AARON T. BECK’S DRAWINGS AND THEPSYCHOANALYTIC ORIGIN STORY OF COGNITIVE

THERAPY

Rachael I. RosnerNewton, Massachusetts

In this essay the author challenges the standard origin story of cognitive therapy,namely, that its founder Aaron T. Beck broke with psychoanalysis to pursue a morepragmatic, parsimonious, and experimentalist cognitive model. It is true that Beckbroke with psychoanalysis in large measure as a result of his experimental disconfir-mation of key psychoanalytic ideas. His new school of cognitive therapy brought theexperimental ethos into every corner of psychological life, extending outward into thelargest multisite randomized controlled studies of psychotherapy ever attempted andinward into the deepest recesses of our private worlds. But newly discovered hand-sketched drawings from 1964 of the schema, a conceptual centerpiece of cognitivetherapy, as well as unpublished personal correspondence show that Beck continued tothink psychoanalytically even after he broke with psychoanalysis. The drawings urgeus to consider an origin story much more complex than the one of inherited tradition.This new, multifaceted origin story of cognitive therapy reaches beyond sectariandisagreements and speaks to a broader understanding of the theoretical underpinningsof cognitive therapy.

Keywords: cognitive therapy, Beck, psychoanalysis, history

Prologue

1961 and 1962 were momentous years forAaron T. Beck. They were the years he made adecisive break with his psychoanalytic past. Heclosed down his large psychoanalytic researchproject on depression, put to rest his applicationfor membership in the American Psychoana-lytic Association that had been rejected twice,and turned his back on the cornerstone of psy-choanalytic theory, the unconscious. He took asabbatical from the psychiatry department at theUniversity of Pennsylvania following a destruc-tive department-wide battle over the future ofpsychoanalysis in psychiatry. He began reading

cognitive and developmental psychology andembarked on a completely new way of thinkingabout depression (American Psychoanalytic As-sociation, undated; Beck, A. T., personal col-lection, A. T. Beck to I. Gregory, ca. April1965, M. Stein, June 7, 1962, A. J. Stunkard,June 22, 1961; G. Piers to A. T. Beck, Decem-ber 16, 1960; S. Schneyer to A. T. Beck,June 13, 1962; I. Sigel to A. T. Beck, March 13,1963; Beck, 1967; Sigel, 1960; Weishaar,1993). He presented his new cognitive theory toclose friends and family as an alternative toclassic psychoanalysis and psychiatry (Beck,A. T., personal collection, notes from “Meetingof the Three,” January 24, 1962, February 7,

This article was published Online First May 30, 2011.This article is based in part on research conducted for a

doctoral dissertation, York University, Toronto, Ontario, Can-ada. Preliminary research was supported by Ontario GraduateScholarships, the Graduate Development Fund and ResearchCosts Fund of York University, a York University President’sDissertation Scholarship and a Post-Doctoral Fellowship fromthe National Science Foundation. The author extends specialthanks to Dr. Aaron T. Beck for providing access to hispersonal collection and for participating in interviews. Theauthor also thanks Dr. Marjorie Weishaar for making avail-

able transcripts of her interviews with Dr. Beck. The authorthanks Drs. Marla Eby, Raymond E. Fancher, Tena T. Ros-ner, and members of the Independent Women’s ScholarsSalon (Drs. Conevery Bolton, Lara Freidenfelds, Joy Harvey,Susan Lanzoni, and Nadine Weidman) for critiquing the man-uscript. Early versions of this material were presented tomembers of Cheiron and the Center for Anxiety and RelatedDisorders, Boston University.

Correspondence concerning this article should be ad-dressed to Rachael I. Rosner, 31 Ripley St., Newton, MA02459. E-mail: rachael@denenberg.com

History of Psychology © 2011 American Psychological Association2012, Vol. 15, No. 1, 1–18 1093-4510/11/$12.00 DOI: 10.1037/a0023892

1

This

docu

men

tisc

opyr

ight

edby

the

Am

eric

anPs

ycho

logi

calA

ssoc

iatio

nor

one

ofits

allie

dpu

blis

hers

.Th

isar

ticle

isin

tend

edso

lely

fort

hepe

rson

alus

eof

the

indi

vidu

alus

eran

dis

nott

obe

diss

emin

ated

broa

dly.

1962; Weishaar, 1993). Two years later heshowcased a new cognitive theory of depressionin two articles in the Archives of General Psy-chiatry: “Thinking and depression: I. Idiosyn-cratic content and cognitive distortions” (TD1,Beck, 1963) and “Thinking and depression: II.Theory and therapy” (TD2, Beck, 1964). Somarks the beginning of cognitive therapy, themost renowned form of psychotherapy to haveemerged in the final decades of the 20th centuryand figurehead to a worldwide cognitive-behavior therapy movement. This origin storyof Beck’s cognitive therapy is one that I havepieced together from archival and other primarysources and it is fully documented.

A variant of this origin story is the one thatBeck likes to tell. It is more abridged and em-phasizes his conceptual break with the psycho-analytic unconscious. He made this decisionafter his experimental study of the psychody-namics of depression—which inspired the cre-ation of the well-known Beck DepressionInventory—failed to confirm his analytic hy-pothesis that depression is a form of invertedhostility (Beck, 1967). He concluded that thewish fulfillment concept was untenable: “Oncewe took the wish fulfillment out of it,” he re-ported, “then there was nothing. If you couldnot depend on motivation, there is nothing inpsychoanalytic theory that can hold any water”(Beck, 1979, p. 18). Between 1962 and 1964 hebuilt a cognitive theory of depression with anew vocabulary taken from cognitive psychol-ogy well-suited to experimental investigation.This theory and therapy were closer to his clin-ical observations and bore none of the concep-tual, linguistic, and technical markers of hispsychoanalytic past (Weishaar, 1993, p. 21).This story encapsulates the moments between1962 and 1964 when he crossed the bridge to acognitive approach.

These two variations of the origin story arecompatible; their differences are a matter ofemphasis. But there is yet another origin storythat Beck tells less frequently and that contra-dicts his assertion that psychoanalytic theory isuntenable without the theory of motivationpredicated on an unconscious. In this story Beckadmits to intellectual descent from ego psychol-ogy, a school of psychoanalysis that flourishedin the 1950s and 1960s and that focused on thefunctions of the ego. He not only admits todescent but speaks to continuity. Consider this

excerpt from a letter from Beck to John Bowlbyin 1981:

It might be a point of curiosity therefore for you toknow that my psychiatric training was completely andexclusively psychoanalytic . . . I would consider mytheoretical work as derivative from ego psychologyrather than from cognitive psychology or learning the-ory. At the present time in fact I am trying to refor-mulate many of the basic psychoanalytic concepts intocognitive terms (Beck, A. T., personal collection,July 29, 1981).

These sentiments appear again in abridged formin an earlier letter to Marvin Goldfried (Beck,A. T., personal collection, November 9, 1978)and yet again in print in 1993: “It may beobvious to spectators in the therapeutic arenathat cognitive therapy has coopted (or beencoopted by) a large sector of the behavior ther-apy approaches to psychopathology. What maynot be so readily discerned are many conceptsderived initially from psychoanalysis . . . corre-sponding, in part, to Freudian notions of pri-mary and secondary processing” (Beck, 1993,p. 197). What is this origin story doing here?How can he both break with psychoanalysis andadmit to continuities?

My aim in this essay is to complicate thestandard origin story of cognitive therapy bygiving this continuity story a weight equal to theother two. The impetus is my discovery of acache of hand-sketched drawings and notes inBeck’s personal files, drawn between May,1964 and January 1965, tucked away inside oneof the filing cabinets that line the walls of aclimate-controlled room in his basement. Thedrawings tell the story of the schema, a theoret-ical centerpiece of the cognitive model thatBeck first introduced publicly in TD2 in early1964 (Beck, 1964). They depict what Beck callsa “bipolar schema” of manic-depression, a ver-sion of the schema Beck does not mention inany of the literature. And what is astonishingabout the sketches is that they contain refer-ences to psychoanalysis. In other words, theyshow Beck drawing on psychoanalytic ideasafter he had published a new cognitive approachthat he himself claims—and that his recordsshow—represented a clean break with psycho-analysis.

What are these drawings doing here? Arethey merely a detour, a cul-de-sac as it were, inhis otherwise linear movement away from psy-choanalysis? Or did Beck’s commitment to psy-

2 ROSNER

This

docu

men

tisc

opyr

ight

edby

the

Am

eric

anPs

ycho

logi

calA

ssoc

iatio

nor

one

ofits

allie

dpu

blis

hers

.Th

isar

ticle

isin

tend

edso

lely

fort

hepe

rson

alus

eof

the

indi

vidu

alus

eran

dis

nott

obe

diss

emin

ated

broa

dly.

choanalysis go underground in a split betweenthe private world of science-in-the-making andthe public world of science-as-justification, toinvoke the discovery-justification distinction ofphilosophers of science like Bruno Latour(Latour, 1987)? In this scenario Beck omittedthe psychoanalytic components of his discoveryfrom his public narratives because they lacked aconvincing truth value for the scientists whomhe wanted to convince. Is this what Beck did?Or perhaps Beck folded his psychoanalyticskills into an intellectual sleeve so that he couldcontinue to reason privately in a mode that nolonger held currency. This idea would be anal-ogous to what historian of science Peter Galisonsuggests happened to the physicist Paul Dirac,who “in-folded” his geometric reasoning whilemaking public only his numerical reasoning(Galison, 2000). Galison’s story invites us toimagine that Beck had at his disposal two dif-ferent modes of reasoning, one of which heplayed publicly and the other privately.

Both Dirac and Beck played in two modessimultaneously. In Beck’s case, however, therewas no split between public and private. Beckclearly left classical psychoanalysis behind buthe simultaneously continued to rely on its ex-planatory power both privately and publicly.We can think of his cognitive and psychoana-lytic ideas as two strains of music, one domi-nant and the other more quiet and tentative,playing within the same composition. From thebeginning of his research Beck had been work-ing at the crossroads of two epistemic commu-nities (Knorr Cetina, 1999): psychoanalyticpsychiatrists and academic psychologists. Psy-choanalytic psychiatrists privileged the subjec-tivism of psychoanalytic theory. They were cli-nicians, not scientists. Academic psychologists,on the other hand, were scientists (see Beck,1991d, pp. 2–3 for a very clear description ofthe epistemic differences between the twogroups). A small but influential number of aca-demic psychologists in North America in thepostwar period who were also clinicians, manyaffiliated with the Menninger Clinic, tried tomarry the objectivist constraints of experimen-talism with the subjectivist constraints of psy-choanalysis (Rosner, 2005). Beck collaboratedwith several academic psychologists at Penn,including a graduate student named MarvinHurvich1 and a social psychologist named Sey-mour Feshbach, to undertake a similar endeavor

with the depression research project. By 1961he had decided that the experimental ethos wasmore to his liking than classical Freudianism.He let go of the unconscious and intensified hisfaith in the greater power of experimentalism(Beck, A. T., personal collection, A. T. Beck toS. Feshbach, October 19, 1965, December 1,1965; A. T. Beck to P. Meehl, March 13, 1968;P. Meehl to A. T. Beck, February 26, 1968).This faith gave shape to the entire opus of hisnew cognitive model, from how he did his sci-ence to how he taught his patients to help them-selves (that is, to become scientists in the lab-oratory of their own lives), to his idea of whatgood living should look like (rational, prag-matic, scientific). His experimental work ulti-mately earned him an Albert Lasker Award forClinical Medical Research in 2006 (Lasker,2006), an award similar to the Nobel Prize inMedicine. He has won or been nominated forsome of the world’s most prestigious prizes.

At the same time, Beck continued to reasonwith what he calls the theoretically neutral ele-ments of psychoanalysis not dependent on theexistence of an unconscious. This is the quieterstrain. It appears as early as September 1961when he writes to his psychoanalytic mentorLeon Saul that “grief, shame, anxiety, disgust,feeling of inferiority or deprivation, and self-rejection”—symptoms of depression—might be“relatively weak in motivational properties.Hence, they do not fit as readily into the moti-vational model as hostility or dependency.” Hecontinues, “I feel that the model has to be re-vised to provide a position for these ‘states’ andto establish their relationships to the needs anddrives. This is as far as I’ve gotten in my think-ing, but I do hope that we will be able to furtherrefine the theoretical structure of psychodynam-ics” (Beck, A. T., personal collection, Septem-ber 28, 1961). This letter establishes the quieterstrain of a motivation-free dynamic theory. Thedrawings of the schema are a continuation ofthat strain. They are a rare moment in hiswork—in the vast collection of his writings thatnow extend far beyond depression to a widerange of disorders including anxiety, phobias,

1 Hurvich was also Beck’s first psychoanalytic patient atthe Philadelphia Psychoanalytic Institute. They collaboratedon Beck’s first dream study (Beck & Hurvich, 1959).

3PSYCHOANALYTIC ORIGIN STORY OF COGNITIVE THERAPY

This

docu

men

tisc

opyr

ight

edby

the

Am

eric

anPs

ycho

logi

calA

ssoc

iatio

nor

one

ofits

allie

dpu

blis

hers

.Th

isar

ticle

isin

tend

edso

lely

fort

hepe

rson

alus

eof

the

indi

vidu

alus

eran

dis

nott

obe

diss

emin

ated

broa

dly.

schizophrenia, and many others—in which bothstrains play at equal volume.

The Black Box of The Schema

Beck first introduced the idea of a schemain TD2 (1964). This definition of the schemaremains the core of the concept in currentpractice:

The schemas are conceived as relatively stable cogni-tive structures which channel thought processes, irre-spective of whether or not these are stimulated by theimmediate environmental situation. When a particularset of stimuli impinge on the individual, a schemarelevant to these stimuli is activated. The schema ab-stracts and molds the raw data into thoughts or cogni-tions . . . In the formation of a cognition the schemaprovides the conceptual framework while the particulardetails are ‘filled-in’ by the external stimuli (Beck,1964, pp. 562–563).

A schema is a hypothetical mental structurecontaining a specific belief such as “life is hope-less” or “I never win” or “everyone is out to getme.” This belief is an enduring feature of aperson’s cognitive makeup. It shapes the con-tent of the temporary and fluctuating evaluativethoughts we experience in a given situation(such as “so-and-so doesn’t like me today”).Bruno Latour might call the schema a “blackbox” (Latour, 1987), an object employed bycognitive therapists whose usefulness is so self-evident that its mechanics are irrelevant.

When we look at the origins of the schema inthe early 1960s, however, we find Beck stillworking out its mechanisms. In 1962 and 1963for instance Beck tries to cluster schemas into“systems” such as “goals and standards,” “self-evaluation,” and “self-disciplinary” (Beck,A. T., personal collection, draft of Thinking andDepression: 2, ca. 1962). At one point he callsit a “concept,” drawing on cognitive psycholo-gists Harvey, Hunt and Schroder (Beck, A. T.,personal collection, handwritten notes June 6,1962, June 24, 1962, July 7, 1962; Beck, 1964;Harvey, Hunt & Schroder, 1961); at anotherpoint he calls it a “construct” in a nod to GeorgeKelly’s personal construct theory (Beck, A. T.,personal collection, handwritten notesMarch 20, 1962, April 2, 1962; Beck, 1991c;Weishaar, 1993). Finally he settles on the wordschema, “extracting” it from Piaget’s develop-mental theory (Beck, 1979, p. 19; Piaget, 1948).

In 1962 Beck asked Hurvich, Feshbach, and adevelopmental psychologist named Irving Sigel

to read first drafts of TD1 and TD2. Each ofthem independently advised him that a struc-tural theory of the schema was inadequate. Theyasked him to explain why a patient’s symptomswould increase and decrease in severity, that is,to account for the dynamics of the symptoms.They urged him to reconsider his rejection ofmotivation (Beck, A. T., personal collection, S.Feshbach to A. T. Beck, January 17, 1963; M.Hurvich to A. T. Beck, February 25, 1963; I.Sigel to A. T. Beck, hand-written comments ondraft of Thinking and Depression, ca. 1962).Indeed Beck faced a difficult theoretical prob-lem. Since the schema was a structure that ac-tivated cognitions, it called for some kind ofenergy variable. What could he use if he did notposit a wish that compelled activation? Thesituation with energy variables in 1962 wastricky. The psychoanalysts, he knew, were in“sharp disagreement” over energy variables.Floyd Allport, in contrast, had used nonpsycho-analytic energy and structures (Allport, 1955;Beck, 1964). Beck detoured around this prob-lem in TD2 by acknowledging the need for anenergy variable without offering a solution ofhis own (Beck, 1964, p. 566). The six drawingsthat follow mark a return to this problem. Theyare visual reasoning exercises exploring a mo-tivation-free dynamic theory of depression.

Transmutation and Topography

The first drawing (Figure 1) actually is not ofa cognitive structure but rather of a volitionalstructure, that is to say a wish structure. It is acontinuum between opposing wishes to aggressand regress in manic depression. Beck labels thestructure “bipolar” because of the oppositionalquality of manic depression. Observe how Beckbrackets the manic and depressive ends to indi-cate where the wishes to aggress and to regressare extreme. A neutral area in the form of asmall circle in the middle of the two arrowsindicates a space on the continuum where nei-ther of the wishes is activated.

Already we are alerted to a strange turn inBeck’s thinking. How could the bipolar schemabegin with a structure not of thoughts but ofwishes—the very concept Beck had rejected?To add to the confusion, there is nothing intrin-sically psychoanalytic in the shape of the struc-ture. It does not, for instance, recapitulate theshape of the ego and id that Freud sketched in

4 ROSNER

This

docu

men

tisc

opyr

ight

edby

the

Am

eric

anPs

ycho

logi

calA

ssoc

iatio

nor

one

ofits

allie

dpu

blis

hers

.Th

isar

ticle

isin

tend

edso

lely

fort

hepe

rson

alus

eof

the

indi

vidu

alus

eran

dis

nott

obe

diss

emin

ated

broa

dly.

his 1923 monograph of the same title (Freud,1923/1960). Nor does it resemble other draw-ings of hypothetical dynamic structures such asthe “life space” drawings of Kurt Lewin(Lewin, 1938).

The answer may be found in a paper onpsychodynamics that Beck wrote in 1960 withMarvin Stein, with whom he had studied at thePhiladelphia Psychoanalytic Institute and taughtat Penn (personal communication, Dr. MarvinStein, April 24, 1997). This paper summarizedmaterial from their second-year course on psy-chodynamics for medical students. The idea of acontinuum of wishes extending from aggressionto regression appears here. Beck and Stein offertheir own way of understanding the dynamics ofpassive needs (defined as “a wish to receivesomething” Beck & Stein, 1961, p. 422C): “Forthe purposes of explanation passivity may beregarded as a continuum, one pole of which isthe wish for absolute rest and at the other end ofthe scale, aggressiveness, self-assertiveness,and productivity. The wish for passive partici-pation or vicarious experience would be closerto the passive end of the continuum” (Beck &

Stein, 1961, p. 422N). Beck invented the con-tinuum to clarify the concept of passive needs.

The continuum idea does not appear inBeck’s schema in TD2. But it suddenly reap-pears in his notes of February 1964. Beck ex-plores two “contradictory attitudes” with a faintovertone of Freud’s pleasure principle: “Ishould work day and night on the paper (be-cause it will bring pleasure)” and “I should notwork day and night on the paper (because it willbring pain)” (Beck, A. T., personal collection,handwritten note, February, 1964). He does notvisualize them but writes that they are structur-alized into a schema that clusters with othersinto “organizations.”

Let us assume that the volitional structure ofFigure 1 visualizes this continuum idea. In Fig-ure 2 we see the drawing that appears on thenext page of his notes. It is a cognitive structure.Notice how the cognitive and volitional struc-tures are identical except now we see two be-liefs: “omnipotent” and “impotent.” Follow thedrawings in sequence. The volitional and cog-nitive structures follow immediately the onefrom the other. The heading on the first page of

Figure 1. Beck’s drawing of a volitional structure (Beck, A. T., personal collection,handwritten note May 5, 1964). The page is entitled “Volitional and Cognitive Structures.”The volitional structure is a bipolar continuum of wishes from aggression to regressionleading to either mania or depression.

5PSYCHOANALYTIC ORIGIN STORY OF COGNITIVE THERAPY

docu

men

tisc

opyr

ight

edby

the

Am

eric

anPs

ycho

logi

calA

ssoc

iatio

nor

one

ofits

allie

dpu

blis

hers

.Th

isar

ticle

isin

tend

edso

lely

fort

hepe

rson

alus

eof

the

indi

vidu

alus

eran

dis

nott

obe

diss

emin

ated

broa

dly.

notes (“Volitional and Cognitive Structures”)suggests that they sit in relationship with eachother. Look at Beck’s handwriting. We findBeck noting a possible similarity between hisbipolar structure and bipolar structures ofGeorge Kelly’s personal construct theory. Thissequence marks the beginning of what I believewas a transmutation of motivational psychody-namics into cognitive dynamics. Remember thatBeck’s first visual act was to create a psychody-namic wish structure visualizing the continuumidea of 1961. Then he drew an identical cogni-tive structure. Beck never returned to thevolitional structure, so we cannot know if heintended to develop the idea. He may haveimagined the cognitive structure as an overlayon the wish structure. But the implication of thissequence is astounding. In two pages Beck con-structed a new psychoanalytic wish structure,put aside the motivational model and installed acognitive model in its place. To speak hyper-bolically, where motivation once reigned cog-nition was now king.

Nonetheless, it is essential to see that thestructure itself did not change. This continuityof structure (from wishes to cognitions) heldopen the possibility of dynamic continuities as

well. The next two drawings, of May 19, 1964,are evidence of these continuities.

Cathexis

In both drawings of May 19, 1964 Beck cutsthe schema in half and treats only the depressiveside. The first drawing (Figure 3) is on a pageentitled “Theory of Personalities.” Beck drawsfive schemas of similar content aligned verti-cally and connected by thick short black lines atequivalent points. The midpoint contains neu-tral thoughts: “I am equal,” “I can manage,” and“things are ok.” The extreme depressive endcontains extreme thoughts: “I want desperatelyto die,” “I hate myself,” and “life is empty.”This drawing depicts in visual form the idea thatschemas of similar content cluster into an “or-ganization” or depressive personality. Note thecloud with diagonal shading hovering verticallyon the far right end of the schemas. This cloudrepresents the spread of energy between theextreme ends. Beck postulates that when theextreme end of a schema is activated energyspills over into the extreme ends of neighboringschemas, provoking a cascade of activation thatconstitutes the depressive personality.

Figure 2. Beck’s first drawing of a cognitive structure (Beck, A. T., personal collection,handwritten note May 5, 1964). Page is entitled “Theory of Cognitive Structures.” Thecognitive structure and the volitional structure are identical in shape.

6 ROSNER

docu

men

tisc

opyr

ight

edby

the

Am

eric

anPs

ycho

logi

calA

ssoc

iatio

nor

one

ofits

allie

dpu

blis

hers

.Th

isar

ticle

isin

tend

edso

lely

fort

hepe

rson

alus

eof

the

indi

vidu

alus

eran

dis

nott

obe

diss

emin

ated

broa

dly.

Two vertical arrows pointing in opposite di-rections appear on the far right side of the page.Note the two barely discernible words tuckedbetween them: cathexis spreads. It is absolutelyastonishing to see the word cathexis here.Cathexis is another word for energy and it isunequivocally a psychoanalytic term. Freud’seditor and translator James Strachey inventedthe word as a translation of Freud’s Besetzung,meaning “filling up” or “occupying” (R. E.Fancher, personal communication, May 22,2010). With rare exceptions cathexis has cur-rency only in psychoanalytic circles. I couldfind no evidence of Beck’s use of the wordcathexis anywhere else in his work extendingback to the mid 1950s. Clearly this was not aconcept he employed regularly. So where doesit come from and why is he using it here?

I propose that when Beck articulated a dy-namics of the schema in early 1964 he tappedinto a deep wellspring in his psychoanalyticpast. There is remarkable similarity betweenthis structure/cathexis exercise and the work ofa midcentury, Hungarian-born psychoanalyst

named David Rapaport. Rapaport was an intel-lectual leader of an influential group of Ameri-can psychoanalysts known as ego psychologists(Gill, 1980; Rapaport, 1967). The ego psychol-ogists wanted to expand the psychoanalytic the-ory of the ego to include cognitive functionslike attention, learning, and memory. In the1950s Rapaport had recently moved from theMenninger Clinic to a small private mental hos-pital in the Berkshire Mountains called AustenRiggs. Beck studied at Riggs on a psychiatryfellowship from 1950 to 1952 (Knight, 1950;Knight, 1951). He attended Rapaport’s seminaron ego psychology (Rapaport, 1951b; Wheelis,1951) and studied with the entire clinical staff(some of whom had moved with Rapaport fromthe Menninger Clinic) including Erik Erikson,Robert Knight, Margaret Brenman, and RoySchafer. They provided Beck with his first com-prehensive exposure to psychoanalytic thinking(Rosner, 1999).

In the early 1950s Rapaport was systematiz-ing theories of the ego to build a new psycho-analytic theory of thinking (Rapaport, 1951a).

Figure 3. Drawing of depressive half of set of bipolar schemas (Beck, A. T., PersonalCollection, handwritten note May 19, 1964). Page is entitled “Theory of Separate Personal-ities.” This drawing shows Beck’s first use of the word cathexis.

7PSYCHOANALYTIC ORIGIN STORY OF COGNITIVE THERAPY

docu

men

tisc

opyr

ight

edby

the

Am

eric

anPs

ycho

logi

calA

ssoc

iatio

nor

one

ofits

allie

dpu

blis

hers

.Th

isar

ticle

isin

tend

edso

lely

fort

hepe

rson

alus

eof

the

indi

vidu

alus

eran

dis

nott

obe

diss

emin

ated

broa

dly.

His aim was to explain the ego’s “reality-adapted secondary processes (in) rational actionand logical—reasonable thought” (Rapaport,1952). Cathexis was a central element of thistheory. Psychoanalysts traditionally emphasizedcathexis as the energy of the id’s drives and theego’s defenses against them. Rapaport believedthat thinking was an ego function with abso-lutely no connection to the id. He proposedcordoning off some cathectic energy solely forthinking and memory. Rapaport, followingFreud, called this energy attention cathexis.Beck does not mention Rapaport by name in hisnotes but Rapaport’s ideas were clearly on hismind at the time. He had been reading Rapaportin 1961 in preparation for a study of the clinicalutility of the digit symbol test (Beck, Feshbach,& Legg, 1962). In 1963 he wrote a letter to RoySchafer, one of Rapaport’s students, about acritique of ego psychology (Beck, A. T., per-sonal collection, October 11, 1963). And Beckreferenced in TD2 Rapaport’s Organization andPathology of Thought (Beck, 1964). The evi-dence strongly suggests— given thatBeck studied with Rapaport, was immersed inego psychology, and had recently been readinghim—that Beck, even in some nondeliberateway, drew on Rapaport’s influence at this mo-ment of transmutation because his ideas bestsuited the problem Beck was trying to solve.Both are theories of thinking predicated onstructure and energy; both privilege a reality-testing and adaptive structure that mediates be-tween internal and external excitations; mostimportantly, both argue that the structure iscompletely independent of unconscious pro-cesses. Cognitive psychologist Nancy Nerses-sian has shown how scientists draw on deeplyheld understandings of source material to thinkcreatively about a target problem even when thesource and target domains are not immediatelycompatible (Nersessian, 2008). She also dem-onstrates the role of visual reasoning exercises,what she calls “imagistic representation,” in thisprocess. Her research helps us to make sense ofthe sudden and simultaneous appearance ofideas reminiscent of Rapaport’s and imagisticrepresentations in Beck’s theory.

The next drawing (Figure 4), also of May 19,1964, articulates a threshold theory of activationof the schema in which the cognitive structureand cathexis interact. Let us begin by noticingthe sequence of arrows. They indicate that this

is a flow chart of the progression of cathecticactivity within a single schema. The first draw-ing on the upper left side shows a moderateamount of cathexis in the midrange of the struc-ture. The shading is light to indicate a moderatedegree of charge. The subsequent drawings il-lustrate how repeated stimulation lowers theactivation threshold and shifts cathexis towardthe extreme end. This leads to a circular feed-back where thoughts themselves begin to evokethe schema and lower the threshold of activa-tion. The thick black lines at the extreme end ofthe schema in the final drawing constitute theenergetic expression of extreme negativethoughts in depression.

This interaction of psychoanalytic and cogni-tive elements appears again in Figure 5, which isan undated drawing, circa 1964. Beck now adds avertical axis to his horizontal topography that linksthe cognitive structures to emotions and rationalthought. We can articulate three distinct strata bynoticing differences in the shading. The bottomlevel of emotion is shaded like a deep subterra-nean sea; the middle level has distinct regions ofcognitive function; the rational level has a uniformlayer of diagonal lines that cuts across all com-partments of the schema. Beck also draws thecognitive functions in a vertical fashion; the moreprimitive and destructive cognitions are in thelower region of the stratum on the left side closerto emotions. He places the more optimistic cog-nitions in the upper region closer to the rationalfaculties.

This drawing also introduces a new role forcathexis. In the preceding drawings the schemawas completely passive in the excitation and flowof cathexis. In this drawing Beck endows therational faculties of the schema with the ability tomanipulate cathexis. Note his handwriting in cap-ital letters just above the drawing. Beck writes,“through use of rational can transfer cathexis fromsad to happy schemas.” Here we see a new mix-ture of psychoanalytic and cognitive sensibilities.A rational agent has entered the structure to har-ness and control the energies that otherwise accu-mulate reflexively in response to stimuli.

Geometry

A final set of drawings appear on January 5,1965 (Figure 6). There are three drawings on apage entitled “Cognitive Structures and Affect.”

8 ROSNER

This

docu

men

tisc

opyr

ight

edby

the

Am

eric

anPs

ycho

logi

calA

ssoc

iatio

nor

one

ofits

allie

dpu

blis

hers

.Th

isar

ticle

isin

tend

edso

lely

fort

hepe

rson

alus

eof

the

indi

vidu

alus

eran

dis

nott

obe

diss

emin

ated

broa

dly.

All the ideas contained in the previous drawingsappear in integrated fashion and are combinedwith a new geometric element: The tubularstructure morphs into squares, rectangles, arcs,vertical lines, diagonal lines, and dashed linesthat allow Beck more specificity of function.

By way of orientation, follow the three draw-ings in sequence from top to bottom. This se-quence depicts the development of the schemaitself. Beck had cultivated a developmental the-ory of the schema in previous notes that resem-bled those of Freud and Jean Piaget. On a pageentitled “Developmental Theory” from May 19,1964, for instance, he wrote:

1. Originally the schemas were dichotomous.2. As child developed he superimposed

checkpoints so that stimulation did not go to theextreme.

3. However the (more recent) acquisition wasmore vulnerable and easily cut through.

4. It is less automatic and involves impositionof judgment.

5. In other words there is a tendency to makean extreme judgment of every situation but theinterjection of judgment tends to arrest the rat-ing at a particular checkpoint.

6. In psychopathology, however, the fol-lowing may happen: (a) the checkpoint hasbeen disrupted as in the case of an extremereaction to trivial stimulus (anxiety, phobia),(b) it is more difficult to impose judgments,(c) the primitive egocentric pole of schema ishyperactive.

(Beck, A. T., personal collection, handwrittennote, May 19, 1964)These six statements find visual expression inthe sequence of drawings in Figure 6.2 Thetop drawing shows the dichotomous “good”

2 Notice that a similar structure appears in very lightshading just above the tubular structure in Figure 2. Theevidence suggests that Beck inserted this more compleximage at a later date.

Figure 4. Second drawing of the depressive half of the schema (Beck, A. T., personalcollection, handwritten note May 19, 1964). This is a flow chart tracking the accumulation andshift of cathexis toward the depressive end of a single schema.

9PSYCHOANALYTIC ORIGIN STORY OF COGNITIVE THERAPY

This

docu

men

tisc

opyr

ight

edby

the

Am

eric

anPs

ycho

logi

calA

ssoc

iatio

nor

one

ofits

allie

dpu

blis

hers

.Th

isar

ticle

isin

tend

edso

lely

fort

hepe

rson

alus

eof

the

indi

vidu

alus

eran

dis

nott

obe

diss

emin

ated

broa

dly.

and “bad” of the childhood form. The bottomtwo drawings show how they have separatedto the right and left to make room for an arc.This arc represents many things at once. Beckhas taken the midrange of the continuum andreshaped it from a vertical line into an up-ward-facing arc. Beck endows this midrangewith the rational faculties that appeared as aseparate stratum in the schema of Figure 4.Finally, he reconfigures it as a mature out-growth out of the primitive structures, thelocation of the “checkpoints” not present inthe primitive childhood structures to which herefers above.

Each of the geometric forms speaks to aquality either of primitive or mature thinking inthis more complex structure. For instance, notethe squares/rectangles of the primitive “good”–“bad” dichotomy. Here Beck has transmuted theextreme end of the tubular structure from a

curve into squares and rectangles. What aresquares and rectangles if not a combination ofstraight lines at right angles? They are rigid andfixed and represent visually a key concept incontemporary cognitive therapy: Primitivethinking in content and in style is fixed andinflexible.

Consider the arc-bridge in the bottom twodrawings. It represents the capacity of our ra-tional faculties to rise above those primitiveevaluations. And what does an arc symbolize? Itis the opposite of a square. An arc connotesflexibility, plasticity, and resiliency. The arc is avisual representation of another key element ofthe mature cognitive model. Mature thinkingnot only is less rigid in content but also moreflexible and resilient in style.

Notice how Beck has taken the emotionalsubstratum of the schema of Figure 5 and givenit its own separate structure. The feelings “sad”

Figure 5. Drawing of schema (Beck, A. T., handwritten note [ca. 1964]). This drawing addsemotions and rational thought to the schematic structure with a new vertical stratification offunctions. The emotions are on the bottom level, cognitions are in the middle, and rationalthought is on the top.

10 ROSNER

This

docu

men

tisc

opyr

ight

edby

the

Am

eric

anPs

ycho

logi

calA

ssoc

iatio

nor

one

ofits

allie

dpu

blis

hers

.Th

isar

ticle

isin

tend

edso

lely

fort

hepe

rson

alus

eof

the

indi

vidu

alus

eran

dis

nott

obe

diss

emin

ated

broa

dly.

and “happy” are now within their own rectan-gular boxes. The dark lines that run verticallybetween the cognitive and emotional structuresare of varying widths to connote greater orlesser strength of connection. Thicker lines in-dicate a strong and close connection to emo-tions. Thinner lines indicate a weak link. Thedashed line indicates no connection. The linesarticulate the idea that primitive thoughts giverise to strong emotions; rational thoughts evokeless emotion because they literally are furtherremoved. Notice too that the midpoint of the arcis exactly neutral with no connection at all toaffect. Beck visualizes this neutrality in themiddle drawing with a vertical dashed line ex-tending downward. The apex is a neutral cog-nitive high ground entirely free from affect.3

There are many questions that can be askedof these drawings. For instance, where wouldcathexis fit? What would happen if we tried toreconcile this drawing with the previous ones?

Remember that when Beck introduced a verticalstratification of function he endowed the newrational faculties with deliberateness. One inter-pretation of the arc-bridge is that the form itselfretains both the quality of deliberateness and anenergy variable. An arc rises up in the middleand touches the ground on each side. Primitivethoughts sit on each end just like in the previousdrawings. But here they sit below the neutralspace, anchored to the ground. Let us explorewhat the dynamics would be of such an arc-bridge.The rational faculties rise out of them withflexibility, strength, and resiliency but a tensionremains between the extension upward and the

3 Beck’s reference to “neuronal connections” in the bot-tom drawing is curious. There are echoes of Freud’s Projectfor a Scientific Psychology, which first appeared in Englishin 1954 (Freud, 1954; Sulloway, 1979), where Freud visu-alized pathways of neuronal connections between ideas.

Figure 6. Final set of drawings of the bipolar schema (Beck, A. T., personal collection,handwritten note, January 5, 1965). These drawings integrate a geometric component intoattributes of the schema that appeared in previous drawings and notes including a develop-mental theory, horizontal and vertical stratifications of functions, and cathexis.

11PSYCHOANALYTIC ORIGIN STORY OF COGNITIVE THERAPY

This

docu

men

tisc

opyr

ight

edby

the

Am

eric

anPs

ycho

logi

calA

ssoc

iatio

nor

one

ofits

allie

dpu

blis

hers

.Th

isar

ticle

isin

tend

edso

lely

fort

hepe

rson

alus

eof

the

indi

vidu

alus

eran

dis

nott

obe

diss

emin

ated

broa

dly.

pull of gravity downward. If we assume thesame kind of energy is at play in this drawing asin the others, then a deliberate act of will wouldbe required to draw the energy up and out ofthese extreme thought gravity wells. Willfuleffort is needed to keep the thoughts at thehigher level from being pulled down by theforce of gravity.

The Cognitive–Psychoanalytic Story of TheBipolar Schema

When we put the topographical, the geomet-ric, the developmental, the energetic, and thetherapeutic components of these drawings to-gether, like stringing words together into a sen-tence and sentences into paragraphs, a story toldin images emerges about our interior world andthe possibilities for growth and change. Like thepublic face of cognitive therapy these drawingsoffer a highly pragmatic vision of what goodliving looks like: The engagement of rationaland creative thinking about one’s self and theworld and the ability to solve problems withflexibility and resiliency. But these drawingscontain an additional component. They speak ofan ongoing conflict to maintain those capacitiesagainst a primitive self that never completelygoes away. They portray a highly charged en-vironment. The most to which we can aspire isto strengthen our innate rational capacities sothat we can resist, time and again, the pull of theprimitive. The drawings tell us that for Beck, atleast in the mid 1960s, the cognitive modelcarried over from psychoanalysis a stark realismabout the struggle of human nature. They are anadmission that we are not entirely in the driver’sseat. But while Freud was pessimistic about thehuman condition, Beck’s belief in an innatecapacity to neutralize our destructive selvessuch that he embodies it within our minds in-vested this model with an optimism not seen inFreud.

Indeed the drawings reveal a complex mix-ture of psychoanalytic and cognitive sensibili-ties. Our mature selves perpetually counteractthe pull of the primitive just as the ego perpet-ually defends against the unwanted impulses ofthe id. The primitive, like the id, is a feature ofchildhood lacking restraint and self-awareness.This quality is akin to Freud’s primary process.The mature faculties, like the ego, grow outof the primitive and mediate between reality and

the primitive experience. This quality is akin toFreud’s secondary process. Beck excised all ofthe motivational elements of psychoanalytictheory including defensive processes of theego that are predicated on the existence of adrive that seeks gratification. So Freudian as-sumptions about destructive impulses and de-fenses are gone. Beck’s insertion of cognitionsand rational faculties in their place also makesthe comparison with Freud incomplete. But theprimitive cognitions and the rational facultiesare in relationship with each other in a wayanalogous to the relationship between the id andthe ego.

Derivatives

The drawings of the bipolar schema are com-pelling evidence in favor of the psychoanalyticorigin story. The problem is that after January 5,1965, Beck makes no further mention of a bi-polar schema either in private or in public. Hedrops the entire exercise. If there are continu-ities, where are they?4 Let us begin by askingwhy Beck might have abandoned the bipolarschema and what he did next. The bipolarschema was a response to criticism from col-leagues that his structural theory was not dy-namic. By late 1961 he confessed to those samecolleagues that certain elements of his theoriz-ing were too far removed from observed behav-ior (Beck, A. T., personal collection, Beck,A. T. to S. Feshbach, December 1, 1965, C.Ward, December 16, 1965). It is probable thathe retreated from the bipolar exercise becausehe could not anchor it in anything he observedclinically.

It is also possible that the bipolar structuredid not support his preference to think interms of categories. Beck simultaneously hadbeen considering a unipolar schema (whichhe never visualized on paper). We find Beckdebating the virtues of the unipolar and bipo-lar ideas in his handwriting in Figure 2. De-liberations of this sort appeared as early asJuly 1962 when Beck established the funda-mental problem of whether his theory shouldbe more elaborate or limited to descriptivecategories (Beck, A. T., personal collection,

4 The author thanks one of the anonymous reviewers forthe very valuable critique of the concluding sections of thisarticle.

12 ROSNER

This

docu

men

tisc

opyr

ight

edby

the

Am

eric

anPs

ycho

logi

calA

ssoc

iatio

nor

one

ofits

allie

dpu

blis

hers

.Th

isar

ticle

isin

tend

edso

lely

fort

hepe

rson

alus

eof

the

indi

vidu

alus

eran

dis

nott

obe

diss

emin

ated

broa

dly.

handwritten note, July 6, 1962). The bipolarversion embodied the more elaborate theory.The unipolar version emphasized categoriesand descriptions. By late 1965 Beck settled onthe unipolar version (Beck, A. T., personalcollection, Beck, A. T. to M. Hurvich, De-cember 16, 1965). The public face of cogni-tive therapy showed no change in the schema.Beck did not begin referring to it as a “uni-polar” structure. Cognitive therapists have noidea that he ever envisioned a unipolar struc-ture in contrast to a bipolar one. Rather, hispublic portrayal of the schema after 1965continued as it had been in TD2: a stablecognitive structure containing a single themethat generates context-specific automaticthoughts.

Beck had been observing for some time thatschemas of related beliefs tend to cluster to-gether such that when one is activated the re-

lated ones also are activated. He had experi-mented with a variety of visual representationsof those clusters as a way of categorizing thecognitive configurations of different personalitytypes. Figure 3 shows what a cluster of depres-sive schemas looked like in the bipolar version.In that drawing he also imagines the energyvariable (cathexis) that causes the cascade ofactivations within the cluster. In September1964, he organized clusters from different per-sonality types in the shape of sectors within apie chart (Beck, A. T., personal collection,handwritten note, September 13, 1964). Figure7 illustrates that on January 11, 1965 (six daysafter the final bipolar drawings), he configuredthe depressive clusters in the shape of a triangle.This triangle does not show a schema but rathersimply establishes the presence of three clustersof schemas in depression. The three cornersidentify the themes of each of the three clusters:

Figure 7. Precursor to the cognitive triad (Beck, A. T., personal collection, handwritten noteJanuary 11, 1965). The themes in the triangle in the upper right hand corner of “loss,”“helpless,” and “self-blame” later became the more familiar triad of negative thoughts aboutthe self, about the world, and about the future.

13PSYCHOANALYTIC ORIGIN STORY OF COGNITIVE THERAPY

This

docu

men

tisc

opyr

ight

edby

the

Am

eric

anPs

ycho

logi

calA

ssoc

iatio

nor

one

ofits

allie

dpu

blis

hers

.Th

isar

ticle

isin

tend

edso

lely

fort

hepe

rson

alus

eof

the

indi

vidu

alus

eran

dis

nott

obe

diss

emin

ated

broa

dly.

themes of “loss,” of feeling “hopeless,” and of“self blame.” He later called this the cognitivetriad.

As Beck refined the cognitive triad (Beck,A. T., personal collection, handwritten notesMarch 10, March 15, 1965) these themes be-came known as negative thoughts about the self,about the world, and about the future. In late1965, he submitted a paper on the cognitivetriad to the Archives of General Psychiatry asthe third installment in his series on thinkingand depression (which was not accepted forpublication). The triad appears in print for thefirst time in Beck’s, 1967 book on depression(Beck, A. T., personal collection, M. Hurvich toA. T. Beck, December 6, 1965; C. Ward toA. T. Beck, December 5, 1965; Beck, 1967). Incontemporary practice the cognitive triad of de-pression functions as yet another black box ofthe model. Cognitive therapists use it to identifythe specific cognitive characteristics of depres-sion. It is not designed to evoke discussions ofthe mechanics of the schema but merely tomake it easy for therapists to identify the sche-mas of depression. As Beck told his biographer,“It was a neat little package that you could sell”(Beck, 1990b, p. 18). We now know that itemerged after TD1 and TD2 as the last in aseries of visual exercises charting possible con-figurations of the schemas.

Even with the cognitive triad in hand Beckstill imported some of his psychoanalytic theo-rizing from the bipolar version into the unipolarversion. Remember that TD2 (1964) introducedthe schema and how it operates in depression.The break with psychoanalysis is evident inTD2. His language is completely devoid of psy-choanalytic concepts. He describes the schemasimply as an “idiosyncratic cognitive structure.”After 1965, however, the schema takes on psy-choanalytic properties. It bears the imprint ofBeck’s thought experiments with the bipolarschema. The red threads that link the bipolarschema with the unipolar version are Beck’spostulates of two levels of thinking akin toFreud’s primary and secondary processes andcathexis. Recall that the bipolar version con-sisted of two opposing primitive qualities with aneutral high ground of mature qualities above andbetween them. The unipolar version postulatedtwo separate sets of unipolar structures, one withprimitive single-themed idiosyncratic schemasand the other with mature capacities. Beck attrib-

uted Freud’s primary process to the “primitive”schemas. He gave the “higher centers” a separatestructure with secondary process.

Some of these psychoanalytic ideas transferredmore completely than others. Beck has been con-sistent in arguing that idiosyncratic schemas areprimitive in a manner analogous to Freud’s pri-mary process (see, e.g., Beck, 1970a; Beck, 1972;Beck, 1984; Beck, 1991a; Beck & Weishaar,1989). In language especially evocative of thedrawings, for instance, he has described the “prop-erties of the primitive schemas” as being “theirbreadth, rigidity, and dichotomous structure”(Beck, 1972, p. 152). He even has hypothesized alink between cognitive and affective structures inlanguage that evokes the black lines of the Janu-ary 5, 1965 drawings, “The intimate connectionbetween cognition and affect may be representedmetaphorically as pathways between cognitivestructures and corresponding affective structures;therefore, a particular cognitive content producesan affect congruent with it” (Beck, 1972, p. 153).

He has been equally clear that the “highercenters” are analogous to Freud’s secondaryprocess. He has been more vague about theirstructural properties, however. He often speaksof the structure as serving a reality testing func-tion. In 1970, for instance, he suggested thatsecondary process was a system of “refined andelastic structures” that test, authenticate, andreject the primitive ideations (Beck, 1970a, p.194). In 1972 he spoke of their role in serving“the demand character of the external stimulussituation” (Beck, 1972, p. 153). In 1984, afterexplaining the similarities and differences be-tween Freud’s primary process and his own, hesimply called “mature thinking” the “higherlevel” corresponding to secondary process(Beck, 1984, p. 120). In 1991, in contrast, heargued that secondary process helped build ma-ture schemas: “Cognitive therapy is designed toproduce more enduring structural change thansimply symptomatic relief (de-activation of theschemas). If the underlying assumptions . . . areundermined and new, more adaptive schemasare constructed then the person will be lessprone to experience later recurrence of depres-sion” (Beck, 1991a, p. 194). The features ofsecondary process structures, however, remainsketchy.

Cathexis also has remained a sketchy concept.Beck often uses energy language such as “grind-ing out,” “powerful streams,” “explosions,” and so

14 ROSNER

This

docu

men

tisc

opyr

ight

edby

the

Am

eric

anPs

ycho

logi

calA

ssoc

iatio

nor

one

ofits

allie

dpu

blis

hers

.Th

isar

ticle

isin

tend

edso

lely

fort

hepe

rson

alus

eof

the

indi

vidu

alus

eran

dis

nott

obe

diss

emin

ated

broa

dly.

on. Beck even has acknowledged that psychoan-alytic energy concepts persist. He told SeymourEpstein, founder of cognitive–experiential self-theory (Epstein, 1994), that “I will draw on psy-chodynamics whenever it suits my purpose! I dofeel that many of Freud’s observations were righton target, and that some of his preliminary formu-lations made sense, but that he simply went toofar” (Beck, A. T., personal collection, October 27,1994). He continued,

I have struggled with the concept of energetics, cathe-xis, charge, and so forth for a long time because I wasnot at all satisfied with the way Freud had mixed in thephysicalistic concepts of energy with the abstract con-ceptualistic concepts of unconscious, and so forth asthough they were isomorphic. However, although Idetoured around these concepts in several of my books,I have found that there is no way I can satisfy myselfwith an adequate explanatory model unless I do bringin these concepts (Beck, A. T., personal collection,November 8, 1994).

But the nature and behavior of this energy vari-able within the model remain unclear.

Finally, a fundamental difference betweenBeck’s psychoanalytic theorizing in the bipolarand unipolar versions lies in his vision of thenature of our interior world. The bipolar modelrecapitulated the conflict between our destruc-tive natures and our adaptive selves of Freud’smodel. In the world of the bipolar schema therewas no possibility of escaping completely thepull of our primitive selves. The best to whichwe could aspire was a kind of energy homeo-stasis, where we willfully use our innate capac-ities for rational thought to outflank the brutestrength of our primitive selves. But the conflictitself was never-ending, just as in Freud’smodel. With the unipolar schema Beck losesthis sense of conflict. From the way Beck de-scribes it, there are simply lower order andhigher order cognitive mechanisms. The goal oftreatment is to deactivate the primitive ones andbuild more mature ones. The conflict may stillbe there but Beck doesn’t venture into thosespeculative waters.

Despite inconsistencies in Beck’s psychoan-alytic theorizing, the historical record allows usto bring into clearer focus the trajectory of thepsychoanalytic origin story. Beck tested exper-imentally in the late 1950s a psychodynamictheory of depression that he operationalized as“inverted-hostility”. His data did not support amotivational theory of depression. He consid-ered revising psychodynamic theory to account

for nonmotivational properties in depression.Instead, he broke with psychodynamics alto-gether and built a cognitive model of depressionin its place in the early 1960s. We now havediscovered a moment from 1964 and early 1965when he returned to a nonmotivational dynamictheory. He revived a psychodynamic concept hehad envisioned in the late 1950s of a continuumof passive-aggressive motivations, structural-ized it, and transmuted it into a bipolar cogni-tive structure. He imbued it with primary andsecondary processes and cathexis, constructsadvocated by a school of ego psychology inwhich Beck trained which proposed that cogni-tive activity occurred in the ego outside therealm of the unconscious. In 1965 he abandonedthe bipolar model and returned to a unipolarone. He de-emphasized cathexis but maintainedprimary and secondary process. This final struc-tural map of the schema persists to the presentday even as Beck has continued to improviseupon it.

Epilogue

Clearly Beck does not want to emphasize thecommunity of classical psychoanalysts withwhom he trained at the Philadelphia Psychoan-alytic Institute, with whom he worked at Penn,and whose national organization refused togrant him membership. In private, Beck hasspoken critically about their culture of loyalty.Writing to Paul Meehl in 1968 he rememberedthat “as time went on, I realized that support for[the psychoanalytic] postulate was ultimatelyderived from the declarative statements of thepsychoanalytic authorities rather than from ev-idence; I began to quaver in my belief that‘20,000 analysts can’t be wrong’” (Beck, A. T.,personal collection, A. T. Beck to P. Meehl,March 13, 1968). His antagonism toward psy-choanalytic dogmatism is palpable in this ex-cerpt from an interview with his biographerMarjorie Weishaar:

The personal elements . . . that got me out of the wholepsychoanalytic framework is the whole notion thatauthorities don’t have to be taken at their face valueand my own data seemed to contradict the authorities;that my own data can be trusted . . . And there were noauthorities that are more powerful in this world exceptmaybe priests—the Pope—but no authority is morepowerful than analysts because they know everything .They have the word. And if you deny their thing, itonly proves that they’re right. Denial is affirmation.

15PSYCHOANALYTIC ORIGIN STORY OF COGNITIVE THERAPY

This

docu

men

tisc

opyr

ight

edby

the

Am

eric

anPs

ycho

logi

calA

ssoc

iatio

nor

one

ofits

allie

dpu

blis

hers

.Th

isar

ticle

isin

tend

edso

lely

fort

hepe

rson

alus

eof

the

indi

vidu

alus

eran

dis

nott

obe

diss

emin

ated

broa

dly.

You’d be surprised how many intelligent people buyinto it (Beck, 1991b, pp. 17–18).

Over the years he has been positioning the ex-perimentalism of cognitive therapy as a directassault on the irrationality of what came before.The story of discontinuity makes clear that ourmemory of his work should begin with his firstvictory of scientific reasoning over psychoana-lytic dogmatism.

But then there is the issue of identity. Theorigin story of continuity tells us not aboutwhere he came from but where he hoped toarrive. Beck needed a new community withwhich to identify after 1962. He had incorpo-rated into his model psychoanalytic ideas com-mensurate with other break-away schools thatemphasized conscious experience in interactionwith the environment, such as those of KarenHorney, Alfred Adler, and Harry Stack Sullivan(see Beck, 1991a, p. 192). He decided to iden-tify himself with them. In 1990, he told PaulSalkovskis that,

First I called this ego psychology, (and then I felt thatthis was) the psychoanalysis of the ‘60s, this is neo-analysis. What I am saying is that (cognitive therapy)is consistent to this day with Adler and Horney and soon. I kind of identified myself with the so-called neo-analytic school (Beck, 1990a, p. 9).

In the early 1960s these groups had coalesced inan organization known as the American Acad-emy of Psychoanalysis. Beck joined the Acad-emy in 1968. He chaired a midwinter session ondepression, spoke at their meetings, served onthe membership committee, and published intheir journal (Beck, 1970b; Rosner, 1999). Hisself-identification as an ego psychologist andneo-Freudian has been a membership badgeshowing that cognitive therapy looks at con-scious activity (rather than unconscious) and theinteraction with the environment (rather than oninternal psychodynamics). This origin storypoints us not to Beck’s origins in classical psy-choanalysis but to his new location among post-classical psychoanalytic communities.

There are a few, like his former patient andcollaborator Marvin Hurvich, who understandintimately the overlay of his model with thesetheories (M. Hurvich, personal communication,April 4, 1997). But these locations are unknownto most psychotherapists. The neo-Freudianswere not quick to accept him as one of theirown. He let his membership in the Academy

lapse in 1976. He tells a similar story in 1997:“When I presented this material before the localanalytic society, I said, ‘this is really neo-analysis.’ They said, ‘Well, Beck, this is nolonger analysis. You better stop calling yourselfan analyst’” (Beck, 1997, p. 7). Beck needed toidentify himself as something other than a clas-sical analyst but it was not at all clear whowould embrace him.

Beck had simultaneously been courting be-havior therapists (see Beck, 1970a). Initiallythey too did not see reason to accept him as oneof their own. Beck continues:

I had to find a new name for this approach. At that timeI was attracted to behavior therapy, so I thought maybeI’d call myself a behavior therapist. I spoke to Dr. Wolpeabout some of my ideas and he said, ‘Well, you’re not abehavior therapist at all.’ So, I ended up with the idea ofcalling my approach cognitive therapy, because it wasbased on the cognitive model of psychopathology (Beck,1997, p. 7).

A split among behavior therapists in the early1970s, however, produced a new cognitivewing eager to incorporate cognitions into thestimulus-response paradigm. That group neededBeck. He needed them, too (Rosner, 1999; seeBeck, 1990b, p. 6 for a detailed story of how henurtured relationships with these therapists).Behavior therapists were sophisticated experi-mentalists. They also had long been critical ofpsychoanalytic subjectivism. Beck shared thissentiment. A group of cognitively oriented be-havior therapists coalesced around Beck. To-gether they charted a future for his new science;this included downplaying the overlaps withpostclassical psychoanalysis (see Clark, Beck &Alford, 1999 for a good example of this sci-ence-as-justification origin story).

There is something scandalous to a contem-porary clinical ear in suggesting that Beckneeded Freud, even if it was a neo-version ofFreud, to flesh out his theory. So many re-sources and so much rhetoric have been in-vested in keeping the wall between him andFreud firm. The training of many cognitive ther-apists does not include psychoanalysis. Conse-quently, they may not be fully conversant in thepsychoanalytic features of Beck’s theory fromtheir own experience. Their training emphasizesthe solid foundation of science upon which cog-nitive therapy rests, which distinguishes it frompsychoanalysis and other schools of therapy.My aim in this essay has not been either to

16 ROSNER

This

docu

men

tisc

opyr

ight

edby

the

Am

eric

anPs

ycho

logi

calA

ssoc

iatio

nor

one

ofits

allie

dpu

blis

hers

.Th

isar

ticle

isin

tend

edso

lely

fort

hepe

rson

alus

eof

the

indi

vidu

alus

eran

dis

nott

obe

diss

emin

ated

broa

dly.

challenge the validity of their science or to takea position on the relative merits of cognitivetherapy, psychoanalysis, or behavior therapy.Rather, I am making public and providing acontext for archival evidence showing that Beckhas been improvising upon both psychoanalyticand cognitive strains in his work nearly sincethe inception of cognitive therapy. The draw-ings and correspondences clarify that these ex-ercises have been a rich data source for some ofthe explanatory metaphors and conceptual mod-eling that appear in his written work. Theysuggest that the creative calculus of cognitivetherapy “in-the-making” has been more com-plex and broader in scope than previouslyknown.

References

Allport, F. H. (1955). Theories of perception and theconcept of structure. New York: Wiley.

American Psychoanalytic Association Papers. (n.d.).Undated and unsigned handwritten document.RG11 Committees, Series 10, Subseries 1 (Com-mittee on Membership 1961–1971), ff. 1962–64.Courtesy of the Oskar Diethelm Library, Institutefor the History of Psychiatry, Weill Cornell Med-ical College, New York.

Beck, A. T. (1963). Thinking and depression: I. Id-iosyncratic content and cognitive distortions. Ar-chives of General Psychiatry, 9 (October), 324–333.

Beck, A. T. (1964). Thinking and depression: II.Theory and therapy. Archives of General Psychi-atry, 10 (June), 561–571.

Beck, A. T. (1967). Depression: Clinical, experimen-tal and theoretical aspects. New York: Harper &Row.

Beck, A. T. (1970a). Cognitive therapy: Nature andrelation to behavior therapy. Behavior Therapy,1(1), 184–200.

Beck, A. T. (1970b). The core problem in depression:The cognitive triad. Science and Psychoanaly-sis, 17, 47–55.

Beck, A. T. (1972). The phenomena of depression: Asynthesis. In D. Offer and D. X. Freedman (Eds.),Modern psychiatry and clinical research: Essaysin honor of Roy R. Grinker, Sr. (pp. 136–158).New York: Basic Books.

Beck, A. T. (1979). Transcript of an interview withAaron T. Beck /Interviewer Bernard Baers, Ph. D.Personal collection, Dr. Aaron T. Beck.

Beck, A. T. (1984). Cognitive therapy, behavior ther-apy, psychoanalysis, and pharmacotherapy: Thecognitive continuum. In J. B. Williams and R. L.Spitzer (Eds.), Psychotherapy research: Where are

we and where should we go? (pp. 114–133). NewYork: Guilford Press.

Beck, A. T. (1990a). Transcript of an interview withAaron T. Beck/Interviewer Paul M. Salkovskis,November 3, 1990. Personal collection, Dr. AaronT. Beck.

Beck, A. T. (1990b). Transcript of an interview withAaron T. Beck/Interviewer Marjorie Weishaar,December 14, 1990. Personal Collection, Dr. Mar-jorie Weishaar.

Beck, A. T. (1991a). Cognitive therapy as the inte-grative therapy. Journal of Psychotherapy Integra-tion, 1, 191–198.

Beck, A. T. (1991b). Transcript of an interview withAaron T. Beck/Interviewer Marjorie Weishaar.Undated, 1990–1991. Personal collection, Dr.Marjorie Weishaar.

Beck, A. T. (1991c). Transcript of an interview withAaron T. Beck/ Interviewer Marjorie Weishaar,August 4, 1991. Personal collection, Dr. MarjorieWeishaar.

Beck, A. T. (1991d). Transcript of an interview withAaron T. Beck/ Interviewer Marjorie Weishaar,October 24, 1991. Personal Collection, Dr. Marjo-rie Weishaar.

Beck, A. T. (1993). Cognitive therapy: Past, presentand future. Journal of Consulting and ClinicalPsychology, 62, 194–198.

Beck, A. T. (1997). The past and future of cognitivetherapy. Unpublished manuscript, University ofPennsylvania School of Medicine, Philadelphia,PA.

Beck, A. T., Feshbach, S., & Legg, D. (1962). Theclinical utility of the digit symbol test. Journal ofConsulting Psychology, 26, 263–268.

Beck, A. T., & Hurvich, M. (1959). Psychologicalcorrelates of depression: I. Frequency of “masoch-istic” dream content in a private practice sample.Psychosomatic Medicine, 21(1), 50–55.

Beck, A. T., & Stein, M. (1961). Psychodynamics. InG. M. Piersol & E. L. Bortz (Eds.), Cyclopedia ofmedicine, surgery, specialties: Revision Service(Vol. XI, pp. 422C –422KK). Philadelphia, PA:F. A. Davis Company.

Beck, A. T., & Weishaar, M. (1989). Cognitive ther-apy. In A. Freeman, K. M. Simon, L. E. Beutler, &H. Arkowitz (Eds.), Comprehensive handbook ofcognitive therapy (pp. 21–36). New York: PlenumPress.

Beck, A. T. (n.d.) Personal Collection [Correspon-dence, notebooks, lecture and course notes, meet-ing notes, research reports, unpublished manu-scripts, drafts of manuscripts].

Clark, D. A., Beck, A. T., & Alford, B. A. (1999).Scientific foundations of cognitive theory and ther-apy of depression. New York: Wiley.

17PSYCHOANALYTIC ORIGIN STORY OF COGNITIVE THERAPY

This

docu

men

tisc

opyr

ight

edby

the

Am

eric

anPs

ycho

logi

calA

ssoc

iatio

nor

one

ofits

allie

dpu

blis

hers

.Th

isar

ticle

isin

tend

edso

lely

fort

hepe

rson

alus

eof

the

indi

vidu

alus

eran

dis

nott

obe

diss

emin

ated

broa

dly.

Epstein, S. (1994). Integration of the cognitive andthe psychodynamic unconscious. American Psy-chologist, 49, 709–724.

Freud, S. (1954). Project for a scientific psychology.In J. Strachey (Ed. & Trans.), The origins of psy-cho-analysis (pp. 347–445). New York: BasicBooks.

Freud, S. (1960). The ego and the id. J. Strachey(Ed.) & J. Riviere (Trans.), New York: Norton.[Original work published 1923.]

Galison, P. (2000). The suppressed drawing: PaulDirac’s hidden geometry. Representations, 72,145–166.

Gill, M. (1980). Rapaport, David. In Dictionary ofAmerican biographies (Suppl. #6, pp. 526–528).New York: Charles Scribner.

Harvey, O. J., Hunt, D. E., & Schroder, H. M. (1961).Conceptual systems and personality organization.New York: Wiley.

Knight, R. P. (1950). Report of the medical director.Stockbridge, MA: Austen Riggs Library.

Knight, R. P. (1951). Report of the medical director.Stockbridge, MA: Austen Riggs Library.

Knorr Cetina, K. (1999). Epistemic cultures: How thesciences make knowledge. Cambridge, MA: Har-vard University Press.

Lasker Foundation. (2006). Retrieved from www-.laskerfoundation.org/2006videoawards/qt_high/clinical/index

Latour, B. (1987). Science in action: How to followscientists and engineers through society. Cam-bridge, MA: Harvard University Press.

Lewin, K. (1938). The conceptual representation andthe measurement of psychological forces. Durham,NC: Duke University Press.

Nersessian, N. J. (2008). Creating scientific concepts.Cambridge, MA: MIT Press.

Piaget, J. (1948). The moral judgment of the child.Glencoe, IL: The Free Press.

Rapaport, D. (1951a). Organization and pathology ofthought: Selected sources. New York: ColumbiaUniversity Press.

Rapaport, D. (1951b). Transcript of Ego PsychologySeminar, Lecture 4. David Rapaport Papers (I: 63,ff: Ego Psychology Seminar Lecture #4). Wash-ington, DC: Library of Congress.

Rapaport, D. (1952, October 29). [Letter to RuthMonroe]. David Rapaport Papers (I: 63, ff: “Out-line of Ego Psychology”). Washington, DC: Li-brary of Congress.

Rapaport, D. (1967). The collected papers of DavidRapaport (M. M. Gill, Ed.), New York: BasicBooks.

Rosner, R. I. (1999). Between science and psycho-analysis: Aaron T. Beck and the emergence ofcognitive therapy. Unpublished doctoral disserta-tion, York University, Toronto, Canada.

Rosner, R. I. (2005). Psychotherapy research and theNational Institute of Mental Health, 1948–1980. InW. E. Pickren (Ed.), Psychology and the NationalInstitute of Mental Health (pp. 113–150). Wash-ington, DC: American Psychological Association.

Sigel, I. (1960). Some thoughts and findings aboutconsistency of cognitive style. Unpublished paper.Personal collection, A. T. Beck.

Sulloway, F. J. (1979). Freud, biologist of the mind:Beyond the psychoanalytic legend. New York: Ba-sic Books.

Weishaar, M. E. (1993). Aaron T. Beck. ThousandOaks: Sage.

Wheelis, A. (1951, January 19). [Memorandum to Dr.Howard, Dr. Schafer, Dr. White, Dr. Young, Dr.Miller, Mrs. Schafer & Dr. Beck]. David RapaportPapers (I: 63, ff: Ego Psychology Seminar Lecture#1). Library of Congress, Washington, DC.

Received September 22, 2010Revision received March 4, 2011

Accepted March 29, 2011 �

E-Mail Notification of Your Latest Issue Online!

Would you like to know when the next issue of your favorite APA journal will beavailable online? This service is now available to you. Sign up at http://notify.apa.org/ andyou will be notified by e-mail when issues of interest to you become available!

18 ROSNER

This

docu

men

tisc

opyr

ight

edby

the

Am

eric

anPs

ycho

logi

calA

ssoc

iatio

nor

one

ofits

allie

dpu

blis

hers

.Th

isar

ticle

isin

tend

edso

lely

fort

hepe

rson

alus

eof

the

indi

vidu

alus

eran

dis

nott

obe

diss

emin

ated

broa

dly.