Handwriting Personality Assessment

Background

Writing by hand has been a major method of communication for centuries. Between the twelfth and thirteenth centuries writing was used by the educated classes, and has evolved today to the point that the written signature is universally recognized as legal identification. Each of us learns to write according to a prescribed school copy, but as soon as we reach writing fluency we alter the form and size ratios in a manner peculiarly characteristic to our writing style. This deviation from school copy, or the way we form our letters forms the basis for handwriting personality analysis. A letter is represented as a geometric form that can be expressed in three ways: the height, the width, and a third dimension which extends from the front to the back of the paper as a result of the pressure exerted by the writer. A fourth dimension in the handwriting that does not enter into the same system of geometric coordinates is called the form of the letter that deals with the contour or shape of the letter.

Historically, there have been two general approaches in evaluating a person's writing: the intuitive or global approach, and an analytical or detailed one. The global approach used by intuitive graphologists uses impressions and looks at the aesthetic features of the writing, paying particular attention to the form of the writing, to make inference about the personality of the writer. The analytical graphologist looks at the small details of the writing - the "i" dots and "t" crossings, the beginnings and endings of strokes, etc., to deduce personality characteristics. Neither of these methods have shown validity, but unfortunately, these are the methods practiced in America and have given graphology a bad reputation.

Around 1895 in Germany a new approach that combined the global and analytical was introduced by two men: W. Preyer and George Meyer. They stressed that it was the movement of the writing that expressed the personality, rather than the writing elements. The act of writing was regarded as an expressive movement, and the movement rather than writing was the connecting link to the personality.

Preyer and Meyers work (find appropriate reference) was synthesized into a system of graphology by Ludwig Klages (find appropriate reference). Handwriting was no longer considered an aesthetic form in which specific signs represented isolated personality traits which when assembled like a puzzle resulted in a personality profile; rather the writing was regarded as an expressive movement in which the rhythm was the most essential characteristic of the writing. He developed a table of contracting and releasing indicators for estimating the strength of the rhythm, but he had no objective method for measuring the rhythm.

Thea Stein Lewinson was born in Berlin in 1907 and studied with Martha Goldberg, one of Berlin's most eminent graphologists, in the late 1920s. She came to New York via Paris in the early 1930s and married Gus Lewinson. While working at the Psychiatric Institute of Columbia University, she became acquainted with a young statistician, Dr. Joseph Zubin, who later became one of America's most preeminent statisticians. Zubin's statistical brilliance combined with Thea Lewinson's intellect and knowledge of graphology provided the foundation for scientific handwriting analysis (Lewinson, 1944).

The genius of the Lewinson-Zubin method is that it presents a hypothesis, that the rhythm permeates all aspects of the writing and can be measured. Lewinson states that the rhythm is the midpoint between the contracting and releasing tendencies, this takes Klages' work one step further in that it provides a scale by which to measure the rhythm, rather than relying only on subjective evaluation. Measurements of the writing were then available for validity and reliability studies for the first time.

These measurements are based on twenty-one factors of the writing as measured on a seven point scale of contraction and release, ranging from +3 representing extreme contraction to -3 representing extreme release, with the midpoint or zero as the point of balance. The twenty one factors that represent a movement tendency that maybe contracted, balanced or released, are divided into four dimensions or components based on the geometric forms of the writing. As stated previously the form is expressed in three ways: the height, width; and depth which is the result of pressure; and a fourth dimension of form or the contour or shape of the letter. The four dimensions are divided in the Form Component, The Vertical Component, The Horizontal Component, and The Depth Component.

In addition to the scientific work of Lewinson there were other serious handwriting analysts who attempted to use handwriting as a projective psychological technique. One was Clara Roman who developed the psychogram (find appropriate reference). She taught her method to students and two of them Florence and Dan Anthony taught the only college accredited courses in the U.S. at the New School in New York City. Unfortunately these courses no longer exist. However, the situation is very different for handwriting analysis abroad. There are scientific societies in England, France, Switzerland, Germany, Italy, and Israel. There are several Universities in Europe that offer advanced degrees in handwriting analysis, and there are many handwriting dissertations and master thesis. Handwriting analysis is an accepted method in Europe for personnel selection.

A number of Ph.D. dissertations were written in Germany on what is called graphometry where every aspect of the writing, except the form elements, is measured. These studies yielded statistically significant results and were accepted in academic circles, but as Lewinson often expressed, these studies omitted a crucial element of handwriting analysis - the form of the letters which to date has to be evaluated by a person. In the Lewinson-Zubin (L-Z) scales there are five factors in this category, but Lewinson gives definitive criteria for their measurements, and while critics point out that inter-scorer reliability could be a problem, in practice when the technicians are carefully monitored their work show great consistency. The proposed research work should more adequately address this problem.

Since the publication of Lewinson's book (1944) statistical research in American handwriting literature that could validate handwriting analysis as a projective technique has been lacking, and there are good reasons for this. One reason is that graphology in America was dominated by the students of the Bunker School of Chicago (find appropriate reference) that offered home study courses in graphology based on the analytical method. After several months of home study one could list themselves as graphologists, and the image of graphology was so bad that no reputable psychologist would deal with it. Another reason is that Lewinson left her work at Columbia to work in censorship with the U.S government because there was a shortage of native speaking Germans with U.S. citizenship. In 1946 she returned to Germany with the U.S. Women's Army corps, and upon her return to the U.S continued her work with the government. What has been classified information until the recent publication of a book on spycraft (Wallace, Melton, & Schlesinger, 2008) is the fact that Lewinson worked for the CIA for eighteen years as a graphologist, and that the CIA used graphology for operations. Thus, although there was no public research on handwriting analysis, there was secret research; Lewinson has overtly published over sixty articles in seven different languages and is internationally recognized a one of the leading figures in handwriting analysis. At the age of ninety Lewinson while presenting a paper in Munich, Germany, let it "slip" she had worked for the CIA. However, this was not of any import because the cover under which she had worked for years was blown during Watergate and any foreign intelligence officer could have tracked this information.

Wallace, Melton, & Schlesinger (2008) state the following: "Five categories of recruitment and agent handling are so universal and fundamental that they can be called the "pillars of tradecraft". These are: assessment, cover and disguise, concealments, clandestine surveillance, and covert communications. Depending on the stage of an operation, one of these disciplines will assume dominance, and every effort will be made to execute it flawlessly. For the CIA, Office of Technical Services (OTS) had the responsibility to develop technical support tools for each pillar that would provide U.S. officers and agents a comparative advantage over their adversaries. The following quoted passages from Wallace, Melton, & Schlesinger (2008) describe the processes.

"Assessment is the first step in recruiting a spy. Selecting the right target from among thousands of individuals who could be potentially help an intelligence service requires identifying the one or two with the motivation and ability to sustain the double life required by espionage. Sound tradecraft requires more than guesswork. Based on the experience OSS had with assessment and testing procedures, the CIA employed a small group of professionally credentialed psychologists (Lewinson was not credentialed) to assist operations officers in winnowing the prospective pool and identifying the most "vulnerable" targets. Like their OSS predecessors, the psychologists of OTS employed a variety of assessment techniques and tests to gain insight into a target's dominant personality traits and potential behavioral responses to specific situations." (Page 364)

"During the Cold war years, when many targets lived in countries with severe travel restrictions, OTS maintained a small staff of handwriting specialists called graphologists. Graphology, a disciplined more respected in Europe than in the United States, seeks to identify characteristics of an individual based on measurable letter formations and line strokes in handwriting. The graphologists measured three dimensions (the vertical, horizontal, and depth strokes or letters) for as many as twenty one different characteristics of the writing. Handwriting has demonstrated the ability to distinguish between mentally healthy and those with mental illness. The OTS graphologists applied the same methodology to identify essential characteristics of persons who were unidentified, would not agree to a structured assessment (such as VIP's), were writers of anonymous letters, or were held in captivity." (Page 371)

After her retirement from the Agency Lewinson attempted to improve the status of handwriting analysis in the United States by founding the American Society of Handwriting Analysis in the hopes it would attract serious graphologists who would do research and establish ethical guidelines for the practice of graphology. Most of the members were recruited from graduates of the program of the New School (find appropriate reference). For a number of years a journal was published four times a year under the auspices of the society and the future of handwriting analysis began to look brighter. However, after Lewinson's death in 1999 the journal has had only a few publications and the membership is dwindling. The aim of the work proposed here is to establish the validity of handwriting analysis - the dream Lewinson brought to these shores seventy five years ago.

Project

The L-Z scales are the scientific handwriting analysis scales proposed by Thea Stein Lewinson and Joseph Zubin and have been used in handwriting analysis over the last half century. One of the vexing problems of these scales is the time it takes for handwriting analysts to measure the L-Z scales. Here we propose a computer assisted L-Z scale extraction for the purpose of handwriting analysis. Most scales can be automatically and objectively extracted once documents are optically scanned. A graphical user interface allows handwriting analysts to extract most L-Z scales more efficiently and objectively.

The most difficult part of this project is the computation of the L-Z scales. This work will continue based on the mathematically defined scales, on the preliminary development by Dr. Cha, and primarily on the earlier project and related work described in Research Day 2009 paper and Research Day 2010 paper. The code for the earlier work will be provided to the team.

Identifying personalities of a writer is an inherently difficult problem, especially for an automated system. One reason for this difficulty is that an automated pattern recognition system for identifying personalities from handwriting would require handwriting samples that are labeled with the personality traits to be identified, and these labeled samples would be used for training the system and for measuring its performance. Although we have some handwriting sampes from know criminals, like the serial killer Ted Bundy, and you can likely find other such samples on the Internet, how would you characterize the personality traits? Is "killer" a trait? Therefore, although identifying personality traits is clearly a problem of great interest, it is recommended that you also work on a more clearly defined problem where it is easy to obtain labeled samples, like the writer's gender (male/female) or handedness (left/right handed).

References

  1. Lewinson, T.S. & Zubin, J., Handwriting Analysis, King's Crown Press, 1944.
  2. Seifer, M., The Definitive Book of Handwriting Analysis: The Complete Guide to Interpreting Personalities, Detecting Forgeries, and Revealing Brain Activity Through the Science of Graphology, New Page Books, 2008, ISBN 1601630255.
  3. Wallace, R., Melton, H.K, & Schlesinger, H.R., Spycraft: The Secret History of the CIA's Spytechs, from Communism to al-Qaeda, Dutton, 2008, ISBN 0525949801.
  4. Lewinson-Zubin Scales.

Fast Agile XP Deliverables

We will use the agile methodology, particularly Extreme Programming (XP) which involves small releases and fast turnarounds in roughly two-week iterations. Many of these deliverables can be done in parallel by different members or subsets of the team.

The following is the current list of deliverables (ordered by the date initiated, initiated date marked in bold red if programming involved, deliverable modifications marked in red, completion date and related comments marked in green, pseudo-code marked in blue):

  1. 9/27 Obtain the earlier project's code from the Vulcan server, review and understand the code.
  2. 9/27 Interact with your customer Jonathan Leet to develop one of two research problems to work on this semester. Search the literature in this area for research problem ideas. Possible research problems are:
  3. 9/27 Plan the semester's work with your customer Jonathan Leet and your instructor.