Online Journal of Ethics, vol. 1, no. 1 (1995- 1997). ISSN 1092-8286 The Ethics of Handwriting Analysis in Pre-Employment Screening by Daryl Koehn An acquaintance of mine was turned down recently for a job. The hiring firm told her that she was rejected because their handwriting analysis had shown that she was not "open to learning." Use of handwriting analysis ("graphology") to screen prospective employees is widespread in France and is becoming more common in both the United Kingdom and America.(2) While job interviews, applications and recommendations remain the preferred screening techniques, more than 5% of American companies used graphology when hiring as of 1990.(3) That number has almost certainly increased during the last few years with well-known firms like Citibank and Bristol-Myers experimenting with the technique.(4) Consultants who advocate graphological screening claim that such analysis is able to reveal important character traits of job applicants. While analysis cannot disclose a person's age or sex, it allegedly can discern (at relatively low cost) character traits of a potential hire--e.g., a candidate's stability, vivacity, creativity, intelligence, imagination, reasoning ability, speed of thinking, and force of character. In addition, graphologists maintain that their skill is extremely useful in identifying paedophiles, sociopaths, persons with cancer, schizophrenics, and epileptics (5). Some go so far as to maintain that the individual's entire personality structure appears in his or her handwriting. The individual is said to have no secrets before the graphologist who supposedly can tell one's private sexual likes and dislikes from a handwriting specimen.(6) While many thinkers have dabbled in handwriting analysis over the years, serious interest in it has emerged only in the last one hundred years. Nineteenth-century Germans applied their customary systemizing of techniques and knowledge and invented the science of graphology. As with lie detector tests, there is enormous controversy over the reliability and validity of graphology as a hiring tool.(7) The technique presupposes a certain psychology of character expression and development. It is widely conceded that successful application of the technique also requires an interpreter highly skilled in psychology. There is the usual professional infighting not merely between psychologists and graphologists but among graphologists themselves over the merits of various analytical techniques. Some research supports a very high correlation between handwriting patterns and specific character traits, while other research suggests a correlation of zero. Controversy over the utility of a relatively new screening technique is to be expected. What is rather more surprising is the dearth of discussion regarding the ethics of using this mode of analysis. Apart from a few debates in state legislatures over use of the technique in job hiring (Rhode Island considered outlawing it; Iowa approved it (8)), the technique has escaped the scrutiny I think it needs. In this paper, I shall assume that handwriting analysis is technically valid. That is, I assume that it reliably reveals a genuine link between certain features of handwriting and some character traits if and when the test is interpreted in a sensitive fashion by persons who understand problems of character development. (Whether such interpretations are likely in the business setting remains an open question to which I shall return). I assume technical validity because, obviously, if the technique is fraudulent, its use to eliminate otherwise qualified job candidates would be unjust. I shall argue what is less obvious--namely, that even if we assume the test to be technically valid, employing graphology to screen job candidates is morally invalid. The technique as a preemployment tool is immoral irrespective of whether the technique is used alone or in conjunction with more standard screening techniques such as the job application, interview, resume and recommendations. Ethical Objections to Graphology Handwriting analysis as a screening tool has tended to be justified on utilitarian, or cost-benefit, grounds. The costs of the test are relatively cheap (9), and the purported benefits are significant: Employers learn about character traits that are linked to success on the job (e.g., "openness to learning," "outgoingness") and about propensities (e.g., a potential for violence) which might be a source of anxiety to employers who are finding it increasingly difficult to fire employees. This cost-benefit analysis is ethically problematic because it fails to examine whether the benefits have been achieved in a moral way.(10) For example, we might reap all sorts of benefits from hanging innocent persons. Yet we do not accept that increased deterrence of criminals and a heightened sense of moral responsibility among the citizenry could ever outweigh the evil of the means employed to reach these benefits.

In the case of graphology, we similarly must look beyond the alleged benefits of graphological screening to any evils involved in employing this particular technique as a means to improving hiring efficiency. If the technique infringes upon exceptionally important rights or values, then, as with the case of hanging an innocent man, the end will not justify the means. Examination of the technique shows that it does indeed violate important values. In particular, the test 1) constitutes an unwarranted violation of privacy by holding job applicants accountable for private intentions and traits when they should be held responsible for public performance; and 2) fails to respect the individuality of persons who are applying to work for a firm. Since the assumed benefits are obtainable through other means; and since the rights of privacy and human individuality are generally conceded by Westerners to be core rights, we ought not to tolerate graphology's violation of these rights. Argument 1: Test Invades Potential Employees' Right to Privacy Employers are certainly well within their rights to desire and to actively seek employees well-qualified to perform the jobs the employer needs to fill. Certain positions (e.g., hitman) are, of course, illegal. But, apart from these limit cases of illegal activities, we generally concede to employers a right to screen employees to determine whether they meet job- related criteria. The key phrase here is "job-related." Whether a person can do computer programming is relevant when one is hiring a systems designer. When technical skills are necessary to successfully perform a job, then an employer may legitimately test for these skills using inhouse tests, such as typing or programming exercises. Character traits or personal preferences unrelated to job performance, however, should be of no concern to the employer. The hiring session with the firm's human resources personnel is billed as a job interview, not as psychiatric therapy. Therefore, probing the job candidate's sexual preferences, taste in food, or evening hobbies is morally impermissible on the ground that the employer is intruding further than is necessary into private matters under the guise of conducting a job interview. I am not saying that the employer may only ask questions regarding technical competence. In many cases, employers legitimately will be concerned about a candidate's verbal and written skills, about her reasons for seeking the job, and about her ability to clue into expectations others hold regarding what counts as appropriate behavior on the job. A candidate who arrives late to an interview or who fails to wear socks might be technically competent yet still fail to be a good corporate employee in an environment where promptness, personal neatness and hygiene are valued by fellow employees and customers alike. It must also be conceded that no hard rule can be drawn against an employer exploring the candidate's "private" motivations, especially with respect to how the prospective employee conceives of the work he will be doing if hired. As I have argued elsewhere at length, a person's reasons for performing an activity and the performance of that activity are not easily divorced (11). Other things being equal, most of us would rather visit a doctor who is genuinely concerned about our health and who loves to heal rather than an expert whose prevailing attitude is "a job is a job." Someone who truly loves their profession or work, is inspired by its aims, and takes pride in its noble history is more likely to bring a higher level of care and attention to their work than someone whose only concern is to get through the day while meeting minimal job criteria set by someone else. Still, even though many character traits and motivations tend to come in packages and do affect job performance, it is one thing to probe motivations directly bearing upon likely job competence (e.g., candidate's attitude toward work) and quite another to investigate matters far afield from the job (e.g., candidate's bathroom habits). The focus of the job interview should be on whether the employee has the technical and the personal skills needed to do the job competently. The graphologist collapses the distinction between the public/professional and the private when he contracts to give a complete character profile of the job candidate. This collapse is worrisome for numerous reasons. First, as John Stuart Mill has argued, the moment one denies any distinction between a private realm and a public realm, the public realm tends to expand uncontrollably. Every action, trait and choice of a person is seen as affecting the public's or corporation's welfare and thus becomes fair game for public scrutiny. Yet the very fact that "private" is a word which makes sense only in contrast to "public" implies that there is a distinction to be drawn between the two realms. Drawing the line may be hard, but we are well- advised to adopt guidelines and attitudes which encourage us to thoughtfully draw it rather than to ignore it. Keeping the focus of the interview on narrowly relevant job skills is one such attitude.

Second, maintaining the focus on job skills is a check on human biases. We all experience more "chemistry" with some persons than with others. If we are thoughtful, though, we also admit that much of what we like falls within the realm of personal taste rather than of ethical judgments. Low-key persons may be put off by highly dramatic persons. Conversely, extremely vivacious persons may be bored to death or feel slightly let down in the presence of less overtly energetic types. Nevertheless, unless one is interviewing for an actress to play the role of an overwrought housewife, a person's expressive energy level will not be relevant to the job performance. Both low-key and highly-strung persons can and do manage persons and sell automobiles. Third, as Americans, we accept the separation of church and state and do not allow corporations to discriminate against persons on the basis of religion or creed. Some attitudes identified by handwriting analysis--e.g., optimism or pessimism--have all kinds of subtle links to the worldviews adopted and encouraged by various religions. Handwriting analysis screening for general attitudes and character traits may easily wind up indirectly discriminating against persons of certain religions.(12) Finally, we should remember that human beings are surprising creatures who choose their character by virtue of engaging in those actions which are consistent with or expressive of the kind of person they would like to be. Character traits are not written in stone. If the candidate has chosen to seek the job, he or she may be trying to turn over a new leaf. In general, persons should be judged by their past and present actions, not by some fantasies or character traits graphology imputes to them but of which they give no evidence in the interview or which is belied by their r‚sum‚. If a firm is not careful, it will find itself turning down a candidate who has an outstanding record of scholastic achievement in high school, who has graduated with honors from a highly respected university, and who asks intelligent questions in the interview because handwriting analysis shows that the candidate is not "open to learning." Combining handwriting analysis with other modes of screening does not resolve the moral problem of invasion of privacy. By law, an interviewing firm already has ample access to transcripts, diplomas, past employment records, as well as complete references to prior job performances. These materials speak to a candidate's past performance on other jobs and document skill levels. The firm also has the means of assessing how a person acts in a variety of public settings. Firms can and do perform a series of interviews with a given candidate, watching his behavior at a meal, at a ballgame, with a variety of in-house employees, etc. Given this opportunity to observe a candidate's behavior and to interrogate him regarding a wide range of beliefs which are clearly relevant to job success, a firm's main motivation for doing handwriting analysis must be to discover hidden personality traits or beliefs not readily disclosed by the candidate's public record of actions and speech. And therein lies the problem. The very use of the test evinces a desire on the firm's part to transgress the individual's right to privacy, a right which should be honored for the reasons I have already stated (13). I thus would argue that it is the firm's character traits and motivations that should be a source of concern here. Not only are firms invading individuals' rights to privacy. They are doing so in a discriminatory fashion to boot. The test currently is being applied primarily to prospective employees. If hidden character traits are so important to job performance, surely current employees at all levels, including senior management and human resources personnel, should be subjected to these tests. The fact that they are not being so tested should raise the suspicion that handwriting analysis is a cover for keeping certain people, or classes of people, out of the job market, a point I will return to in the next section. A firm might try to save the use of the test by applying it uniformly to all employees and by using it only to disclose certain character traits relevant to the job. This response, though, is not acceptable for numerous reasons. Insofar as the test by its very nature is kind of invasion of the right to privacy, expanding its application and restricting its scope does not stop this evil. Violating one person's right to privacy does not become less morally objectionable if we violate the privacy rights of all of a firm's employees. Such an action merely compounds the original offense. Moreover, it is extremely dubious whether we can neutralize the test's evil by limiting the scope of that for which handwriting analysis tests. Insofar as the test is designed to look for hidden traits, the test is objectionable at its core. Tinkering with its scope does not address this core problem. Furthermore, graphology is not like lie detector tests whose scope can in principle be restricted. The administrator of a polygraph test can be told to ask about past job performance but not to gather any data regarding candidates' dreams about robbing banks or fantasies about sleeping with persons other than their spouses. These explicit limitations on questions arguably ameliorate the invasion of privacy in the case of the

polygraph because it does not reveal character traits, aspirations, and preferences. At best, it discloses merely whether a given candidate feels anxious about certain matters. However, if my handwriting really does reveal my attitudes toward life, the world and learning, there is no in principle reason why it should not equally reveal my attitudes toward my body or my sexual preferences. If the activity of writing a few sentences reveals a host of what are clearly interrelated character traits, there is no reason why this same activity should disclose only one-half or one-quarter of the complex of traits and not the whole. In other words, in gathering data about some parts of the character, we will secure other data about the character as well, data unrelated to job performance. Someone who writes double "e"'s close together may like to go home after work and cross-dress. But I fail to see why a potential employer should be allowed to gather such data in the first place. Even if those doing the handwriting analysis are super-scrupulous and manage to block out such bits of errant data as must register in passing as they evaluate the test, others with access to this data may not be so observant of the rules. Gathering the data invites further scrutiny and constitutes even more of an invasion of privacy than other tests (e.g., polygraph tests) outlawed because of their intrusiveness. Objection 2: Graphology Fails to Respect Human Individuality There is quite a lot of information employers might like to obtain in order to better judge an individual's character, yet we legally prohibit employers from asking certain questions. Thus, while many employers believe that married people make more stable workers than unmarried persons, states (e.g., Illinois (14)) have made it illegal to discriminate in hiring on the basis of marital status. As a result, asking job candidates about marital status is legally suspect. Our respect for the individual grounds this prohibition. We think that individuals should be judged as such, not categorized according to their membership in some marital, racial, sexual, gender, or religious group. Handwriting analysis circumvents the law by trying to determine an employee's traits (e.g., stability) according to some handwriting group- stereotype to which he or she belongs. (Indeed, some graphologists have so little respect for the law and so much confidence in their stereotyping that they have proposed using the technique in lieu of court proceedings to identify and prosecute criminals! (15)) The analysis works by comparing the speed, size, slant, form, pressure, layout, and continuity of an individual's handwriting with various patterns and typologies and assimilating this person's script to these types. As a result, the individual judged ceases to be an individual and becomes little more than a composite of traits. This end result differs little from judgments based on race, sex, religion, etc. Granted, no individual is totally unique. Any evaluation of character or, for that matter, skills, turns, in some measure, on employing generic ideas about virtue, vice, and technical competence. Still, there is a human individuality which manifests itself in our imagination and in the innovative arguments we choose to advance. Standardized handwriting analysis is far less respectful of individuality in this latter sense than other modes of screening. The individual who is asked to write a personal essay describing her qualifications in her own terms; and who is given an opportunity in an interview to describe her motivations in seeking a particular job retains far more of what makes her the distinctive person she is. This more personalized format gives the individual an opportunity to express unusual or provocative opinions the employer may not have previously considered. Upon reflection, the employer may think these comments so pertinent and insightful that she awards the job to this candidate. Handwriting analysis, though, is ostensibly purely formal. It does not provide the candidate with any opportunity to distinguish himself or herself in this substantive fashion. At best, graphology will yield some vague assessment such as "the candidate is highly creative." In addition, when the screening is interactive as it is in an interview, the candidate is given an opportunity to challenge statements the employer makes. If the interviewer volunteers interpretations of what the candidate has said, the latter can explicitly confront and rebut claims she thinks are false or unfair. The candidate, however, almost never sees the handwriting analysis report. He or she stands condemned by a faceless, anonymous accuser who has lumped the candidate in with others whose handwriting has established the standard for analysis. It is worth remembering in this connection what the driving force is behind graphological testing. Handwriting analysis, like automated telephone prescreening, is increasingly being used early in the hiring process because it purports to deliver salient, accurate information cheaply. Yet precisely because these techniques are standardized, the data has reduced value. Judgments about the precisely relevance of some perceived character traits to a job are rarely straightforward. Good interviewers learn through training

and through the interaction itself to qualify previous judgments. Perhaps the candidate who fails to make eye contact has a guilty conscience as is standardly assumed. On the other hand, perhaps the candidate is a recent immigrant from a country where direct eye contact is considered rude. Alternate interpretations sometimes suggest themselves in a face-to-face encounter with individuals who are fully present in their living, acting, speaking personhood. Handwriting analysis done at a distance by an expert who has never even met the candidate will not stimulate the evaluator's imagination in the way the in-person interview or personal essay may. On the contrary, the cheapness of the technique stems from its elimination of the important human activity of hypothesizing about the case at hand. Even if we assume, as I have done, the general technical validity of graphology, the firm will have to take care to make some allowances for special cases if it is to avoid the charge that it grossly disrespects human individuality. For example, because we learn much behavior from our particular communities, the ethical employer and graphologist will correct for any possible effects of this learning on a candidate's handwriting. Persons trained in the United States using the Palmer method may write quite differently than those who learned via other methods or those whose first language employed ideograms rather than letters. While some graphologists do ask for additional information such as country of origin, others do not. Why not? Well, these complications mean increased costs and hence reduced sales of the graphological technique to firms. Given that firms are drawn toward the analysis in the first place as a cheap form of screening, we legitimately may question whether these tests are going to be administered with the requisite sensitivity and interpreted with the necessary care. Finally, any and all correlations to character will still need to be very carefully interpreted. Given some correlation between handwriting slant and a particular character trait, anyone who is serious about respecting another's individuality will consider whose character exactly has been revealed. Teachers can and do force left-handed children to write with their right hands. Or they change students' handwriting to resemble their own. In addition, peer pressure induces students to modify their writing. Girls may make their writing more curvy so they will not suffer the humiliating taunt of writing like a boy. Fads, too, play a role. In a given year, students will begin to dot the "i" with a smiley face or a heart. The child may very well take this habit into adulthood. In these cases, exactly whose character are we evaluating--the teacher's? the student's? the peer group's? the generation's? (16) The handwriting analyst may try to compensate for these influences by choosing to discount or ignore certain features of the handwriting. This strategy, though, merely raises another troublesome question: Why aren't these and other exclusions merely prejudices of the analyst's own character? Since all handwriting is learned behavior, each of our scripts necessarily reflects the influences of the character of our teachers, peers and generational cohort. These influences, too, are part of our character. Excluding them seems an arbitrary product of the analyst's whims, while including them would make the test much more complicated and costly and would no doubt once again require an interview in order to determine what the possible influences are. If so, the technique loses any value it might have as a tool for pre-screening candidates for an interview or as a substitute for an interview. The graphologist might try to counter this concern regarding influences and societal biases by arguing that people choose to conform. Insofar as they do so, we do not have to worry about discounting for conformity. The choice to conform, too, is part of the character reflected in the individual's handwriting. Though clever, this response is far too simple- minded. It is unquestionably true that children learn through imitation. It is also true that in some loose sense of the term, the imitation is "chosen." Children do not imitate every action or person; their imitation is selective. It is hard to see how handwriting analysis will be able to discriminate which character trait the child found attractive enough to imitate. A shy child may greatly admire a flamboyant aunt who happens to be an embezzler. She may model her handwriting on letters she receives from this aunt. In the process, her handwriting may come to resemble that of a highroller. While we surely would not want to claim that the child is predisposed to larceny, how will the graphologist separate out the child's love of mere flamboyance from the aunt's desire to steal, given that both traits may have entered into the child's handwriting by dint of the choice of model? Handwriting analysis appears to be an incredibly clumsy mode of character assessment which seriously underestimates the complexities of the processes through which we learn individuating behavior. As a result of these oversimplifications, graphological analyses may result in employers unfairly imputing others' character traits to the job candidate. One final worry: While I have many doubts regarding the theory of character assumed by graphology, I have no doubt about possible correlations between handwriting and motor control. Handwriting is a motor

process involving hand-eye co-ordination. A change in handwriting or difficulties in controlling one's strokes is one symptom of those diseases that affect htis co-ordination. While an interviewer might not discern that a candidate is in the early stages of multiple sclerosis, handwriting analysis might very well disclose the presence of the disease. In an age in which corporations are desperate to control the rising health costs associated with providing insurance to employees, the temptation may grow to use handwriting analysis as a means of screening out those prospective employees likely to incur high medical expenses down the road. Under the American Disabilities Act, discriminating against the disabled in hiring is illegal. Using graphology amounts to a surreptitious accumulating of evidence of given job candidate's motor skills and health. Therefore, it may be a viiolation of the ADA and yet another sign that the employer is unwilling to respect human individuality by abiding by the laws designed to protect this individuality. Conclusion The use of handwriting analysis to screen prospective employees is immoral because the technique constitutes an unwarranted invasion of individual's privacy and fails to respect human individuality. These two values are important ones in our community. Violating them is a known evil. Alleged benefits of using graphology as a screening technique are speculative at best and do not begin to justify the violation of human rights and dignity the technique entails. As we have seen, combining graphology with other screening techniques does not make it any less invasive or disrespectful. I would add in closing that making the test voluntary does not lessen the technique's evil either. Voluntarily consenting to an invasion of one's rights or to an abuse of one's dignity does not make either harm acceptable. There are some rights that one cannot forfeit without losing one's humanity. No just legal system honors "voluntary" contracts or arrangements predicated upon such forfeiture (e.g., a contract to sell oneself into slavery). Therefore, neither we nor our legal system ought to honor highly abusive and intrusive pre-employment testing procedures. Moreover, we must wonder whether any job candidate's agreement to undergo handwriting analysis qualifies as truly "voluntary." Candidates are not stupid. They will reason that the firm is using graphology because the firm thinks the procedure worthwhile and valuable. To decline the test amounts to a judgment by the candidate that the firm's reasoning is, at best, unsound and, at worst, unjust. The rare employer may be willing to endure the implied criticism and to discuss openly the candidate's reluctance to submit to the test. But in most cases, the employee who declines to undergo the test will correctly surmise that he thereby places himself at a disadvantage vis-a-vis other job applicants. Submission to the firm's will becomes a hidden prerequisite for getting the job. The "voluntary" test is, in reality, coercive. As a free society, we try to reduce coercion. Job candidates should not have to be subjected to it as part of a routine job application. Since the technique of handwriting analysis is in itself immoral; and since it remains such irrespective of whether it is used by itself or in combination with other techniques or with or without the candidate's "consent", I conclude that the technique ought not to be used to screen potential job applicants. Endnotes 1. I am grateful to Sanni Judy, Laura Pincus and the journal's outside reviewers for their comments on earlier drafts of this paper. 2. Alasdair Palmer, "Reading Between the Lines," The Spectator, October 2, 1993. 3. Power of the Written Word," The Economist, June 16, 1990. 4. "The "Watch Out, John Hancock," Fortune, January 4, 1988, 82. 5. "Power," op.cit.; "Reading," op.cit. 6. "Reading," op.cit. 7. See, e.g., Baruch Nevo, "Validation of Graphology Through Use of a Matching Method Based on Ranking, Perceptual and Motor Skills (December 1989), vol. 69 (3, pt.2), 1331-1336; Efrat Neter and Gershon Ben-Shakhar, "The Predictive Validity of Graphological Inferences: A Meta-analytic Approach," Personality and Individual Differences (1989), vol. 10 (7): 737-745; Anat Rafaeli and Amos Drory, "Graphological Assessments for Personnel Selection: Concerns and Suggestions for Research," Perceptual and Motor Skills (June 1988), vol. 66 (3): 743-759; Adrian Furnham and Barrie Bunter, "Graphology and Personality: Another Failure to Validate Graphological Analysis," Personality and Individual Differences (1987), vol. 8 (3): 433-435; Richard J. Klimoski and Anat Rafaeli, "Inferring Personal Qualities through Handwriting Analysis," Journal of Occupational Psychology (September 1983) vol. 56 (3): 191-202. 8. I could not find specific figures on the average cost per applicant of a screening test. However,

businesspersons quoted in the literature think that the costs are not particularly high, at least not high in light of the perceived benefits. See "Reading," 21. Some firms like S.G. Warburg require that every hire from managing director to janitor provide handwriting samples. 9. "Power," op.cit. 10. The point is analogous to Brenkert's argument that we must consider the morality of the way in which we get information as well as the desirability of having this information. George G. Brenkert, "Privacy, Polygraphs and Work," in Tom L. Beauchamp and Norman E. Bowie, Ethical Theory and Business (Englewood Cliffs, NJ: Prentice-Hall, Inc.), 280-286. 11. Daryl Koehn, The Ground of Professional Ethics (London: Routledge Press, 1994). 12. Handwriting analysis may be considered a facially neutral assessment which, when implemented, has a disparate impact on one protected group or another. If this is the case, that group would have a Title VII claim against the employer or potential employer for disparate impact. See, generally, Dawn BennettAlexander and Laura Pincus, Employment Law for Business, Ch. 3 (Irwin Publ. Co.: Burr Ridge, IL, 1995). 13. I would note that some ethicists have argued that the lie detector test, too, is essentially immoral irrespective of whether one limits the questions as employer is allowed to ask. See Brenkert, op.cit. 14. Illinois Human Rights Act. 15. The graphologist Margaret White claims that graphologists "can tell paedophiles simply by looking at their handwriting. You don't need a whole legal investigation for that! We could save the country millions." White is quoted in "Reading," 21. 16. Graphologists do exclude some signs on the grounds that they are "particular" or "idiosyncratic." These signs may be important, yet the analyst may choose to count them. Again I ask: What justifies this choice?