Visit Your Local PBS Station PBS Home PBS Home Programs A-Z TV Schedules Watch Video Donate Shop PBS Search PBS
SAF Archives  search ask the scientists in the classroom cool science
scientists from previous shows
cool careers in science
ask the scientists

Photo of Barry Beyerstein Paper Personality -- Barry Beyerstein

Neuroscientist Barry Beyerstein investigates the use of handwriting analysis as a tool for psychological measurement. Find out more about Barry's research into graphology by reading his answers to questions from Frontiers viewers.

QSince there's no scientific basis for assessing personality based on handwriting, do you personally think it's ethical for companies to use results from graphology when considering individuals for jobs or loans?

I think it is scandalous that a pseudoscientific "character reading" method like graphology should be used to make decisions that can seriously affect people's reputations and life prospects. It is highly unethical that many of the purchasers of graphological services do not even tell the unsuspecting writers that their script is being subjected to this thoroughly discredited procedure, so they have no opportunity to object if their reputations are blackened or they are denied a job, promotion, loan, etc., on these spurious grounds. Some graphologists even crow in their advertisements, for instance, that you can have your spouse's handwriting SECRETLY analyzed to see if he/she is sleeping around! They say they will secretly advise employers on employee honesty, drug use habits, sexual deviance, violence prone-ness, etc., and the employee need never know. In our book, "The Write Stuff" (B. and D. Beyerstein, eds., Prometheus Books, 1992), we also provide instances where police, the judiciary, and even a member of the Canadian National Parole Board, had consulted graphologists regarding guilt of suspects, severity of sentencing required, and suitability for release of convicts on parole. Another ethical lapse we have noticed frequently is that many graphologists we have interviewed or debated were quite ignorant of confidentiality requirements -- i.e., they were willing to show around, in public forums, their potentially damaging, pseudoscientific analyses with names of the individuals still prominently displayed.

When my brother Dale and I assembled our book, mentioned above, we asked two lawyers, one Canadian and one American, to survey the relevant statutes, case law, precedents, constitutional protections, and administrative edicts in their respective jurisdictions that might be useful if someone, alleging they had been harmed, were to mount a lawsuit against a handwriting analyst. Both lawyers came back with fairly pessimistic outlooks in the chapters they wrote for our book. In both countries, libel and slander are always actionable, but graphologists usually stop short of doing this outright. In certain governmental settings, there are statutes that require all instruments used to assess people for personnel purposes to be able to demonstrate suitable scientific backing -- breach of these requirements, where they exist, would be actionable. Also, there are provisions in some jurisdictions against using personnel selection procedures that discriminate against women, visible minorities, etc. Graphology passes on this requirement because it works equally badly for everyone. Other than that, in both the US and Canada, in the private sector, there is very little to prevent a believer from using graphology, astrology, tea leaf reading, or any other equally-discredited method for choosing employees, marriage partners, or loan grantees.

To answer the question directly, I think it is highly unethical to use a procedure that has failed numerous fair scientific tests of its validity, especially where important decisions are being made that could result in an unfair advantage or disadvantage to anyone.

Q One thing that wasn't clear on the show is just how widespread the practice of using graphology is. Job and loan applicants were mentioned - are there others? Is this criteria used by a large percentage of companies in the USA or Canada?

When my brother and I began working on "The Write Stuff," we intended to include a chapter documenting how widespread the use of graphology really is. In the end, we abandoned the exercise because the data were so difficult to obtain and so unreliable. For example, we were told at one point that the Vancouver School Board was secretly hiring a graphologist to identify the actual (AND POTENTIAL!!) child molesters in the local teaching ranks. When we inquired, we were told by senior spokespersons that nothing of the sort was occurring. Further investigation revealed that what we had feared was indeed happening, but that it was not an official policy of the School Board or its senior administration. An over-zealous and ill-informed employee of the school district's personnel department was doing an unofficial "study" with his friend, a local handwriting analyst (the whole sad story is contained in the introductory chapter of our book). If we had simply taken the honest, but mistaken, word of the first officials we contacted, we would have been misled into thinking graphology was not being used. We have found this to be the case quite often -- e.g., a lower ranking personnel officer will use graphology and not tell his/her superiors. Thus the company would erroneously end up in the "negative" column if we simply asked the vice-president for public affairs whether they were using graphological services or not.

In other cases, we have simply been lied to by company officials who realized that a public acknowledgment that they were using graphologists' services could lead to ridicule and/or outrage -- e.g., in one instance, we had seen actual documents that angry company employees had "leaked" to us -- showing that graphologists WERE being consulted by their firm -- but when we approached the officials whose names were on the leaked documents they flatly denied any such thing was happening. Thus the figures gathered by any survey that simply asked companies if they currently use handwriting analysis, or have used it in the past, would obviously be distorted by this wish to avoid what we call "the giggle factor."

On the other side, graphologists like to boast about the big companies, agencies, etc., they have worked for, and unfortunately, these claims often turn out to be true. However, these claims are often inflated too -- for instance, in some of the cases we checked out, all that had happened was the graphologist came by offering a "free trial offer." The company accepted but found the advice worthless and declined to purchase any further services. The graphologist still claimed, to the company's dismay, that he had been "a consultant" to the firm. In other instances, companies had been paying users, but quit when they learned the facts about graphology. They weren't anxious to own up. Occasionally, I have been told off the record that the company uses graphology, but told the respondent would deny it, if I quoted him or her.

That's a long-winded way of saying we don't really know how prevalent the use of graphology is in industry. The problem is: "figures don't lie, but liars can figure." We're pretty sure the prevalence is higher in France, Germany, and Israel where, for historical reasons, the "giggle factor" for graphology is less pronounced in the population at large than it is here in North America. But it's safe to say that there are quite a few graphology firms earning a good living on this continent too. The exact number of clients they have, we don't really know.

Q Is it possible that the questions the subjects are asked to write about -- how they met a challenge, etc. -- are what REALLY give the graphologists clues to personality?

There are two primary reasons that handwriting analysis SEEMS to work in everyday settings while it fails routinely in properly controlled scientific tests. One is the "subjective validation" effect demonstrated by Ray Hyman (when he did the palm reading) in the SAF episode on which we appeared. I.e., we unconsciously "read in" particulars we know about ourselves in response to what are in fact quite vague and variously interpretable assertions by palmists, graphologists, astrologers, etc. -- this makes the reading seem highly relevant when it really isn't. That is why, on the program, Alan Alda and I tried to find our own analyses from a pile of anonymous graphologist's reports. If the reports had really been specific and accurate, we should have been able to recognize our own descriptions out of the pile.

In our book, "The Write Stuff," we show that, in a couple of hundred well-conducted experiments, graphologists have never been able to pass this kind of test (which controls for spurious accuracy due to the subjective validation effect -- also known as "The Barnum Effect" after the great circus showman).

More to the point of the question, the other reason handwriting analysts seem to "get it right" in unscientific tests is that they infer things about the writer from the contents of the writing itself (though they claim they don't). I think it is very suspicious that they have all kinds of lame excuses why they can't use a page of writing copied out of an encyclopedia, for instance -- it wouldn't have all the raw material for shrewd guesswork. Graphologists typically demand, instead, what we critics call "rich texts," i.e., personal letters or written answers to probing questions they themselves pose. They swear they pay no attention to the contents, but there is strong evidence to the contrary. In the "test" that Alan and I participated in, we had to answer about a half dozen questions, the answers to which would give lots of useful information about the writer.

This was apparent to researchers such as the Dutch psychologist, Abraham Jansen. Jansen had graphologists analyse letters of application from job applicants (which obviously should contain lots of information about the kind of person the writer is). With this kind of input, the graphologists scored somewhat above chance expectancy on Jansen's scales. (Even at that, they didn't do a very impressive job.) Then Jansen had the same letters typed out on sheets of blank paper. He then gave the typed versions to non-graphologist volunteers to read and asked them to make the same ratings he had asked the graphologists to make. These volunteer raters were just as accurate in their ratings as the graphologists had been, with nothing but the information contained in the typed versions of these "rich texts" to go on. Our book contains many similar examples from the research literature.

Knowing this, I laid a trap for the graphologists in our test. I objected in advance that they were stacking the deck in their favour by demanding that we do their work for them (by telling them in advance all about ourselves in the way we answered their questions). I then proceeded to answer their questions truthfully, but using language that is open to different figurative interpretations. I assumed they would take the more conventional interpretation in each case and crafted my answers so that if they were paying attention to the contents, rather than just the handwriting, they would get a systematically misleading picture of what I am really like. This should in no way have harmed them if graphology is valid, because the contents of the writing should be irrelevant to the attributes they ascribed to me. I wrote down the errors I predicted they would make in advance -- before I sent them my handwriting sample. The graphologists fell for it. For instance, they asked us to describe a major adversity we had overcome. I said getting over an illness was such an instance and that my faith had carried me through. In my personality description they said (as I predicted) that I had health problems. In fact, I never said the illness was my own or that it was a conventional medical ailment (though that is what I expected them to assume). The illness was depression and it occurred in someone I care about deeply, not myself. The faith I referred to, which was amply rewarded, was my faith in my own scientific discipline of psychopharmacology. The antidepressant medications that were given to the person to whom I was referring worked very well, just as my faith in science had led me to expect. You would never have guessed that all this was true from the analysis I got from the graphologists, however. I also laid out some false suggestions that I myself suffer from frequent depression (which is not true) and they fell for that too, making me out to be quite the opposite to the kind of person I really am.

On another note, these allegedly scientific personality readers had the audacity to suggest that there were indications in my handwriting that I am a substance abuser and to suggest that my parents might have molested me as a child. Both, fortunately, like their assessment of my health are categorically untrue, but it should be obvious the kind of harm such allegations could produce if they reached the wrong hands. The unfalsifiable, and hence pseudoscientific, nature of what graphologists do is obvious from the following. For instance, they stated: "The subject is withdrawn and emotionally aloof." (I think my friends and family will agree this is untrue). However, like the palm readers Ray Hyman described on the program, they cover themselves for times when they completely blow it in the following way: In my graphological analysis they then go on to say that I "can compensate by putting on a pleasant facade." So if they get it right, they take credit for being accurate, but if I say they're wrong, they say that, underneath, I am in fact like they say I am -- it's just that I'm good at covering it up. There were many other inaccurate things in their reading of me that could have been quite harmful had this reading gone to a potential employer, for instance.

Q Is the study mentioned on the show (the one at William and Mary) being continued to test the effectiveness of graphology? I'm curious if they have yet compiled enough data to come up with a conclusion. Thanks for any info you have.

I had been aware for some time that these studies were under way, but none to my knowledge have appeared in the appropriate peer-reviewed scientific journals. It's possible they have come out very recently and I just haven't seen them yet, but I don't think so. From what the professor in question said on this episode of SAF, he has nothing that offers any real support for graphology yet. I should say, however, that even if he did have a positive result, this would have to be taken in the context of nearly two hundred previous studies that have shown that graphologists cannot live up to the claims of practitioners to accurately describe personality, moral attributes, abilities, etc.

No important question in science is ever settled with one or two studies. It is always the overall picture derived from many experiments, done by different researchers, using different methodologies, that should be used to decide where the truth in a scientific dispute is most likely to lie. There will always be a few outliers or flukes in the literature that contradict the story supported by the overwhelming mass of converging studies. When that is the case, its a pretty safe bet that the small number of contradictory studies will be due to chance variability, undetected errors in the apparently contradictory studies, etc. In our book, "The Write Stuff," an Australian researcher, Geoffrey Dean, did a meta-analysis of virtually every study to date that purported to evaluate handwriting analysis. In technical terms, Dean concluded, "the effect size is essentially zero." In lay terms, the scientific literature overwhelmingly supports the notion that handwriting analysis is pseudoscientific bunk.

Q In a previous answer you mentioned the "giggle effect" and said it less pronounced in other countries than in the USA. I'm wondering just what the giggle effect is and why other nationalities would have a different view of graphology than Americans.

This was just my own cutesy way of saying that, for some reason, the average European and the average Israeli don't seem to find the idea that personality traits could be encoded in handwriting as humorously naive as many North Americans do (unfortunately, there are PLENTY of people on this continent who still think graphology is a reasonable tool for revealing what people are really like, just the same). I was thinking of Martin Gardner's famous phrase that sometimes a good horselaugh is worth a thousand syllogisms.

As to why graphology is considered more respectable in Europe and Israel, I don't have a definitive answer, but I think it's an accident of history. Joe Nickell summarized the history of graphology in our book, "The Write Stuff" (Prometheus Books, 1992) and he shows how this form of divination acquired a patina of respectability thanks primarily to a few forceful, well-connected personalities who championed its cause among the "establishment" in those countries. This gave it an enduring aura of authority that still lingers on today. In the US and Canada, graphology has always been more of a fringe practice, rightly associated with astrology, palm reading, and circus sideshows in most people's minds. For some reason (probably because its advocates weren't as socially well-connected here as there), graphology never penetrated the powerful precincts of those who influence public opinion to the same degree here as in Europe. North American psychology books, for instance, have always been much more likely to hold up graphology as an example of a bogus personality test.

As for the Israeli-connection, I think graphology got its foothold when a number of European refugee graphologists immigrated there and set up shop. Amidst the social flux and upheaval of founding the new state, there wasn't as much of a pre-existing establishment for graphologists to crack -- they had the opportunity to get in on the ground floor, so to speak, and become the establishment. Once graphology gets a toe-hold, it is very hard to eradicate, for reasons I alluded to in my previous answers. The "subjective validation effect" (a.k.a. "The Barnum Effect") makes informal demonstrations of graphology very convincing, despite the fact that proper tests that control for the subjective validation effect show graphology has no scientific validity. If there is less of a likelihood that someone will giggle at you when you suggest consulting a graphologist, you are more likely to give it a try, even if you are somewhat dubious about it to begin with -- if that initial barrier is broken, you are more likely to get "hooked" by the subtle psychological factors that make it SEEM to work when it really doesn't.

Q I want to know if graphology can be used for evidence in criminal cases or court.

In "The Write Stuff" and in my chapter on graphology in Gordon Stein's "Encyclopedia of the Paranormal" (Prometheus Books, 1996), we distinguish between the professions of "graphologist" and that of a "questioned document examiner." Unfortunately, these two occupations are often confused in the public mind -- the former is a practitioner of pseudoscience and the latter is a perfectly respectable member of a branch of forensic science. Questioned document examiners (QDEs) often testify in court regarding the probability that a defendant was the writer who forged the cheque, wrote the ransom note, etc. (Remember the "Ivan the Terrible" case in Israel where an American citizen was accused of being the vicious death-camp guard? Joe Nickell was one of the document examiners who showed that the SS guard's ID card and its signature were real, not a forgery -- they never even considered trying to say whether the defendant was guilty or not, however -- that's the crucial difference between the two kinds of practitioner). QDEs deal with the authenticity and authorship of written documents. Graphologists, on the other hand, claim to be able to advise the court on the believability of witnesses, the guilt or innocence on the accused, and even the severity of the sentence that should be passed, simply from reading supposed signs of "trustworthiness" or "evil" in the handwriting of the suspect. The vast majority of police, lawyers, and judges see this for the nonsense that it is, but a few ill-informed judges, here and there, have accepted graphologists' testimony in court. Fortunately, this has been quite rare, as far as I have been able to determine.

The picture gets a bit muddy, however, when a few questioned document examiners also dabble in graphology (the vast majority don't). They may testify in court, legitimately in their capacity as QDEs with respect to the provenance of a given document, but they then let on to an unsuspecting public that the graphology work they also do is equally legitimate. When our book came out, I had many nice letters from questioned document examiners, thanking us for pointing out the huge difference between what they do and what graphologists do. The vast majority of QDEs are just as upset about being confused with a graphologist as an astronomer would be if she were mistaken for an astrologer.

QHow do graphologists decide what a "normal" writing trait is when we are all taught to do it in the same way and then proceed to do it our own way?

Graphologists don't worry about what the "normal" (i.e., average) way of producing the written "sign" is -- they simply interpret the way we form these various features on the page in much the same way ancient oracles interpreted the entrails of oxen or smoke in the air. I.e., it's a kind of magical divination or fortune telling where "like begets like." In other words, whatever the symbolic thing you are interpreting reminds you of is what you attribute to the person for whom the reading is being done. As I said on the SAF program, graphology is all metaphorical or allegorical interpretation and therefore magical, not scientific. For instance, if your letters slant forward, you are a forward, outgoing person; if you write big, you "think big" (i.e., you are expansive and extroverted), if you make big capital "I"s, you are egotistical; if your letters are spaced far apart you are stand-offish and shun close contact with others, and so on. My favorite, which really shows the metaphorical nature of handwriting analysis, is that how sexy you are is supposedly revealed by how big and bulbous the loops of your letters are that dangle below the line! I mentioned that one during the taping but I guess it was a bit to lascivious to make it into the final version of the show. If you want more examples of this sort, see my chapter, "The Origins of Graphology in Sympathetic Magic," in our book, "The Write Stuff."

The big dangers and unfairness come when graphologists use these allegorical methods to attribute abnormal traits to innocent people. For instance, one very prominent graphologist says he can identify potential thieves because they have "acquisitiveness hooks" on their letters and they are therefore likely to snag other people's property. How would you feel if someone (who claims to be using a scientifically valid method of character reading) were to say something like that about you to a potential employer, potential spouse, the police, or the Internal Revenue Service?

QIf handwriting doesn't tell your personality, how do you explain that everyone's handwriting is different even though we were all taught to write the same?

One of the things that gives graphology a certain appeal to many people is that we all have unique personalities and, despite the fact that we are all taught the same handwriting moves, we all develop quite idiosyncratic "hands." Indeed, we can usually identify who a letter is from merely by glancing at the handwritten address. If writing is so individualized that the QDE's referred to in my answer above (about graphoplogy and the courts) can pin a piece of written work with a fair degree of certainty on a particular writer, it seems plausible to some people that we might encode some of our individual personality traits in our writing too. Unfortunately, scientific tests overwhelmingly say otherwise. Our faces are unique enough to serve as personal identification too and the old and abandoned pseudoscience of physiognomy once claimed that you could read character traits from faces, just like the graphologists claim for writing. It's up to the graphologists to prove that unique features of our writing correlate with unique features of our psychological makeup. This they have failed to do.

Why do we have such individualized handwriting? In my chapter, "Handwriting is Brainwriting. So what?," in our book, "The Write Stuff," I review the scientific literature on writing that suggests that the individuality of our handwriting arises primarily from biomechanical differences in our bones, muscles, tendons, etc. We just "fall into" certain patterns because they "feel right." Some of us do embellish our writing voluntarily as a form of self expression, but there is no evidence that anyone can reliably infer anything but completely obvious and general attributes from this. For instance, a flamboyant, artistic person might write with a more artful hand than someone else, or a neat person may write more neatly and precisely than a real slob, but there are so many exceptions to this sort of relationship that it is practically worthless for attributional purposes. At any rate, you wouldn't need a graphologist to tell you this (or that a stingy person might write closely compact lines to avoid wasting valuable stationery) and secondly, there are lots more valid predictors of these kind of traits to rely on. If this kind of very general attribution were all graphologists were claiming to, I wouldn't waste much of my time criticizing them. They do much more dangerous and unsupportable things, however. They claim that they can detect potential thieves, sex offenders, adulterers, drug abusers, tax evaders, etc., from "signs" in handwriting. There is no more scientific evidence for this kind of trait being encoded in handwriting than there is for the trait of vegetarianism being revealed by "signs" in one's handwriting.

QDo you know long has graphology been going on or how it got started?

In our book, "The Write Stuff" (Prometheus Books, 1992), Joe Nickell recounts the history of graphology. He found antecedents to modern handwriting analysis in ancient Hebrew, Chinese, and Roman texts. In the Middle Ages it was a popular craft among itinerant fortune tellers. In a chapter in "The Write Stuff" that I co-authored with a Chinese collaborator, Zhang Jing Ping, we delved more deeply into the ancient Chinese roots of graphology, which developed independently (but quite similarly, nonetheless) from its Western counterparts. Basically, graphology is probably as old as writing itself, because people have always grabbed ahold of anything that is unique about one person vis-a-vis the rest in their age-old search for anything that will reveal what a person "is really like." (So you they can predict and manipulate their fellows who are always trying to send self-serving false signals about themselves).

The Chinese idiographic script is ideal for this kind of auguring (i.e., divining). The principles of calligraphy analysis devised by the Chinese were very much like those that emerged independently in Western cultures. Before writing existed, facial features were used for this kind of "character reading." In the process called "physiognamy," facial features were interpreted in much the same way that graphologists interpret written features. These are both examples of magical divining processes -- basically, the oracle or diviner attributes, by metaphor or analogy, the traits she is reminded of by looking at the thing being analysed to the person being "read." (See my earlier answers in this forum for examples of this "like-begets-like" form of magical thinking). Unfortunately, none of these processes of metaphorical or allegorical attribution of traits, graphology included, works when put to a scientific test.

The modern history of graphology (pretty much unchanged to the present day) begins with the 17th century physician Camillo Baldi who lived in what is present-day Italy. Several other noblemen from the Italian peninsula re-worked Baldi's treatises until the focus of interest shifted to France. Several influential French Catholic clergymen (including the Abbe Flandrin and the Abbe Michon) formed, in the 1800's, the basis of what is still practiced as graphology today. It was the political and social prominence of these churchmen that is probably responsible for the fact that handwriting analysis remains more publicly acceptable in France than in North America today (see my answer from an earlier set in this forum, regarding why the popularity of graphology varies widely in different countries today).

QI have heard that some people claim that they call spirits when they write (automatic writing). Can you think of an experiment that would prove it is the person, not the spirit, who is doing the writing?

Ray Hyman, who appeared on the same episode of SAF that I did, is one of the world's leading experts on this phenomenon which is known as "ideomotor movement." This refers to the fact that parts of our brains that are actively processing information, but working outside of our present awareness, can initiate muscle movements that FEEL as if they are not caused by ourselves. I.e., there is no sense of willing the movement or any sense of effort or direction as it unfolds. It may therefore feel like it is being caused by some outside force. This is the primary explanation for dowsing, the ouija board phenomenon, apparent spiritual possession, and the pendulum experiment that Ray Hyman did with Alan Alda on our SAF episode.

To show that the information really comes (unbeknownst to him or her) from the person doing the "automatic writing," I would try a couple of things. One would be to ask the supposed spirit being contacted to identify himself or herself in detail and then ask him or her some very pointed and detailed questions. I.e., questions that someone who had really lived in the time and place in which the spirit supposedly did would likely know -- but most ordinary people today would probably not. That is to say, you would use the same technique that counter-espionage police would use to try to break the "cover" of a suspected spy. A good spy would have been drilled on lots of details about the time and place of his adopted background story, but good interrogators can usually begin to spot errors, inconsistencies, contradictions, etc., if they keep asking probing questions and carefully record and cross reference the answers.

So, if the "entity" supposedly moving the pen during the automatic writing was supposed to be the spirit of a Civil War soldier, for instance, it would be easy to ask details about the equipment, regulations, drills, battles, etc., that anyone who had really "been there" would know but that would soon exhaust the knowledge of the average person holding the pen. Likewise, if the spirit was supposedly an accountant during his life in this world, ask arcane questions about double entry book keeping, marginal rates, interest compounding, auditing practices, certification programs, . . . .

If you wanted to get fancier, you could put well-sealed earphones on the person handling the pen. You could then feed questions into the room where the "spirit" could hear them along with the onlookers and into the earphones at the same time. On some trials, the same question goes into the room and over the earphones, on other trials the room gets one question and the earphones get a different one. If it's unconscious ideomotor activity that is really driving the pen, the holder of the pen should answer the question that came through the earpiece rather than the one that was voiced in the room when the two questions differ. Obviously, for this to work, they would have to be questions that couldn't simply be answered with a "yes" or "no."

The physicist Michael Faraday did a very clever experiment with sitters at a seance to see if the table that was supposedly moved by spirit entities was being moved by outside forces or by the sitters themselves (who were all eminent, honest pillars of society). As you might expect, Faraday showed quite conclusively that the sitters were moving the table -- though they vehemently (and probably quite honestly, as far as they knew) denied it.

QI have five different handwritings (I invent them when I'm bored.) One is really big, dark and sloppy, one is small light and very slanted, one is very round neat and clean with strange A's and I have two kinds of cursive as well. What happen if they tested me?

This question nicely points out one of the major problems with the whole idea of handwriting analysis. We all write somewhat differently, depending on haste, writing posture or surface, or the purpose of what we are writing (a love letter versus a shopping list, for instance -- which one reflects the "real you"?) Graphologists say you can't disguise your handwriting -- the "real you" will poke through for the "expert" to reveal correctly anyway. In that case, quite different scripts from the same person would result in the same traits being ascribed to that individual. However, if several different people were each to write like one of the different scripts used by the multi-script writer described earlier, the graphologist would say that these different styles were diagnostic of different traits in each of these different writers. Doesn't make much sense, does it? Also, if someone really does have several different writing styles, she should exhibit fundamentally different personalities when she writes in one way as opposed to another. Not too likely. (Some graphologists actually believe this, though -- they call it "graphotherapy.") The claim is that if you have an undesirable personality trait, you can eliminate it by finding the sign for it in your writing and changing the way you write it. The bad character trait will magically disappear once you stop using its associated sign in your handwriting.

To answer your question more directly, if you sent samples of each of your different script styles, with different names on them, to the same graphologist, I predict that you would get back very different personality readings from each submission. I've actually done this myself and that's what happened. (Others have sent the identical sample of writing several times to the same graphologist over a period of months and got wildly different reports each time. Still others have sent the same sample to different graphologists and got different interpretations.) You saw one trap I laid for graphologists on our SAF episode -- I led them predictably astray by putting misleading cues in the content of what I wrote, but I didn't disguise my writing that time. In other demonstrations that I have done for other media-run "tests" of graphology, I have changed certain aspects of my normal script deliberately in the sample that was to be analyzed. Because I know what graphologists look for, it is child's play to get any reading you want by inserting the right features. What I would do in these send-ups is insert "signs" (as graphologists like to call them) that are supposedly indicative of a group of traits that are quite the opposite of what friends would say I am really like. I would then (like I did on our SAF show) predict in advance how the graphologists would go wrong. You should hear the nasty language aimed at me when they are told what I have done and they are left with egg on their faces. I am very careful, however, never to do anything that the graphologists themselves say would impede their ability to do an accurate reading.

QDo people who do handwriting analysis go through any kind of formal training? I am wondering if there are actually courses that teach this skill of reading a person's personality.

Another big problem for graphology is that there are different systems of graphology and they often disagree with one another. In our book on graphology, I present numerous examples of writing "signs" that one branch of graphology says are benign and another says are indicative of terrible traits. How would you feel knowing your job prospects or coveted promotion hung on something that arbitrary and discriminatory? The fights among competing schools of graphology are often more heated than the fights with us skeptics. It doesn't really matter, though, because none of the competing systems does any better than any other in scientifically-controlled tests. I.e., they all do equally badly in ascribing characteristics accurately to writers. Because the basis of graphology is metaphorical interpretation of "signs" in writing, there tends to considerable overlap among most of the systems (within a given culture, we all tend to be reminded of somewhat similar things when we look at the same stimulus and free associate), but each school has its unique aspects and that is what they fight over. There is also a long-standing dispute between those schools that advocate a "holistic," impressionistic interpretation of the overall "feel" of a person's script, as opposed to the competitors who interpret numerous small features of many individual letters, etc.. In "The Write Stuff," we point out the flaws in both approaches.

Since no reputable university in North America will offer a degree program in graphology (any more than they will give degrees in tea leaf reading, palmistry, or astrology), the "certification" process for graphologists comes primarily from correspondence courses, night school courses, or a few professional graphology schools. They simply see if someone has memorized the set of invalid correspondences they believe in and then certify him or her as competent. None of them does any serious research to see if their methods actually work. The International Graphoanalysis Society of Chicago is the biggest of these institutions and it offers impressive-sounding "certification" in its method of handwriting analysis. The IGS claims to be the most scientific of the schools and it is very disdainful of its competitors. In my chapter on graphology in Gordon Stein's "Encyclopedia of the Paranormal" (Prometheus Books, 1996), I take apart the teachings of this very pompous school and show it to be the same pseudoscience as all the rest. When I was asked to debate graphologists at a conference in the IGS's home town, they refused to send a spokesperson -- draw your own conclusions. In studies where graduates of the IGS program are compared to devotees of other schools, we have found that they all do equally badly.

Probably the majority of graphologists selling their questionable practices to the public are still self-taught with the help of a few books from the local library.


Scientific American Frontiers
Fall 1990 to Spring 2000
Sponsored by GTE Corporation,
now a part of Verizon Communications Inc.