Apr 30 2013


Imagine applying for a job, a position you really want and feel is a good match for your skills, and during the interview process you are seated in front of a psychic. The psychic is wearing full regalia, with a turban, crystals, and mystical garb. They proceed to give you a psychic reading – a reading which will be used to decide whether or not you will be hired for your dream job.

You can substitute any number of techniques for the psychic reading – a tarot card reading, palm reading, astrological chart, or phrenological analysis. Would you feel comfortable with such techniques deciding your fate? Would you feel outraged?

That is exactly what is happening in many corporations today, particularly in France. The technique that is being used, however, is graphology. It is as legitimate as any cold-reading technique (that is, not at all) but retains a veneer of scientific legitimacy. Graphology, or handwriting analysis, is a psychic cold-reading dressed up for the corporate world.

Graphology was first developed by Jean-Hippolyte Michon, a French priest and archaeologist. He published his first journal of graphology in 1871. The idea is that the particular aspects of a person’s handwriting reveals their character. Graphologists study the size, slope, pressure, connections, and other tiny details of handwriting, with each detail revealing an aspect of personality.

Like iridology, palmistry, and astrology, there is a complex system of graphology that can take years to master. That in itself, however, does not say anything about the legitimacy of graphology. People are industrious and we are good at developing complex systems based on nothing at all, except our imagination. Complexity alone is not a sign of validity.

The beginning of exploration is doubt. Skeptics learn about the many mechanisms of self-deception so that we understand that just because something seems to be real, that does not mean that it is. This is the motivation for scientific analysis – controlling for all of those mechanisms of deception and bias. Only then will we know if a phenomenon is real or not.

Handwriting analysis has been subjected to properly blinded experimental tests. Graphologists are given samples of text that are neutral, meaning that the content of the text does not reveal anything about the person writing it. They are also blinded to the target subject, and given the task of analyzing the handwriting. Their results are then compared to standard personality profiles of the subject, and to other graphologists examining the same samples.

The results of such studies, not surprisingly, show that graphology provides no information to the graphologist. Their readings do not match the personality of the target, nor do they even match each other. Graphology does not work.

In a related study the graphologists were given autobiographical texts, and their readings were far more accurate. However, laypersons were also given the same text and they did just as well. This convincingly shows that graphologists get their information from means other than analyzing handwriting.

One interesting study showed handwriting samples along with fake personality profiles. They found that naive subjects rated the match between the two more highly when the rules of graphology were followed. The authors conclude that part of graphology’s continued popularity and use may be due to the inherent biases in perceived correlation. In other words – flowery script equaling a flowery personality makes superficial intuitive sense and give graphology a legitimate feel (even though that feeling is completely illusory).

This is a common theme that we see in many such pseudosciences – they tend to follow naive but universal human biases. Sympathetic magic is one example – things have a function that mirrors what they look like. Powdered rhino horn is therefore used as an aphrodisiac. A long life line equals a long life (an analogy between linear dimension and temporal dimension).

The real world, however, is often counter-intuitive and does not comfortably conform itself to our biases.

The recent BBC article about graphology in France contains some very revealing quotes from graphology proponents.  The sentiments expressed can be, and have been, used to defend every type of pseudoscience you can imagine.

Graphologist Bertram Durand defends his craft by saying:

“And just because we cannot measure its success rate using mathematics or statistics – that doesn’t mean it is not a valid tool. In all our client studies, there is an extremely high satisfaction coefficient. People use it because it works.”

Sorry, Durand, but the fact that you cannot measure its success in controlled studies absolutely means that it is not a valid tool.  That’s why we do controlled studies.

He then makes the argument from popularity – a blatant logical fallacy. Contained in that fallacy is extreme (perhaps willful) naivete about human perception. He is essentially arguing that people could not be fooled into thinking graphology works if it didn’t. This is demonstrably false. People can be fooled into believing almost anything. Psychologists have documented numerous sources of bias and flawed thinking that lead to confident but false conclusions.

Graphologist Geoffroy Desvignes says:

“I have no idea how it works, but to me it is obvious: the handwriting of a marketing guy is not the same as the handwriting of a sales guy, which is not the same as the handwriting of an artist or of an accountant at Deloittes!”

The, “I don’t know how it works, it just does,” defense of the implausible. The evidence shows that it does not work, but Desvignes appeals to the intuitiveness of graphology. He nicely demonstrates what the researchers found above – that graphology has an intuitive feel, even though it has no basis in reality.

When all else fails, appeal to a conspiracy. Durand is quoted as saying:

“There is a big psychology lobby that has it in for us. Companies that produce recruitment personality tests have a big interest in undermining what we do, and they have a lot of means.”

Right – Big Psychology is keeping graphology down so they can corner the market in corporate recruitment analysis.


Graphology does not work. It provides no unique information or insight to graphologists, who cannot even agree with each other when reading the same handwriting.  Like many systems that developed in the 18th and 19th century (phrenology, homeopathy, iridology) it is sciencey but lacks a true scientific background. It is naively appealing, but there is no scientific evidence-base to back it up. When subjected to controlled observation, like N-rays, it simply does not exist.

Graphology can sometimes function as a cold reading, but most often it is a hot reading. Graphologists get their information from other sources, mostly the content of the writing (rather than the form of the writing).

Graphology is no more legitimate than a psychic in a turban, but has maintained a superficial respectability that allows it to continue in the corporate world. Fake psychics are savvy to this marketing angle of graphology.

Mark Edward wrote a revealing article on Skepticblog describing how he sells cold readings to corporations by putting on his tweed jacket and doing handwriting analysis.  Perhaps the night before he was wearing a turban and doing a psychic reading. It’s the same shtick with a different costume and props.

34 responses so far