History of the Research Center for Mental Health, New York University (NYU)

Robert R. Holt

[This paper was presented on June 10, 2006, at the Annual Meeting of the Rapaport-Klein Study Group]

[See the video of this presentation]


What follows is the account of a unique institution, in which I had the privilege of playing a fairly prominent role. In recent years, I have become convinced that the Research Center for Mental Health deserves at least a footnote in the history of clinical psychology, and there is no one alive who knows this tale better than I do. RCMH (as we familiarly called the "Research Center") is noteworthy, I believe, because it was the first university-based center for systematic, scientific research dedicated to the clarification and testing of psychoanalytic theory. So far as I know, it is the only one of its size and longevity with such a central focus. As I hope to show, it produced a respectable body of work, pioneered in a few respects, and helped train a generation of psychoanalytically oriented researchers.

This institution was conceived by the late Stuart W. Cook, head of the department of psychology in the Graduate School of Arts and Science, New York University (NYU). He was trying to build up a first-rate department of psychology at a then rather under-financed, privately-owned university by the device of setting up a series of research centers for several of the major divisions of the discipline. The idea was that the leaders of these centers would get grant money to support the research work of bright young Ph.D.s, who could teach part time and create a setting in which graduate students would be drawn into ongoing projects. Working as research assistants, they would get investigative training, could get started on dissertation projects, and might find some needed financial support. Cook himself founded one for social psychology, the Research Center for Human Relations, and staffed it with a brilliant and outstanding group of colleagues. His plan was to match it with similar institutions devoted to research in clinical and in industrial psychology. Accordingly, in 1952 he persuaded a major figure in his department, Isidor Chein, to found the Research Center for Mental Health as an institutional home for his large, multidisciplinary project on heroin use in New York City (see Chein, 1964). And a few years later, Raymond Katzell founded the Research Center for Industrial Behavior.

Though of major stature, Chein was not a clinical psychologist, and Cook thought of what he was doing as only a holding action until he could get someone from within clinical psychology to take over. At the American Psychological Association (APA) meetings in 1952, George Klein introduced me to Stuart Cook, who offered both of us jobs. I liked him immediately, and the subsequent years made my fondness and respect for him steadily increase. A gentle but strong person, he proved to be an almost ideal democratic leader: devoted to encouraging everyone's ideas and participation in group decisions and skilled in doing so, but also tough and effective in dealing with deans and fighting for our rights. He deserves a great deal of credit not only for developing the basic plan but for creating an ideal departmental climate in which the fledgling center could grow and take wing.

George and I had met and become good friends at the Menninger Foundation, where we were both members of David Rapaport' s Research Department and the Psychological Staff. We decided that when we left Topeka -- that we would leave, like most of our colleagues -- we would try to work together to put into effect some of our common ideals about research in clinical psychology. We wanted to use the best statistical, experimental, and psychometric methods to confront psychoanalytic propositions with data, to confirm or refute them, and to enlarge the theory. Moreover, we realized that the theory was in a chaotic enough state that it was necessary to do a good deal of preliminary work to clarify and systematize it so that the propositions we tested would really make a difference. Our initial impression was that the theory of thinking was a good place to start: it could be made operational and testable fairly easily; it seemed central to Freud's metapsychology; and we already had ideas about cognitive studies that would engage it.

We decided to run the Research Center for Mental Health jointly. Klein took the title of Director of Research, and I had the administrative job of Director. We began in the fall of 1953.

In a number of ways, George and I made a good team, for we had relatively complementary abilities. His charismatic intensity and infectious enthusiasm for new ideas drew followers to him and inspired his students. Like a sparkler, he constantly threw off suggestions for new investigations, many of them brilliant. My style was steadier; I supplied critical judgment, picked up the pieces, tried to make sure that things got finished, and smoothed the feathers he occasionally ruffled. Yet we found it difficult to work collaboratively on the same projects, turning out only a few joint publications. We remained good friends to the end, but both had to struggle against tendencies to be competitive and envious, I perhaps more than he. If he had happened to be the survivor, writing this retrospect, no doubt it would look somewhat different.

During its short lifetime, RCMH was located in three different buildings: first, in the Bible House, an old and dilapidated structure that was torn down in a few years. By then we had moved to a small factory building at 21 Washington Place, near the center of the university, along with the rest of the psychology department. The Research Center for Mental Health had the entire fifth floor, but soon it began to be too small. Finally, George and I got a research space construction grant from the National Institute of Health (NIH), which enabled us to reconstruct another former factory quarters in a nearby building into a fine new set of offices and laboratories. There was a gala opening celebration in March 1965. With our foot in the door, the rest of the department gradually acquired the rest of the 12 large floors of what is now known as the Psychology Building, at 6 Washington Place.


For about a decade, George Klein and I had good luck with our research grant requests. Luckily, the 1950s and 1960s were a time of huge expansion in the research budget of the National Institute of Mental Health (NIMH). Congress constantly pressed more money on the institutes of health than they asked for. In one dizzy moment of early success, our requests to both NIMH and the Ford Foundation for several years of program support were simultaneously granted. Since each gave us half of what we had asked for, at first we thought we could accept both; but we settled for Ford. When that was over, NIMH gave us a couple of multi-year program grants; and though we had one dry year when the University had to carry our operation with little outside help, we also got substantial grants from several other sources. When the NIH announced its program of Research Career Awards, which in effect created a research professorship for the recipient as long as he stayed at his university, we both applied and got them, in 1962.

It was just the right time in history to build up a center of research on the psychoanalytic theory of thinking, carrying on theoretical and empirical studies simultaneously. The prestige of psychoanalysis was great enough so that foundations and the Institutes of Health were eager to support attempts to make it more scientific. And with funding for a program of interlinked studies, we were able to build up a first-rate staff fairly rapidly. We took in several post-doctoral fellows from other universities -- Irving Paul, Fred Pine, Donald Spence, Paul Wachtel, and David Wolitzky, most of whom stayed for several years as staff. Since our main teaching assignments were to sponsor doctoral thesis research, we could pick out some of our most talented students and give them staff jobs: e.g., Morris Eagle, Leo Goldberger, and Nancy Goldberger. We could even make a place in this non-medical setting for psychiatric psychoanalysts: Walter S. Boernstein, Hartvig Dahl, Merton M. Gill, and Robert J. Langs (not all of them at once!). At one time or another, the staff included Judith Antrobus, Ilse Bry, Harry Fiss, Harriet Linton (later Barr), and Sydney J. Segal. On a number of occasions, we were joined for a year by distinguished foreign colleagues, such as Gudmund J.W. Smith from Sweden (twice), Pinchas Noy from Israel, Douglas Kirchner from Australia, and (for shorter visits) Gian Vittorio Caprara and Nino Dazzi from Italy and Bjorn Christiansen from Norway. Though I cannot mention all of the gifted and productive graduate students who did their dissertation research at the Research Center, many became professionally well known, e.g., Sheldon Bach, Phebe Cramer, Yael Danieli, Carol Eagle, Rita Eagle, Steven Ellman, David Fitzgibbons, Stanley Grand, Ricardo Hofer, Helene Kafka, John Kerr, Leah B. Lapidus, Paul Lippmann, Judith Rabkin, John Munder Ross, George Silberschatz, Doris K. Silverman, Hariette Weintraub Kaley, and Carl Zlatchin. Throughout, our work could not have been carried on without the able help of shop technicians, and a clerical staff headed by a paragon of competence, William T. Francis.

In the 1950s and 1960s, we experienced the hypomanic excitement of a new institution in its initial growth spurt. Morale was excellent; spirits often ran high at our regular staff meetings, where students and staff alike could present work in progress for critique or make suggestions about the research of others. At one time, we were affluent enough to hire Melanie Chussid to combine the disparate roles of librarian and cook. I hoped to invoke again the solidarity-making magic of breaking bread together at work, which I had known at the Harvard Psychological Clinic with Murray. For a few years, it worked: we had staff lunches daily, where the conversation was at times as good as I remembered it from my graduate student days. At some staff meetings, distinguished researchers, theoreticians, and other stimulating intellectuals addressed us. A few names come to mind: the philosopher Michael Polanyi, the artist Ad Reinhart, Peter Swales (whose presentation to our staff was his first public lecture), Abram Kardiner, Hy Witkin, Roy Schafer, David Rapaport, and Sheldon Korchin.

Aside from the fact that its staff steadily published work done at RCMH, beginning in the late 1950s, the center sponsored a few important publication outlets. Most notable of these was the monograph series Psychological Issues, which George Klein founded (at David Rapaport’s urging) and edited from its first issue in 1959 until his untimely death in 1971. Its remarkably competent managing editor, Suzette Annin, was an employee of the research center for about a decade. For most of its brief life, the Mental Health Book Review Index was published from RCMH by Ilse Bry, whose editorials in it virtually established a new field, sociobibliography (Afflerbach & Franck, 1977). And the annual, Psychoanalysis and Contemporary Science, was in a number of ways an outgrowth of work at RCMH; four of its original editorial board of 12 were staff members.

After about a decade, George and I decided to let our titles (Co-Director) reflect the actual state of affairs, and we took on my former student and collaborator Leo Goldberger as Associate Director. Beginning in 1971, after I had resigned and Klein had died, Goldberger became the Research Center's last Director.

George's sudden and wholly unexpected death in the spring of 1971 threw RCMH into a crisis. Coincidentally, Merton Gill and Hartvig Dahl had both decided, for different reasons, to leave at the end of the semester, while Paul Wachtel and one or two others of the junior staff received good job offers elsewhere. A year earlier, when it was time to submit an application to NIMH for another five-year program grant, I had resigned the co-directorship in the expectation of leaving New York (though actually I turned down opportunities elsewhere and stayed on). Now Leo Goldberger took over the directorship. When news of George Klein's death reached NIMH, that institute requested a re-application, and granted it but at a lower level of funding and for a shorter time. During the next couple of years, the staff -- now decimated of its senior members -- drifted away. Funding, once so easy, was now very difficult to find. The Research Center for Mental Health died a slow, anticlimactic, and hardly noticed death, which is therefore difficult to date: sometime in the early 1970s. 

The research program of RCMH

Subliminal influences on cognition

The Research Center began with a small grant from the NIMH, to pursue a lead George Klein had turned up at Harvard. Working there together, he and Gudmund Smith rediscovered metacontrast: the fact that when a short flash of a visual figure (the "subliminal stimulus") is immediately followed by a longer exposure of another (the "supraliminal stimulus"), the first one is not reportable though it can be seen if presented alone. The grant enabled us to hire a junior staff person, Harriet Linton (later Barr) and two graduate student research assistants: Susannah Gourevich and Donald P. Spence. Substantively, it launched our center on a program of research on the effects (on various kinds of cognition) of perceptual inputs that were below the threshold of conscious recognition. We hoped in this way to develop an experimental method of studying unconscious cognitive functioning -- which we thought implied the primary process -- in statu nascendi.

Our first study had a complex design and somewhat puzzling results, which it took several years to analyze and publish (Klein, Spence, Holt, & Gourevich, 1958). We were looking for direct and symbolically transformed effects of crude drawings of male or female genitals, presented subliminally, on judgments about the supraliminally shown drawing of an ambiguous or androgynous figure. Some subjects produced the expected data, but others gave paradoxical, contrast effects, apparently as a result of defensive processes. (Parenthetically, I find it ironically amusing that a later paper [Krosnick, Betz, Jussim, & Lynn, 1992] reports similar findings but without any reference to this or any other paper from the Research Center for Mental Health.) That was followed by a considerable series of other related studies, using a variety of techniques to input various kinds of information without the subject's focal awareness, and various dependent variables.

Shortly after coming to New York, George and I became friends with Charles Fisher, the psychoanalyst who did so much to start the wave of interest in subliminal phenomena by replicating the Poetzl study (see Fisher, 1960). Our staff too used this technique, of giving a person such a brief glimpse of a pictur e or verbal message that he or she could report nothing more than a flash with no content. Subsequently, dreams were collected from the subject, sometimes in our sleep laboratory, and were quantitatively scored for evidence of the subliminal input. Or, subjects were asked to produce visual images and draw them; these were scored, using various controls.

Proceeding from the general hypothesis that the critical variable in subliminal effects was the lack of focal attention to the input, various members of our staff developed a series of other techniques for achieving input without focal attention. They included distraction, making the critical stimulus part of the ground rather than the figure, retinal rivalry, dichotic listening, the color-word interference effect, incidental learning, and prolonged attentional fixation. Pine (1960) made ingenious use of the fact that rooms in our quarters at 21 Washington Place were poorly sound-proofed by having an accomplice play an audiotape loudly in an adjoining room. The experimenter would sigh, apologize to his subject, and ask the latter to ignore the intruding noise (the incidental input), while proceeding to tell stories about TAT cards. Pine obtained the usual kinds of transformative effects.

For about a decade, research on subliminal influences on thought was a prominent part of the RCMH program. We were not the first in this field, but the dozens of publications that came out of our laboratory did a good deal to help make it a focal issue in the cognitive psychology of the time. We learned a good deal about the technical parameters of the effects, but made our most distinctive contributions in exploring the influence of states of consciousness, defenses, abilities, cognitive styles, motives, and other aspects of personality on these elusive phenomena (see, for examples, Eagle, 1962; Klein, 1970; Pine, 1961; Spence & Ehrenberg, 1964, Wolitzky & Wachtel, 1973).

Yet, by the late 1960s, work in this area had virtually ceased, mostly because we became interested in other problems. We were probably influenced also by the rise of the signal detection movement. I remain convinced that the "improved" threshold procedures that were introduced missed the point. We had not been particularly concerned to show that the subject had no conscious awareness of any aspect of the stimulus input, by whatever means, only that he or she could not report its content or major meaning. Moreover, there were striking similarities in the types of effects (e.g., symbolic transformation of the unnoticed input) whether the critical stimulus was subliminal or incidental; while focal presentation of the same picture, text, or sounds produced qualitatively different results. Nevertheless, much of the work that had been done was misleadingly discredited by the argument that the inputs were not totally subliminal.

At least brief mention should be made of the remarkable program of research carried on by Lloyd Silverman (1982, 1983) during these years. Although he set up his own laboratory at a nearby hospital of the Veterans Administration, he needed office space, so we worked out a deal in the late 1960's: he had an office at RCMH and use of lab space in return for supervising the subliminal research of some of our graduate students, but without an appointment. In the fall of 1976, he joined the faculty of the same clinical program in which we did our teaching and drew on the same pool of graduate students for his research assistants, but maintained his autonomy and originality. Therefore, though we remained in close touch with his work and he with ours, his striking findings with his subliminal psychodynamic activation technique (which confirmed so many propositions in the clinical theory of psychoanalysis) cannot be claimed among the Research Center's achievements.

Measuring primary process thinking

Before leaving Topeka, I had begun work on a method of scoring Rorschach test responses for manifestations of primary process thinking. In the new research center, one major task was to develop this technique into a usable instrument for measuring, not only the primary process, but also a person's means of controlling and defending against it. I scoured Freud's writings for hints about concrete operational indicators of the primary process as well as studying his theoretical remarks about it and the secondary process, and looked for ways to concretize them, usually with the aid of a body of experimental data. The result was a scoring manual, which went through 10 mimeographed editions, which I made available to any interested researcher at cost. In 1983, it was published by Boria in an Italian translation (Holt, 1983; see also Holt, 1977). Only now (Holt, 2005) has the final revision been published, together with a critique and restatement of the psychoanalytic theory of thinking, and a review of the large body of research generated.

Much of my own work and that of my doctoral students during the first 15 years of RCMH was devoted to developing the scoring manual and applying it to a broader range of data, first to free verbal texts, with Leo Goldberger (1961), who also contributed the first method of distinguishing adaptive from maladaptive regression—that is, well-controlled primary process usable for adaptive purposes vs. poorly controlled manifestations constituting thought disorder. Then in her doctoral dissertation, Carol Eagle (1964) modified and extended the method so that it could be used to score dreams and TAT stories.

I will omit a description of the method here. It is complex and not easily conveyed in a short time. Suffice it to say that I was able to train a good many graduate students in the scoring,

which reached respectable levels of rater reliability, so that it was available as a dependent measure in a number of our substantive research endeavors. Word spread about this operationalization of a central Freudian concept, and work using it got under way in a good many other countries. The two main centers of research using my primary-process manual were in Montreal and in Rome, thanks to the efforts of Germain Lavoie (Canada) and Gian Vittorio Caprara and Nino Dazzi (Italy), who were responsible for French and Italian translations of the scoring manual. (See, for example, Caprara et al., 1977; Lavoie, &&) Caprara and I also collaborated on a developmental study of primary process manifestations in Italian children's Rorschachs (Caprara, Holt, Pazielli, & Giannotti, 1986).


Thanks to the work of Kris (1952), the alleged role of primary process thinking in artistic creativity was well known when the work at RCMH began. Hence, a good deal of work on this topic was undertaken almost from the Research Center's beginning. One of the main reasons for interest in the primary process scoring scheme was that its measure of adaptive vs. maladaptive regression yielded frequently replicated correlations with creativity and resistance to stress. For unknown reasons, adaptive regression is a positive predictor of creativity, but almost exclusively in male subjects. In most of our work on this topic, we used available college students and the creativity tests of Guilford (1967) or others similar to them (see, for example. Pine & Holt, 1960).

Studies of cognitive styles

Another aspect of the RCMH research program which began in Topeka and continued until the center's end can also be looked on as focused on developing instruments for measuring aspects of cognition. When he came to NYU, George Klein (1970) and his students (Gardner, Holzman, Klein, Linton, and Spence, 1959) had already devised a group of techniques for measuring cognitive styles such as leveling vs. sharpening, tolerance for unrealistic experiences, equivalence range, focusing (Schlesinger, 1954), and constricted vs. flexible control. These characteristic modes of acquiring and processing information were stable and were related to types of defenses and abilities. We thought of them as personal-structural parameters of many traditional phenomena and effects in cognitive psychology. For example, there was every reason to suppose that people of differing cognitive styles would differ in their tendency to produce either adaptive or maladaptive primary process in our various experiments. So we routinely assessed subjects' cognitive styles in most of our research, in addition to a good deal of dissertation work that explored the relevance of these variables to many aspects of personality and behavior. We were already using the interactive designs that Cronbach (1957) called for when he published his important paper on the two traditions of scientific research in psychology, experimental and psychometric.

Perceptual isolation (sensory deprivation)

In the early days of the Research Center for Mental Health, one of the most exciting new developments in psychology was the work done independently by John Lilly in Bethesda and by Donald Hebb and his group at McGill on what became known as sensory deprivation. It seemed as though remarkable, regressive effects on cognitive processes could be achieved by keeping subjects in as totally bland and undifferentiated an environment as possible. When a graduate student who had transferred from McGill, Leo Goldberger, approached me about sponsoring dissertation research in this area, I enthusiastically agreed. Consistent with our overall conception of the Center's work as an exploration of the psychoanalytic theory of thinking, we thought of this special type of experimental situation as an interference with reality contact, and reasoned that the strange phenomena reported by prior researchers could best be understood as intrusions of primary process thinking. It was a short step to the hypothesis that the nature of a person's response to being deprived of ordinary reality's support for secondary process (rational, realistic) thinking could be predicted by his handling of primary process material while responding to Rorschach inkblots. Leo Goldberger predicted that those who were comfortable with wish-fulfilling or autistic modes of thinking in the Rorschach situation would withstand a period of perceptual isolation (as he preferred to call it) better than those whose primary-process responses were distorted, and were accompanied by unpleasant affect or pathogenic defenses, all of which was summarized by negative ratings of Defensive Effectiveness. The results supported all hypotheses.

In Goldberger's dissertation study, the first of several on which we collaborated, subjects lay on a couch in a sound-proof room with halved ping-pong balls fastened over their open eyes for 8 hours (or less, if they asked to be released—as a minority did). To our surprise, the changes in cognitive processes expressed in their tape-recorded verbalizations during this ordeal by insipidity only partly seemed well-characterized as intrusions of primary process. Instead, much of the thinking was free of wishful content or the classical distortions of the dream work; yet it was fuzzy, or full of errors, or loosely wandering, or the like. So we found that we could reliably rate (in addition to good secondary process thinking, adaptive, and maladaptive primary process) what we called regressed secondary process (Goldberger, 1961; Goldberger & Holt, 1958).

During the next few years as a post-doctoral fellow, Leo Goldberger followed up various leads in his thesis study, mostly supported by Air Force money. He and I replicated the first study using quite a different sample of subjects—no longer college students but unemployed members of Actors' Equity--with many of the same results but with some interesting differences, somewhat too complicated to go into here in full detail. In both samples, however, we found two contrasting patterns of reaction, which we called adaptive and maladaptive. Though they were related to different personality traits in the two samples, in both samples indicators of ego-strength were positively correlated with the adaptive pattern, and measures of hypochondriasis were negatively correlated with that pattern of response (Goldberger & Holt, 1961a). One measure that did not hold up was the Rorschach measure of adaptive regression, which had worked well in the original sample but trended in the opposite direction with the actors. It did, however, prove a sturdy predictor in several studies of sensory deprivation conducted outside our own laboratory (Wright & Abbey, 1965; Zuckerman, Persky, Hopkins, et al., 1966; Zubek, 1969).


For several years, a substantial part of the RCMH's program was devoted to research on dreams. Our new laboratory contained two sound-proof rooms specifically designed for this work, with beds, self-contained bathrooms, and built-in wiring to convey many channels of data to an external electroencephalographic machine. The work is extraordinarily time-consuming, requiring the experimenter to stay up an entire night while his/her subject sleeps, being awakened at various times to report mental content. Consequently, we did not complete a large number of studies, though George Klein made this one of his own main interests for severa years, and Harry Fiss worked extensively on dreaming (see, e.g., Fiss, Klein, & Bokert, 1966).

Despite their reputation for bizarreness, most manifest dream texts are disappointingly prosaic. Even dreams from the night following LSD were far from pervaded with signs of the primary process at work. We did, however, learn something about how the content of dreams can be influenced, how the control of primary process is related to color in dreams, the effects of being awakened during periods of rapid eye movements (when most dreaming occurs) and at other times, and a few other such topics, though we found no royal road to unconscious cognition.

Effects of a psychedelic drug on cognition

In the earliest days of our collaboration, George and I worked together on a pilot study of mescaline. My analyst vetoed my taking any, so George was one trial subject and the novelist Waldo Frank a second. The results were a bit disappointing; we decided to shift to a newer drug about which a small literature was already accumulating: lysergic acid diethylamine (LSD). We obtained a supply of 100 microgram doses from the manufacturer, and a couple of junior staff members tried it out, with promising results. So we got a grant and set up a rather elaborate study, one that involved every member of what was by then a rather large staff of the research center. Since the drug's effects lasted for at least a full working day, it was possible for many simultaneous experiments and exploratory studies to go on. We gathered data for the entire academic year 1958-59, which happened to be shortly before the general public learned about LSD and before the era of its street use. Only a couple of years later, our study would have been impossible. 

For subjects, we had to use mature adults instead of the usual college students, and George hit upon a ready source of under-employed people: the theatrical union, Actors' Equity. We screened them carefully, rejecting any who seemed as if they might be too upset by the experience, and obtained their informed consent. Accepted subjects first underwent a day of personality assessment and pre-testing. On the experimental day they were randomly assigned experimental or control status by a staff member who administered the drug or placebo and otherwise had no contact; the rest of us were blind. Testing began early in the morning, and continued all day, ending with a night spent in the dream laboratory of Donald Goodenough and David Schapiro at the Downstate Medical Center in Brooklyn. There the subject was festooned with scalp electrodes and awakened by Carol Eagle or Helene Kafka (1963) when there was EEG evidence of dreaming so that they could get the data for their dissertation researches. Finally, subjects came back to the lab for retrospective discussions and post-testing.

Sixteen of the 20 LSD subjects were later recruited for a day's stay in the perceptual isolation room, for the above-mentioned replication of Leo Goldberger's first experiment. That made it possible for us to make direct comparisons between LSD and isolation for their potency in disrupting secondary process thought. LSD won hands down, and there was very little similarity in the effects of the two conditions (Goldberger & Holt, 1961b).

The two or three years during which we prepared for the LSD work, gathered data, and completed the personality assessments were the high points of the Research Center for Mental Health. For once, everyone was working together on interlinked projects, using the same subjects and studying fascinating phenomena. Every staff member (including graduate students) was assigned as biographer for one of the 30 subjects, being in principal charge of assembling, analyzing, and integrating the various available materials for personality assessment. Every subject wrote a short autobiography, was interviewed several times, took Rorschach, TAT, WAIS, and various other projective tests and structured inventories. Following the model of the Harvard Psychological Clinic's procedures, we agreed on a set of personality variables to be rated in a series of Q-sorts (drawing on the work of Jack Block, 1961), and then went over the distributions of scores across subjects and made final consensual judgments. The result was a rich body of test scores and rated traits; many of us made good use of them in exploring the personological correlates of variables that interested us (see for example Holt, 1966).

There is an odd contrast between the sense of having had a great experience of joint work on a problem of great practical as well as theoretical interest, and the meager output of publications that resulted. For a few years, we all worked intensively on the analysis of the huge mass of data; several dissertations were completed and 10 papers were published on the findings, but it took a long time to get out a slim book on the overall results (Barr, Langs, Holt, Goldberger, & Klein, 1972). By then, LSD was old hat, and doses of 100 gamma were considered small.

We were becoming aware that the concept of primary process could be applied to almost any kind of primitive, pathological, or regressive thought, but that doing so did not help in distinguishing among the varieties.


One thing that virtually all of our attempts to study the primary process had in common was a concern with mental imagery. That prompted me to choose, as the topic of my presidential address to the APA Division of Clinical Psychology, a historical survey of the literature on imagery with some comments about its then-current renaissance (Holt, 1964).

One of the things I did with the LSD subjects was to administer a test of deliberate imaging, with the interesting finding that the drug did not help: the imaging process under LSD, though sometimes florid, is out of the subject's conscious control. In another study, I found that when subjects were interviewed about their experience of many types of imagery and the results were intercorrelated, there was very little generality. Knowing that a person had had eidetic imagery as a child, for example, said nothing about how vividly he experienced pressure phosphenes or how well he could summon up a picture of a car and make it move down an imaginary street (Holt, 1972, 2005).

For several years, Sydney Segal (1972) pursued some interesting variations on the classical experiment of Perky, in which she showed that subjects could not always distinguish between a mental image and a faintly projected picture. The cognitive style of the subject turned out to be an important parameter of this effect, as well as the subject's motivational state and state of consciousness.

Clinical research

Under this heading I group several kinds of investigations. At various times, our clinical interests and training led us into research on empathy, clinical sensitivity, and clinical listening. Our intense focus on psychoanalysis led to research on countertransference, on interpretation, and on various other aspects of the psychoanalytic process. Merton Gill experimentally studied the effect of deliberately inaccurate interpretations. In one of our sound-proof rooms, Hartvig Dahl (1972) conducted the first complete and successful psychoanalysis to be tape-recorded, and began collecting a library of recorded (partial) analyses. While he was at RCMH, he began his current program of quantitative research, in which he has shown how fruitfully statistical techniques like factor analysis and computer content analysis can be applied to recorded psychoanalyses. George Silberschatz (1986) used some of Dahl's data in his pioneering research on the psychoanalytic process.

Psychoanalytic theory

George Klein and I had always felt that Freud's core ideas have tremendous value. Yet his theories contain so much that is vulnerable even to relatively uninformed critics that we agreed that the best way to preserve what is lastingly valuable would be to purge the theory of its errors and fallacies.

We began with the psychoanalytic theory of thinking, studying the relevant texts together in a staff seminar led by Benjamin B. Rubinstein as well as separately. David Rapaport (1951, 1960) had provided us with a model and a method, though I doubt that he would be happy with all that we did in attempting to emulate him. At first I focused on finding operational statements that might be translated into scoring rules for my Rorschach scoring system. But the more we read the volumes of the Standard Edition as they were published, the more troubled we became over contradictions, confusions, factual statements that have been proved untrue, and methodological errors of various kinds. It began to be plain to us that first, the doctrine of psychic energy, and then all of metapsychology, was fatally flawed. At first, we hoped to remove its errors, revise, and revitalize it, but soon realized that more radical steps were necessary. George Klein and I did not agree on what those should be; but I think it is fair to say that the theoretical work done at the Research Center for Mental Health contributed considerably to the present near-consensus that metapsychology is dead, and that psychoanalysis is in a theoretical crisis. Most of that work is reported in two books (Klein, 1976; Holt, 1989; see also Gill & Holzman, 1976). It would be much more agreeable to be able to report that we produced a viable alternative to Freud's metapsychology, but the fact is that only the negative task was essentially completed. (For a beginning on the task of recasting psychoanalytic theory in testable form, see Holt, 2005, Vol. 1.)


Clearly, it is difficult for me to evaluate the RCMH, having been so much a part of it for so many years. Nevertheless, let me try, by asking first what progress we made on the objectives we originally set for ourselves. Plainly, we vastly underestimated the theoretical task, which should logically have preceded empirical work. Yet if we had attempted to get the theory into a logically tight and coherent form, with key propositions stated in terms that lent themselves to test against specifiable data, we might never have done much research. Our failure to design experiments that were in any way crucial for the theory is not so much our fault as that of the theory: No one else had ever done so, either. As Grünbaum (1984) has convincingly argued, many of Freud's propositions are indeed testable, though we did not find any in his theory of thinking that we attempted to verify directly. Instead, we retreated to a defensible position which, however, did not enable us to refute anything: we used psychoanalytic theory as a general source of inspiration and a guide in choosing topics and methods. (See, however, Holt, 2005.)

When it comes to assaying the importance of our research output in its own right, so to speak, or without regard to its usefulness in testing theory, I must really defer to later historians. We published a lot; when I complete a bibliography of research done at RCMH, I am confident that it will run to a few hundred items. Yet, as Paul Meehl (1978) has sagely commented, it is very difficult to discern much progress in any part of the softer branches of psychology, of which psychoanalytic research is surely one. People work for awhile on a set of problems, and research reports pile up, but who can say whether or not they cumulate in the way it is supposed to happen in science? Researchers turn to other topics, not so much because they have definitively answered their original questions or exhausted what is interesting about a set of phenomena, but because the grass begins to look greener in other pastures. So it was with us. Meehl (1978) adduces a large number of basic reasons for this discouraging state of affairs; indeed, it may never be drastically improved. 

Therefore, I don't feel bad about our lack of brilliant breakthroughs or truly crucial experiments. Not many research institutions in soft psychology or psychiatry had them either; indeed, I believe that any search for crucial experiments in most of psychology is misguided (Holt, 2003, 2004).

In the history of psychoanalysis, it may eventually be seen as important that a sustained attempt was made to test a critical part of the theory. I do not think that the theoretical work we did would have been possible if we had not been trying simultaneously to find ways of verifying or refuting what psychoanalysts say is the case. It is surely too early to evaluate my recent reformulation of the psychoanalytic theory of thinking, which grew directly out of work done at RCMH. Nevertheless, the efforts we made there were definitely worth while, a rich learning experience for many of us.

In the process, we created a training ground for a generation of researchers in psychoanalysis and clinical psychology. That may turn out to have been the most important contribution of the Research Center for Mental Health at New York University.


Afflerbach, L. & Franck, M. (Eds.) (1977). The emerging field of sociobibliography: The collected essays of Ilse Bry. Westport, CT: Greenwood Press.

Barr, H. B., Langs, R. J., Holt, R. R., Goldberger, L., & Klein, G. S. (1972). LSD: Personality and experience. New York: Wiley.

Block, J. (1961). The Q-sort method in personality assessment and psychiatric research. Springfield, IL: Charles C Thomas.

Caprara, G. V. e altri (1977). Studi sulla personalita; Quattro ricerche con il metodo di R. Holt [Studies in personality: Four researches with the method of R. Holt]. Rome: Boringhieri.

Caprara, G.V., Holt, R.R., Pazielli, M.F., & Giannotti, A. (1986). The development of primary process in children's Rorschachs. Journal of Personality Assessment, 50, 149-170.

Chein, I. et al. (1964). The road to H. New York: Basic Books.

Dahl, H. (1972). A quantitative study of a psychoanalysis. In R. R. Holt & E. Peterfreund (Eds.), Psychoanalysis and contemporary science. Vol. 1 (pp. 237-257). New York: Macmillan.

Eagle, C. J. (1964). An investigation of individual consistencies in the manifestations of primary processes. Unpublished doctoral dissertation, New York University.

Eagle, M. (1962). Personality correlates of sensitivity to subliminal stimulation. Journal of Nervous and Mental Disease, 134, 1-17.

Fisher, C. (1960). Introduction to Preconscious Stimulation in Dreams, Associations, and Images, by O. Potzl, R. Allers and J. Teler. Psychological Issues, 2(3), Monograph 7.

Fiss, H., Klein, G. S., & Bokert, E. (1966). Waking fantasies following interruption of two types of sleep. AMA Archives of General Psychiatry, 14, 543-551.

Gardner, R., Holzman, P. S., Klein, G. S., Linton, H., & Spence, D. P. (1959). Psychological Issues, Monograph 4. Cognitive control: A study of individual consistencies in cognitive behavior. New York: International Univ. Press.

Gill, M. M., & Holzman, P. S. (Eds.) (1976). Psychology versus Metapsychology: Psychoanalytic essays in memory of George S. Klein. Psychological Issues, 9(4), Monograph 36.

Goldberger, L. (1961). Reactions to perceptual isolation and Rorschach manifestations of the primary process. Journal of Protective Techniques, 25, 287-302.

Goldberger, L. & Holt, R. R. (1958). Experimental interference with reality contact (perceptual isolation): Method and group results. Journal of Nervous and Mental Disease, 127, 99-112.

Goldberger, L., & Holt, R. R. (1961a). A comparison of isolation effects and their personality correlates in two divergent samples. USAF ASD Technical Reports, 1961, No. 61-417, 46 pp.

Goldberger, L., & Holt, R. R. (1961b). Studies on the effects of perceptual alteration. USAF ASD Technical Reports, 1961, No. 61-416, 20 pp.

Grünbaum, A. (1984). The foundations of psychoanalysis: A philosophical critique. Berkeley: University of California Press.

Guilford, J. P. (1967). The nature of human intelligence. New York: McGraw-Hill.

Holt, R. R. (1964). Imagery; The return of the ostracized. American Psychologist, 19, 254-264.

Holt, R. R. (1966). Measuring libidinal and aggressive motives and their controls by means of the Rorschach test. In D. Levine (Ed.), Nebraska symposium on motivation, 1966 (pp. 1-47). Lincoln: University of Nebraska Press.

Holt, R. R. (1972). On the nature and generality of mental imagery. In P. E. Sheehan (Ed.), The function and nature of imagery (pp. 6-33). New York: Academic Press.

Holt, R. R. (1977). A method for assessing primary and secondary process in the Rorschach. In M. A. Rickers-Ovsiankina (Ed.), Rorschach Psychology (rev. ed.) (pp. 375-420). Melbourne, FL: Krieger.

Holt, R. R. (1983). Il processo primario nel Rorschach e nel materiale tematico. [The primary process in the Rorschach and in thematic data.] Rome: Borla.

Holt, R. R. (1989g). Freud reappraised: A fresh look at psychoanalytic theory. New York: Guilford.

Holt, R. R. (2003). New directions for basic psychoanalytic research; implications from the work of Benjamin Rubinstein. Psychoanalytic Psychology, 20, 195-213.

Holt, R. R. (2004). A few dissents from a magnificent piece of work. [Commentary] Applied & Preventive Psychology, 11, 43-44.

Holt, R. R., with Mayman, M. and Eagle, C. J. (2005). Primary process thinking: Theory, measurement, and research (2 vols.) Psychological Issues, Monograph No. 65-66.

Kafka, H. (1967). The use of color in projective tests and dreams in relation to the theory of ego autonomy. Unpublished doctoral dissertation, New York University.

Klein, G. S. (1970). Perception, motives, and personality. New York: Knopf.

Klein, G. S. (1976). Psychoanalytic theory: An exploration of essentials. New York: International Universities Press.

Klein, G.S., Spence, D.P., Holt, R.R., & Gourevitch, S. (1958). Cognition without awareness: Subliminal influences upon conscious thought. Journal of Abnormal and Social Psychology, 57, 255-266. 

Kris, E. (1952). Psychoanalytic explorations in art. New York: International Univ. Press.

Krosnick, J. A., Betz, A. L., & Jussim, L. J. (1992). Subliminal conditioning of attitudes. Personality & Social Psychology Bulletin, 18, 152-162.

Meehl, P. E. (1978). Theoretical risks and tabular asterisks: Sir Karl, Sir Ronald, and the slow progress of soft psychology. Journal of Consulting and Clinical Psychology, 40, 806-834. (Reprinted, with commentaries, in Applied & Preventive Psychology, 2004, 11.)

Pine, F. (1960). Incidental stimulation: A study of preconscious transformations. Journal of Abnormal and Social Psychology, 60, 68-75.

Pine, F. (1961). Incidental versus focal presentation of drive related stimuli. Journal of Abnormal and Social Psychology, 62, 482-490.

Rapaport, D. (ed.) (1951a). Organization and pathology of thought. New York: Columbia Univ. Press.

Rapaport, D. (l960). The structure of psychoanalytic theory: A systematizing attempt. In S. Koch (ed.) Psychology: A study of a science. Study 1. Conceptual and Systematic. Vol. 2. Formulations of the person and the social context (pp. 55-183). New York: McGraw-Hill, 1959. Also in Psychological Issues, 1960, 2 (2), Monograph No. 6.

Schlesinger, H. J. (1954). Cognitive attitudes in relation to susceptibility to interference. Journal of Personality, 22, 354-374.

Segal, S. J. (1972). Assimilation of a stimulus in the construction of an image: The Perky effect revisited. In P. W. Sheehan (ed.), The function and nature of imagery (pp. 203-230).

Spence, D. P. (1964). Conscious and preconscious influences on recall: another example of the restricting effect of awareness. Journal of Abnormal & Social Psychology, 68, 92-99.

Spence, D. P. & Ehrenberg, B. (1964). Effects of oral deprivation on responses to subliminal and supraliminal verbal food stimuli. Journal of Abnormal & Social Psychology, 69, 10-18.

Wolitzky, D. L. & Wachtel, P. L. (1973). Personality and perception. In Benjamin B. Wolman (Ed.), Handbook of general psychology. Oxford, England: Prentice Hall.

Wright, N. A. & Abbey, D. S. (1965). Perceptual deprivation tolerance and adequacy of defenses. Perceptual and Motor Skills, 20, 35-38.

Wright, N. A & Zubek, J. P. (1969). Relationship between perceptual deprivation tolerance and adequacy of defenses as measured by the Rorschach. Journal of Abnormal Psychology, 74, 615-617.

Robert R. Holt, Ph.D.
Box 1087
Truro, MA 02666
Tel. 508-3492120
E-Mail <>

Web Editor: Paolo Migone. For suggestions or corrections, please e-mail to: