Introduction
“Somebody's looking into your skull” and trying to find evidence of “paranoia,” “anxiety,” “sex violence”, and “neuroses” . . . . A young man with a yen to stand behind a police badge in Billings must first submit himself to a psychologist who will dig inside his skull. The digging frightens some. The instrument used for this probing of the secret nooks and crannies of a candidate's subconscious is a 32 year-old test called the Minnesota Multiphasic Personality Inventory (MMPI).
So ran the opening sentences of a 1975 article in the Billings Gazette Sunday Magazine. The Minnesota Multiphasic Personality Inventory (hereafter MMPI), according to the Billings Gazette, was the most popular psychiatric test the world-over and the article's author had no wish to laud that achievement (Clawson Reference Clawson1975). The Billings Gazette Sunday Magazine had, however, arrived rather late to the debate about psychometric testing and its role in promoting a psychology of adjustment in American society (Buchanan Reference Buchanan1994; Buchanan Reference Buchanan1997; Buchanan Reference Buchanan2002). Indeed, the rapid uptake of psychometric testing and adjustment psychology, and the anti-expert reaction that testing eventually engendered, appeared throughout the 1950s and 1960s in the pages of The New York Times, The Washington Post, and sundry other media outlets. In those sources, descriptions of methods for “Separating Curables from Incurables” (Anon. 1947), reviews of “hair-raising reports” about the “manipulation of buyers, voters” (Spingarn Reference Spingarn1957), portraits of the putative normal man in whom “imagination is limited” (Harrison Reference Harrison1961), and the deployment of tests in criminal cases (e.g. Robinson Reference Robinson1969) accompanied many commentaries arguing that such tests were invasions of privacy (Corrigan Reference Corrigan1965). Said one critic in a New York Times article, it is “amazing how easily we have relinquished our constitutional rights of protection against search and seizure when it applies to our minds instead of our properties” (quoted in New York Times, Anon. 1966b). Time Magazine, likewise, reported that the MMPI had received “strong congressional criticism on the ground that it requires answers that are nobody else's business” (Anon. 1965b) with ultimately personality tests being banned in federal government hiring, affecting “90 Per Cent of Federal Jobs” (Causey Reference Causey1965).
Many critics and historians have recounted the emergence of psychological culture and the role of testing technologies in helping individuals to begin to “understand themselves” and to adjust to the dominant order (Foucault Reference Foucault, Martin, Gutman and Hutton1988, 18; Herman Reference Herman1995; Rose Reference Rose1996, 101–15). The rise of applied psychology in the interwar period and after heralded a growing faith in America that psychologists, working often in concert with large bureaucracies, could “facilitate” the adjustment of individuals to a “fixed environment” (Napoli Reference Napoli1981, 13, 38- 41, and 120; Capshew Reference Capshew1999, esp. 50–51; Gibby and Zickar Reference Gibby and Zickar2008). While the asylum or state hospital remained important, there was ever-growing recognition that psychiatrists and psychologists worked on problems beyond those institutions that were associated with neurology, human relationships, substance abuse, mental disorder, mental retardation, and even social and moral deviance (Grob Reference Grob1983; Grob Reference Grob1991a; Farreras, Hannaway, and Harden Reference Farreras, Hannaway and Harden2004). One result was Emil Kraepelin's new nosological system (Engstrom Reference Engstrom2003, 140–142); another was the ascendance of psychoanalysis in psychiatry (Hale Reference Hale1995; Burnham Reference Burnham2012). The last was the rise of psychometrics.
Although the origins of psychometric tests traced back to Francis Galton's anthropometry, the introduction in America of the Stanford-Binet intelligence tests and the subsequent use of intelligence tests by the American Armed Forces during the First World War, meant that by the 1920s the testing of individuals had become a mainstream enterprise (Kevles Reference Kevles1968; Gould Reference Gould1981; Sokal Reference Sokal1987, 1–17). Indeed, by the close of the interwar period “American investigators, in their almost exclusive reliance on tests, seemed to have substituted technology for science” (Danziger Reference Danziger1990, 165).Footnote 1 Technologies of assessment thus became a marked feature of the narrative landscape of those fields’ advances in understanding, intervening, and adjusting people with mental illnesses (Buchanan Reference Buchanan1997). As Nikolas Rose observes of their deployment by governments, militaries, industry, and other disciplinary institutions, “the enquiries and judgments” of psychometric tests were psychological, individual, reformatory, and “not made in terms of rule and its transgression, but in terms of a norm and an assessment of normality” (Rose Reference Rose1989, 140). In this sense psychometrics were technologies of power, and they eventually engendered heightened public reactions, not least of which was the anti-expert movement called anti-psychiatry (e.g. Szasz Reference Szasz1960; Laing Reference Laing1967; but also see Burnham Reference Burnham1988, 187–192).
Psychometric technologies offered competing strategies for individual adjustment, and their deployment paralleled a variety of other ornate statistical practices and technologies that came ever-more to the foreground in Cold-War America (Rose Reference Rose1985; Danziger Reference Danziger1990; Hale Reference Hale1995). As opinion polls, for example, became a fixture of the American landscape, a growing faith in averages and individual desires to be within the “norm” began to appear in populations. People used such ‘technologies of the self’ to refine and understand themselves against a hypothetical bell curve. It was, however, another matter to let others use these technologies as the tools of managed democracy. Such uses marked the emergence of a pattern of “distrust and dependency” on such technologies (Igo Reference Igo2007, 155). While defenders of personality tests, especially the MMPI, routinely claimed that psychometrics revealed “fit individuals for particular occupations” (Anon. 1965a) it did not go unnoticed that its users had no adequate response to critics’ complaints that: “there seems to be no evidence . . . that an impressive percentage of those tagged as potentially unstable in, say 1956, have actually become unstable ten years later” (Friendly Reference Friendly1966). Such comments foreshadowed the emergence of widespread suspicion of the ends to which psychological and psychiatric expertise was directed, and both fields became cultural pariahs in the Vietnam War-era (Herman Reference Herman1995, 255).
What do these changes in public and professional attitudes toward psychometrics signal? On one hand, that psychometrics generally, and the MMPI specifically, could elicit such cultural and political controversy suggests that ongoing efforts to standardize psychiatric and psychological approaches through psychometric technologies had begun paying professional dividends in postwar America (Coon Reference Coon1993). On the other hand, it is possible to observe that the rising backlash against psychometrics parallels a cultural phenomenon historian Paul Forman identifies as a cultural transition from the historical epoch of modernity to the epoch of postmodernity (Forman Reference Forman2007; Forman Reference Forman2010). Following Forman's chronology through the case of the MMPI, it is possible to see that the spirit of the interwar and postwar advance of psychometrics was embedded in the research laboratory. In modernity, psychometric tools emerged as a means and method for standardizing, linking, and hardening formerly “soft” knowledge claims in psychiatry and psychology. Psychometrics served initially as lots and lots of means for validating therapeutic standards and unifying scientific approaches to personality. Yet in the 1950s psychometric means became inextricable from psychometric ends as the technology moved from the laboratory into society and culture. In the epoch of postmodernity those ends in day-to-day applications by government and industry epitomized (at least in cultural imagination) the suspect technologies of power and stoked the anti-expertise ethos that followed.
The remainder of this essay enlarges upon this argument by focusing on the MMPI (Dahlstrom Reference Dahlstrom1992). Derived from published and unpublished sources, this essay first revisits the origins of the MMPI in 1930s-modernity, and then traces its subsequent popularization and the controversy it ultimately engendered in the era of the Vietnam War into postmodernity. Psychometrics, as the second section seeks to make more explicit, first found life in modernist clinical and scientific searches for standardized means, methods, and procedures for analyzing personality. From there the MMPI as a technology that could make psychiatric and psychological knowledge morphed into a tool of applications for generating specific policy or industry-determined ends. The case of the MMPI clarifies this mid-twentieth-century shift for psychiatry and psychology in preferences from science towards technology, from means to ends, and from loyalty to institutions, procedures, and methods towards entrepreneurial, transgressive, and speculative behavior patterns in American culture (Forman Reference Forman2012). In short, where the authors of the MMPI defined modernist norms through variation within putative transgressions and deviances, postmodern culture fashioned modernist transgressions and deviances as so typical as to be the normative cases to which society simply had to adjust.
Both Means and Measures: The Development of the MMPI Reconsidered
Originally the creators of the MMPI – Starke R. Hathaway (1903–1984) and John C. McKinley (1891–1950) – intended their inventory to resolve intractable methodological and procedural challenges that had become commonplace in the interwar mind and brain science and medicine, or so Hathaway later recalled (Hathaway Reference Hathaway1964a, 204–205; idem 1974, 112–114). In the 1930s, the time when Hathaway and McKinley first began constructing their novel personality inventory, psychiatry and neurology had become established clinical specialties in the United States, yet there was little unity between both clinical fields and with psychology (see esp. Hale Reference Hale1995; Farreras, Hannaway, and Harden Reference Farreras, Hannaway and Harden2004; but also Casper Reference Casper2008 and Gavrus Reference Gavrus2011). The patient groups seen by psychiatrists and neurologists (and later by psychologists) were, moreover, hardly easily characterized, the question of the etiology of mental illness was rarely raised, and treatments were wide-ranging and of questionable efficacy. Talk therapy, education and training, asylum management, tranquilizers and stimulants, insulin, and malarial shock therapies largely constituted the available regimens (Pressman Reference Pressman1998; Shorter and Healy Reference Shorter and Healy2007; Evans Reference Evans2011).
Neurologists, psychiatrists, and psychologists at the University of Minnesota Medical School found a creative, local institutional means of acknowledging these ambiguities in the 1930s and 1940s (Hathaway and Meehl Reference Hathaway and Meehl1951, iv). As Roderick Buchanan has noted in his history of the development of the MMPI, the Minnesota legislature had founded a psychiatric unit at the University of Minnesota in 1936 and thus seemingly blessed a sharper division between neurology and psychiatry (Buchanan Reference Buchanan1994, 150; Hathaway Reference Hathaway and Krawiec1974, 110). In practical terms the Minnesota Medical School's A. B. Baker, a broad-minded neurologist, and psychiatrist-neurologist John C. McKinley, trained their medical students in Minnesota's Division of Nervous and Mental Diseases from a neuropsychiatric perspective, one that embraced new trends in psychology, social work, and occupational therapy, and provided Hathaway, then a young experimental psychologist, with ample teaching opportunities in the medical school (Hathaway and Meehl Reference Hathaway and Meehl1951, iv; Hathaway Reference Hathaway and Krawiec1974, 112).Footnote 2 Nowhere is their joint neuropsychiatric perspective more apparent than in An Outline of Neuropsychiatry, a textbook jointly edited by McKinley, Baker, and Hathaway (and later others) that ran to multiple editions. Best described as an approximately 250-page bound typescript, this volume provided students with an indexed resource for the quick – the entire book is schematic – diagnosis of nervous and mental diseases (McKinley, Baker, and Hathaway Reference McKinley, Baker and Hathaway1940). Hathaway's own approach was decidedly physiological, a fact reflected in his textbook Physiological Psychology (Hathaway Reference Hathaway1942).
Efficiency and standards in diagnosis was but one of the motivations for the creation of the MMPI (Buchanan Reference Buchanan1994). Hathaway and McKinley were also concerned by the lack of standards in mental disease diagnosis or of any means for evaluating scientifically the efficacy of psychological and psychiatric therapies (Hathaway Reference Hathaway1964a, 205). Their research questions and studies focused on resolving the intractable problems caused by a lack of diagnostic and therapeutic standards (more broadly, Marks Reference Marks2000). Indeed Hathaway and McKinley appear not at first interested in broader social applications of personality inventories, an attitude that only began changing dramatically in the years leading up to McKinley's death in 1950 (Hathaway Reference Hathaway and Krawiec1974, 115).Footnote 3
Through research in the 1930s and early 1940s, Hathaway and McKinley succeeded in creating a tool for delineating patient personality and monitoring treatment efficacy (Hathaway and McKinley Reference Hathaway and McKinley1940). Their research produced data through a number of clinical scales (commonly recognized psychiatric classifications), including social introversion, hypochondriasis, depression, hysteria, psychopathic tendencies, masculinity-femininity identity issues, paranoia, psychasthenia, schizophrenia, and hypomania (ibid.; McKinley and Hathaway Reference McKinley and Hathaway1940; McKinley and Hathaway Reference McKinley and Hathaway1942a; McKinley and Hathaway Reference McKinley and Hathaway1942b; McKinley and Hathaway Reference McKinley and Hathaway1944). To this largely pioneering work on the psychometrics of personality types, Hathaway and McKinley supplemented a further, novel dimension, one that redressed their inventory's putative rigor. They introduced validity scales that measured test subjects’ exaggeration of positive personal virtues, their tendency to deviate from community norms, and their inclination towards dishonesty. It was these validity scales that would later account for the MMPI's rapid translation into industry, government, and education (Dahlstrom Reference Dahlstrom1992; Buchanan Reference Buchanan1994).
Hathaway and McKinley published numerous articles from the period of their scientific collaborations, roughly 1937 to 1950. Although there is little in the way of archival materials from that same period, a few of their published papers capture the intellectual development of the inventory (Hathaway and McKinley Reference Hathaway and McKinley1940; McKinley and Hathaway Reference McKinley and Hathaway1940; McKinley and Hathaway Reference McKinley and Hathaway1942a; McKinley and Hathaway Reference McKinley and Hathaway1942b; McKinley and Hathaway Reference McKinley and Hathaway1944). As these papers make clear, the development of the inventory – although inextricable from practical ends – was at first a largely experimental program of research in psychiatry and psychology. In 1939 Hathaway had, for example, noted that abundant psychological research had focused on aberrant, unstable personality characteristics, but he also cautioned that there were other groups which he termed “supernormal” who had received scant attention and could be characterized as abnormal (Hathaway Reference Hathaway1939). Yet identifying groups who were “supernormal,” or “normal” and “below normal” required a means of detection that could compare responses to the inventory against a scaled, statistically averaged range of population responses. Hathaway and McKinley's early research thus statistically defined normal and abnormal character profiles through the construction of a new questionnaire system, and in this way their research was little different from the personality research common throughout the interwar period and after and was never directed towards any particular end, even if clinical deployment was intended (Danziger Reference Danziger1990, 85–87).Footnote 4
The MMPI, as Hathaway and McKinley constructed it, measured “normal” thoughts, actions, emotions, and desires. Their professional collaboration uniquely merged imperatives in both research psychology and psychiatry. In their efforts to define normal and abnormal responses, Hathaway and McKinley relied upon patients and visitors to the University of Minnesota Medical School Hospital. Often these patients were recovering from surgery or were family members or friends visiting patients undergoing hospital treatments (see, for example, Hathaway and McKinley Reference Hathaway and McKinley1940). For the control (i.e. “normal”) population, the co-collaborators concluded they would deem visitors to the hospital psychologically normal, provided they were not under the care of a physician. In addition, they also adopted other normative characteristics: being married, between ages 26 and 43, possessing an eighth grade education or better, being somewhat religious, and of moderate economic background – these they took to define the typical and thus “normal” Minnesotan (the reified aggregated subject who became the normative standard for everyone else, see: McKinley and Hathaway Reference McKinley and Hathaway1940; McKinley and Hathaway Reference McKinley and Hathaway1942a; McKinley and Hathaway Reference McKinley and Hathaway1942b). Later Hathaway observed that his normal test subjects were typically of northern European descent and mainly white (Hathaway Reference Hathaway1964b).
How Hathaway and McKinley actually determined the way to define and measure the MMPI's experimental populations is unclear in both their published papers and unpublished manuscript papers. Indeed one immediately puzzling feature of the MMPI is that it adopted, analyzed, and statistically described diagnostic categories in psychiatry and psychology that were increasingly contested. Hathaway in 1974 identified the Kraepelian system of diagnosis as one source (Hathaway Reference Hathaway and Krawiec1974, 115) and claimed as well that the chief aim of the inventory had always been to discriminate between diagnostically identified groups and normals (ibid., 116). However, hysteria, the best example, was a clinical diagnosis that had practically disappeared by 1940 (a point acknowledged by Hathaway Reference Hathaway1971; but see Micale Reference Micale1993). Categories like “psychopathic deviate” and “schizophrenia,” as Bonnie Evans (Reference Evans2011) has shown, had equally varied acceptance and interpretation by psychiatrists and clinical psychologists. The most likely conjecture is that McKinley and Hathaway, perhaps in consultation with A. B. Baker, relied upon the diagnostic definitions set in their own neuropsychiatry textbook (vaguely Kraepelian) to define the conditions and to isolate their primary experimental populations (a claim made because there are evident overlaps between Wilson n.d., and McKinley, Baker, and Hathaway Reference McKinley, Baker and Hathaway1940).
Equally unclear from the record is how McKinley and Hathaway determined what “true” or “false” questions would best capture these scales – this is perhaps one reason explaining why the original inventories had more than one-thousand test questions.Footnote 5 In any case, they explored various normal and abnormal patterns in response to those questions through comparison of responses taken by control and test subjects (the latter were largely comprised of patients clinically diagnosed with the conditions). In all of these papers, McKinley and Hathaway compared responses from both groups in order to generate scales that would capture below normal, low normal, normal, and supernormal responses (Hathaway and McKinley Reference Hathaway and McKinley1940; McKinley and Hathaway Reference McKinley and Hathaway1940; McKinley and Hathaway Reference McKinley and Hathaway1942a; McKinley and Hathaway Reference McKinley and Hathaway1942b; McKinley and Hathaway Reference McKinley and Hathaway1944). Eventually the later practice of diagnosing mental illness from the results of the exam rendered any abnormal response suspect, but in effect the conceit at the heart of the MMPI was that the inventory measured normal personality as a healthy range of “normal” mental disorder. There was, for example, a normal range for paranoia, just not too little or too much.
The papers published from 1942 through 1944 dwelt not only with the construction of the inventory but contended with deeper limitations that threatened the tool's overall efficacy and legitimacy as a means of producing empirically rigorous definitions and therapeutic standards. The “depression” scale, for instance, relied upon characteristics at the time of testing rather than any definition of depression in general. In this instance, Hathaway and McKinley charged hospital workers with identifying depressed patients for the study from within the patient population at the University of Minnesota Hospital. Often the normal individuals tested in the study were visiting sick patients at the hospital, suggesting that everyone might be mildly-depressed (McKinley and Hathaway Reference McKinley and Hathaway1942a). “Psychasthenia,” which McKinley and Hathaway characterized as excessive doubt, compulsions, obsession and unreasonable fear, was no less problematic. It was a condition so rare (or so rarely diagnosed) that Hathaway and McKinley's paper could rely on a sample size of twenty patients only (Reference McKinley and Hathaway1942b). In the instance of “hysteria,” Hathaway and McKinley observed in a 1944 study that the test conditions actually seemed to cause elevated responses in normal test subjects. In almost all instances, McKinley and Hathaway argued that statistics provided a means of overcoming these limitations. Each scale could be adjusted to account for anomalies, small populations, or the test conditions themselves. Statistical science offered a dispassionate method for circumventing these conditions.
By 1944 – a year after their inventory was first published – Hathaway and McKinley published a study in the Journal of Applied Psychology which discussed the “hysteria” scale and also explored possible ends for the MMPI's “hypomania” and “psychopathic deviate” scales. Hypomania they defined as a mild form of manic excitement in manic-depressive psychoses; sufferers had unstable moods and were flighty and had a tendency towards being over-talkative, distractible, enthusiastic or gloomy, and sometimes too happy or too frustrated. Psychopathic deviance was marked by asocial behaviors only. They speculated that juveniles with hypomania would likely benefit from counseling. Meanwhile their study of psychopathic deviance relied upon a survey of responses from 17–22 year-olds from “broken homes,” “college students,” and a population of prisoners. The differences were startling, and the authors’ commentaries speculated that the differences offered predictive possibilities. In the case of psychopathic deviance, McKinley and Hathaway noted that abnormal individuals sometimes had supernormal test scores and did not admit to having common flaws. The co-collaborators suggested that the inventory provided clinicians with a means for identifying these individuals, because it was often difficult to see beyond the “halo effect” such individuals projected (McKinley and Hathaway Reference McKinley and Hathaway1944). Other applications would soon become manifest.
The MMPI was a product largely of the research reported in these papers (and many others). In the 1950s and 1960s, the life of the MMPI changed dramatically as it became a highly ambitious psychometric tool, one that provided a means of evaluating the whole of personality and engaging in research into populations (Buchanan Reference Buchanan1997). In a 1939 paper, Hathaway observed that he had been dissatisfied with many other personality tests, including the Bell Adjustment Inventory, the Bernreuter Personality Inventory, and the Humm-Wadsworth Temperament Scales (Hathaway Reference Hathaway1939, 112–113). These competitors, he and McKinley had sniffed later in 1940, were too small in scope, limited to one particular reaction, and often relied upon language far too complex for diagnosing low IQ patients (Hathaway and McKinley Reference Hathaway and McKinley1940; Hathaway Reference Hathaway and Krawiec1974, 114). The MMPI as they published it in its first version was broader, simpler, and highly adaptable to varied populations (Hathaway and McKinley Reference Hathaway and McKinley1943b).
It was in such early discussions as these that the inventory began its journey from the research laboratory of psychology and psychiatry into more public domains. Like many in wartime America, American psychologists had begun searching for ways to aid the war-effort, and those trends influenced discussions about the MMPI as an experimental tool towards questions about its possible application, including its method of forecasting and also detecting ‘at risk’ individuals (see Hathaway to Blegen, 2 July 1942, Hathaway Papers). In fact, the MMPI would be used experimentally by the US military for “screening and selection” purposes and as part of a battery of aptitude tests (Kennedy, Hughes, and McNeil Reference Kennedy, Hughes, McNeil, Kennedy and Zillmer2012). It is noteworthy, too, that sources in Hathaway's personal papers (mainly from the period after 1943) largely direct attention away from the research involved in developing the MMPI and move towards showcasing the means that rendered it suitable to the ends of the mass market. Many of the primary sources consist of various productions of the inventory, much of it testifying to the engineering ingenuity that underlay Hathaway's design of the test in a pre-computing era.
To be sure, the MMPI's introduction of validity scales – four scales that could assess for example the honesty of the test taker – accounted for its widespread use (Dahlstrom Reference Dahlstrom1992). It was those measures of validity that also led to increasing public mistrust and congressional disquiet about applying the MMPI beyond clinical settings. In a significant sense the validity scales forced test subjects to comply with the central logic of the inventory (Friendly Reference Friendly1966). The inventory invited subjects to give honest responses about their fantasies and fears or risk having to retake the test, an activity that was itself construed as indicative of mental disorder and aberrant personality (King, Good, and Brantner Reference King, Good and Brantner1961, 11–15). Although individual answers were never the point, many questions were of an extremely personal nature, no doubt inspiring fear in test subjects of the potential for blackmail, retaliation, or gossip and innuendo, and Hathaway very much wanted the patient to speak truthfully for him or herself (Hathaway Reference Hathaway and Krawiec1974, 115). Whereas Hathaway and McKinley – and those psychologists, psychiatrists, neurologists, and physicians who began using the exam – likely regarded the MMPI as one means of diagnosing and categorizing patients, evaluating treatment approaches, and as a part of a wider “psych” profile, the public – as reflected especially in newspapers and transcripts from congressional hearings – increasingly began focusing on the ends of the MMPI, suspecting that it was abetting a potentially more alarming political project organized by the “brain watchers” (Bard Reference Bard1966). What's more, since the method of the test enforced compliance and obedience with its own internal logics, it gave the people who administrated the test both power and position to treat below normal and supernormal variations from acceptable community norms punitively. Thus if the MMPI origins derived largely from scientific questions about standards for the diagnosis psychological and psychiatric conditions and measures of the efficacy of treatments, the public's subsequent response to the inventory was all about ends to which imagined others might and actually did direct it.
Of Inextricable Ends: Psychometrics in Modernity and Afterwards
The MMPI was a technology developed in modernity, a period when scientists and policy-makers cultivated expertise, methods, and procedures that illuminated scientific principles and facilitated the pursuit of further knowledge. Its development coincided, however, with the cultural emergence of a more incredulous Atomic Age, one that critics and historians later claimed as demonstrating the impressive primacy of technologies and the variety of the ends to which they could be directed (Lyotard Reference Lyotard1984; Forman Reference Forman2010). Increasingly Cold War American culture aggrandized more entrepreneurial and transgressive behaviors; increasingly also Americans evinced a growing dislike of procedurist states and cast regulations, rules, and expectations of obedience and compliance as obstructing individual self-expression and economic behavior.Footnote 6 While many lauded and celebrated the ends and applications of technologies in-and-of-themselves, the ability to use technologies as a means of furthering future science, maintaining social order, adjusting individuals or governing individual behavior, met growing cultural resistance in the Vietnam era (Igo Reference Igo2007). Technology was for empowerment and enhancement; it was not for regulation and control.
While psychometrics was part-and-parcel of the modernist process by which methodological uniformity began emerging in the mind and brain sciences, psychologists’ distinctive contribution to these trends was in their growing consensus that the mind and brain sciences could meet industrial, military, or government needs (Danziger Reference Danziger1990, 101; Herman Reference Herman1995). Psychometricians generally, and those administrating the MMPI in particular, contended with the reality that the ends to which the tests might otherwise have been laudably directed by experts also paralleled the means with which the modernist state democratically regulated social, economic, and moral behavior (Rose Reference Rose1989).Footnote 7 How these broader patterns were explicitly realized in the instance of the MMPI can be explored through the life of the “thing” itself.
As Hathaway and McKinley developed their clinical and validity scales to quantitatively measure personality, they originally appear to have intended the inventory as a means of evaluating personality and affect over time and for assessing the efficacy of psychiatric treatments (Hathaway to Blegen, 19 May 1941, Hathaway Papers). The results from the MMPI were supposed to be a snapshot of personality, taken to be a dynamic, changing thing, and in An Atlas for the Clinical Use of the MMPI, published in 1951, it was explicitly acknowledged that the results of the MMPI varied over time (Hathaway and Meehl Reference Hathaway and Meehl1951).Footnote 8 In this sense the MMPI was the logical extension of standardizing hospital technologies that recorded heart rate, blood pressure, and micturition, only it provided a formal readout of the changing mind rather than physiological changes (see Howell 1995, esp. 56). Hathaway's early career had been marked by such interests in such devices of revealing; he had even invented a “lie detector” while in university (Hathaway Reference Hathaway and Krawiec1974, 107).
Crucially Hathaway and McKinley recognized that their research subjects, who had often come from surgery or were in in-patient psychiatric care, tended to be unhappier than patients who were questioned after a few weeks of recovery. The point of the questions underpinning the inventory was not that patients would answer them the same way each time but rather that when they answered them they would do so honestly, which would allow the inventory to capture changes in affect from the last assessment (ibid., 114–115). Treatment efficacy could also be measured, and in a perhaps too-honest moment expressing his dissatisfaction with psychiatric therapies, Hathaway described a patient before and after frontal lobotomy as showing no change in result in the MMPI (Hathaway Reference Hathaway1942, 276–7).
Such examples showcase the fact that Hathaway and McKinley did not originally regard their inventory as a one-time measure of personality – that end emerged later when industry, government, and counselors began applying the test as one-time assessment vehicles (King, Good, and Brantner Reference King, Good and Brantner1961, 3). Instead, for them in the late 1930s and early 1940s, the inventory established a measure of and a baseline for personality against which the same test subject could be measured again subsequently (Anon. c. 1947; Hathaway Reference Hathaway and Krawiec1974, 115–116). Likewise, the inventory's baselines functioned as a means for establishing a starting point for future comparative research, which would necessarily examine a variety of populations in order to provide ever greater scientific understanding of personality and mental illness. This research occurred from 1937 to 1949 and appeared in various atlases of MMPI profiles thereafter (Hathaway and Meehl Reference Hathaway and Meehl1951; Hathaway and Monachesi Reference Hathaway and Monachesi1961; Hathaway Reference Hathaway1971, 2).
Developing the MMPI furthermore raised a number of technical and statistical challenges; solving these effectively rendered the inventory a signal achievement of Hathaway's career as medical psychologist, especially after McKinley's death in 1950, and the result of solving them also clearly aided its widespread dissemination. Here certain early details of Hathaway's biography are suggestive.Footnote 9 Starting out as an undergraduate in electrical engineering, Hathaway had turned away from that subject to majors in psychology and mathematics, earning his Bachelor's degree in those subjects at The Ohio University in 1927. He subsequently pursued a Masters degree in psychology and statistics at Ohio State University and then completed his doctoral training in psychology and anatomy at the University of Minnesota, Minneapolis, in 1932, where he continued his career conducting neurophysiology research and teaching (Hathaway Reference Hathaway and Krawiec1974; Anon. 1986). Although Hathaway saw psychology as a “more challenging” field of work, his early interests in radio, engineering, electricity, and vacuum tubes never fully vanished and indeed become most visible in photographs from his private life in Minnesota. In these photographs Hathaway fashioned images of himself at home as an inventor, comfortable in a well-equipped private workshop, and tinkering with metals and circuits. In his autobiography, he observed that his childhood had been spent in “workshops” (Hathaway Reference Hathaway and Krawiec1974, 105). In other comments, he supposed that some of his attitudes, especially his view that technology could provide a solution to psychology's woes, emerged from a desire to engineer tools to quantify the human mind. As he noted: “I moved into psychology mainly basing my interest upon my knowledge of electronics and physics and at first was concerned with building of apparatus for measurement of the electro-dermo responses . . . and the development of other apparatus useful to such aspects of psychological measurement” (Hathaway Reference Hathaway1963).
Thus, for Hathaway as psychometrician the creation of the tools of measurement and the making of the means of their rapid interpretation was as much of an achievement as the scientific research the tools enabled. These motivations appear throughout his archive at Minnesota. Many of the sources in his personal papers capture a material history of his inventory in the way obvious problems were uncovered and resolved. Scoring, one such technical problem, lay in the challenge of efficiently revealing patterns in the “true” or “false” responses to the questions. Initially, McKinley and Hathaway had created a set of cards that patients could group by “true”, “false”, or “no answer.” The cards had letters and numbers in the upper right-hand corner indicating the scales to which the answer applied. Once the test subject had answered all of the questions, the examiner counted the “true” responses and organized them by scale (Ricks to Hathaway, 25 March 1966, Hathaway Papers). Thereafter with some simple arithmetic, accompanied by hand-drawn pencil graphs in a standardized form with clearly labeled scales, a picture could be generated for interpretation. Below normal, low normal, normal, and supernormal responses were reified into a graphical portrait recording the spikes and troughs of personality – this was a personality profile.
Clever test-subjects, however, might spot patterns in the card-based system and intentionally moderate their answers and it was eventually only used for low intelligent participants (King, Good, and Brantner Reference King, Good and Brantner1961, 4). Hathaway's way of solving this problem lay in introducing test subject blindness. Thus a second form of the inventory was born, this time a bubble sheet with “true” and “false” responses in which the subject penciled-in responses. This new strategy made the subject less aware of the existence of patterns in the test. Whereas in the prior situation the subject had possessed some limited agency, the new method introduced greater opacity to the testing situation. As one respondent to the questionnaire sarcastically scrawled across its cover: “if this is an aptitude test, answers cannot be made in a sixteenth inch square; if it is a sanity test, you’re off on the wrong foot” (Hathaway and McKinley Reference Hathaway and McKinley1943a). Yet the answer sheet presented Hathaway with a new technical challenge: how to tabulate the answers to 556 questions without making mistakes? His solution, although simple, was ingenious. Thirteen templates, each representing a clinical or validity scale were made to fit over the bubble-sheet, with holes open over the relevant answers that any clerical assistant could tabulate (King, Good, and Brantner Reference King, Good and Brantner1961, 54–64).Footnote 10
These matters of the material career of the MMPI were but one facet of the way in which means and methods were found to aid the examiners’ evaluation of the MMPI's results. Another was the inventory's validity scales which could identify dissimilation through omission, outright lying, or random answers, and here Hathaway's evident pride can be gleaned even from autobiographical sources, where he wrote, for example, “the F scale was my invention” (Hathaway Reference Hathaway and Krawiec1974, 116). The simplest scale of validity was checked to make sure that inventory omissions had not invalidated the results (Hathaway and McKinley Reference Hathaway and McKinley1943b). The scale that measured lying was more sophisticated and sought to identify individuals who were presenting themselves in a too-positive light or also too-negative light (Wilson n.d.; but also see McKinley and Hathaway Reference McKinley and Hathaway1944). The MMPI relied upon questions scattered throughout it such as “I do not read every editorial in the newspaper every day” and “If I could get into a movie without paying and be sure I was not seen, I would probably do it” (Hathaway and McKinley Reference Hathaway and McKinley1943a). Both questions when answered truthfully by most people were expected to be “yes”. In this way, people who were attempting to give responses that reflected themselves in an overly positive light would be revealed. A high score on this scale did not invalidate the test, but it did indicate that scores elsewhere on the scale were higher too. The final scale explored the attitude of the test subject by comparing test responses to 90 per cent of other test takers. This scale was a means to demonstrate that the test subject had understood the test, had taken the time to read it carefully, but it furthermore created stronger internal validity between responses about physical symptoms versus mental ones (Anon. c. 1947; Wilson n.d.).
The important point is that the validity scales and technical solutions to scoring the test were evermore tied to the challenge of disseminating the MMPI to a wider professional audience. The Psychological Corporation, apparently after some cajoling, had begun disseminating the inventory, and by the 1950s the paper version, test sheets, templates, and interpretive guides had become common features of the psychological and psychiatric clinic, its rapid acceptance due largely to its validity scales (Memorandum of Agreement, 1942, Hathaway Papers; Wulf to Hathaway, 31 May 1983, Hathaway Papers). By 1966, the Psychological Corporation was distributing “500,000 copies” per year (Feburary 8, 1966).
The interpretive guide that developed with the inventory outlined the various profiles and offered therapeutic recommendations. It even called attention to recognizable differences across populations. As one example: normal test subjects would usually receive a mean score of 4 on questions pertaining to an unlikely degree of virtue. But the interpretive guide intimated that these findings needed to be qualified: the mean for members of the clergy, for example, was six (Wilson n.d., 3). Thus, the blunt instrument of psychometric validity was necessarily checked by the examiner's awareness of mitigating factors. At the same time, the MMPI was easy to use, convincing, and had the aura of statistical responsibility.
As yet another indicator of its rapid uptake and dissemination was the publication in 1961, with, the authors stated, much advice from Hathaway, of a short guide for physicians, “adaptable to many other settings . . . in such diverse fields as counseling, industry, and internal medicine” (King, Good, and Brantner Reference King, Good and Brantner1961, i, 3). Republished and slightly enlarged in 1974, the volume included easily understood explanations of the test's utility, a description of the purposes of the validity scales, and a cook-book of descriptions of the twenty-eight personality “codes” that most frequently appeared (ibid., 38). It also included a number of additional scales, including for instance a “caudality scale” that could measure the “unsuccessful” application of a lobotomy (ibid., 37).
While Hathaway's involvement in producing the guide was marginal, the guide succeeded because of the “codes” that he and others invented in the late 1940s and which appeared in the 1951 Atlas. These codes were equal parts research achievement and innovation. Originally Hathaway and McKinley had utilized letter abbreviations for each of the clinical scales. “Mf,” for example, had been originally used for Male/Female Inverts but was rendered as Scale 5 or just “5” in the new coding system (1 through 9 represented the clinical scales). The coding procedure was involved, and need not be described here exhaustively. Suffice it to say that the codes were ordered in terms of their highest supernormal peaks, then their normal range, and finally their below normal range, with various other signs – underlines or apostrophes – indicating important information. There was no intention of calling attention to high peaks over low peaks, but rather the profile was to be understood as a whole (and when appropriate graphed that way in terms of an adjusted raw score as well). The advantage of the codes was that they allowed for comparison of an individual's personality with that of other like groups, and furthermore allowed strategies for adjusting the individual's personality to others as was sometimes done in marriage counseling.
As has already been discussed, the instrument's popularization also coincided with the emergence of growing controversy about the validity of psychometric testing. While scientists and administrators saw the MMPI in administrative terms, the press, the public, and policy makers and the public viewed it as an invasion of personal privacy (see many press clippings in Folder 2, Hathaway Papers). It was in the validity scales especially that a means and an end merged together. To both the public and to politicians the ends appeared to be subject compliance, and in the instance of the MMPI, the means were inextricable from this ultimate end. In order for the MMPI to function as a technology supporting scientific research into personality and to evaluate treatment efficacy, test subjects had to produce valid results. The tool forced compliance. It was this feature of the MMPI that fundamentally elicited public controversy. For no matter how much those who used it for mere data collection, scientific inquiries into personality, or therapeutic evaluation protested that it was simply a tool with no end in mind other than understanding personality and personality changes, the end, or the inventory's many potential ends, became manifest. As Minneapolis Tribune newspaper columnist David Mazie put it: “What does a sex life have to do with a job?” (Mazie Reference Mazie1963). Hathaway's answer was that “jobs are not really appropriate for the terribly maladjusted or the highly prejudiced,” and he thus expressed little concern that his inventory could become a tool of social surveillance (ibid.; also see Hathaway Reference Hathaway and Krawiec1974). Quoted in The New York Times during a moment of scandal in 1966, Hathaway proffered that “To attack tests is to a certain extent comparable to an attack upon knives. Both good and bad use of knives occurs because they are sharp instruments” (Anon. 1966b).
By the 1960s many industries and government agencies had embraced the MMPI. They had often done so as a one-time measure of personality fitness, even though it had been originally intended for multiple uses. Export-Import Bank, Bonneville Power Administration, the Peace Corps, and the Department of Labor were among those who began screening prospective employees and a number of studies of juvenile populations in Minnesota and other places commenced as well (Ervin to Bennett, 21 April 1966, Hathaway Papers). Another critique of the MMPI that emerged was that its use in hiring practices was in some sense permanently determinative of personality in terms of fitness for employment. Many believed personality changed in time, especially in moments of distress. Many also believed personality could be adjusted, and such punitive interpretation of the test's results conflicted with optimistic psychological theory. There was also a sense that elements of the test were discriminatory, suited to the ends of hiring, especially when it used answers to questions about religious conviction as a means of selecting a personality.
While clearly even prior to McKinley's death, he and Hathaway had begun considering possible social applications for their inventory, it is also clear that Hathaway was somewhat ambivalent about its widespread use. In answering the criticisms that emerged about the MMPI, he often fashioned himself as a dispassionate scientist (Hathaway Reference Hathaway and Krawiec1974, 117). He argued, on the one hand, for the utility of personality tests, while, on the other, he admitted their limitations. In 1971 he wrote: “If another twelve years were to go by and we have not gone on to a better instrument or better procedure for the practical needs, I fear that the MMPI, like some other tests, may have changed from a hopeful innovation to an aged obstacle” (Hathaway Reference Hathaway1971, 3).
Conclusion
In this way, Hathaway never escaped the constraints of modernist values, even as Hathaway, engineer cum inventor remained enthralled by the ends of his psychometric means. At the same time, the public had begun its own form of escape from modernity – it was not the technology that they had rejected but rather its deployment by experts in pursuit of allegedly well-adjusted personalities. Where psychologists saw transgression, the public increasingly adopted pluralism and equivalence and saw only different groups and different practices (Igo Reference Igo2007).
One place where this was obvious even to Hathaway was in the case of gender identity interest patterns. The inadequacy of that category diagnostically was so obvious to Hathaway that by 1951, scale 5 was left out of the code and indicated only parenthetically alongside the other numbers in his first atlas (Hathaway and Meehl Reference Hathaway and Meehl1951; King, Good, and Brantner Reference King, Good and Brantner1961, 25–26). But it became even more obvious in the 1970s that the results of people “whose cultural or ethnic background is very different from the Minnesota norm group” required a qualified interpretation, especially for “hypotheses about personality and adjustment . . . of blacks, American Indians, Spanish Americans, the extremely wealthy, seminarians, artists, poets and dancers, isolated mountain folk, and other similar groups” (King, Good, and Brantner Reference King, Good and Brantner1974, 91).
In the 1950s and 1960s the MMPI emerged as a significant technology. At least at the University of Minnesota Medical School, and despite considerable rivalries there, the MMPI brought together clinical psychologists, psychiatrists, and neurologists (Hathaway and Meehl Reference Hathaway and Meehl1951, iv). The test aided identification of organic neurological conditions (King, Good, and Brantner Reference King, Good and Brantner1961, 60), and Hathaway himself professionally bridged all three worlds at Minnesota (Hathaway Reference Hathaway and Krawiec1974, 117) and he believed that the test united those specialties more than it divided them (Hathaway Reference Hathaway1942, 315–316). While his instrument emulated rigorous demography, its material output – the graph – paralleled in shape the oscilloscopic readouts from electrophysiology. The inventory thus crystallized in a familiar configuration a major subject within the science and medicine of the mind and brain, and it did so in ways that suggested it aided psychologists, psychiatrists, and neurologists in finding a unifying means of translating and comparing their research, clinical findings, and therapeutic results.
The MMPI was a technology that provided many means and for many groups, so much so that when they revised their The Physician's Guide to MMPI in 1974, Patricia King, Ellison Good, and John P. Brantner re-titled it A Practical Guide to the MMPI: An Introduction for Psychologists, Physicians, Social Workers, and Other Professionals (1974). The ends of the MMPI – like so much of applied psychology – were at once mundane and novel. While they aided the standardization of mental health diagnoses and the evaluation of treatment efficacy, they also hinted at the reasonable end of detecting individuals who deviated from population norms in their attitudes and character. In their Atlas of Juvenile MMPI Profiles, Hathaway and Elio Monachesi suggested that their atlas might prove useful “to predict the direction and nature of changes” in juvenile personalities (Hathaway and Monachesi Reference Hathaway and Monachesi1961, viii). In this way, the MMPI emulated its modernist precursors that evaluated intelligence, which had been introduced to categorize lower functioning children and thereby became part of the educational history of the United States (Sokul Reference Sokal1987).
Stepping away from these narrower details, it is perhaps right to end by returning to the broader chronological currents that have framed this story (Forman Reference Forman2010; Forman Reference Forman2012). It is easy to see that the project of psychometrics embodied the many ideals of modernity, but despite the popular ire it generated afterwards, psychometrics can also be seen to have been one of modernity's more innocuous expressions. Personality assays – the written tests, their penciled graphs, ribbon-printed paper, typewriters, and card-punch computers – appear now rather innocent technologies of biopower in comparison with the actual medical scandals that came to public attention in the 1960s and afterwards and which so contributed to backlash against expertise that characterized the Vietnam-era and afterwards. More specifically, the MMPI, one test among many of personality, lacked an alluring novelty and resisted categorizing people in meritocratic terms. If the MMPI had predictive powers, then it is also true that no truly draconian public policies seem to have emerged from it that stigmatized and hurt populations (individuals, of course, are another matter).
Fundamentally public reaction to the MMPI pointed towards an emergent, different cultural relation in American society to transgression and to technology. The MMPI was so inextricably embedded in that transformation that the moderns who made and used it, and the publics that critiqued it, inhabited different cultural worlds. The possibilities of psychometrics were limited precisely by a culture that had sought a means of understanding and limiting personality rather than producing personality through transgression. In sharp contrast, designer pharmacological therapies, cyborg-imagery, green-florescent proteins, transgenics, and neuro-imaging technologies – all of which began emerging in the same period – set that new culture in sharp relief. Those technologies offered the same normative possibilities as psychometric assessments, but where psychometrics was seen as restraining personality, its “tech” progeny claimed also to empower and enhance human design. Where modernity's tools of standardization provided psychiatry, neurology, and psychology (and also in consequence government and industry) the means of defining normality, the emergent culture of America in the 1960s and 1970s was opposed to exactly that kind of definition. By the 1980s that culture would extol radical individuality, link transgression with consumerism and social progress, and market and mainstream deviance. In short, where moderns sought a means of defining “normality” for the purposes of adjustment, postmoderns celebrated their transcendence from adjustment.
