidence that she could reliably identify the guilty parties in an anonymous pile composed of scripts of convicts and upright citizens (and providing, of course, the contents of the scripts contained no useful clues, which they typically do).

It is bad enough that one might lose a coveted position on the basis of bogus advice, but to have one's standing in the community, and possibly even freedom, jeopardized in this way is frightening indeed. How would you feel being branded a thief because you have "desire-for-possession hooks" on your S's? Bunker , the founder of Graphoanalysis, the self-proclaimed most scientific school of graphology, seriously contends that these "acquisitive hooks" reveal a disposition to snag others' belongings.

The most transparently absurd claim in the whole field, one so bizarre that not even all graphologists endorse it, is that of "graphotherapeutics." According to believers, not only does handwriting unerringly reveal personal attributes, but if you should dislike any of the traits it discloses, you can expunge them by removing their diagnostic signs from your script . The oh-so-scientific Bunker devotes a whole chapter to showing how "changing handwriting will change personality." This merely underscores the contention that sympathetic magic is the real rationale behind graphology, for the essence of magical thinking is that causes resemble their effects and are therefore interchangeable. Case in point: graphotherapists insist that personality causes writing causes personality. What better evidence of this could we seek than Bunker's  assertion: "He [Bunker's client] had made a few changes in his writing -- not major changes, and he had achieved results." In this case, the writer, with minor retrenchments in his penmanship, was miraculously redeemed from his previous persona, that of a suicidal spendthrift. Here we see another common attribute of crackpot science, namely that effects are posited which are dramatically disproportionate to the magnitude of their alleged causes. If it sounds too good to be true, it probably is.

In philosophy, any doctrine can be dispatched merely by showing it to be internally inconsistent. Graphology is so vague and self-contradictory that devotees have ample room to explain away blatant errors. On the one hand (no pun intended), they say writing is such a sensitive psychological barometer that it varies, moment-to-moment, in response to subtle mood changes. But in the next breath, they will tell you writing is so impervious to change that you cannot hide your true nature by intentionally falsifying your script -- the real you will still shine through. Even though normal and disguised script from the same person look different, they still denote the same traits for the graphologist; but if those same disparities were found in the scripts of two different people the graphologist would say they were indicative of different traits. Graphologists also reply to those who say their writing varies in response to haste, writing posture, desire to make an impression, etc., that, though the writing is obviously different on those occasions, it still denotes the same personal attributes. If you try to deceive the graphologist by disguising your handwriting, your rigid personality stubbornly keeps the graphological signs intact, but if you change your writing at the behest of a graphotherapist, your malleable personality will realign itself to reflect the new, improved script. One manual for aspiring graphologists I reviewed cautioned neophytes not to become discouraged, because not everyone with a given sign has the suggested trait and not everyone with the trait has the sign. How could the system ever fail? This ability to be all things to all people makes graphology essentially unfalsifiable. On that ground alone, it can be excluded from the house of science.

Before presenting graphology's dismal record in empirical tests, let me first dispose of several oft-heard rationalizations for why handwriting analysis ought to work.

Handwriting is brainwriting. Yes it is, but walking is also controlled by the brain, so should we henceforth refer to it as "brainstepping," as Karnes and Leonard  wryly suggest? Why would we think that just because something is controlled by the brain, it necessarily correlates with any other traits, aptitudes or propensities? That is a claim to be supported with evidence, not glibly assumed. Vomiting has an associated center in the brain too. Does that justify using individual regurgitation styles to assess someone's intimate make-up?

In an earlier chapter , I maintained that research into the neural substrates of writing and personality actually supplies some of the best arguments against graphology. For instance, brain damage can alter either writing or personality, independently. There is no evidence that if a head injury affects personality, writing will necessarily change too -- as it should if graphology were valid. Furthermore, there is no reason to suspect that the brain mechanisms responsible for writing and those for temperament and aptitudes could be linked in the lockstep fashion necessary if graphology were to be taken seriously. Research on the physiological correlates of personality shows that individual traits are not localized in circumscribed brain areas that could conceivably be mapped, one-to-one, onto the minute muscle programs that create particular writing features. In the same chapter, I also noted that the graphologists' naive notions of how the brain determines personality (not to mention their outmoded conceptions of personality itself  are virtually identical to those of the discredited system of phrenology. Graphology would require a brain organization akin to that posited by the phrenologists to make it remotely plausible. For this necessary but unlikely brain organization to exist, it would either need to have evolved (and thus be inherited), or be acquired early in life. Either way, the implications for graphology are daunting.

If natural selection shaped brain structure such that it could allow connection of every minute character trait with a unique writing movement, graphologists should be able to suggest what possible survival advantages this profligate use of biological resources might have conferred. So far, no graphologist I am aware of has even realized that this is a serious impediment to scientific acceptance of graphology. Moreover, since the ability to write is, at most, 6,000 years old and the brain evolved to essentially its modern form eons before that, the putative circuits that would eventually link personality and writing must have evolved for some other purpose. What was this stupendous number of presumably dormant neural pathways selected for and what were the selective pressures that pushed the brain in that odd direction so long before humans got around to inventing writing?

If, instead, one views expression of personality in writing as an acquired skill, the difficulties for graphology are equally grave. Since writing is obviously a learned behavior, how does the brain unerringly modify every learned writing movement to make it congruent with each of the numerous traits a child will grow up to express? What kind of mechanism could conceivably ensure that everyone who is destined to be devious will acquire the same neural program, say, to make l-loops in the same way? Do parents ever say, "Susie, you are obviously gifted with leadership talents, be sure to form your capitals in this way, instead of the way your teacher showed you"? Writing also varies across language groups. What differences in early experience in the various linguistic communities ensure that the infant's brain will develop into the appropriate variant so that it will attach emerging personality traits to quite different writing movements if the child happens to learn the Chinese as opposed to the Russian, Arabic, Hebrew, or Roman alphabets? Writing in all of these scripts admittedly becomes personalized, but that individuality arises from biomechanical factors quite different from, and far more interesting than, the graphologists' parochial conjectures .

In sum, the graphologists' "brainwriting" argument is true but irrelevant to their claims. This rationale would only be necessary if there were a need to explain a proven relationship between writing and other personal attributes. Unfortunately for graphology, much empirical research, reviewed below, says such correlations are illusory in the first place.

Writing is individualized and personality is unique, so each must reflect the other. Aside from the obvious logical flaw in this argument, why should we accept, without good evidence, that any two admittedly idiosyncratic aspects of a person will necessarily bear any particular relationship to one another? True, forgeries have been exposed and writers of extortion notes convicted on the basis of distinctive penmanship, but does that imply that each of these identifiable writing features is reliably tied to another unique attribute of the author (in that person alone, not to mention everyone else who writes similarly)? Faces are sufficiently different to serve as personal identification on a driver's license, but the state still requires that you take the road test before certifying your driving skills. At one time, though, it was believed that facial features could reveal other personal characteristics. The pseudoscience of physiognomy held there were uniquely honest faces and criminal faces, generous faces and miserly faces. None but the woolliest New Ager could swallow this today. To show that mere uniqueness is no proof that every noticeable earmark is necessarily emblematic of something else, a colleague of mine relates the following story. As a quirky hobby, he trained himself to identify individuals from the distinctive sound of their footsteps in the hallway leading to his office. Many a visitor was un-nerved by being welcomed by name, long before he or she came into view. Here is a subtle, but reliably detectable, personal feature. It may be good for winning bets and amazing one's friends, but would any reasonable person seriously think that this toe-tapping signature would be a good basis for hiring an employee, rejecting a mate, or accusing someone of pilfering from the stockroom?

Writing is a form of expressive movement, so it should reflect our personalities. Elsewhere, I have reviewed the literature on expressive movements and facial expressions and shown why attempts to shore up graphology by appealing to these data fall short [14]. While it is true that there are legitimate studies linking a few global aspects of temperament to certain gestural styles, these data offer no comfort to the graphologists who attempt to ride on their coattails. The kinds of personal styles found to be loosely related to expressive body movements are much more general than the narrow traits the graphologists claim to infer from writing. A tendency to be forceful, irascible, or domineering might be readable from body language but, even there, the correlations are too weak to be useful in making the kind of detailed ascriptions graphologists attempt. And no reputable scientist has ever suggested that something as abstruse as piety or good business sense is any more likely to be encoded in gestures than, say, vegetarianism. Nonetheless, graphologists proudly note that handwriting analysis is mentioned in Allport and Vernon's 1933 classic, Studies in Expressive Movement. They are less quick to tell us, however, that in the estimate of those respected psychologists, "...the [graphological] terms employed often seem to obscure rather than reveal the personality." .

Similarly, facial expressions may sometimes betray a lie as it is being uttered, but no competent expert thinks everybody who tends to be duplicitous has a gestural tag for this or any other general inclination of the sort graphologists claim to detect. Moreover, the body and facial movements studied by credible researchers are biologically-based social signals. They are unlearned and convey useful information only because they are the same for everyone. Writing is quite the opposite, a learned skill that graphologists think mirrors individuality because it is different for everyone. And good actors are quite good at faking body and facial signals, something allegedly impossible with writing, according to graphologists. Perhaps most damaging of all to the graphologists' cause is the evidence that such information as can be gleaned from unconscious movements and facial expressions is readable by anyone without formal training. There is no need to pay anyone a fee to interpret what they mean.

The police and courts use graphology, so it must be valid. I am tempted to say, "Ronald Reagan used astrology, so it must be valid" and leave it at that, but there are a few other useful lessons to be derived here. Yes, unfortunately, some misguided officials have employed handwriting analysts in forensic settings, but the practice is not as widespread as graphologists imply. As a group, police officers, lawyers, and judges are no more or less prone to erroneous beliefs than anyone else. Faced with difficult decisions where no other method offers certainty (an ideal breeding ground for superstitions), some in the criminal justice system occasionally get swept up in hopeful nonsense, just like the rest of us. The vast majority do not endorse graphology or psychics, however. Graphologists occasionally offer their services to the police and get a polite hearing, as any citizen is entitled to. And for reasons related to the subjective validation effect, discussed below, the recipients may well have been more impressed than the performance warrants. Of course, like psychics who claim to solve crimes, a few chance hits and reliance on conventional clues, boosted by a bit of embellishment and unabashed self-promotion, can establish a high but unearned reputation.

The artificially inflated reputation enjoyed by handwriting analysis is largely due to the tendency to confuse the profession of graphologist with that of a questioned document examiner (QDE). As Dale Beyerstein  has observed, nonsense often rides piggyback on sensible knowledge, and graphology, though it bears only the most superficial resemblance to scientific document examination, misappropriates the latter's well-deserved prestige. Both fields analyze handwriting, but that is where the similarity ends.

A QDE is a scientifically-trained forensic investigator who also has considerable knowledge of the history of papers, inks, writing implements, systems of penmanship, and styles of expression . QDEs are respected experts who are frequently consulted by the police and the courts. Their modus operandi is quite different from that of a graphologist, however. The job of a QDE is to establish the provenance and authenticity of documents, some of which are handwritten. Unlike a graphologist, a legitimate QDE would never attempt to discern the personality of the writer from the script he or she examines. Where appropriate, the QDE will compare the writing in disputed documents to known samples from the hand of the putative author. Thus a typical question for a QDE might be, "Is this an authentic letter from Mozart to his patron or a clever forgery? Or, "Did the defendant in the dock write this ransom note?" By comparison, a typical question for a graphologist might be, "Does this writer harbor a secret resentment of authority?"

If need be, a QDE will chemically analyze the ink, microscopically examine the fibers and watermarks of the paper, and look for distinctive marks left by different kinds of writing instruments. In addition, he or she might compare grammar, style, and punctuation to social or historical norms, all for the purpose of establishing when, where, and by whom a given document was written. The exposure of the infamous "Hitler Diaries" as forgeries showed QDEs working at their best . As consultants in litigation or historical disputes they are asked only to rate the probability that a given person wrote the document in question, not to pass on the guilt, innocence, or any other psychological trait of the alleged author. That a few QDE's also practice graphology on the side also leads to confusion in the public mind. Most QDEs are just as unhappy at being confused with a graphologist as an astronomer would be if mistaken for an astrologer.