The EGS Bulletin, No.8: Special supplement, Nov. 1997,  

"Between amusement and factor analysis; the scientific aspects of graphology"

by Renate Joos

Lilienthalstr, 12, D-70193 Stuttgart.

This is a precis of an article by a distinguished German graphologist. prepared and translated at the request of Ted Barnett, from the German text (the original is 24 pages long, plus 11 pages of notes).

"...arising from the insatiable curiosity of mankind about his own nature, and as an outcome of everyday psychology... despite faults and shortcomings... out of the beginnings of psychology as a science, an instrument is arising that will slowly develop a more precise scientific and critical method." (Heiss,1950).

The question as to whether one can draw conclusions from handwriting in psychologically interesting sense, was raised in Europe at the start of the 17 the century by Prosper Aldorisius (1611), and later by the Bolognese doctor Camillo Baldi (1622). The Frenchman F. de Mells had also written about signatures and forgeries and the "art" of detecting the latter, already in 1604. In the past, even Ovid had warned his readers that one's own writing could reveal features of character. But it was not until the 17th century that serious investigations were started, when the two first-mentioned authors started regarding handwriting analysis as a "science”.

In 1928, Bührig divided the history of the subject into three phases: l. intuitive,2. analytical and 3. scientific.

The 18th century added little, since at the time of “Sturm und Drang', there was a revolt against the increasing rationalization involved in the Renaissance. Lavater (1741-1801) commented in volume 3 of his physiological observations: “I notice a strange analogy between speech, gait and handwriting, in most people." He agreed with what Klages would later write: “All bodily movements are modified by people's temperament and character". Lavater noted national differences and other details. His friend the poet Goethe, who collected handwriting specimens, enthused about Lavater's "Fragments” and visited him in Zurich. Lavater's works were translated into French. In 1812, he encouraged Hocquart to write a book on graphology.

Abbe Flandrin, around 1830, inspired his pupil J.H. Michon (1806-1881), to develop a new system, which in 1875 called "Système de Graphologie", the first use of the term 'graphology' as such. It was based on a large collection of handwritings.

Michon's pupil, the watchmaker and apiarist J. Crépieux-Jamin (1858-1940) developed his own system which layed much stress on whether the writing as a whole was harmonious or inharmonious. He tended to categorize human beings.

A lively exchange of letters between Michon and Madeleine Schorer (1881) showed that Germany also was beginning to take graphology seriously. A. Henze (1814-1883). believing himself to be 'inspired', from 1851 onwards, assessed 75,000 handwritings! This is according to a Leipzig newspaper, in which he also gave advice based on graphology; he was actually the first person to make any money from the subject!

Binet, in 1906, wrote that graphology "is one of the most beautiful experiments that a psychologist could wish for". Yet he pointed out that graphological factors should be better defined and that there were errors to be avoided. But he saw the subject as an art. Today, people wonder whether graphology is genuinely scientific.

The point, of course, queries its reliability. Science implies order , and one of the troubles is that it also implicates a set of rules, which have constantly to be checked in practice. Anyone who holds to the same theory when it has constantly been found wrong, is liable to be seen as a charlatan. Object and subject, theory and practice, 'independent' variables, even in controlled experiments, tend to be dependent on the activities of the scientist. Also, an abstract language with vague terms is hardly to be trusted.

Ludwig Klages, one of the great scientists, counted for a long while as the founder of scientific graphology. Heiss commented however (1964) that Klages' philosophical teaching was "fruitful but also speculative".

If graphology is to be considered a science, definite relationships between personality and handwriting need to be established. It is not just a matter of doubting the statements of particular graphologists, who of course think they are right, or the experience of participants; what really matters is the extent to which graphological assessments are basically verifiable. Writers such as Beyerstein and Beyerstein (New York,1992), authors of the most critical of recent books, say that if in doubt, graphologists rely on 'intuition' or on "mystical sources of enlightenment". So it is simply a matter of finding whether personal characteristics can be specifically linked to people's individual handwriting features or not.

G. Meyer (1899) tried to form a link between his patients and their handwriting, while also noting their emotions, and distinguishing the 'hypokinetic' from the 'hyperkinetic'. But Bachmann (1935) disputed this as being too 'experimental'.

Lockowandt (1993) maintained that graphological methods had made progress with psychological diagnosis. Graphometry, in the 1960s, was written about by Fahrenberg and also Fischer, who investigated degrees of changes in writing due to illness or to age. Schroeder had written in 1880 on the subject of the writing of the mentally ill, but had concluded that scientific accuracy would never be attainable. Wundt (1832-1920) and his follower, the Munich psychiatrist Kraepelin (1856-1926), also worked on the pathology of handwriting. At Leipzig University Psychological Institute. F. Krueger (1874-1948) and J. Rudert (1894-1980) made further investigations on validity. Foerster in 1927 (according to wallner, 1998) laid the foundation- stone for a new approach to reliability.

Lewinson and Zubin (1942) made a further attempt to measure variables in handwriting. The former defined the bridge between graphology and Graphometry. At the Salzburg congress (4/10/1967), she stated:

"We are now at the critical point where our technique, which for years, counted as empirically correct and useful, must be subjected to statistical validation... It should be our communal target to try for more precise examination and validation, without upsetting the basic principles of scientific analysis as established in Europe since the start of the century."

The term Graphometry arose in 1900 for the measuring process of legal (forensic) handwriting expertise. Two writers dealt with this, both in 1906: A. Binet, and the child psychologist A. Gesell. But it was not until the 60s that Bergström's method came from Sweden to Germany, via Wallner and to the Freiburg Psychological Institute of R. Heiss.

Graphometry became the basic discipline of graphology. It undertook to make psychometry useful for graphology as well, and derived most of its ideas from Freiburg. Research was done into measurements and classification of graphological factors.

In the last thirty years or more, research has tried to follow psychological principles of checking, avoiding what the graphometrists call “speculative” concepts. Lockowandt reported on the results of analysis in Müller and Enskat's book "Graphological fliagnosis” (severely criticized by Beyerstein and Beyerstein).

Foreign writers have almost no knowledge of the thorough, detailed work begun by the German-speaking medical research workers and the educational and forensic workers as well, using new technical methods of analysis. These also concern, variability between national handwritings, different professions, family and class differences also age and illness, sex influences and different personality signs. Also the influence of medicines. Alcohol etc. Gender, state of health, biographical details, intelligence, gerontology, all have been considered in research. Doubrawa believes strongly in graphometry.

Lockowandt discusses the Freiburg school's results (which were not available to recent American investigators).

Wallner formulated four work hypotheses, the verification of which gives the basis for diagnosis. In 1959, he laid the foundations for a "Unification of graphological terminology", with a systematic division of variables, for giving a complete and objective description of handwritings (1959 and 1972). Other authors followed.

It is now a fact that human handwritings (in cases of normality) are individually unique and individual. Otherwise they would not be legally binding. Exceptions occur only through alcohol, drugs, electric shock, hypnosis or special training.

Reliability of handwriting features would appear to be 80-90% or in many cases higher still so that one can rely on features of character being constant.

In 1929, Bobertag followed up the question of graphology's reliability, and Raech also, in 1957, asked “Has graphology established itself?" Both said 'yes'. Admittedly, many former assessments count nowadays as outdated. Lockowandt holds that the question of reliability has never been finally answered.

Investigations into reliability started as early as 1919. Hofsommer (1973) reviewed the matter.

Attempts at simple correlation were modified by the fact that it is never possible to draw graphological conclusions about individual criteria based on one sign alone.

Multiple correlations were often used. 

(The first part of this booklet, here summarized, concerns the history of Graphology. The next part is several pages devoted to accounts of the various doubts of critics as to the accuracy of graphology. The rest, including the author's own views, is again summarized from this point.)

In 1988 on BBC TV, a scientific documentation was begun concerning the usefulness of graphology in the selection of personnel, which was broadcast from March 1989 onwards with the title "The Write Stuff". It was clear to those with inside knowledge, that the program was prejudiced from the start. The kernel of the book published in 1992 by Beyerstein and Beyerstein was the report by G.A. Dean ("The bottom line: Effect size") about graphological results. This summarized the search for information frown literature, which was made difficult by the fact that before 1970 there was no computerized index of the available books and articles. Many had been written for magazines which were not available in public libraries. Dean stated that many inconsistencies occurred even in often quoted works, not helped by the fact that reproductions were often too small to be useful. Out of 47 studies, based on graphological signs in relation to a criterium such as business success, only 56% were really valid. But it was alleged that it made no difference if amateurs or professionals were involved, thereby contradicting European experience.

Dean's criticisms may have appeared to be just, mainly because research corresponded late in the day to the demands of science and even then too little research was done. Undoubtedly there are far too many pseudo-scientific publications and too few studies of reliability, etc. There was too much random concentration, e.g. on the writings of celebrities. Good work was seldom reproduced, and results could thus not be compared. Dean concludes: "Yes, graphology Is valid,... but not valid enough... or reliable enough... to be useful." If untested new techniques exist and give better results, it has to be shown: "Nothing else will do".

Wallner, however, accuses Dean of superficiality, subjectivity and generalization. Even the crass comments of the Beyersteins include the admission that "The graphologists clearly did better than the psychologists" (Edwards, 1992).

(From here onwards, the authoress pays more attentaon to her own views on graphology.)  

It is not possible here to weigh up critically all the arguments of the defenders and opponents of graphology.

The reproach "Non-graphologists are generally as good as graphologists" can be dismissed by one experiment after another. In tests, a "lay" group is compared with a group of psychologist with some graphological training. But when one knows that only years of intensive tuition make it possible to detect handwritten peculiarities and remember them, (Cohen, 1969, says writings are far more difficult to remember than faces) , one realizes that the above-mentioned set-up is no basis for clear judgment. The validity of the assessments depends on the competence of the judges. For instance, Dean is a chemist, born in England and living in Australia, who has changed from chemistry to journalism. Reichold (1969) maintains that only a team of graphologists can produce a completely reliable report.

Unfortunately it is seldom possible to get a writing judged by several experts. This, however, is particularly valuable if, for instance, a writing is in an unfamiliar language.

It is not unusual for graphologists to obtain clues from the contents of a writing, but well-trained graphologists should be able to draw more than just haphazard conclusions even from a neutral text, written in a language they do not personally know. Of course, conclusions as to intelligence and level of culture can be drawn from spelling mistakes or from remarks on the subject's life, but it is better that there should be some basic information given in advance by means of a questionnaire. One reproach is that in the English language, very few books can really count as scientific. There are many in German which are not available to English readers and have not been translated. Dean listed 201 authors, only 13 of whom wrote in German. Even these were not read in the original, only quotations being used. But in German universities, particularly Freiburg-im­Breisgau, important scientific experiments have been made.

Cohen (1969) writes that the less information there is about candidates, the greater is the influence of stereotyped expectations about the combination of different factors. Another difficulty is the use of information that is not comparable, such as spontaneous handwritten specimens, or copied texts.

Yet another trouble is that the character of the graphologist can sometimes influence his or her objectivity. Those with rigid writing tend to emphasis irregularity more than other graphologists do, and writers of complex handwriting tend to differ from thorns who write with a simpler style. The introverts judge differently from the extroverts.

What is Known as the "Barnum-effect" is the resentment caused by criticisms directed against readers of their own assessments ,which they suppress, meanwhile accepting exaggerations (as in a circus) when flattering comments are provided to tone down the effect of negative remarks.

There are also doubts as to the fairness of graphologists regarding the opposite sex (Klages was not always perfect in this respect). This should be looked into, as well as tables for the assessment of particular population groups (age. education. profession, racial origin), and eventually also of situations.

Graphometry had hoped to become the basic discipline of graphology, by attempting to found a diagnostic method of analysis of handwriting, using modern psychological methods. But in 1970, Angermann made the reproach that it had grossly underestimated the structural complexity of the graphological process. Dilthey (l924) felt that "divination" did not exclude logical and methodical analysis. Bachmann and Wallner both felt that forming and testing hypotheses should not exclude logic and method.

A high degree of intuition is frequently demanded from graphologists. Anyone who is called upon to balance writing characteristics with psychological concepts, must be able to synthesize much data in a brief amount of time. Observation, consideration and conclusion is a process which can be learnt. Thus graphology has less to do with art than with capacity, though one might assume that those who have artistic leanings will develop different insights from those who do not have them. According to the investigations of Dreyfuss and Dreyfuss, professional people of all types should go through a five-stage training: the novice, who learns facts; the more advanced, who uses rules; the competent, who recognizes the essentials; the accomplished, who can consider alternative decisions; and lastly, the expert, who works intuitively: "Particularly important is that this succession should mean progress: from the analytical approach to an unfamiliar subject, dividing its matter into recognizable elements and following the rules; finally, expertise, based on earlier experience and on the unconscious recognition of similarities between new situations and those met with in the past".

The same authors wrote: "Competence is rational, but cleverness characterizes the transition to expertise. which is not merely rational ("arational"). 

What is noticeable is the acceptance that graphology has always received in newspapers, as well as among so many members of the public. Yet Eysenck (British journal of Psychology) wrote In 1945: ".. the boundlessness of the graphologist's faith, the enormity of his claims, must make the cautious scientist hesitate."

The Heiss pupil L. Michel (who died in 1997) warned graphology it was not enough to rely on appeals to authority, or on intuition. A. Müller (1994) writes: "The graphologist should not only sit behind his desk, but should also get to know the structure of the firm for which he is working, the atmosphere, the fellow-workers, the organizational system. Such information is often not available to those who carry out tests. Graphologists should recognize that it is unreliable to diagnose specific suitability for a profession, purely on the basis of handwriting."

To sum up, one dare say that for diagnosis of various aspects of personality, graphology is definitely useful, provided the examiner is really competent. Variables in handwriting measure qualities of personality that standard tests do not measure. Significantly better results come from the additional information thus obtained, as has been proved by carefully documented studies from recent years.

(Note 153. "Graphology belong, oddly, to those subjects where there is hardly ever an objective attitude, but always an emotional one - either adherent or opponent. Kroeber-Keneth, L., 1968.) For the consolation of Crépieux-Jamin's girl-friend, people still amuse themselves about graphology. But they overlook what a comprehensive analysis can add to the understanding of a personality: that is, the improvement of the validity of other processes.

The value of systematic graphology should be confirmed by scientific proof. It could be used in many departments of selection. It is usually not underestimated by those who are ready to collect relevant experiences over a long period.

W, H, Müller, as long ago as 1935 , on the subject "Is graphology a science? ", wrote: "What is lacking is not scientific graphology, but a sufficient quantity of scientifically trained graphologists" This remark must be seen as a spur for research and for teaching .

Finally, perhaps Binet was right to state, a hundred years ago: I feel that graphology, which while developing like a wild plant has already given us pretty flowers, is worth our trying to rationalize it. It is an art of the future.

Renate Joos (translated in a summarized version by Ian Stirling)