Albert Ellis,rational therapy and the media of ‘modern’ emotional management

Introduction

Cognitive Behavioral Therapy (CBT) is today one of the most common non-pharmaceutical courses of psychological treatment available for anxiety, depression and a number of other common mental health complaints. CBT has largely replaced traditional psychoanalytic techniques as a standard form of therapy in the Anglo-American world: all psychodynamic therapies ask patients to actively renarrate the relationship between thinking, feeling and acting, but CBT is often structured as a course of treatment that makes this renarration finite in length, and focused on particular troublesome behavior or other symptoms of mental distress. The National Alliance on Mental Illness notes that CBT is ‘problem-focused, and goal-directed’, and observes that ‘because CBT is an active intervention, one can also expect to do homework or practice outside of sessions’ (NAMI, 2012). Spawning an apparatus of professional techniques and practices, cognitive behavioral therapy has thus become a staple therapy for individual control and management of difficulties with personality.

One of the key historical antecedents of CBT is the 1962 book Reason and Emotion in Psychotherapy, by New York psychologist Dr Albert Ellis. Ellis developed his theory of what he termed rational therapy (RT), later variously renamed to ‘rational emotive therapy’ and ‘rational emotive behavioral therapy’, between the late 1940s and early 1960s. ¹ Ellis’s ongoing work on and promotion of RT were instrumental in supporting and shaping the later efforts of cognitive psychologists such as the University of Pennsylvania’s Aaron T. Beck to define and consolidate what is now known as CBT – and Beck’s work is the more famous today (Rosner, 2014). Yet rational therapy, which Ellis noted in later published accounts had begun as a response to the apparent failures of traditional psychotherapeutic techniques to help his patients regain control over their neuroses, was by the late 1960s and 1970s its own fully fledged clinical and popular therapeutic system distinct from what would become known as CBT, and one which earned Ellis international popular and professional prominence.

Despite Ellis’s diminished public reputation since his death in 2007 at the age of 94, he remains a key figure in the history of postwar American psychology. Understanding the development of Albert Ellis’s theory of rational therapy within the broader history of the postwar spread of individual psychodynamic therapy in the United States reveals that RT exemplified several interrelated shifts in professional practices, social structures, scientific norms and use of media technologies that were critical to both the popularization and instrumentalization of therapeutic self-help. Together, these shifts made rational therapy not only successful, but also indexical of a technical, material and social redefinition of the emotional self in the eyes both of psychologists and of their patients. This self was construed as legible to both groups through a ‘clarified’ rational language which appealed to seemingly common-sense definitions of everyday behavior.

As Nikolas Rose notes of this postwar period, ‘psychological expertise played an active role in the formation of new images of human subjectivity…linking private realms into calculations concerning the wellbeing of the nation’ (Rose, 1996: 148–9). Rose’s use of the term ‘calculation’ is apt: rational therapy was the foremost of a wave of psychological practices in post war America that sought to unite approaches to clinical therapy that understood the human mind as a discrete motivator of conscious and unconscious behavior together with the dominant model in experimental psychology of the day, namely behaviorism – according to which human behavior could be understood, and potentially calculated in advance, without resorting to an understanding of the mind or mental activity. Ellis, like other psychologists of the period, sought to combine these two strains of thought in a clinical practice that would help patients put their disordered thoughts and feelings into a rational order – but Ellis’s familiarity and ease with the mass media propelled RT to a prominence that was unparalleled, and which is instructive for understanding the contemporary socio-technical effects of the digital media increasingly used for therapy today.

Original research in the Ellis archive and a review of Ellis’s correspondence and other papers make it clear that Ellis was both intensely concerned with obtaining scientific legitimacy for his new therapy, but was also distrustful of existing techniques for testing and quantifying mental disorders. After presenting an overview of Ellis’s biography and the context for his development of rational therapy, I argue that the form of the therapy itself was particularly suited to trends in postwar American clinical psychology. Though Ellis’s famously powerful, even aggressive, personality and ambition had some role in rational therapy’s success, Ellis’s focus on the capacities of a standardized language to shape and manipulate the practices and outcomes of patient self-narration was ideally suited to a popular media landscape defined by a so-called ‘middle brow culture’ of self-education and self-improvement. In order to synthesize cognitive and behavioral models of the disordered self, Ellis’s RT developed a common-sense language of rationality, a solution to the problem of defining the emotional self without resorting to numbers that was resonant in its postwar middle-class milieu. Psychology has long relied on a common language to help define its coherence as a discipline and its epistemological force as a therapeutic intervention: as Kurt Danziger observes, ‘there are profound affinities between [psychological] language and method’ (Danziger, 1997: 158). Moreover, the history of psychology’s popular dissemination had long been (and continues to be) tied to the publication of ‘self-help’ guides and descriptions of systems that ordinary Americans could follow for the purpose of self-improvement. Ellis brought together these two strains of therapeutic culture into one package: for Ellis, the language of RT as a structuring and simplifying method of therapy was transformative because it brought both a systematic order to self-narration in therapeutic practice, and a more legible way of exerting psychological expertise through popular media mechanisms.

As such, rational therapy’s therapeutic techniques cannot be separated from the media through which they were presented. Rational therapy’s tenets were supported by Ellis’s use of media objects of the postwar period such as homework assignments, self-help texts and the apparatus of mass media publicity, that facilitated, and ultimately shaped, a widespread popular assumption that the self could be articulated through the embodied rigors of a more scientific language of treatment espoused both by the therapist and eventually the patient (Brandt, 2013). Ellis’s system of a clarified therapeutic language also resonated with the socio-technical changes of the 1950s: it mobilized a number of intellectual influences both to construct and remediate a model of the human patient as a rational elucidator of emotion, amenable to the modernist ethos of administration and expertise that characterized social, economic and technological discourses in postwar American life.

These two aspects of rational therapy – a rational ordering of clinical language and its dissemination via multiple media forms – resonated with broader social and political ideologies of the day in the United States and other western countries, and cemented its popularity and utility in the postwar period. Other scholars have examined the connections between psychology and the social, technological and cultural movements that made up what is now known as ‘modernism’ at the turn of the 20th century. Modernism, as Dorothy Ross defines it, is a set of discourses that revolved around the exploration of subjectivity and its relationship with a rapidly shifting material world (1994) – Freud, Jung, Adler and the other pioneers of psychoanalysis and psychology were caught up directly in this project (Burnham, 2012; Micale, 2004; Richardson, 2006). In a similar vein, Ben Singer suggests that scholars be attentive to the emergence of a ‘neurological conception of modernity’ in the late 19th century: the sense both from contemporary thinkers and through subsequent historical analysis that ‘modernity must…be understood in terms of a fundamentally different register of subjective experience’ (Singer, 1996: 72). Singer thus directly connects the socio-technical conditions of the exterior world to the interior subjective life of its inhabitants.

A broader implication I draw from Singer’s thesis is that the development of ‘neurological modernity’ was not confined to a particular point within the history of modernism. I argue that Ellis’s development of rational therapy was inextricably tied to what James C. Scott has termed a ‘high modernist ideology’. Scott describes this particular form of modernism as ‘a strong, one might even say muscle-bound sense version of the self-confidence about scientific and technical progress…and above all the rational design of social order commensurate with the scientific understanding of natural laws’ (Scott, 1998: 4). In architecture and urban planning, high modernism flourished in the work of architects and planners such as Le Corbusier, but I suggest that Ellis’s rational therapy sought to organize patients, and people in general, along lines of legibility and rational ordering that are consistent with Scott’s definition. Ellis’s novel brand of clinical therapeutic practice translated the high modernist impulse for rational order out of the world of the built exterior and into the interior sphere of the self – echoes of which can be seen in the digitally mediated therapeutic culture of the present.

Rational therapy and its socio-technical milieu

The Second World War and its aftermath were a heyday for psychology in the United States (Capshew, 1999; Hale, 1995; Herman, 1995): the former had destabilized the conditions under which ordinary Americans understood the normality of their emotional lives (Horwitz, 2013: 98). Psychologist and philosopher Rollo May gave an explicitly socio-technical reason for this new sense of uncertainty in his 1950 book The Meaning of Anxiety: ‘from…the birth of the bomb’, May observed, ‘anxiety shifted from a covert to an overt problem’. In May’s view, ordinary people were increasingly confounded not simply by the prospect of profound and potentially destructive technological change on a societal scale, but also by the more personal, individuated effects of new technologies and social mores born of the world conflict: ‘confusion, alienation, psychological disorientation, and uncertainty with respect to values and acceptable standards of conduct’ (May, 1996: 3-4). This iteration of ‘neurological modernity’ was an existentially unsettling one for ordinary people, even as the American government entered a period of cold war ‘high modernist’ consensus around scientific and technical solutions for social problems, and on the necessity of battling communism overseas (Orr, 2006).

Owing to psychologists’ perceived effectiveness during the Second World War both at treating soldiers with ‘war neuroses’ and aiding the United States’ propaganda efforts, psychologists prospered, along with psychiatrists and other experts in the realm of what prior to the war had been called ‘mental hygiene’ but which afterwards became known as ‘mental health’ (Bertolote, 2008: 113). As Nathan G. Hale Jr notes, therapeutic techniques, including psychoanalysis, reached new heights of popular and institutional prestige in the immediate postwar period, and psychologists sought to cement their status as a scientific profession on par with their psychiatrist cousins (Hale, 1995: 300). Some of these psychologists, including Albert Ellis, began to treat clients in private practice using psychotherapeutic and other mixed methods. This expansion came in part because the number of psychoanalysts in the postwar United States was increasingly dwarfed by popular interest and attention in mental health concerns and, consequently, by sheer demand for therapy: there were fewer than 1,000 psychoanalysts practicing in the United States in 1957 (ibid.: 289).

Albert Ellis’s own background was itself personally difficult and professionally eclectic, leading to a bedside manner that was sometimes more showman than academic but also providing Ellis with a knowledge of popular culture and salesmanship which would serve him well in his later promotion of RT. Ellis’s career as a psychologist was in itself made possible in part by the upheaval brought about by the war and the Great Depression before it. Born in 1913 to working-class parents who would divorce when he was 12, and subsequently raised in New York City, Ellis graduated from the City University of New York with a business degree in 1934. According to his own unpublished 1957 biographical account, Ellis spent the 1930s doing ‘some editorial work for a picture magazine’; he also ‘helped manage several business enterprises, and ingeniously managed to never work more than twenty hours a week at remunerative activity’. Ellis became fascinated during this period by what he termed ‘sexuo-amative research’ (Ellis, 1957a), the then risqué study of human sexual behavior popularized in the public’s mind by studies conducted by the University of Indiana’s Alfred Kinsey. Ellis’s interest in sexual and marriage counseling, which he claimed to have been called upon to perform for friends and acquaintances, prompted him to return to college for further training in psychology.

Ellis entered the Teachers College at Columbia University in 1942, and began private practice in sex and marriage counseling immediately after being awarded his Master’s degree in 1943; at the time, he was compiling material for a manuscript entitled ‘The Case for Sexual Liberty’ ([Albert Ellis Institute], 2014). For Ellis, as for Kinsey and many of the era’s ‘sexologists’, an expansive and liberal view of human sexuality was grounded in the empirical rigors of a scientific method that relied on numbers. Sarah Igo observes that the power of Kinsey’s famed reports on human sexuality came from ‘the figures Kinsey unleashed, [which] carried a great deal of weight because they were numbers: spare, clear, and direct’ (Igo, 2007: 247). Trained in both Rogerian and Freudian forms of analysis while at Columbia, Ellis was also influenced by the impacts of Kinsey’s work, and was eager to bring this methodological clarity to individual clinical therapy. By 1950 Ellis had set himself up in New York City in private practice as a counseling and marriage psychologist. ² Famously, even aggressively loquacious, Ellis preserved masses of his correspondence from the late 1940s onward, an archive which presents Ellis engaging in good-natured epistolary debates with a vast array of colleagues, patients and personal friends – many of which focused on Ellis’s dissatisfaction with the clinical practices of his chosen profession.

Even at the height of its popularity, some psychologists had begun to attack the clinical utility, and worse the fundamental scientific validity, of psychoanalysis. In 1952, the British behaviorist psychologist H. J. Eysenck launched a much-discussed attack on psychotherapeutic methods in the Journal of Consulting Psychology. Eysenck’s paper, which compared a number of extant studies on the remission rates of individuals in psychoanalysis with the ‘spontaneous recovery’ of neurotic patients in American mental hospitals, concluded that ‘roughly two-thirds of a group of neurotic patients will recover or improve to a marked extent within about two years of the onset of their illness’, and that, as a result, psychotherapy seemed to have no demonstrable effect whatsoever (Eysenck, 1952). The psychoanalytic community hotly contested Eysenck’s methodology and results, both because of flaws in the particular data Eysenck had used but also in terms of the experiment’s design. Proponents argued that while psychotherapy was undoubtedly a medical treatment, it required a scheme of evaluation fundamentally different from other medical procedures, which were subject to assessment through quantifiable and measurable numerical outcomes. Because of the individuality of each patient’s psyche, psychoanalysts argued, controlled studies of psychotherapeutic effectiveness needed particular care and sensitivity to be carried out, and even then might not capture the mix of factors (including patient attitudes, external social factors and the therapist’s skill) that made such treatments a success (Hale, 1995). Even as psychoanalytic techniques enjoyed popular vogue and respectability, voices like that of Eysenck advocated for not only a more eclectic mix of therapeutic techniques, but also the necessity of holding clinical psychoanalyses to the same set of statistical norms that were extant in laboratory-based experimental psychology (Erwin, 1980).

Having been led by his interest in sexuality into private practice, Ellis was describing his psychoanalytic method by 1957 as ‘violently anti-orthodox’ (Ellis, 1957a). Ellis was among the new, more unconventional clinical practitioners who shared Eysenck’s interest in holding therapeutic methods accountable to what he understood as scientific standards of rigor. Ellis was by no means the first American psychologist to seek to modify Freud’s theories to be more rigorous and purportedly scientific. In the 1930s, Harold Lasswell had developed machines that automatically recorded a patient’s physiological responses for later comparison with the clinician’s notes (Malin, 2014: 164), and Ellis’s contemporaries, such as neo-Freudians like Karen Horney and ego psychologists like David Rapaport, were also invested in creating new methodological languages for psychotherapy. However, Ellis’s exploitation of popular media forms, and his endless proselytizing for his new approach in various public settings, helped set rational therapy apart as a particularly accessible and public approach. According to the origin story for rational therapy presented in Reason and Emotion, Ellis had never fully subscribed to the psychodynamic theories of personality espoused by Freud, nor to the psychoanalytic techniques in which Ellis had been trained (Ellis, 1962: 3); his description of the genesis of rational therapy in Reason and Emotion chronicles a succession of methodological disillusionments, first with classical Freudian analysis, and subsequently with the welter of alternative techniques popular in individual psychotherapy at the time. Perhaps most conclusively, Ellis wrote that psychotherapy methodologically unsound: he averred that the prescribed and regimented techniques of Freud – including its limitations on the therapist’s ability to intervene in the patient’s process, and the strict physical positioning of analyst and patient in the analytic milieu – constraining and irrational. Psychoanalysis, Ellis declared, was significantly helpful to his patients in only about 50 to 60 per cent of cases (ibid.: 5).

Ellis’s letters to his friends and colleagues reveal how the evolution of rational therapy was shaped by his sexual liberalism, his enthusiasm for what he considered scientific rigor, and his training in a postwar ‘mental hygiene’ tradition that had dissolved ‘conventional distinctions between positive mental health and social welfare’ (Herman, 1995: 25). In a letter to Dr Meyer Solomon in November of 1953, 10 years before the publication of Reason and Emotion, Ellis was already advocating for a more ‘scientific psychoanalysis’, bemoaning the fact that the term ‘analysis’ was by that time so widely used in the United States to describe Freudian methods that any attempt to differentiate between different therapeutic techniques in the popular press was useless. A more ‘scientific’ psychoanalysis, according to Ellis, ‘would keep all the indubitably valid points of the Freudian and other schemes, and throw away the trash’ (Ellis, 1953b). The bulk of Ellis’s clinical work in the early 1950s, to a much greater degree than he notes in Reason and Emotion in Psychotherapy, was focused on sexology and marriage counseling, including analyses of the depictions of human sexuality in popular culture. Ellis’s attitude towards the utility of psychoanalysis as a therapeutic technique was thus in part shaped by the specific challenges posed by his métier, including the fact that frank discussions about sexuality were not the public norm at the time, and that anxiety and neurosis about sex that Ellis identified in American society through its cultural products were widespread. Observing a wide range of patient anxieties around sex, relationships and intimacy, Ellis increasingly classified these concerns as being less than pathological. Indeed, Ellis questioned whether many of his patients were in fact mentally ill at all. ‘How seriously disturbed are the people who generally come for “marriage counseling”?’ he asked rhetorically in October of 1951 in a letter to his friend and confidant Robert Harper. ‘Are most of them distinctly neurotic?…How many marriage problems are truly caused by cultural or situational factors, unaccompanied by serious personal disturbances?’ (Ellis, 1951a). In the same letter, Ellis also expressed parallel dissatisfaction with the psychoanalytic methods he had been trained in. ‘Does the “non-directive” technique [of talk therapy], rigorously applied, stink so badly as I think it does?’ he wondered.

If Ellis was no fan of Freudian analysis, he was also critical of the more eclectic, and to Ellis almost as unscientific, therapeutic techniques being explored by many of his contemporaries. Correspondence between Ellis and Carl R. Rogers in the summer of 1952 provides one illustrative example. Rogers and Ellis were both trained in the mental hygiene tradition, and Rogers would eventually become one of the most famous American psychotherapists of the century, rivaled only by Ellis himself. ³ A decade older than Ellis and firmly established at the University of Chicago, Rogers had achieved fame by the early 1950s with his theory of Client-Centered Therapy (CCT), later known more widely as person-centered or Rogerian psychotherapy. At a presentation by Rogers to New York marriage counselors, Ellis had criticized Rogers’s studies evaluating the results of client-centered therapy so savagely that Rogers felt compelled to respond on personal terms. ‘I have always felt that your dislike for me and everything I stand for is so strong that you are much more interested in tearing down what we do than in attempting to understand it’, Rogers wrote to Ellis; ‘I have no objection to your tearing this [paper] to pieces also but I suppose I do have a vague hope that before you do so you might read it once with no other purpose than to see what we are doing’ (Rogers, 1952). The paper to which Rogers referred examined the effectiveness of CCT using assessments of patient self-reporting tabulated using a form of factor analysis known as the Q technique, which measured patient self-response across a number of variables in an attempt to locate common subjective traits (Kampen and Tamás, 2013).

Ellis’s critique of Rogers’s evaluation of his client-centered therapy, and by extension claims for the efficacy of the method itself, was grounded in a distrust of the data reported by Rogers’s patients that formed the foundation for the Q technique’s statistical analysis. Ellis’s criticism focused on the objectivity of patient motivation in answering Rogers’s evaluative questions. ‘The Q technique’, Ellis wrote to Rogers, ‘is merely an objective method of scoring various things which the client says about herself (just as personality inventories are an objective method of scoring similar self-sayings’ (Ellis, 1952). Ellis objected not so much to Rogers’s interest in pursuing a standardized evaluation as to what Ellis considered a lack of meta-analysis of the Q technique itself, and a further lack of critical assessment regarding CCT’s reliance on previous psychodynamic theories. ‘You seem determined, as ever’, Ellis wrote testily in the same letter, ‘to believe that you…are originating and experimentally proving various personality theories, instead of, as is actually true, re-discovering and re-proving [them].’ Ellis’s concern with the validity of psychotherapeutic techniques was central to his own development of rational emotive therapy, but so was his questioning of who – therapist or patient – was best equipped to articulate a legible sense of self. In contrast with Rogers’s interest in subjective numerical assessments drawn from the patient herself, Ellis sought a different paradigm not only for measuring success in his clinical practice, but also for reinforcing psychotherapy as a scientific pursuit both for the benefit of patients and for his professional colleagues: that of a clear, easily comprehensible language for therapy.

Clarified concepts and straight thinking

Central to Ellis’s vision of a new, more ‘scientific’ therapeutic method was a mode of clinical intervention grounded in a standardized therapeutic language accessible to everyday consumers of therapy (which in Ellis’s case meant middle-class New Yorkers, most of them white). Despite his fervent empiricism, Ellis was not enamored with methods for evaluating the efficacy of therapeutic treatment that relied on numerical quantification: in numerous pieces of correspondence in the 1950s, Ellis made clear his dislike for the popular psychometric tests and rating scales of the time. For instance, Ellis questioned the general efficacy of the Minnesota Multiphasic Personality Inventory (MMPI), a standardized personality test developed in the early 1940s by Stark R. Hathaway and John C. McKinley of the University of Minnesota and already popular by the early 1950s for its statistical reliability (Hathaway and McKinley, 1940). Ellis approved of the utility of the MMPI as an assessment tool able to diagnose clusters of answers defined by the test itself as indicating psychopathology, but felt the scale was a poor instrument for ascertaining the particulars of a patient’s history and the wider context for a patient’s complaint. In a 1953 exchange with Dr James McConnell of the University of Texas regarding studies exploring the scientific accuracy of the qualitative interview method, Ellis noted how difficult it could be to determine whether the MMPI accurately assesses a patient’s psychological condition to the same degree as an in-person interview, and under what conditions (Ellis, 1953a). In another letter, Ellis averred that he could ‘not get overly enthused’ about use of the MMPI, ‘except in certain situations’ (Ellis, 1955b).

As Ellis worked to develop an organized conceptual basis for his therapeutic methods that did not rely on extant psychological rating scales, he was preoccupied by epistemological questions regarding what both the patient and the therapist could understand about a patient’s psychological state through language, and how to systematically transform that knowledge into an effective clinical technique relying on common linguistic definitions. A central crisis in this development came with Ellis’s participation in a conference on ‘Psychoanalysis and the Philosophy of Science’ in 1954 at the University of Minnesota’s Center for the Philosophy of Science, after which Ellis was invited to contribute to the first volume of a new publication, Minnesota Studies in the History of Science, on the ‘Foundations of Science & the Concepts of Psychology and Psychoanalysis’. As he was preparing the manuscript for Reason and Emotion in Psychotherapy in the winter of 1955, Ellis was also editing his contribution for the Minnesota volume, and entered into a correspondence with Michael Scriven, an Australian mathematician and research associate at Minnesota who would go on to become a major authority in the history and philosophy of science on the epistemology of evaluation (Kochen, MacKay, Maron et al., 1964; Scriven, 1960). In the course of editing Ellis’s contribution to the new series, the two engaged in a spirited debate as to the conceptual underpinnings of rational therapy and the development of a particular kind of ‘operationalized’ or ‘clarified’ psychoanalytic language for clinical purposes.

In the resulting article, ‘An Operational Reformulation of Some of the Basic Principles of Psychoanalysis’, Ellis laid out what would, with some semantic changes, become the core linguistic principles of RT as elucidated in 1962’s Reason and Emotion in Psychotherapy. Ellis justified his call for a change in clinical therapeutic language by advocating for the theory called ‘operationism’, which had been formulated by physicist Percy Williams Bridgman in his 1927 book The Logic of Modern Physics. By the 1950s Bridgman’s book had also become influential within experimental psychological circles; Ellis now sought to apply its tenets to clinical therapeutic practice. As Ellis described it, operationism meant that ‘to be operationally meaningful, a statement must be confirmable at least in principle: that is to say, a scientific theory must be tied to observables at some point’ (Ellis, 1956: 132). Ellis sought to draw a clear distinction between his own view of operationism in psychotherapeutic treatment and that of behaviorists such as B. F. Skinner. Instead of an orthodox behaviorist view, which dismissed mental processes as lacking experimental relevance unless they were manifested in external behavior, Ellis advocated for what he called ‘modified operationism or empiricism’, in contrast to the ‘extremes’ of Skinner’s model.

Ellis’s goal in advancing this ‘modified operational approach’ was to translate concepts from psychoanalysis that he considered unscientific into ‘lower-order’ terms. ‘Psychoanalytic principles’, Ellis wrote, ‘should be stated in terms so that they are, in some final analysis, in principle confirmable in terms of some ultimate observables’ (Ellis, 1956: 135). In practice, this process consisted of redefining Freudian terms like ‘libido’ into a simplified language of human perception and response drawn from the biological sciences of the day. Ellis grounded his axioms in the assumption that ‘every human organism (a) perceives (observes, sees, senses, or feels) and (b) responds (acts, performs, or behaves)’, with thinking, learning and evaluation of good and bad all stemming from these innate processes (ibid.: 139). While Ellis allowed in his conclusion that ‘rigid operationism would, at the present time, unduly restrict scientific thinking and investigation’, he nonetheless advocated for making psychoanalytic principles ‘more operational [and] observable’ where possible (ibid.: 152).

The correspondence between Ellis and Scriven in the winter and spring of 1955 suggests that Ellis was not inclined initially to make these later concessions around the utility of scientific abstraction. The gist of the disagreement between Ellis and Scriven concerned Ellis’s epistemological assumptions, particularly the distinction between phenomena that were ‘observable’ versus phenomena that were ‘confirmable’. For Ellis, ‘observable’ and ‘confirmable’ were synonymous, and he disliked and dismissed the category of what he called ‘implicitly definable terms’ – terms that Scriven defined as ones ‘used in an attempt to unify a particular body of knowledge’ in the sciences without a direct connection to observable phenomena. Scriven suggested in response that ‘in psychoanalysis, concepts can be clarified’, but that only ‘hypotheses can be confirmed or denied’ (Scriven, 1955a).

Ellis enthusiastically seconded Scriven’s suggestion that the conceptual language of psychoanalysis be ‘clarified’, wondering whether in doing so therapeutic language could be made coextensive with the (to his mind) empirical fidelity of observable patient evidence. ‘If psychoanalytic constructs were clarified’, Ellis wrote to Scriven, ‘I believe, the clarification would be in terms of clinically observable material; and if clinically observable material were properly employed as a basis for theory construction, the constructs thereby derived would be essentially the same…as the “clarified” psychoanalytic constructs.’ Ellis’s understanding of the term ‘clarification’ is explicitly lexicographical: drawing the right definitions from empirical observation would in turn reproduce ideal clinical outcomes. In seeking an empirical standard for his new mode of therapy, Ellis was searching for a reliable scientific metric not in the numerical standardization of the rating scales of such as the MMPI, but at the level of a ‘straight-forward’ or ‘clear’ language of rationality as he understood and defined it through his own observations (Ellis, 1955a). ⁴

In reply to Ellis’s broader argument, Scriven attempted to impress upon him the centrality of abstract thought as a tool of scientific inquiry – and by implicit extension, as an element in therapeutic practice. Scriven suggested that his own definition of ‘confirmable’ read as ‘observable or leading (perhaps very indirectly) to observable consequences’, in effect arguing for a capacious, though common-sense, understanding of observability which would allow followers of Ellis’s new theories to include concepts, such as thinking, that were only implicitly observable. This openness to the abstract stood, as Scriven noted, in contrast ‘to the extreme behaviorist position’, one that ‘insisted that the idea of a thought in another man’s mind was illegitimate since it could not be observed’ (Scriven, 1955b). To press his point, Scriven compared psychoanalytic concepts such as the libido to the electric field: ‘to say that an electric field is observable is simply false in the ordinary sense of the term’, he wrote, noting that such a field was nonetheless ‘confirmable [via] observable consequences’. Scriven further pressed the distinction between the possibility of, in his words, ‘the libido as being unobservable (although this is not a drawback to its scientific status); the existence of the libido as being unconfirmable (if true a devastating drawback); and the concept of the libido as being vague, wholly mathematical, etc’. As such, Scriven noted, ‘“observable” refers to entities and processes; “confirmable” refers to hypotheses, assertions; and concepts are, properly speaking, neither observable nor confirmable, but useful, clear, vague, etc.’ (Scriven, 1955b).

In other words, Scriven’s concern was that Ellis was committing a category error. The utility of the term ‘libido’ as a vague clinical concept might be debatable, but its status as an abstract concept that was not directly observable in Ellis’s consulting rooms was not itself a problem for its potential relevance to science (the phenomena of sexual attraction were evidently observable to Ellis). Scriven confessed himself ‘a little uneasy’ at this prospect: as he noted, ‘strictly speaking, it is not the construct that is observable but the designatum (or reference) of the construct’ (Scriven, 1956). Drawing on discussions and research with the philosopher Carl Gustav Hempel, who was a colleague at Minnesota at the time (and himself a logical positivist and no metaphysician), Scriven noted the following: ‘it appears we must have [scientific] theories because 1) they are heuristically indispensable in the context of discovery; 2) they are mnemonically indispensable in the context of education; and 3) they are logically indispensable to theories of any complexity’ (ibid.). While Ellis eventually acknowledged this point in the finished article, he nonetheless continued to practically confuse the operational naming of observable phenomena with their confirmation, reifying in language a certain set of purportedly ‘straightforward’ ‘observable constructs’ that simply swapped out obscurities and misconceptions at one level of comprehension for those at another. In other words, instead of creating a ‘clarified’ empirical language stripped of abstract theories about knowledge and human behavior, Ellis incorporated his own abstract assumptions about empirical observations, grounded in operationism and behaviorism, into the structure of RT as a technique.

This debate, through which Ellis accepted some but not all of the younger Scriven’s suggestions regarding the language in which to express his theories, had direct relevance for the model of human subjectivity assumed by the mechanics of RT. In Reason and Emotion, Ellis identifies 1954 as the year when he realized the importance of ‘language and the symbol-producing facility that goes with language’, in distinguishing humans from the animal subjects of the experiments of Skinner and Pavlov (Ellis, 1962). This recognition meant that Ellis was thus able to develop a language for both therapy and assessment in clinical psychotherapy that was appealingly vernacular, and while heavily influenced by Skinner, was not so dogmatically inflexible as to be useless in the everyday world of doctor–patient relations. This practical step in itself was valuable for the popular uptake of Ellis’s theories, as it allowed Ellis to popularize behaviorist concepts that had previously had a veneer of scientific impenetrability and even, paradoxically, abstraction.

More significant were the ways in which Ellis’s attitude towards language itself and its relationship to the observable world were incorporated into rational therapy’s clinical techniques. Language and its logic were fundamental to this new theory, and Ellis advocated for their close and consistent connection in clinical practice: in Reason and Emotion in Psychotherapy, he wrote that ‘the rational therapist is a frank propagandist who believes wholeheartedly in a most rigorous application of the rules of logic, of straight [forward] thinking, and of scientific method to everyday life’ (Ellis, 1962: 102). Ellis understood this language of ‘straight thinking’ as the stable medium for clinical progress, and also the chief means for performing an evaluation of the efficacy of therapy. ‘Evaluating’, Ellis writes, ‘is a fundamental characteristic of human organisms and seems to work in a kind of closed circuit with a feedback mechanism, since perception biases response and then response tends to bias subsequent perception’ (Ellis, 1956: 139). Intervening at the level of everyday language meant that rational therapy would change the course of this ‘closed circuit’, making it legible to the therapist and reordering it for the benefit of the patient.

As part of his project of ‘clarification’, Ellis took aim at the same ‘implicitly definable terms’ (in the finished piece described as ‘heuristic concepts’) over which he had debated with Scriven. Ellis was explicit in his desire to restate Freudian theories in plain language – in ‘molar terms’, to use Ellis’s chemical analogy – in part for the

…practical advantages of eliminating dogmatism and fuzzy thinking, encouraging new psychoanalytic hypotheses and experiments, making for a closer rapprochement between general psychological theory and psychoanalysis, and making analytic principles more acceptable to practicing therapists and their patients. (Ellis, 1956: 136)

Yet Ellis left unexplored the broader societal values implicit in his embrace of a more operational terminology for psychotherapy. Ellis’s ‘lower-order concepts’ for ‘straight thinking’ themselves held implicit a particular set of hierarchizing assumptions not only regarding human cognition, emotion and behavior, but also regarding epistemology itself: what could be known scientifically about individuals, and by those individuals themselves, from their expression, behavior and actions. James C. Scott observes that one of the tenets of high modernist ideology is its embrace of ‘scientism’: ‘a faith that borrowed…the legitimacy of science and technology’ (Scott, 1998: 4). While Ellis sought to create a language that was consonant with behaviorist principles, rational therapy nonetheless steered dangerously close to an ideological program. The ‘straight’ thinking promoted by RT techniques entailed, and via CBT still entails, acceptance and internalization of operationist assumptions about the primacy of externalized, expert evidence over other forms of self-knowledge. This attribute of RT enabled the therapy to act as a conceptual mechanism for regulating language in ways that introduced hierarchies of value, and that was extensible across many individuals to produce a unified system of what could be known and articulated about the self.

Media objects for and of therapy

Ellis’s development of rational therapy took place not merely in a historical moment that understood the restructuring of language as a solution to problems of mental hygiene. The postwar period was also the golden age of American technological optimism, with projects of expert modernization and improvement propagated and circulated in a variety of new mass media forms such as television, popular magazines and cultural exhibitions (Spigel, 2012; Turner, 2012). Ellis’s therapeutic methods and emphasis on language were thus not simply influenced by the milieux in which they were formulated, but designed to succeed in them: not only the postwar shift in ideas about the utility of psychoanalysis, but also a postwar bureaucratic material culture of paperwork and printed matter, mass broadcasts and middle-brow culture. The material constraints and affordances of Ellis’s life and work were connected to broader economies of professionalism, the circulation of knowledge, and the interaction of ‘expert’ and ‘popular’ culture, all of which shaped both Ellis’s trajectory as a thinker and the ultimate popular success of rational therapy. As such, both the media objects through which Ellis sought to develop RT, and those through which he promoted it, shaped the parameters of his thinking and the development of the therapy itself. These objects for and of therapy changed the structures of power and knowledge production among therapists and clients, making the emotions defined as legible by RT’s language widely accessible to a consuming public, patients and their doctors.

Foremost among the media influences that shaped rational therapy was Ellis’s familiarity with mass-market publishing. Prior to his turn to psychology, Ellis’s early professional career was shaped by his relationship with the New York print media. By his own account, Ellis’s professional writing career started as a teenager, when he ‘began to win prizes in the “Biggest News of the Week Contest”’ run by Joseph Pulitzer’s New York World tabloid. Among his later literary accomplishments, Ellis also noted that he had ‘published scads of comic verses in New York newspapers’ (Ellis, 1957a). Ellis wryly observed that his editor criticized his first psychological work for being ‘too literary for [its] medico-psychological audience’, and that the piece needed ‘to have the colons and semi colons replaced by commas and periods’ (ibid.). Becoming a psychological professional was a matter of writing style as well as intellectual substance, but Ellis’s facility with popular genres, as well as his prolific output and taste for publicity, served him well in his career as psychologist straddling academic and popular publishing.

Prior to the publication of Reason and Emotion in Psychotherapy in 1962, Ellis had already authored or coauthored 13 books, many aimed at a popular audience. Notable among these were 1957’s How to Live with a Neurotic, 1958’s Sex Without Guilt and 1960’s The Art and Science of Love, intended for a mass public readership. A New York Times advertisement for Sex Without Guilt emphasized the book’s broad relevance and, by extension, the appeal of Ellis’s methods: ‘Countless men and women must fight a lonely battle against guilt for doing those sexual things which are neither harmful to themselves nor to others’, the ad copy read. ‘They number in the millions…One of the purposes of this book is to help you understand sex guilt so that you may throw off its shackles.’ Ellis’s focus on clarifying therapeutic language in both clinical and popular contexts was made plain by the advertisement, which reassured readers that ‘always the talk is easy-to-read; plain-talk born of a healthy guilt-free attitude’ ( New York Times, 1958). As noted above, Ellis’s focus on ‘plain-talk’ as a sexologist signaled his wider interest in a clarified language of therapy that was a signifier of mental health: the rational patient, one able to confront his or her neuroses with the appropriate vocabulary, would be well on the way to recovery.

Ellis was not just familiar with print media, but also with broadcast formats such as radio and television. He prepared extensively for his frequent appearances on the radio throughout the 1950s: his notes for an appearance on New York AM talk radio station WEVD in early December of 1951 to promote his first book, The Folklore of Sex, suggest both Ellis’s long-standing interest in evaluating human behavior, and his desire to present a ‘scientific, logical, [and] socially intelligent’ view of sex to the public. Here, as in later work, Elis was particularly concerned with the evaluation of (in this case sexual) norms; he sought to inform his listeners that ‘we cannot…even agree on a definition of the word normal’. Ellis’s espoused solution to the problem at this point in his career was a fairly straightforward cultural relativism, but Ellis’s preoccupation with the definition of the word ‘normal’ points to the later emphases of RT. According to Ellis, social norms about sex prompted cultural definitions of the normal and abnormal, ‘a hard fact for professional moralists and do-gooders to swallow’ (Ellis, 1951b). As he developed rational therapy, Ellis appeared more frequently on the radio to promote it: three times on different New York City radio stations in 1960, and eight times in 1961 (including two appearances on Los Angeles radio stations as part of a trip to California). Ellis also began to appear on the new medium of television to discuss rational therapy, including on New York’s Channel 5 in September of 1961.

Ellis’s prolific writing and public appearances were less in the service of winning patients for himself and his colleagues (such as friend and collaborator Dr Robert Harper), though they accomplished this end too. More important to Ellis was the popularization and extension of rational therapy as a recognized and popular psychotherapeutic technique, and the popularization of media forms to facilitate RT, including his own books, worksheets and other media published by the Institute for Rational Living to scaffold the work of his patients. Ellis noted in a 2003 paper that RT’s emphasis on homework assignments, through which patients used a variety of worksheets to ‘counterattack their irrational beliefs’, had been a standard of the therapy since the publication of Reason and Emotion (Ellis, 2003). Rational therapy as it developed in the 1950s was further enabled, refined and articulated for Ellis through the mediated viewing and recording of its performance. Ellis and his collaborators gave numerous workshops for other practitioners, either by invitation of some institution or organized around a large professional gathering such as the annual meeting of the American Psychological Association (APA). Ellis also performed a number of public demonstrations of rational therapy. In preparation for a workshop at the 1963 meeting of the APA, Ellis and his colleague Dr Leah Gold Fein discussed a plan to have therapists from different methodological schools treat patients for 10 minutes each in front of workshop participants – ‘real, live clients’, Ellis suggested, would be ‘much better [than] role-playing ones’ (Ellis, 1963).

Exemplifying his awareness of the mass media’s potential as a tool of popularization, Ellis also taught rational therapy as a set of therapeutic techniques by circulating audio recordings of sessions with his patients to colleagues. The circulation of these tapes, and preserving their quality, was a major topic of Ellis’s professional correspondence in the late 1950s. In a December 1957 letter to a St Louis colleague, Dr Jeffrey Caine, Ellis elaborated at length on the utility of the recordings. ‘Several therapists’, Ellis noted,

…after they had heard these interviews…[had] themselves benefited from hearing them. I have also found, during the last few years in which rational therapy has been developing, that I have benefited immensely from applying it to others – since, consciously and unconsciously, I also apply it at the same time to myself. (Ellis, 1957b)

Ellis understood that the combination of a codified linguistic system and accessible media for therapy had the potential to extend the reach of the therapist to patients unable to visit the office in person, foreshadowing later CBT techniques and contemporary forms of digitally mediated therapy provision such as Talkspace, a service that connects patients to therapists via text message. Ellis conceived of RT as a mediated therapy, both in the particulars of its techniques and in the mode of its popular dissemination. And in contrast to the tradition of American self-help books, Ellis did not confine his ambitions to one form of media. Instead, he responded to the potential inherent in both print-based and electronic channels to serve as mutually reinforcing media of emotional management under the broader ambit of RT as a conceptual system. The rational language of the therapy itself enabled the interoperability of RT’s concepts across various media, and the relative simplicity of RT’s directives towards ‘straight’ thinking made the therapy legible at both a popular and an abstract level.

Conclusion: High modernism and the legacy of rational therapy

Ellis’s multimedia mode of therapy was consistent with a wider project of high modernist cultural work in the American liberal tradition that typified the Cold War (Turner, 2013). James C. Scott observes that this form of making legible is the hallmark of high modernism, and his description neatly describes RT’s desired effect on a patient’s emotions:

Certain forms of knowledge and control require a narrowing of vision [that] brings into sharp focus certain limited aspects of an otherwise far more complex and unwieldy reality. This simplification…makes the phenomenon at the center of the field of vision more legible and hence more susceptible to careful measurement and calculation. (Scott, 1998: 11)

By the mid-1960s, RT was firmly established as an alternative therapeutic approach to traditional psychoanalysis. As rational therapy became better known through its popularization and uptake via other therapists, Ellis was increasingly deluged with letters from current and prospective patients, readers of his books and listeners to his radio appearances, and other individuals and groups interested in rational therapy. Many of these letters were from patients who felt rational therapy had been successful and who wrote to Ellis to express their thanks. ‘I think I have gotten the real gift of therapy’, one patient wrote – ‘the ability to think my own way out of my problems’ (Anonymous, 1958). Others were more effusively descriptive in their accounts of how rational therapy had helped them. ‘I have your book Reason and Emotion in Psychotherapy in front of me’, one woman, a nurse, wrote to Ellis, ‘[and] when I thought about your book and your sentences this big anxiety I felt rising in my throat for a few minutes disappear [sic] like a charm.’ After explaining how she had used rational therapy to cope with her fear of death and her superstitious anxieties regarding a deer she had hit with her car outside a local cemetery, the woman remarked at how her anxiety about her feelings was at the root of her ongoing distress. Rational therapy, she wrote, had let her say to herself ‘I am what I am’, and accept her emotions as being subject to her own rational persuasion (Anonymous, n.d.).

Yet the idea that the most effective techniques to deal with anxiety and uncertainty were linguistic, a reshaping of the language people thought and spoke, was just one of several competing ones for making emotions legible, classifiable and evaluable during the period; and like high modernism more broadly, rational therapy fell from favor as a fully fledged system. RT was in some ways a victim of its own success, and of an ongoing de-emphasis on language, and a commensurate re-emphasis on quantification that took place in the 1980s and 1990s. As Rachael I. Rosner suggests, the biomedicalization of psychology exemplified by the publication of the Diagnostic and Statistical Manual (DSM-III) gave new impetus to Aaron Beck’s iteration of cognitive therapy (CT), exemplified by a heavier emphasis on numerical tracking and self-evaluation than Ellis’s approach (2014). Beck had been heavily influenced by Ellis: Beck contributed to Ellis’s publications on RT during the 1970s, and as Rosner observes, ‘Beck borrowed so heavily from Ellis in creating therapeutic techniques for CT that an analyst, reviewing one of Beck’s books, lamented that he could no longer tell whether Beck was a psychoanalyst or a devotee of Ellis’ (ibid.: 753). However, Beck’s position at the University of Pennsylvania and his connections to governmental granting agencies gave CBT an implied sense of scientific authority, and funding sources, that Ellis’s mass-market brand increasingly lacked.

Nonetheless, RT’s subsequent incorporation into contemporary Cognitive Behavioral Therapy extended many of Ellis’s initial therapeutic methods and techniques, including his emphasis on a finite course of treatment, and on teaching patients to become self-sufficient through positive reinforcement – in effect, practicing self-therapy. Even as Ellis’s desire to create a language of ‘straight thinking’ in the service of making his patients more rational was superseded professionally, his goals succeeded conceptually, and his legacy is vital to the history of emotion’s leap into the digital domain. Contemporary modes of self-help and self-improvement, ranging from clinical CBT therapy workbooks to smartphone-based digital applications, mix an emphasis on numerical quantification with an ordinal schema of rationality, clarity and self-fashioning that Ellis would recognize as stemming from rational therapy. Like the mixed media of Ellis’s day, these current psychological technologies are also performing work of therapeutic remediation in a vernacular that Ellis helped to invent and disseminate to the public. As Ellis quipped to his friend Robert Harper, ‘Fame – or at least infamy – is my goal; otherwise I certainly wouldn’t be a psychologist’ (Ellis, 1951a). Ellis’s renewed fame, or infamy, would be well deserved. As the tenets of RT are folded into new forms of digitally mediated therapy, the uses and abuses of a ‘scientific’ language of self-help and self-control remain as present, and problematic, as ever. CBT succeeds as a digitally mediated form of therapy today precisely because of its participation in discourses of rationality and emotional control alongside its clinical efficacy (Golumbia, 2009). The successful integration of Ellis’s theories into contemporary digital therapeutic platforms attests to their continuing power as a form of emotional management aimed at the feeling public.