Merton Gill, Psychotherapy, and Psychoanalysis: A Personal Dialogue

Robert R. Wallerstein

[Paper presented on June 10, 1995, at the Annual Meeting of the Rapaport-Klein Study Group]


Merton Gill occupied a unique place in my intellectual and emotional scheme of things, as close friend and colleague, as scientific ideal and moral gadfly, and as research mentor and supporter, although - despite mutual wishes that would have had it otherwise - we never worked collaboratively in the same setting. I first met Merton in the early 1950s - now more than four decades ago - a time when I, at the very start of my career as a psychotherapy researcher, was fashioning, together with some colleagues, what would evolve into the Psychotherapy Research Project of the Menninger Foundation, an ambitious thirty-year study of the processes and outcomes of both psychoanalysis and the psychoanalytic psychotherapies. At the time, we were applying to the Foundations' Fund for Research in Psychiatry for our first extramural funding.

Merton, whom I knew then only through his early writings on the nature of psychoanalysis and psychotherapy and on their relation (writings that had played a seminal role in the conceptual organization of our project), came as one of the Fund's site visitors. At our first meeting he was his characteristically incisive and lucid self, probing (often uncomfortably so) and at times acerbic, but at the same time always curiously respectful. After the visit none of us could be sure of Merton's take on the project, but in due course we were awarded the grant, which helped launch our project on its long-term career. And it was in that project that my own lifelong preoccupation began with these same issues of the psychoanalytic psychotherapies vis-à-vis their psychoanalytic parentage, issues that had already so engaged Merton's interest and to which his early contributions in 1951 and 1954 had already given an almost defining cast.

Over the next four decades, Merton and I, though we were engaged in a variety of other theoretical and technical concerns, each maintained an abiding interest in these issues, which have been a major preoccupation, both conceptually and technically, for the entire field of American psychoanalysis for half a century now, and yet remains today a still unsettled and ambiguous range of issues, theoretically and clinically, as will become clear enough over the course of this essay. My format in this tribute to Merton will be a close focus, via extensive quotation, on the development and evolution of Merton's thinking in this area over that time span, which will highlight his central role both in defining, in the earlier period, the emerging dimensions of these conceptualizations, and in formulating, in the later one, the sharp revisions arrived at by a significant portion of the field. My own views, both agreement and disagreement, will emerge via my organization of Merton's thinking and my responses to it, which include, in the early nineties, a published exchange of four letters seeking to clarify our respective understandings (Gill and Wallerstein, 1991). Hence the subtitle of my essay, "A Personal Dialogue."

Two comments are in order before we begin. First, a good number of analysts besides Merton Gill, most notably Robert Knight, Edward Bibring, Leo Rangell, Leo Stone, and Anna Freud, played major roles in formulating the consensus that emerged in American psychoanalysis during the late forties and early fifties regarding what constitutes psychoanalysis proper, what constitutes the range of psychoanalytic psychotherapies, and how their relationship is to be understood. The reaching of this consensus, crystallized in the contributions of these and other authors to a series of panels in the early fifties, and published together in one issue of the Journal of the American Psychoanalytic Association, is a story I have recounted at length elsewhere (Wallerstein 1989, 1995b). Here I will trace only Merton's role in these developments, and while this risks exaggerating his place in the overall picture, it is no exaggeration to place him only after Knight, who in the late forties first framed the conceptions denoting psychoanalytic psychotherapy as psychoanalytic but nonetheless distinct from psychoanalysis proper, and who, within the demarcated range of the psychoanalytic psychotherapies, first formulated authoritatively the proposed fundamental distinction between expressive and supportive therapeutic approaches. With the single exception of Knight, Merton Gill played a larger role than anyone else in delineating the similarities and differences between psychoanalysis and the derived psychoanalytic psychotherapies.

My second comment is to point out that I have long felt-though Merton never agreed, and indeed this was the focus of our 1991 exchange-that over the years Merton's evolving views on psychotherapy vis-à-vis psychoanalysis underwent a radical shift. To my mind, by the late seventies he had reversed himself almost totally, yet curiously his early and late views have been (and remain) almost equally influential. But that has also been a feature of Merton's influence on our field.

This is a paradox I noted in a 1984 appraisal of Merton's 1982 monograph Analysis of Transference. I began with an admiring paragraph locating him, and his influence in American psychoanalysis - in my view - as follows:

"It has been Merton Gill's fate to be an acknowledged leader in (American) psychoanalysis almost from the beginning of his career. An uncommon percentage of his writings have been almost instantly - and justly - hailed for their critical influence at particular points in the history of analysis in America, in relation both to psychoanalytic clinical practice (and its relation to the derivative dynamic psychotherapies) and to psychoanalytic theory formulation (in the ascendancy and then. the decline of ego psychology - the apotheosis of metapsychology)" (Wallerstein, 1984, p. 325).

I then went on to cite a succession of Merton's contributions which I felt substantiated this assertion, ending the paragraph with this final sentence:

"Curiously, the signal influence of each of these successive landmark contributions has endured despite Gill's own subsequent turning away from and even disavowing positions that he had so persuasively earlier espoused" (Wallerstein, 1984, p. 326).

Merton did indeed acknowledge these profound changes of mind readily in regard to first his impassioned espousal and later his equally passionate repudiation of the ego psychological metapsychology paradigm, saying once to me, regarding his influential 1963 monograph Topography and Systems in Psychoanalytic Theory, that he disavowed every word in it and wished he had never written it. He by no means acknowledged, however, what I felt to be an equally great turnabout in his views on the nature of the psychoanalytic therapeutic enterprise-but that is part of the unfolding story to which I now turn.

Gill's Early Views

Gill's first article on this subject, published in 1951, came in the wake of Knight's seminal formulations published in the 1940s. Gill began with a revision, in the light of those formulations, of Freud's definition of psychoanalysis as stated in a well-known passage in On the History of the Psychoanalytic Movement (1914). There Freud had stated, of "the facts of transference and resistance," that

"any line of investigation which recognizes these two facts and takes them as the starting-point of its work has a right to call itself psycho-analysis even though it arrives at results other than my own. But anyone who takes up other sides of the problem while avoiding these two hypotheses will hardly escape a charge of misappropriation of property by attempted impersonation, if he persists in calling himself a psycho-analyst" (Freud, 1914, p. 16).

Gills modification was as follows:

"In point of fact the designation 'psychoanalysis' is reserved for that technique which analyzes transference and resistance. Psychoanalytic therapy is any procedure which recognizes transference and resistance and rationally utilizes this recognition in the therapy, though this may be done in many different ways, and part or all of the transference and resistance may not be analyzed" (Gill, 1951, p. 62).

It is these "many different ways," in which "part or all of the transference and resistance may not be analyzed," which represent the various distinctions made among the expressive and supportive modes delineated originally by Knight. Gill stated these differences in terms of goals and strategies. Of the first he said,

"In psychoanalysis the goal is relatively clear: a progressive analysis from the surface to the depth; analysis of the defenses and the motives for defense; the development and analysis of the transference neurosis; a resolution of symptoms, and as complete a 'structural' alteration of the neurotic aspects of the personality as possible. In psychotherapy the goal may be anything from as quick relief of a symptom as possible, with the restoration of the previous integrative capacity of the ego, through a whole range of more ambitious goals up to analysis, the most ambitious of all. The choice of therapy may be divided into that which determines the minimum necessary to restore the ego to functioning, and that which strives for the maximum change that is possible" (Gill, 1951, p. 63; emphasis added).

In terms of strategy, Gill said,

"The gross major decision is whether the defenses of the ego are to be strengthened or broken through as a preliminary toward a reintegration of the ego… The decision to strengthen the defenses is made in cases in which this is all that is necessary, or in those in which this is all that is safely possible" (Gill, 1951, p. 63; emphasis added).

At the other end of the spectrum,

"Analysis then is clearly the procedure for a middle range in which the ego is sufficiently damaged that extensive repair is necessary, but sufficiently strong to withstand pressure" (Gill, 1951, pp. 63-64; emphasis added).

And then determinedly at the end, in connection with the question of the inevitability and necessity of the transference neurosis and Freud's famous statement that an enemy cannot be slain in effigy, Gill stated,

"Psychotherapy must determine how thoroughgoing an alteration in personality can be achieved by only partial development of the transference neurosis" (Gill, 1951, p. 70).

Of course, Gill also made room for common enough admixtures of technique:

"While the two poles of either strengthening the defenses, or of analyzing them as first steps toward reintegrating the damaged ego, stand as the gross opposites of two theoretical modes of approach, the psychotherapy of any specific case will show intricate admixtures of both" (Gill, 1951, p. 65).

And finally, in this rather brief paper, Gill added to the sophistication of the conceptualization of supportive techniques through his elaboration of three different ways of "strengthening the defenses":

"A first principle, then, for techniques of strengthening defenses is to encourage, praise, or in general to give narcissistic support for those ego activities in which defense is combined with adaptive gratifications [e.g., compulsive hard work], and to discourage by subtle or direct techniques those activities which are maladaptive gratifications, whether or not they are combined with defense (p. 66). A second principle ... is that one must take great care not to unwittingly attack an important defense [e.g., the denial of dependency longings] (p. 66). [The third] way in which the theory of strengthening the defenses may be formulated is the one proposed by Glover. He suggests that artificial neuroses may be set up which offer a partial discharge for derivatives of instinct... For the theory of the results achieved by such a technique Fenichel suggests that 'by this partial discharge the instinct becomes relatively weaker and the work of defense against the remainder becomes easier.' The defense may then be said to have been relatively strengthened" (Gill, 1951, pp. 66-67).

This 1951 delineation of the modi operandi of three specific supportive techniques was the first such psychoanalytic conceptualization. The simplistic assumption prevailing to that point had been that support meant no more than varieties of reassurance, encouragement, etc. - this despite the clear knowledge that simple reassurance is often not at all reassuring.

Gill's next paper on the subject is his best known, a contribution to the debates published in the Journal of the American Psychoanalytic Association in 1954. The impetus for this concentrated outpouring of articles on a single theme was severalfold. An important aspect was the intense ferment stirred up within psychoanalysis by the bold technical proposals of Alexander et al. (1946), proposals that met a wide range of reaction, from great enthusiasm for having pointed the way to more effective and speedier psychoanalytic: treatment, to severe dismay that the dilution of proper psychoanalysis into an amorphous psychotherapy, all the more dangerous because it masqueraded as real analysis and so could hoodwink the uninformed, was under way, perhaps irreversibly. The latter development was regarded as the logical but extremely unhappy outcome of the growing closeness in America between psychoanalysis and psychiatry. Alexander himself participated in two of the panels, as did Frieda Fromm-Reichmann and Edith Weigert, who held similar if not identical positions. At the same time, a substantial portion of the presentations of those with opposite views (Gill, Rangell, Stone) was devoted to a rebuttal of the Alexander innovations, as no longer psychoanalytic (though possibly effectively psychotherapeutic), and to a redefining of Alexander's (1946) concept of the corrective emotional experience to bring it more into line with properly psychoanalytic precepts.

An even more encompassing impetus to the publications of 1954 was the growing controversy over how the similarities and differences between psychoanalysis and psychotherapy were to be understood. Here the central confrontation occurred between two major viewpoints on the most appropriate way to conceptualize the relationship between dynamic psychotherapy and psychoanalysis proper. Basically, the issue lay between the position advanced by Alexander, Fromm-Reichmann, and their colleagues (but altogether a distinct minority), who saw the historical trend as blurring, if not ultimately altogether obliterating, the technical distinctions between dynamic psychotherapy and psychoanalysis, and the viewpoint espoused by the great majority of analysts, for whom Bibring, Gill, Rangell, and Stone served as major spokesmen. The latter conceived the scientific problem to be the clarification and sharpening (and thereby the preservation) of the conceptual and operational distinctions between the two. It was these diametrically opposed views that clashed at the panels published in 1954. And of all the contributions to those debates, it was Merton Gill's that emerged as the most incisive, the one that most clearly defined the dimensions of the emerging majority consensus on the relation of psychotherapy to psychoanalysis-that of an established diversity of goals and techniques (the spectrum of psychoanalytically based psychotherapies, ranging from the most supportive to the most expressive, psychoanalysis itself) within a unity of theory (that of psychoanalysis). He thereby ushered in an era of consensus that lasted for approximately twenty years, though with increasing strain over that time span.

It must be remembered that Gill's response to Alexander's technical innovations, particularly the concept of the corrective emotional experience, was written from the ego psychological standpoint of the 1950s, considerably before the evolution of his own later thinking in terms of an interactive "two-body psychology" wherein the transference was reconceptualized as a plausible construction by the patient of the ongoing analytic interaction. This caveat in mind, let us see how the Merton Gill of 1954 tackled the conceptual issue:

"there is little doubt that Alexander is correct in stating that by overt behavior toward the patient one can more quickly get him to change some aspects of his behavior. But what is the meaning of such a change? It is an adaptation to this particular interpersonal relationship-as it exists between patient and analyst. But this is not the goal of analysis. The goal of analysis is an intrapsychic modification in the patient, so that for example his dependent behavior is given up not because he has learned that if he acts too dependent he will be punished by a loss of therapy hours, but because despite the invitation to regress and the maintenance of the frequency of his hours he has come to feel and understand his dependency in such a way that he no longer needs it or wants it-and that this is a conclusion valid not simply for this particular interpersonal relationship but has more general applicability, in short has the status of an intrapsychic change" (Gill, 1954, p. 781).

Though Gill did not state it explicitly, his conclusion was clear: Alexander's innovations had taken him beyond the realm of proper analysis. Despite his claim that they were merely part of a more effective analysis, one better geared to the needs and requirements of his patients, they were to Gill's mind a variety of psychoanalytically oriented psychotherapy. Though granting that such a therapy could well be appropriate, depending on the clinical circumstances, Gill deemed it important to maintain the distinction. Gill ended his discussion of the issue by redefining what he considered a true corrective emotional experience:

"Certainly to meet the patient's transference behavior with neutrality is to give him a corrective emotional experience without the risks attendant on taking a role opposite to that which he expects, as Alexander suggests" (Gill, 1954, p. 782).

But the larger part of Gill's essay was focused on the central issue of the overall similarities and differences between psychoanalysis and the derived psychoanalytic: psychotherapies, all from the perspective of conceptual clarification and a sharpening of distinctions. Gill began by offering a definition of psychoanalysis, the parent therapy, a succinct statement that came to enjoy very wide currency: "Psychoanalysis is that technique which, employed by a neutral analyst, results in the development of a regressive transference neurosis and the ultimate resolution of this neurosis by techniques of interpretation alone" (Gill, 1954, p. 775). He went on for several pages to explicate each phrase of this sentence in extended detail.

Clearly, this definition circumscribes psychoanalysis far more precisely than did Freud's definition of 1914, which stated simply that any therapy that recognizes the two facts of transference and resistance and takes them as the starting point of its work can call itself psychoanalysis. It is also consistent with Gills earlier modification (1951, p. 62) of Freud's definition whereby Gill drew a distinction between psychoanalysis (which analyzes transference and resistance) and psychoanalytic psychotherapy (which recognizes transference and resistance and rationally utilizes this recognition in the therapy), though that might be done "in many different ways."

In the 1951 paper, Gill (following Knight) bad already made the point that the "gross major decision is whether the defenses of the ego are to be strengthened or broken through as a preliminary toward a reintegration of the ego" (p. 63). In 1954 he elaborated on this idea. "Methods of psychotherapy…" he wrote, "are primarily supportive or primarily exploratory, with all grades in between… Exploratory psychotherapy may be brief or it may be long, even longer than the usual psychoanalysis" (p. 772). He focused in the later paper on this exploratory (or expressive) psychotherapy as an entity on the spectrum of therapies "in between" supportive therapy and analysis proper.

This spectrum he arrayed as follows. A primarily supportive therapy is the preferred mode for patients too ill for analysis (or for the expressive therapies). These are patients whose egos are not strong enough to withstand pressure, whose failed psychic equilibrium is to be restored via that "strengthening of defenses," to be carried out through a variety of supportive techniques spelled out by Gill, Knight, and others. "A major contraindication" to expressive therapies with such patients "is the presence of severe regressive factors or the danger that such may develop. The instances of rather sudden onset of psychosis shortly after beginning psychoanalysis are to be attributed to the regressive pressure of the technique per se in a precariously balanced personality" (p. 780). A primarily expressive therapy is the preferred mode for patients whose illnesses are less severe-circumscribed neurotic problems, acute reactive disorders - or who need help negotiating transitional life states. Their egos, not unduly deformed, can tolerate efforts at "analyzing the defenses" to the extent necessary, via established methods of interpretation leading to insight. These are patients on the other side, neurotically ill but "too well" for psychoanalysis, in the sense that they neither require nor warrant a treatment so ambitious and far-reaching.

This way of conceptualizing the spectrum of therapeutic modes and their differential indications puts expressive psychotherapy exactly in that "intermediate" position-certainly as regards technique-between supportive therapy and psychoanalysis. Two quotations from the 1954 paper indicate the delicate balancing Gill had in mind.

"In contrast to our analyst…," he wrote, "the psychotherapist is willing to permit many transference manifestations to remain unresolved on the one hand and on the other to behave in ways which would make it more difficult to resolve if he were to attempt it" (Gill, 1954, p. 784). But he also believed that "we have failed to carry over into our psychotherapy enough of the non-directive spirit of our analyses" (p. 786). "My stress on this point," he added, "arises from my feeling that discussion of therapeutic results in psychoanalysis and psychotherapy too often views them as polar opposites, with psychoanalysis regarded as producing structural changes, and psychotherapy as unable to produce any significant intrapsychic change, but only altering techniques of adjustment through transference effects and shifts in defensive techniques" (p. 787). By way of further explication, he raised the question "whether there is not more to be said on results and mechanisms in prolonged psychotherapy with more ambitious goals by a relatively inactive therapist and in intensive [expressive] psychotherapy" (p. 789).

In this connection, Gill made the statement, surprising for the time, that after all, psychoanalysis was not such a totally definitive therapeutic reconstruction as its theory - as then understood - perhaps promised:

"It is generally," he wrote, "the more experienced analysts who are not so optimistic about the sweeping character changes often hoped for from psychoanalysis. And there is no doubt that we can still recognize our friends and colleagues, even after they have been analyzed. In Analysis Terminable and Interminable, Freud quotes wryly the observation that 'Every advance is only half as great as it looks at first'" (Gill, 1954, pp. 786-787).

Given this, Gill could more readily see the similarities and closenesses (in goals, methods, and results, though with real differences) between psychoanalysis and the expressive psychotherapy that he now called an "intermediate type of psychotherapy":

"What we might call more intermediate types of psychotherapy are relatively recent. This is the psychotherapy done by people analytically oriented or trained whose goals are intermediate between rapid symptom resolution [i.e., supportive psychotherapy] and character change [i.e., psychoanalysis], where techniques are in a sense intermediate - for example, relative neutrality and inactivity; transference dealt with, though not a full regressive transference neurosis; interpretation the principal vehicle of therapist behavior - and, I suggest, where results are likewise intermediate" (Gill, 1954, p. 789).

By this, he wrote,

"I wish not to be misunderstood. I am not suggesting that psychotherapy can do what psychoanalysis can do; but I am suggesting that a description of the results of intensive psychotherapy may be not merely in terms of shifts of defense but also in terms of other intra-ego alterations" (Gill, 1954, p. 793).

This theme he emphasized again in his final summarizing statement:

"I have tried to say that techniques and results in psychoanalysis and intensive, relatively non-directive psychotherapy are not the polar opposites which they are often declared to be, and that a more positive and detailed description of changes both in psychoanalysis and in psychotherapy which will take account of our newer formulations in ego psychology and include descriptions in terms of intrapsychic alterations and techniques of adaptation, will help to make this clear" (Gill, 1954, pp. 795-796).

But, of course, Gill's balancing caveat, already quoted, is apposite here, that "the psychotherapist is willing to permit many transference manifestations to remain unresolved… and… to behave in ways which would make it more difficult to resolve if he were to attempt it" (p. 784).

All of this adds up to the statement that in contrast to the maximalist goals of psychoanalysis in terms of fundamental conflict resolution and character reorganization, the goals in the other psychotherapies range from the most minimal to the increasingly more ambitious, with no agreement as to how asymptotically close to those of psychoanalysis they can reach in the different kinds of patients treated psychotherapeutically. In 1954, Gill put it similarly to the way he had in 1951:

"The goals of psychotherapy extend over a very wide range. To take first the goals of psychotherapy with a relatively strong ego. The goal may be the resolution of a crisis, assistance through a troubled period, or symptom amelioration.... But goals may range up to more ambitious aims in cases where there is no pressing problem, but where psychoanalysis is impossible or not used for external reasons ... goals are much more ambitious than in palliative psychotherapy, and ... more important results are achieved than I believe is often admitted" (Gill, 1954, p. 785).

Here is the full exposition of Merton Gills perspectives in the 1950s, upon psychoanalysis and the dynamic psychotherapies and the clear distinctions posited between them, even while acknowledging all the areas of relatedness and of overlap. However old-fashioned or even rigidly mechanical these early views of Gill may seem when placed against the more complexly interactive and fluid relational conceptions and two-body perspectives of today, it should be remembered that they prevailed as accepted doctrine on these issues, despite their gradual erosion, well into the 1970s, and that in somewhat muted form they undergird much of the thinking and teaching on psychotherapy vis-à-vis psychoanalysis even. today, and that Merton Gill more than any other individual was their architect.

The 1979 Revision

Gill had not revisited this specific subject for a quarter century, when in 1979 the Southern Region Psychoanalytic Societies sponsored a symposium in Atianta to which three of the central protagonists in the 1954 debates - Gill, Rangell, and Stone - were invited to update their views in the light of all the changes that had occurred within psychoanalysis in the intervening period. Many of these changes bore clearly on issues of psychotherapy vis-à-vis psychoanalysis and called out for expression at some point in the evolving discourse around these issues. I refer here to such trends as (1) the delineation of the therapeutic or working alliance as distinct from the transference evolution, with a postulated interactive relationship between the two; (2) a growing focus on the affectively charged therapeutic relationship as a coequal and interactive determinant (together with veridical interpretation leading to working through and insight) in bringing about therapeutic change; (3) a growing shift from a "one-body" to a "two-body" psychology, with an increased focus on an interactional, interpersonal, and subjectivist paradigm explanatory of the transference-countertransference interplay within the analytic situation; (4) the increasing acknowledgment and acceptance of theoretical diversity or pluralism within organized psychoanalysis worldwide, a situation in which the question inevitably arises of where the boundaries between psychoanalysis and the derived psychotherapies can properly be drawn, and in which one person's psychoanalysis can readily become another's psychotherapy; and (5) the new diagnostic and therapeutic conceptualizations being advanced in regard to the widening scope of patients appearing in psychoanalytic consulting rooms (i.e., the borderline personality organizations and the narcissistic personality disorders), the conceptual and technical considerations being proposed for their therapy, and the extent to Which these should be considered still psychoanalytic or clearly psychotherapeutic.

Gill, Rangell, and Stone had essentially been in full agreement during the debates of the early fifties; they had been representatives of the "classical" position, then the majority view, that posited a continuous spectrum of psychotherapies-one marked, however, by an articulated and distinctive crystallization of differentiated therapeutic modalities along its length. These ran from psychoanalysis through expressive psychotherapy (the "intermediate therapy") to supportive psychotherapy, each with its specific and differentiated applicability to a particular nosological segment (reasonably coherent) of patients specifically amenable to it. It is this appearance at the Atlanta panel of three contributors who, a quarter of a century earlier, had spoken with so united a voice, that rendered all the more impressive the significant divergence of views - on precisely the same questions and issues - that marked their discourse in 1979.

Of the three, I believe, the one whose views had shifted most radically was Merton Gill (though he did not agree with this assessment). Gill, following upon Knight, had been clearest in the early fifties in his distinct delineation of psychoanalysis, expressive therapy, and supportive therapy, each with distinctive characteristics, different techniques, and different goals, and each differentially indicated for a different segment of the psychopathological spectrum. The very radical shift that I detect in his views has followed - to my mind - as a direct consequence of his preoccupation over this same period with his evolving conception of the overriding primacy of the interpretation of the transference as the criterion of psychoanalysis and of what is psychoanalytic. These newer views on the nature and piace of transference interpretation include a distinction between resistance to the awareness of transference and to its resolution; an insistence on the earliest possible interpretation of the transference, including a searching out of all possible implicit transference allusions and making them explicit from the very start of the treatment; a focus on the here and now as against the genetic thrust in the transference interpretation; and the need to elaborate all the implications of the two-person as against the one-person view of the two participants' contributions to the transference. All of this was presented in detail by Gill in a succession of publications (1979, 1982, 1984; Gill and Hoffman, 1982; Gill and Muslin, 1976) and is discussed by me elsewhere (Wallerstein, 1984). Here I will pursue only the implications of these views for my present theme - Gill's now very significantly altered ideas on the nature and relationship of psychoanalysis and psychotherapy. These implications he made specific in his contribution to the 1979 symposium, published in a revised version in 1984 as "Psychoanalysis and Psychotherapy: A Revision."

Gill developed these new views with his customary logical precision, summarizing his 1954 paper in a single sentence: its thrust "was to insist on the difference [between psychoanalysis and psychotherapy] and at the same time to recognize that the two are on a continuum" (Gill, 1984, p. 162). He then reviewed both the "intrinsic criteria" by which analysis is customarily defined - "the centrality of the analysis of transference, a neutral analyst, the induction of a regressive transference neurosis and the resolution of that neurosis by techniques of interpretation alone, or at least mainly by interpretation" (p. 161) - all strikingly consistent with his 1954 definition, and the commonly recognized "external criteria" or markers - "frequent sessions, the couch, a relatively well integrated patient, that is, one who is considered analyzable, and a fully trained psychoanalyst" (p. 161). It was then that he came to his taking-off point:

"The question of the relationship between psychoanalysis and psychotherapy is even more important in practice today than it was in 1954 because of the practical difficulties in maintaining the ordinarily accepted extrinsic criteria of analysis... The question becomes: How widely can. the range of extrinsic criteria be expanded before the analyst must decide for psychotherapy rather than psychoanalysis?" (Gill, 1984, p. 162).

After reviewing the sociocultural and economic difficulties that analysts face today in maintaining the external criteria of psychoanalysis, Gill went on:

"The changes I will propose are more radical than a simple extension of the recommendation I made in 1954 that we carry more of the non-directive spirit of psychoanalysis into our psychotherapies" (p. 163). Rather, "I will argue that with the definition of analytic technique at which I will finally arrive, it should be taught to all psychotherapists and that how well it will be employed will depend on their training and natural talent for the work" (p. 163). And, "I mean that analytic technique as I will define it should be employed as much as possible even if the patient comes less frequently than is usual in psychoanalysis, uses the chair rather than the couch, is not necessarily committed to a treatment of relatively long duration, is sicker than the usually considered analyzable patient and even if the therapist is relatively inexperienced. In other words, I will recommend that we sharply narrow the indications for psychoanalytic psychotherapy and primarily practice psychoanalysis as I shall define it instead" (Gill, 1984, p. 163).

What we see here, of course, is a proposal to assimilate to psychoanalysis what Gill in 1954 had taken such pains to demarcate as the related but nonetheless distinct arena of expressive psychotherapy - that "intermediate form of therapy" - or, in other words, to now blur, perhaps obliterate altogether, the differences he had once felt it so vital to maintain. To me, this is in effect a revival of the position, held by Alexander and his followers, that Gill had once led the effort (largely successful) to reject.

Gill clearly acknowledged the basis for these radically altered views: "The recommendation I am proposing is an outgrowth of my changed views on transference and its analysis…" (p. 164). He then went on for ten pages. reviewing his newer views on the overriding primacy in psychoanalytic work of transference interpretations in the here and now and ending with a summarizing statement:

"My reconceptualization of these intrinsic criteria of analysis, namely transference and its analysis, the neutral analyst, the regressive transference neurosis, free association, and the role of experience in addition to interpretation lead me to the conclusion that the centrality of the analysis of transference, as I have defined transference... is alone the distinguishing characteristic of analytic technique… it is the distinguishing feature of psychoanalysis. It is what distinguishes it from psychotherapy. It remains for me to try to show that it can be maintained even in an expanded range of external criteria" (p. 172; emphasis added). "My changed conceptions of the intrinsic criteria of analysis, some beginning efforts to apply them, and experience in teaching them to therapists of varied experience [have] led me to believe that the attempt to use analytic technique in this broader range of circumstances is justified" (Gill, 1984, p. 173).

Gill then went on to develop his idea that even a necessary frequency or use of the couch may be dispensable in proper psychoanalytic work - depending on the patient and, implicitly, the patient-analyst dyad:

"While the couch is ordinarily considered to be conducive to regression, it may enable an isolation from the relationship which has a contrary effect. No universal meaning of any aspect of the analytic setting may be taken for granted. It follows that no universal prescription can be given for this or that type of case. One may generalize that analytic work goes better with healthier patients lying down and sicker patients sitting up and with frequent sessions for both kinds of patients but a particular patient may not conform to the rule. The meaning of the setting must be analyzed in each instance. Nor is degree of pathology the only variable which determines a patients response to the analysis of transference. Apart from pathology, some take to it like a duck to water and can work despite infrequent sessions, while others never seem to find it congenial" (Gill, 1984, p. 174).

Gill, then, dealt specifically with the impact of each of the usual external parameters of the psychoanalytic situation. As for frequency,

"It would seem obvious that one can accomplish more with greater frequency simply because there is more time to work. But if greater frequency is frightening to a particular patient, frequent sessions may impede the work despite interpretation. One cannot simply assume that more is better. The optimal frequency may differ from patient to patient. We must not confuse optimal frequency with obligatory ritualized frequency" (Gill, 1984, p. 174).

As for the couch,

"The argument that psychoanalytic technique cannot be used in the face-to-face position because of the reality cues afforded the patient is a variant of the mistaken idea that it is the external stimulus rather than the patients interpretation of that stimulus that matters. Discussions of this issue seem to gravitate to quantitative terms as though there are fewer stimuli if the patient cannot see the analyst. It may be that the patient facing the analyst is exposed to a wider range of stimuli but the patients response is to the quality not the quantity of stimuli" (Gill, 1984, p. 175).

As for the experience of the therapist,

"Therapists use the technique to whatever degree their skill and comfort in exposing the transferential experience of the relationship permits" (Gill, 1984, p. 175). "Finally, as for the relatively long duration and open-endedness of analytic work, I do not know the lower limit in duration for the use of the technique I advocate, but I have used it successfully in once-a-week therapy preset to last no longer than nine months" (p. 175).

Gill declared that central to the views he was challenging was an

"implicit assumption about the psychoanalytic process which I would like to question. It is that analysis is a kind of all-or-none proposition, yielding its positive results only if carried through to the end. It is this belief which may sustain patient and analyst through long periods of apparent stagnation and stalemate, but this belief is often a vain illusion. Freud compared interrupting an analysis to the interruption of a surgical operation. I suggest, on the contrary, that in the changed way of conducting it which I am proposing, analysis may be a process with progressively cumulative benefits, interruptible at various points without necessary loss of what has been gained" (Gill, 1984, p. 176).

Surely this characterization of the prevailing view of psychoanalysis as an all-or-nothing endeavor is something of a straw man, since analysis can assuredly rest or stop at all stages of incompleteness, quite aside from the understanding that in theory analysis is never competed in any but an asymptotic sense. None of this, however - and this is my point - is the same as assimilating every expressive-interpretive psychotherapy to psychoanalysis, as Gill has here proposed we do, his only stipulation being that the therapeutic thrust be focused unswervingly on interpretation of the transferences (implicit as well as explicit) in the here and now.

In this entire uncompromising statement of his radically altered perspective on the relation of dynamic psychotherapy to psychoanalysis, Gill made only one concession to the generally prevailing views he was challenging. At the very beginning of his article he had written,

"Although I will propose that psychoanalysis… as I define it is applicable across the whole range of psychopathology, my convictions are strongest for its application to patients ordinarily considered analyzable for whom issues of time and money preclude the usual setting of analysis" (Gill, 1984, p. 161).

Where, then, would all this leave the issue of psychotherapy vis-à-vis psychoanalysis? Gill posed the question this way: "Should the use of the technique I describe in less than the optimal setting for an analysis still be called psychoanalysis?" (Gill, 1984, p. 176). He obviously felt that the answer to this query was yes, though he did say that "other things being equal, obviously an analysis conducted at lesser frequency cannot accomplish what otherwise could be" and that "there ought to be different names for an analysis carried through as fully as it could be and one which is partial and incomplete" (p. 176). The only alteration of name that he proposed, however, was of "psychoanalytic psychotherapy," which he now said "should be reserved for a technique which does not deal with the transference in the way I have suggested is the essential criterion of analytic technique" (p. 176; emphasis added). In other words, "psychoanalytic psychotherapy" should be the designation for that therapeutic approach that is not psychoanalysis (i.e., that does not systematically interpret transference). That is what Knight originally, and Gill after him, called in the 1940s and 1950s psychoanalytically oriented supportive psychotherapy. And, in what can be considered a further retrograde step, there is more than a passing allusion to the idea that this psychotherapy that is not psychoanalysis rests fundamentally on "witting and unwitting suggestion" (p. 177) - a return to what I have elsewhere called the first era in the developmental history of dynamic psychotherapy in relation to psychoanalysis, the era of the simple assumption of Freud (and of Jones and Glover) that the sole operative mechanism in all therapy other than "strict and untendentious psychoanalysis" is basically hypnosis or suggestion.

A 1991 Exchange of Views

I have presented these perspectives on Gills early and late views within the context of an historical overview of the relation of psychotherapy to psychoanalysis that I published in 1989 and subsequently expanded in my book on that topic (1995b). Merton, in a letter to the editor of the journal in which my article appeared, took serious exception to my representation of his views. That initiated our public dialogue on our differing assessments of his new position, a dialogue we had already been privately engaged in over a considerable length of time. Gill's letter (Gill and Wallerstein, 1991) began with the statement that my assessment of his views in my 1989 article, "however well intentioned, is a serious misrepresentation" (p. 159); he asked, a bit further on, "How can as thoughtful, conscientious, and well-versed a man as Wallerstein go so far astray?" (p. 159). My seeming misunderstanding became clearest to Gill when I compared his revised views to Alexander's. He then granted that indeed there was a similarity to Alexander, in that both of them espoused a two-person view of the analytic situation, but insisted that there was also a significant difference, in that Alexander advocated the manipulation of the interaction (the corrective emotional experience), while he, Gill, advocated interpreting the interaction as fully as possible. He then said that I had presented neither the similarity nor the difference correctly. The reason he gave for this putative misunderstanding was his feeling that I did not really appreciate the analytic situation to be interpersonal and interactional, despite the fact that I might say I do. From there he went on to say that my view of the analytic situation remained essentially within the framework of a one-body psychology, and that though I might regard psychotherapy as interactional, I did not view psychoanalysis as such, however beset it might be by unavoidable countertransferences and "infiltrations." This meant, according to Gill, that I regarded "interaction in analysis as advertent and inadvertent infiltration, not as part and parcel of the situation" (p. 159).

Asked by the editor to respond to Gill's letter, I began my response (Gill and Wallerstein, 1991) with the statement,

"Despite the strenuous efforts that both Merton Gill and I characteristically make towards the utmost of clarity in the presentations of our views and despite the mutual goodwill and serious intention that we bring to understanding the nature of our agreements and our disagreements, it seems that we each feel that the other seriously misunderstands and misrepresents his viewpoint. Let me explain" (Wallerstein, in Gill and Wallerstein, 1991, p. 161).

I then said that I agreed fully with Gill's description of both the similarity and the very significant difference between his position and that of Alexander; where the latter would manipulate the interpersonal interaction in analysis, Gill would interpret it. However, in my linking of Gill's current views with Alexander's, I had been referring neither to the particular similarity nor to the particular difference that Gill described in his letter. I was referring rather to the efforts of Alexander in the fifties and of Gill in 1979 (completely contra to Gill in the fifties) to array both expressive (or exploratory) psychotherapy and psychoanalysis proper along a continuum in which differences are merely quantitative and all distinctions based on the usual criteria for psychoanalysis are declared not critical. Alexander had in fact called for the "unification" of psychoanalysis and dynamic psychotherapy (see Wallerstein, 1989, p. 569), a position Gill had strongly opposed in 1954 but now, in a complete reversal of his earlier view, espoused a quarter century later. Gill's letter had clearly not convinced me that I had misread him. When I called his position on this issue a return to the position of Alexander, which in 1954 Gill had led the effort to counter, I had in no way tried to equate Gill's views with Alexander's on any other salient matter. Nor, incidentally, did I see any contradiction between this assessment of Gill's position (and Alexander's) - as blurring if not completely obliterating any distinction between expressive psychotherapy and psychoanalysis - and my other assessment that on the issue of the nature of supportive psychotherapy Gill had reverted to the early position of Freud (and of Jones and Glover), viewing it essentially as nothing more than simple suggestion.

I also addressed Gill's claim that I viewed psychotherapy as interactional but stili adhered to a one-person view of transference that would render psychoanalysis itself non-interactional. My counterclaim was that I considered both modalities interactional and that that was not where I located their essential difference. Rather, to quote my letter,

"My view is that I see psychoanalysis simply as the therapy that rests centrally (though not exclusively) on… the effort to systematically analyze, in so far as one can, including analyzing: the psychoanalytic interaction and all the suggestive elements that enter into it. I see psychoanalytic psychotherapy (varyingly supportive and expressive) as resting partly, where possible, on an interpretive-analytic base, but also, in varying degrees, depending on the characteristics and needs of the particular patient, on many other kinds of (supportive) technical interventions-which I have adumbrated and illustrated at great length in my book, Forty-Two Lives in Treatment (1986)" (Wallerstein, in Gill and Wallerstein, 1991, p. 163).

That is, I view both psychoanalysis and analytic psychotherapy as interactional; I see the difference in the comprehensiveness of the effort to analyze (i.e., interpret) the interactions and in the elaboration of a wide array of kinds of interaction (called in my book supportive interventions), rather than subsuming them all under the single rubric of suggestion.

When Merton received my response to his letter, he wrote to the editor-and to me-asking the special dispensation that he be allowed to respond to my response, with the understanding that I would in turn be able to respond to that; at that point the exchange would come to a close, with no further effort on his side, and the four letters would then be published. This was agreed to. Gill began his second, briefer letter with the statement that the first exchange had indeed clarified where we agreed and disagreed. He further expressed his satisfaction that I did adhere to "the centrality of the interpersonal aspect of the psychoanalytic situation" (p. 164). He also acknowledged that he was using the term suggestion far more broadly than 1, meaning by it "any interpersonal influence whether direct or indirect" (p. 164; emphasis added). Where he correctly thought we disagreed was in his feeling that he was sharpening the distinction between psychoanalysis and all other psychotherapy by insisting that it is always possible to readily and simply distinguish psychoanalysis and psychotherapy, as follows: "If the therapist is attempting to deal with the analysis of the transference as completely and thoroughly as he can, the therapy is analysis. If he is not it is psychotherapy" (p. 164). He said my confusion in this matter lay in the idea that his "belief that psychoanalytic technique can be done at a lesser frequency and sitting up" necessarily allied him with Alexander. "I have attempted," he said, "to clarify that these are extrinsic, not intrinsic criteria, and I have insisted that analysis must be defined in terms of its intrinsic criteria" (p. 165).

My equally brief response to this last letter agreed that indeed much had been clarified about our respective views. After indicating the areas in which we were in full accord, on the centrality of the interpersonal aspects of the psychoanalytic situation, and on the fact that I never intended to impute to him Alexander's 1946 conception of the corrective emotional experience, I then stated the two main issues on which we agreed that we disagreed. The first was my feeling that he had indeed returned to Alexander's earlier efforts to assimilate expressive psychotherapy to psychoanalysis:

"The nub of our difference is Gill's current belief that any therapy that tries to analyze the transference as completely as possible is psychoanalysis, no matter what the conditions or dimensions of the treatment (what he calls 'extrinsic criteria'). I believe that it can only be analysis if the conditions of treatment are such as to properly facilitate and enable a psychoanalytic: process to unfold… Otherwise it is an expressive psychoanalytic psychotherapy in which transferences can indeed be analyzed to the extent possible under less-than-optimal conditions" (Wallerstein, in Gill and Wallerstein, 1991, pp. 165-166).

My other disagreement was with Gill's predilection for using the term suggestion more broadly than I considered useful. i felt that this return to the usage of Freud conflates varieties of distinctive interventions under one general rubric that diffuses out specific meanings and differentiated understandings (as well as guides to interventions).

After these 1991 letters our subsequent and aborted exchange is doubtless anticlimactic. It came in connection with a request from the Journal of the American Psychoanalytic Association that I review Gill's capstone book, Psychoanalysis in Transition: A Personal View (1994). Overall, I wrote an appreciative review (Wallerstein, 1995a) elaborating why I was calling it "a 'must read' for seasoned analysts and neophytes alike, indeed for anyone concerned to be au courant with the status of American analysis and Merton Gill…"; but I also stated that "yet, as is usually the case when Gill is concerned, I must enter my caveats" (p. 598). I went on to state where I felt that once again - not surprisingly, in the self-same area of the nature of psychotherapy vis-à-vis psychoanalysis - Gill in expounding his own viewpoint misunderstood or misrepresented mine.

In chapter 5 of this final 1994 book, "Psychoanalysis and Psychotherapy," Gill declared that the Psychotherapy Research Project that I headed at the Menninger Foundation (my final clinical accounting was reported in Forty-Two Lives in Treatment) buttressed his position that the erstwhile ideal of psychoanalysis as a therapy is unattainable and that interactive elements enter necessarily into every analysis. But he then went on to claim that I had not given up the ideal in principle but had concluded only that it was unattainable in practice and that analysis therefore often "necessitated the introduction of interactive contaminants into the analytic situation" (Gill, 1994, p. 62). Though Gill stated explicitly that contaminants was his word, not mine, he went on to argue the difference between his position and mine on the basis of my viewing interactive components of the analytic process only as necessary contaminants (no longer identifying this as his word) rather than as inherent, ubiquitous, and vital to the full understanding of the process.

While Gill and I do have significant differences on these issues - witness our 1991 exchange - I disagree with their characterization in Gill's final book. Rather, I see as a major thrust of the Menninger project and of Forty-Two Lives in Treatment the demonstration that there is an inevitable (and inherent, not accidental or contaminating) interpenetration of what had once been considered mainly psychotherapeutic components or modes into every psychoanalysis-and, incidentally, vice versa-and that this has made the once very distinct conceptual differentiations between the two modalities of psychoanalysis and psychoanalytic psychotherapy much more ambiguous in both theory and practice than they were in earlier years felt to be. On this issue I consider myself fully in accord with Gill's position; I too view psychoanalysis as both a one- and a two-person situation, accepting all the implications of that position, including the fact that every analysis is inherently both interactive and interpretive.

My actual differences with Gill on these issues, while substantial, are quite other. The main ones are two. (1) For all the greater overlap consensually acknowledged today (as compared to the 1950s) between psychotherapy and psychoanalysis - I have called this their mutual infiltration or interpenetration - there are still to me, and I think to most of those involved with these issues, though clearly not to those adhering to Gills views, conceptual and technical distinctions between the two modalities that I feel important to retain in the interest of clarity and precision in our work with patients. (2) Corollary to this, since neither psychoanalysis nor psychotherapy exists anywhere in its pure form, since interpenetrating elements are always present (necessarily so if the interests of patients are to be served), it follows that the interactions inherent in analysis are not always interpreted, don't always need to be, and in fact should not always be. It is a special sadness to me that I was informed of Merton Gill's death the very day I was finishing my review of his final work; my effort might have led to yet another friendly but critical exchange of views, as again we would try to clarify precisely how we understood each other, and where we agreed and disagreed.


I have devoted this essay to such a detailed exegesis of Merton Gills always formative impact on our continually evolving conceptions of the nature of psychoanalysis and the psychoanalytic psychotherapies, and of their ever altering relation to each other, because clearly it has been such a central concern of his as well as of mine, but also and most importantly, of our whole discipline as a purposeful clinical healing endeavor. If in the 1950s it seemed that a majority consensus was being formed regarding a definition of psychoanalysis and the range of psychoanalytically based psychotherapies, as well as an understanding of their similarities and differences, together with a promise of ever increasing clarification and specification as theory advanced and clinical experience accumulated, this was as much as any single persons the doing of Merton Gill. And if in more recent decades this convergence of psychoanalytic perspectives has progressively unraveled and the once seemingly crystal clear conceptual boundaries and relatednesses have become less clear, more problematic, replaced now by a range of reconceptualizations responsive to the varying tenets of a psychoanalytic world now more relational, constructivist, and pluralistic, then it can be said of Merton Gill that he has been a leading influence in staking out one of the significant contending positions on these issues, though it is not mine. But whatever our differing individual takes on these issues, we can all agree that Merton Gill's influence is writ large and will endure.


Alexander F., French T.M. et al. (1946). Psychoanalytic Therapy: Principles and Applications. New York: Ronald Press. Internet edition of chapters 2, 4, and 17: The corrective emotional experience: (Italian transl: Psicoterapia e scienze umane, 1993, XXVII, 2: 85-101. Internet edition: The corrective emotional experience:

Freud S. (1914). On the history of the psycho-analytic movement. Standard Edition, 14: 1-66. London: Hogarth Press, 1957.

Gill M.M. (1951). Ego psychology and psychotherapy. Psychoanal. Q., 20: 62-71.

Gill M.M. (1954). Psychoanalysis and exploratory psychotherapy. J. Am. Psychoanal. Ass., 2: 771-797.

Gill M.M. (1963). Topography and Systems in Psychoanalytic Theory (Psychological Issues, 10). New York: International Universities Press (Italian transl.: Il modello topico nella teoria psicoanalitica. Torino: Boringhieri, 1979).

Gill M.M. (1979). The analysis of the transference. J. Am. Psychoanal. Ass., 27 (Suppl.): 263-288.

Gill M.M. (1982). The Analysis of Transference. Vol. 1: Theory and Technique (Psychological Issues, Monogr. 53). New York: Int. Univ. Press (Italian transl.: Teoria e tecnica dell'analisi del transfert. Roma: Astrolabio, 1985).

Gill M.M. (1983). The interpersonal paradigm and the degree of the therapist's involvement. Cont. Psychoanal., 19, 2: 202-237 (Italian transl.: Il paradigma interpersonale e la misura del coinvolgimento dell'analista. Psicoterapia e scienze umane, 1995, XXIX, 3: 5-44).

Gill M.M. (1984). Psychoanalysis and psychotherapy: a revision. Int. Rev. Psychoanal., 11: 161-179. Internet edition: (Italian transl.: Psicoanalisi e psicoterapia: una revisione. In: Del Corno F. & Lang M., a cura di, Psicologia Clinica. Vol. 4: Trattamenti in setting individuale. Milano: Franco Angeli, 1989, pp. 128-157. Internet edition:; discussion:

Gill M.M. (1994). Psychoanalysis in Transition: A Personal View. Hillsdale, NJ: Analytic Press (trad. it.: Psicoanalisi in transizione. Milano: Cortina, 1996).

Gill M.M. & Hoffman I.Z. (1982). The Analysis of Transference. Vol. 2: Studies of Nine Audio-Recorded Psychoanalytic Sessions (Psychological Issues, Monogr. 54). New York: Int. Univ. Press.

Gill M.M. & Muslin H.L. (1976). Early interpretation of transference. J. Am. Psychoanal. Ass., 24: 779-794.

Gill M.M. & Wallerstein R.S. (1991). Psychoanalysis and psychotherapy (An exchange of letters to the editor). Int. J. Psycho-Anal., 1991, 72, 1: 159-161, 164-165.

Wallerstein R.S. (1984). The analysis of transference: A matter of emphasis or of theory reformulation? Psychoanal. Inquiry, 4: 325-354.

Wallerstein R.S. (1986). Forty-Two Lives in Treatment. New York: Guilford Press.

Wallerstein R.S. (1989). Psychoanalysis and psychotherapy: an historical perspective. Int. J. Psycho-Anal., 70: 563-591.

Wallerstein R.S. (1995a). Review of Merton Gill's "Psychoanalysis in Transition: A Personal View." J. Am. Psychoanal. Ass., 43: 595-600.

Wallerstein R.S. (1995b). The "Talking Cures." New Haven, CT: Yale University Press.

[Note: This paper later appeared in D.K. Silverman & D.L. Wolitzky, editors, Changing Conceptions of Psychoanalysis: The Legacy of Merton M. Gill, Hillsdale, NJ: The Analytic Press, 2000, pp. 199-218; European distributor: Eurospan. We thank for the permission]

Robert S. Wallerstein, M.D.
290 Beach Road
Belvedere, CA 94920
Tel. 415-389-9323, Fax 415-789-9530

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