Intersubjectivity and dialecticism


To restore silence is the role of objects.(Samuel Beckett, Molloy, p. 13)


There has been a psychoanalytic civil war being waged in the USA over the last 30 years between what has represented the mainstream tradition of American psychoanalysis, ego psychology, and an upstart that goes by a variety of names but which I will label here intersubjectivity. It may appear at this point that we should declare intersubjectivity the winner. Certainly the range of theorists writing with this outlook have attracted a wide following, and much that is written in contemporary ego psychology can be understood as a response to the various challenges posed by these revisionists. Yet, as AndréGreen (2000) observed, intersubjectivity must be seen as more of a corrective movement than an outright revision of psychoanalysis. Psychoanalysis is and has always been intersubjective. ‘Intersubjectivity’, as a group of contemporary theories, addresses issues that have been present in our theory since Freud, but were inadequately theorized. The division between drive and object, internal and external, has been problematic throughout our history. Green believes that Freud emphasized drive and internality to the detriment of the object in order to focus attention on what was new in his theory. Beginning with Klein, Fairbairn and Winnicott, many theorists have sought to rectify that situation. Over the last 30 years we have seen that process intensify as intersubjective theories have come onto the stage and, at times, pushed drive and the intrapsychic to the wings. Any theory that eliminates one side of the intrapsychic–intersubjective, drive–object axes creates particular difficulties if it still wishes to be considered psychoanalytic. In their effort to provide the corrective psychoanalysis has long needed, these theorists call attention to the object side, but some also call for a dialectic between the two sides. Critics tend to omit this dialecticism from their remarks and present intersubjectivity as destructive of the concept of drive. Ultimately, it is not in terms of metapsychology that these writers are most innovative or have had the greatest impact, but rather in terms of clinical technique.

There is no singular theory of intersubjectivity; there are a variety of schools of thought that are intersubjective in nature. Leaving aside contemporary Kleinian theory, these schools go by names such as relational, constructivist, interpersonal, and intersubjective systems theory. It is the relational school that represents the largest group and competes with intersubjectivity for having its name on the banner of the rebels. I have chosen to use the word intersubjectivity as the umbrella term because it best designates a theoretical orientation, whereas relational primarily designates a community of analysts with political and social ties, and a theoretical orientation only secondarily. Relational psychoanalysis is a movement: There is an institute with a ‘relational track’1, an official journal2, an international organization3, and various meetings considered relational. Intersubjectivity lacks this kind of infrastructure. It is a collection of ideas generated by a number of independently working theorists, and has no such structure to oversee its identity. Depending on one’s perspective intersubjectivity could be considered a type of relational theory, or relational theory could be a type of intersubjectivity. I will take intersubjectivity as the broader term within which relational theories fall but will include references to both to encompass the range of relevant issues.

Relational theory can be dated to the 1983 publication of Greenberg and Mitchell’s Object Relations in Psychoanalytic Theory, and the creation of the relational track at the NYU Postdoctoral Program in 1988. While Greenberg and Mitchell used the term to refer to an historical trend in psychoanalysis, it soon came to be regarded as a contemporary school of thought (Greenberg, cited in Mitchell and Aron, 1999). These authors take Sullivan and Fairbairn as the first relational theorists, and Aron traces its origins as far back as the work of Ferenczi and Rank (Aron, 1996). The origins of psychoanalytic intersubjectivity cannot be as precise. There are no specific foundational publications, and the ideas must be seen as emerging gradually across not just a range of writers, but a range of disciplines as well. Consideration of the concept of intersubjectivity first arose in philosophy, as a reaction against Descartes, four centuries ago. Two centuries later Hegel’s Phenomenology of Mind based self-consciousness in a kind of intersubjectivity. But it was in the phenomenology of Edmund Husserl, a contemporary of Freud’s, that intersubjectivity became a specific focus of philosophical inquiry. Though many other post-Cartesian philosophers engaged with this topic, the term itself only entered into philosophical usage with Husserl in 1931. Within psychoanalysis intersubjectivity is implicit in the Freudian model of treatment, but it did not enter psychoanalytic terminology until Jacques Lacan introduced it in 1953. In the second half of the 1970s the word began appearing in a sense approximating to its current usage in the work of infancy researchers, especially Colwyn Trevarthen, and the psychoanalyst Robert Stolorow and his colleagues. Since that time the term has come to be used to describe the work of many others, including Jessica Benjamin, Thomas Ogden, Irwin Hoffman, Christopher Bollas and Charles Spezzano. A group of analysts originally associated with more classical theory like Arnold Modell, Dale Boesky, James McLaughlin, Owen Renik, Theodore Jacobs, Warren Poland and Judith Chused (referred to by Dunn (1995) as interactionists) became associated with intersubjectivity early on. Today the ideas of intersubjectivity are so ubiquitous that it is not possible to make a comprehensive list. Prominent relational analysts include Stephen Mitchell, Philip Bromberg, Lewis Aron, Donnel Stern, Jay Greenberg (somewhat disassociated from this group in recent years), Adrienne Harris and Emmanuel Ghent.

Intersubjectivity can be understood in broader or narrower senses. The early infant researchers who first employed the term tended to use it quite broadly, referring to it as “a deliberately sought sharing of experiences about events and things” (Trevarthen and Hubley, 1978, quoted in Stern, 1985, p. 128). Stolorow et al. (1978) provided these definitions, which also cast a wide net: “The interplay between transference and countertransference in psychoanalytic treatment can be conceptualized as an intersubjective process” (p. 249), and at a later date, “[P]sychoanalysis is pictured by us as a science of the intersubjective, focused on the interplay between the differently organized subjective worlds of the observer and the observed” (Stolorow, 1984, p. 643). These definitions employ the term as a means to shift attention away from isolated units of action, i.e. the analysand (or infant) as sole object of observation, and toward regarding the dyad as the unit of study. In the years since, some theorists have stressed a view of the psychoanalytic setting as a single system in which nothing happens to one person without it producing an effect on the other, while others have stressed the relative independence of each actor in his subjectivity. In the latter case intersubjectivity may refer to the recognition possessed by each participant of the differences between them or, more simply, the mere fact of interaction between two subjectivities. Relational analysts tend to use the word intersubjectivity in its most limited sense to describe the work of Stolorow and his colleagues and, at times, that of Jessica Benjamin.4 Their definition of intersubjectivity is based on the theorist, not the theory, although they consider this work to also be a part of relational theory.

Clearly there is a great deal of overlap between what would be called intersubjective and what would be called relational. The differences between them have less to do with theoretical concerns and more to do with history and institutional alliances. The explosion of interest in intersubjectivity since the 1980s marks an elaboration of developments that were under way in psychoanalysis since at least the 1950s, when Heimann, Racker and other Kleinians led the way in bringing attention to countertransference as a central element of clinical practice. Intersubjectivity represents only one of the many outgrowths of that development. The relational school, though incorporating much from the wider world of psychoanalysis, was created by analysts who had trained in interpersonalism, and it seems reasonable to locate its origins there. As such, it came onto the psychoanalytic scene from what was considered the fringes by mainstream American analysts who tended to view interpersonalists as non-analytic. We can assume that history has contributed to the conflicts between these groups.


To understand intersubjectivity, and the orientation and implications psychoanalytic thinkers on this topic assume, we must start with its historical emergence in philosophy. This account is meant to provide a brief, general orientation to the topic through five key thinkers.5 Since Descartes’ (1596–1660) description of mind–body dualism, Western philosophy has been preoccupied with the question of subjectivity. Cartesian subjectivity is that of the isolated mind, able to be certain only of itself, of its own thinking and self-awareness. Lacking direct access to anything other than thought, everything else (with the exception of God) is subject to doubt, including the minds of others. Descartes effectively invented the concept of the subject as a self-contained monad, divided from all that is external to it. Many philosophers following Descartes challenged this notion, but it was 200 years before the most significant of these arose with Hegel. For Hegel (1770–1831) subjectivity, or self-consciousness, required an encounter with an other. In his master–slave dialectic he described a sort of primal myth in which self-consciousness arose out of a struggle between two individuals vying for recognition by the other. Each acts out of the conviction that recognizing the other would diminish himself, and so conflict ensues. With victory the ‘master’ comes to discover he is ultimately dependent on the loser, his ‘slave’, since he now needs the slave for recognition. The slave, on the other hand, gains in self-awareness through his work, from his own productions. Seeing his industry materialized provides a kind of mirror. This allows him to acquire a greater self-consciousness than the master, and eventually to go on to challenge and dominate the master, making him the slave. As this cycle repeats itself they gradually come to realize they are dependent on each other: without mutual recognition neither can achieve adequate self-consciousness. In these two philosophies we witness a shift from the Cartesian, solipsistic, one-person model to the Hegelian dyadic, two-person model of mind and self-consciousness.

Edmund Husserl (1859–1938) was the founder of phenomenology. Through the ‘transcendental phenomenology’ of his Cartesian Meditations he pursued the question of how one could gain awareness of the existence of other subjectivities, i.e. the problem of intersubjectivity. At the same time, he reinforced many of the problems created with Descartes’ isolated mind. Reminiscent of Plato’s ideal forms, Husserl sought an awareness of the world as it truly is. To do this he proposed the elimination of all assumptions and beliefs that could bias any pure perception by ‘bracketing’ them out. In this way the subject could gain a certainty about the external world of objects that Descartes left in doubt. As Sass (1989) explains, this “does not imply a reduction of experience to something like sense data” (p. 457). Rather, “we turn our attention away from the object being referred to (and away from our psychological experience of being directed toward that object), and ‘turn our attention to the act’ [quoting Dreyfus, 1982, p. 6]... to the meanings and objects-as-meant of normal experience” (p. 457). In other words, the external is experienced via the internal meaning of the experience with that external: “the subject is discovered to have ‘ontological priority’ over the world” (p. 458). An awareness of other subjects emerges in a similar manner, out of what Husserl called ‘empathy’. By empathically imagining our own experiences as occurring in another we are able to recognize their subjectivity. Paradoxically, Husserl’s intersubjectivity is a decidedly one-sided affair, occurring in the mind of only one person. On the other hand, it provides for the presence of others as subjective entities like oneself, unlike Descartes, and it allows for such awareness implicitly through the human capacity for empathy, without passage through Hegel’s struggle for recognition.

The revision in hermeneutics that began with Husserl’s student, Martin Heidegger (1889–1976), took a very different approach. Heidegger challenged the idea that it was possible to bracket out presuppositions and thereby gain access to the world through mind. Instead he saw the subject as existing in and through its action in the world. The Cartesian division of subject and object was false according to him, since each was implicit in the other. The Heideggerian subject [Dasein] is not really a subject at all, it is a ‘being-in-the-world’ for whom it is impossible to conceive of subject or object, self or other, independently of the other, and neither has priority. Dasein exists in a ‘unified field’ (Sass). This has profound significance for the idea of intersubjectivity since the problem is shifted away from intersubjectivity and to subjectivity. The mind as isolate is the anomaly rather than the norm, and intersubjectivity is the basic human condition. Now it is self-awareness that becomes problematic: what is it we recognize when we experience ourselves as separate beings? For Heidegger the individual we sense is something of a construction, a fiction we create secondarily to our being-in-the-world. Heidegger defines Dasein by this very question of being: Dasein is that which questions its being, a question that depends on an awareness of the end of being, i.e. death and temporality. Furthermore, Dasein is never able to truly know itself because there is no greater medium against which to contrast itself. Dasein exists within a horizon of all that makes it who and what it is. Like the proverbial fish’s ignorance of the water, Dasein has little idea of this horizon due to its very plenitude. This is a kind of unconscious, but one composed of presence and immanence, not repressed or hidden away. Jacques Lacan used these ideas in developing his theory of primal repression and the impossibility of ever knowing the “truth about truth” (2006, p. 737). For both Heidegger and Lacan language is the medium of this unknowingness, structuring thought at the same time that it makes it invisible to itself.6

Paul Ricoeur and Hans-Georg Gadamer were the most important philosophers to advance Heidegger’s ontological hermeneutics through the second half of the 20th century, and both had direct influences on psychoanalysis. Ricoeur (1970) is largely known through Freud and Philosophy, his book on the metapsychological structure of Freud’s work. Gadamer (1998), whose major contribution was Truth and Method, did not discuss psychoanalysis as extensively as Ricoeur, but his work explores the subtleties of intersubjective communication, making it very relevant to the clinical encounter in psychoanalysis. But it is to Maurice Merleau-Ponty (1908–1961) that I will turn last because, even though he has been less recognized by psychoanalysts than Ricoeur and Gadamer, his work could be considered a philosophy of intersubjectivity. Starting from Gestalt psychology, he broadened its idea of holism from the realm of psychology to an outlook on ontology in general. In his major work Phenomenology of Perception (first published in 1945) Merleau-Ponty (2002) devotes a chapter to ‘The Body in Its Sexual Being’, which is largely on psychoanalysis. There he pays considerable respect to the field of psychoanalysis while also correcting some distortions he felt Freud made in conceptualizing the body and mind in overly scientistic, objectifying ways. A similar sensibility appears 15 years later in the work of the analyst George Klein as he led the movement to disencumber psychoanalysis from its mechanistic metapsychology, and elevate the place of clinical theory and analytic praxis.

Merleau-Ponty’s philosophy is based on the idea of a unity of subject and object. He felt the Cartesian separation of mind and body created an illusory division, and that the two can no more be divided than thoughts can be divided from the things thought about. In this way his philosophy follows from Heidegger’s being-in-the-world, but it also returns to Husserl – at least Husserl’s late work in which he tries to reconcile his work with that of Heidegger. Agreeing with Heidegger that the human being is defined through his presence in the world, Merleau-Ponty shifts the substrate of Dasein from the world to the body, making the body the manifestation of unity of subject and object. Perception by means of the physical body links outside to inside and self to other. The body is also the site of the unconscious, but an unconscious somewhat different from Freud’s. Merleau-Ponty’s unconscious is not hidden under consciousness, but rather surrounds consciousness: “In perception there is something perceived and something not perceived ... In perception there is a perceiver ‘unaware’ of his perceiving” (Romanyshyn, 1977, p. 216). Perception requires a kind of ‘forgetting’ of the perceiver, and so creates a ‘figure-ground phenomenon’ in which the perceived is figure and the perceiver the ground. This is the Merleau-Pontian unconscious located between self and other, and based in perception. The body is also a kind of history, a repository of past experience, and this is how Merleau-Ponty accounts for the repressed. The body may repeat past behavior in the present as if unaware of the new circumstances because it retains a history. This situated the unconscious “not at the center of a psychological life that exists within and beneath one’s conscious life, but at the center of one’s life in the world, which one is and always is as an embodied history of desire” (Romanyshyn, 1977, p. 221).

Our psychoanalytic tradition from Freud is a psychoanalysis of the subject, that is to say, an isolated, monadic, Cartesian being that is in conflict with the objects that surround him. The Hegelian subject too is confined in his isolated subjectivity and can only achieve a degree of intersubjective recognition through struggle. With Heidegger and Merleau-Ponty we cross an ontological threshold where selfhood is understood as a later development that follows from a primary state of being-in-the-world, or intersubjectivity. Merleau-Ponty, with his interest in the embodied subject, provides a specific medium for intersubjective relatedness. The body is the site where being and the world are manifest as one.

These 20th century philosophers represent the hermeneutic and phenomenological traditions that have been influential in psychoanalysis since the 1960s. Sass (1989) points out that this trend has at times been confused with the humanistic tradition. He cites Roy Schafer (i.e. his work on action language and narrativity) and Heinz Kohut in particular as examples of that confusion, but I would add, in this era of intersubjectivity, that much of that same confusion has persisted in the writing of our more contemporary intersubjectivists. Sass explains that: “[T]he most characteristic feature of humanistic psychology is the emphasis it places on the conscious human subject” with “emphases on four purported qualities of the subject, what we might call: (1) freedom; (2) privacy; (3) uniqueness; and (4) self-transparency (the latter refers to a certain faith in the certitude and clarity of subjective experience)” (1989, p. 436). One immediately recognizes that these are qualities that are inconsistent with core concepts of psychoanalysis, such as the unconscious, psychic determinism, the potential for self-deception, interpretability, and certain basic commonalities of human motivation. Modern hermeneutics is based in Heidegger’s efforts to “disabuse man of his illusion of being at the center or foundation of the experiential universe”(1989, p. 455), while humanism’s aim is just the opposite. The humanistic outlook is present in Husserl’s transcendental phenomenology in which the subject finds the ability to overcome the distortions of his prejudices to see the world in its essence. We could also call this a subjectivism, in the sense that the individual subject is the ultimate source of knowledge. Post-Heideggerian hermeneutics asserts that this is never possible, and that knowledge is always partial and contingent on the horizon within which one exists. Self-knowledge derives not through a process of isolation from the world, but rather from involvement in the world.

When traditionalists challenge intersubjectivists it is often due to their use of these principles of humanism. They assert that this reflects an inadequate appreciation for the constitutional, mechanistic qualities of the mind, and too great a reliance on consciousness. But there is a certain irony in this since traditionalists often fall into the same trap. In modern liberal society we are conditioned to accept the independence and self-determination of all, and psychoanalytic practice constantly confronts the conflict between those values and antagonistic traits the mind betrays. Some intersubjectivists, in order to guard against these traps, go out of their way to stress the dialectical nature of their theory. Dialectic can mean a number of things in this context but it includes the appreciation for both humanistic and non-humanistic outlooks. Analysts in general, not just intersubjectivists, engage with the modern hermeneutic tradition far too rarely. Notable exceptions to this can be found in the work of Hans Loewald, Donnel Stern, who has placed Gadamer’s work at the core of his theory, and Madeleine and Willy Baranger for whom Merleau-Ponty had an important influence. Paul Ricoeur has certainly had a more pervasive, if less specific, influence on psychoanalysis, and his approach espoused a dialectical view7 very much in keeping with the intersubjective dialecticians.

History, thinkers, controversies


Although geographically inappropriate to this essay, Jacques Lacan deserves mention here for a couple of reasons. It is not only that he was the first analyst to use intersubjectivity as a technical term, but also that the evolution in his view of intersubjectivity anticipates a central concern in the controversies around intersubjectivity today. Lacan’s use of the word began in 1953 in what might be considered his manifesto, Function and field of speech and language in psychoanalysis. For Lacan intersubjectivity was a characteristic of the symbolic register insofar as it described the transindividual character of language. Language went beyond the individual person and was intersubjective because it involved shared, unconscious rules for communication between subjects. The Lacanian unconscious is a product of language that is common to all who speak, rather than something uniquely residing within each individual. Humans are subjects insofar as they are subject to language and its laws; subjectivity is a consequence of language. In his Seminar on ‘The purloined letter’ (Lacan, 2006), first published in 1956, Lacan provides a classic example of the intersubjectivity of language. There he describes a structure of three subjective positions8 that are dictated by the signifier (represented by the purloined letter, which also represents language as a whole). He shows how the various characters of the Poe story rotate through each of these positions. The characters are powerless to determine their fate within this structure and instead become prey to it, uniting them intersubjectively within the structure determined by that mysterious letter – cogs in a mechanism larger than themselves.

By around 1960 the meaning of intersubjectivity shifted for Lacan from an aspect of the symbolic register to that of the imaginary. Subjectivity then replaced intersubjectivity as the object of psychoanalytic interest. Intersubjectivity came to represent the idea of sameness, an effect of the ego, which aims to eliminate differences between subjects. The intersubjective is then the ego-to-ego relationship based in mirror-like illusions one person generates to believe he is similar to another (Lacan, 1998). Lacan’s shift in usage contains a dichotomy we encounter repeatedly with intersubjectivity: that of difference vs. sameness, two vs. one, the otherness of the unconscious vs. the unity of consciousness. These add to the dichotomy of the hermeneutic vs. the humanistic. Much of the theoretical controversy among American analysts rests on these dichotomies, and perhaps ultimately on a dichotomy in Freud’s work: that between his topological and structural models. The topological model was proposed before he developed his concept of transference, and it mapped the path of the treatment: from consciousness to the unconscious and back again. With the structural model the psyche is conceived less in terms of finding what lays hidden, and more toward charting the relations between and within its various components. Those internal relations become manifest externally in the transference. Thus the two models present a dichotomy that is related to the others: whether it is the unconscious or the transference that guides the treatment. In the former case we come closer to a hermeneutics of difference. In the latter case the humanism of transference relations takes precedence.

Infancy research

In the 1970s research techniques were developed that led to a new view of the human infant. The longstanding assumption that infants lacked the capacity to differentiate self from other came to be seen as erroneous. The merged mother–infant unit gave way to a view of interacting beings with different levels of self-awareness. With two subjects of interest, albeit at different levels of organization, it followed that their intersubjective relatedness was also of interest. Communication was now understood as going in two directions: not just from a mother to an infant, but also from a thinking infant to a mother recognized as a separate being. Intersubjectivity began to appear in the work of Colwyn Trevarthan in 1974, and, in the following years, Daniel Stern made intersubjectivity a central aspect of what he called the “subjective sense of self” (D.N. Stern, 1985, p.124). For Stern this meant that the infant “can sense that others distinct from themselves can hold or entertain a mental state that is similar to one they sense themselves to be holding” (p. 124). Stern’s subjective sense of self began to develop around 7 months, but Trevarthan viewed intersubjectivity as innate and used a broader definition, saying it involved “a deliberately sought sharing of experiences about events and things” (1978, cited in D.N. Stern, 1985, p. 128). The word ‘deliberate’ suggests that the infant does sense that the mother possesses a similar mental state, and thus implies an earlier onset. Infancy research and intersubjective theory have maintained an active cross-fertilization and cooperative relationship since that time. Work on mentalization by Fonagy and Target has been particularly conducive to such work, as has attachment research by Main, Lyons-Ruth, and D.N. Stern. Morris Eagle, Susan Coates, Arietta Slade and Doris Silverman have also made considerable use of infancy research in their psychoanalytic writing.

Stolorow and self psychology

Robert Stolorow can be credited with introducing the word intersubjectivity into American psychoanalysis in 1978. His use of the term is closely allied to Husserl’s as can be seen in this quote:

Psychoanalytic treatment requires a capacity on the analyst’s part to empathically understand the subjective meanings implicit in his patient’s expressions. Probably the single most important source of such empathy resides in the therapist’s ability to identify with his patient and find analogues in his own life of the experiences which are presented to him. But from this same source comes one of the chief obstructions to analytic understanding; namely the distinctive configurations of self and object representations which organize and pervade the therapist’s own subjective world... The interplay between transference and countertransference in psychoanalytic treatment can be conceptualized as an intersubjective process in which two basic situations repeatedly arise; representational conjunction and representational disjunction.

(Stolorow et al., 1978, p. 250)

The authors conclude:

If the analyst possesses self-knowledge and a capacity to decenter from the central themes of his own personal life, then episodes of representational conjunction and disjunction need not give rise to interferences in the treatment process.

(p. 253)

In this view the skilled analyst is able to accurately empathize with his patient to become aware of the patient’s mind while also not allowing his personal issues interfere. In later years Irwin Hoffman (1999) would criticize Stolorow for maintaining what he considered a conventional attitude toward the analyst’s ability to attain such purity of knowledge. Indeed, Hoffman makes clear that this is the very attitude intersubjectivity9 aims to replace.

For Husserl, empathy provides the means to achieve intersubjectivity. It allows for true knowledge of the other – enough to erase any doubt of the other. In this early paper Stolorow and colleagues invoke empathy as the way to achieve understanding of the other. At times the analyst’s subjectivity will interfere rather than assist this process and misunderstanding will occur instead. Stolorow’s intersubjectivity describes both the success and failure of understanding, so that it becomes interactions of virtually any sort. If transference and countertransference are present, then intersubjectivity is as well. More recently Stolorow has expressed the breadth of meaning the term holds for him more clearly:

For my collaborators and me, intersubjectivity has a meaning that is much more general and inclusive, referring to the relational context in which all experience, at whatever developmental level–linguistic or prelinguisitic, shared or solitary takes form (Stolorow & Atwood, 1992). For us, an intersubjective field–any system constituted by interacting experiential worlds–is neither a mode of experiencing nor a sharing of experience. It is the contextual precondition for having any experience at all (Orange, Atwood, & Stolorow, 1997). Our intersubjective perspective is a phenomenological field theory or dynamic systems theory that seeks to illuminate interweaving worlds of experience. Experiential worlds and intersubjective fields are seen as equiprimordial, mutually constituting one another in circular fashion.

(Stolorow, 2002, p. 329–330)

It becomes difficult, when the term is presented in this way, to see much difference between intersubjectivity and mind itself.

Stolorow’s emphasis on empathy reflects the strong current of self psychology in his work, though he has been rather ambivalent about his allegiance to it in his writing (Stolorow, 2004; Stolorow et al., 1999). He has described himself as “someone who uses this framework [self psychology] extensively in his own work” (1986, p. 387), while also retaining a critical perspective. But self psychology represents one of the important sources of intersubjectivity not only because of Stolorow’s work, but also for its more general influence on American psychoanalysis in the 1970s and 1980s. Its model of psychopathology is derived from the parent–child relationship, not from drives or constitutional factors in the individual. It also emphasizes the need for scrutiny of the analyst’s role in the process insofar as he is able to empathize or not with his patient. In these particular ways it looks less to the individual patient in isolation and more toward the individual in his relationships, presenting a significant shift away from a ‘one-person’ classical model. Though still not fully a ‘two-person’ model, this new perspective opened the door to the two-person theorizing intersubjectivity would present in the ensuing years. It is also important to note that self psychology led the way in producing important cultural shifts in the American psychoanalytic establishment. The American Psychoanalytic Association (APsaA), which had routinely rejected non-classical theorists from its journal and from the program of its meetings, altered its policy in the case of self psychology. The reasons for this are varied and beyond the scope of this paper, but this toe in the door of the establishment provided a path others have been able to take since that time, giving them greater attention and respect than they would otherwise have had. Intersubjectivity has benefited from this. As it entered the psychoanalytic scene in the USA the hurdles it had to overcome to gain some legitimacy were lower than would have been the case prior to self psychology. This undoubtedly contributed to it becoming widely recognized beyond its adherents and practitioners.

Mitchell, Greenberg, Aron and the relational movement

A few years after Stolorow began using this term, Jay Greenberg and Stephen Mitchell (1983) published their groundbreaking Object Relations in Psychoanalytic Theory. It was here that relational theory was first introduced through a ‘comparative psychoanalysis’ of how object relations had been conceptualized since Freud. In their view it was object relations that ultimately organized all psychoanalytic theories and so could be used to clarify the similarities and differences between them. Starting with Freud’s own struggle with how to maintain drive theory while also accounting for the motivational qualities of the social world, they went on to examine the interpersonal theory of Sullivan, Melanie Klein and her followers, Hartmann’s ego psychology, and the more recent efforts to integrate object relations theory with ego psychology by Mahler, Jacobson and Kernberg, as well as the ‘mixed model’ of drive and object in Kohut and Sandler. They highlighted Sullivan and Fairbairn who they said provided “the major systematic alternative to drive theory” (Greenberg and Mitchell, 1983, p. 4), and called them “radical revisionists of the relational model” (p. 351) because of their rejection of any drive-based model. It should be noted that Fairbairn’s rejection of drive was not as complete as Sullivan’s. He retained drive as a phenomenon that developed secondarily from innate object relatedness. But these differences were of such fundamental importance that Greenberg and Mitchell believed it made the idea of integrating the two impossible. As Mitchell and Aron (1999) later put it, “[R]elational concepts do not provide understandings of different phenomena from those explored by the drive/defense model; relational concepts provide alternative understandings of the same phenomena” (p. xiv). It is from the relational models of Sullivan and Fairbairn that their subsequent work developed, and which then attracted the large following that has come to be known as the relational school.

In their book Greenberg and Mitchell used ‘relational’ to refer primarily to Fairbairn and Sullivan, but with the theory they derived from those writers relational soon became the term used to name an educational track at the NYU Postdoctoral training program. The Relational track was started in 1988 and was meant to provide an alternative from the drive-based theory of the Freudian track, and the focus on external relations of the Interpersonal–Humanistic track (Bromberg, 2009). By ‘relational’ the five founders10 meant to include internalized as well as external interpersonal relations (Aron, 1996). In a few short years the small group of faculty that had founded the track had expanded considerably, and they had started the journal Psychoanalytic Dialogues. Their ideas quickly spread far beyond New York. Relational analysis has continued to grow and now represents a major part of psychoanalysis in the USA. Considerable theoretical diversity exists within the group in spite of its origin in the opposition to drive theory. It now includes theorists with views consistent with a drive-based theory, as well as interpersonalists and self psychologists. As Aron (1996) says, “Relational analysis is not a unified or integrated school of thought, nor is it a singular theoretical position. Rather it refers to a diverse group of theories that focus on personal, intrapersonal, and interpersonal relations” (p. 19). It is difficult to see how this would distinguish it from most other schools of psychoanalysis. Even when drives are seen as the primary motivational sources, it is how the drives are expressed via internal and external relations that is the substance of any analysis. Ghent (2002) states: “There is no such thing as a relational theory, but there is such a thing as a relational point of view, a relational way of thinking, a relational sensibility” (p. 7).

In the spirit of pluralism some writers have argued against the ‘dichotomizing’ view of Greenberg and Mitchell, and called for a more ‘dialectical’ approach. These tend to be non-relational analysts who would like to see these ideas integrated into more traditional theory. A case can be made for this given the range of interpretations of the drive concept in recent psychoanalytic history. But the two founders, Mitchell (until his death in 2000) and Greenberg, remained strong advocates for this separation. Mitchell maintained a drive-free model in his work, but Greenberg (1993) soon began writing about the need to retain a drive concept. He also became a critic of certain assumptions found in the relational school. Greenberg (2001) views these analysts as united by their clinical theory, not by their adherence to an object-based theory over a drive-based one. He has described the relational approach to clinical technique as eschewing belief in any single correct technique and as promoting flexibility in technique that is unique to each analytic dyad. With this loss of frame such an approach will inevitably draw greater attention to the analyst’s contribution to the transference, and there will be a complementary diminution of attention on analysand’s contributions to it. The analysand may then, reasonably, seek solutions to his problems in the analyst’s behavior rather than himself. Greenberg cautions that this approach leads analysts to want to behave well, like ‘good’ people, and in the process to no longer function as analysts.

As many have written (Bachant et al. [1995], Kernberg, Bollas), the elimination of drive from theory also may have an effect on the analyst’s engagement with unconscious process. As the relationship with the analyst in the here and now becomes the focus, the inference of conflict in a dynamic unconscious of the there and then is more difficult. Mitchell saw the individual psyche as one part of a larger relational matrix. His was an ontological theory in its unrelenting focus on the social nature of man. As such, patients could only be understood within that context. He acknowledged the role of inborn endowment, but only as it was expressed to, and affected by, the relational surround. Interpretation, then, always included that environmental element, and the isolated unconscious was de-emphasized. As with self psychology, this is essentially a trauma-based theory. Yet relational theory is diverse, and Mitchell’s drive-free model no longer defines the community. Harris (2005) has discussed a shift in orientation with respect to the concept of conflict among relationalists. In considering conflict these analysts think less of an incompatibility that demands resolution, and more of a dialectic that can be maintained between opposing terms. The implication is that health has more to do with awareness and acceptance of conflict than with its elimination.

This raises one more important difference between traditionalists and many relational analysts. Defense and repression have faded from their concepts and been replaced by dissociation and vertical splits in the psyche (Bromberg, 1998; D.B. Stern, 1997). Multiple self states are perceived by these analysts where others would recognize shifting levels of consciousness and types of defense. Although integration of these selves may result from analysis, the treatment is more oriented toward bridging the gaps in awareness due to dissociation.

Another way to understand what unites relational analysts other than their metapsychological or clinical theory is in terms of psychoanalytic communities, specifically their independence from APsaA. As Greenberg (2001) once feared would happen, relational analysis has become a ‘movement’, For 70–80 years APsaA had been the only large psychoanalytic organization in this country. With few exceptions, it was an organization comprised exclusively of physicians. Psychologists who were admitted had to sign a document stating they would use their training for research purposes only, not for clinical work. In 1985 four psychologists brought an anti-trust lawsuit against APsaA, the International Psychoanalytic Association and two New York training institutes, to challenge this exclusionary practice. The settlement that was reached before a trial ensued required APsaA, the IPA, and one of those institutes (the New York Psychoanalytic Institute) to allow psychologists to be clinical members and candidates. The other institute (Columbia), which is part of a university, was exempted in deference to the university’s right to set its own admissions criteria (Simons, 2003). In the years following the lawsuit many psychologists have been accepted at training institutes of APsaA, but the pre-existing distaste for its practices continues for many. The arrival of the relational school during those same years offered a welcoming alternative. It is a group that was created by psychologists and has been consistently led by psychologists. Division 39 of the American Psychological Association provided a nationally-based organizational home for the group early on, before it had grown to the point of establishing its own national and international organizations, and Division 39 continues to serve as one the major settings for national meetings of relational analysts. Given the theoretical diversity of the relational group, it could well be seen as more of a political division than a theoretical one. Its identity could be understood in relation to those politics as the un-APsaA, or the un-medical-analysis organization. Such perceptions must also contribute to the disagreements that get played out under the guise of theoretical debates.

Interpersonalism and American psychiatry

Interpersonalism has been of central importance not only to the origins of the relational school, but throughout American psychiatry and psychoanalysis where its influence has been less acknowledged. It is rarely studied in psychoanalytic institutes belonging to the American Psychoanalytic Association. While the American psychoanalytic establishment came under the sway of Europeans fleeing to the USA in the 1930s and 1940s, Harry Stack Sullivan developed his interpersonal psychiatry which was then further developed by others11 into interpersonal psychoanalysis (Levenson, 1992). This could be considered a more American psychoanalysis in that it looked outward to the social surround as the means to reach the intrapsychic. Sullivan’s ideas were generally dismissed by these new arrivals with many of the same criticisms we hear today toward relational psychoanalysis. Sullivan’s ideas emerged out of the American philosophical traditions of pragmatism12 and operationalism13 from earlier in the 20th century. These were theories that meant to ground philosophy in “lived experience, practical reality, life as it is sensed and felt” (Greenberg and Mitchell, 1983, p. 82).

Interpersonalism was initially intended to counter Kraepelin’s views of schizophrenia. Sullivan (1962) sought to bring understanding to psychotic thought that had been seen as meaningless, and he used Freud’s ideas to provide a theoretical base. Schizophrenia was the result of experiences in past relationships according to Sullivan, which caused a deficit in the capacity to relate to others. As he developed his theory further he sought to correct what he saw as the impersonal, mechanistic qualities of Freud’s drive theory, and he and his followers argued that Freud had neglected the significance of the social and cultural context as motivational forces. They sought a theory close to what could be observed and consciously experienced, rather than in experience-distant theories that could never be substantiated directly in the clinical setting. He also stressed that the therapist is never in an objective position and as a ‘participant–observer’ needs ‘consensual validation’ from the patient to draw conclusions. Together they create an interpersonal field of their own, and it is the interpersonal, not the individual, that forms the basic unit of therapeutic interest. Sullivan accepted the existence of an unconscious but aimed his therapeutics at that which could be put into words by the patient, rather than what was inferred by the analyst without concrete, conscious data. He called the unconscious the ‘immutably private’.

In spite of its overt rejection within established psychoanalysis, interpersonalism thrived within the teaching centers of American psychiatry for the period during which psychoanalysts held most of the leadership positions (from approximately the 1940s, after European analysts arrived, until the 1980s, when biological psychiatry began to assume priority). It made sense in hospital settings where clinicians struggled to find useful interventions for psychotic disorders, but was snubbed where loyalty to Freud was paramount. Principles of interpersonalism were routine in teaching psychiatric trainees, but were ignored in the analytic institutes they attended. Since psychoanalysis was predominantly practiced by physicians, it is difficult to imagine that interpersonal theory did not find a place in the clinical practice of American psychoanalysis, even if its principles were derided on paper by those same analysts.

Among the more interesting and influential interpersonalists was Harold Searles. Searles wrote on the intensive psychotherapy of schizophrenics, particularly with respect to countertransference and psychotic processes that occurred within the analyst. He gained recognition and notoriety with an early paper (Searles, 1959) in which he discussed the emergence of loving, erotic feelings for his analytic patients,14 and his belief that it was important to inform them of those feelings. A courageous and original thinker, his ideas have been taken up by some intersubjectivists, particularly in the work of Irwin Hoffman.

Gill, Hoffman and constructivism

Perhaps the most influential American writer on intersubjectivity has been Irwin Hoffman. Early in his analytic career Hoffman began a collaboration with Merton Gill and, like Gill, his original sources derive from American ego psychology. Gill had worked closely with David Rappaport early in his career, and then went on, in his later work, to theorize the analytic process as a pure culture of transference. Gill’s was a newly considered transference that was provoked by the analyst as much as emerging from the patient’s history. As a result, his call for greater emphasis on transference interpretation was directed particularly at the here and now situation between the analytic dyad. The transference, according to him, was present continuously in all the patient’s associations, and was not solely based on the patient’s distortions, but on legitimate perceptions of the analyst as well: “How the patient experiences the actual situation is an example of the role of the actual situation in a manifestation of the transference” (Gill, 1979, p. 265). He goes on: “[T]he current situation cannot be made to disappear – that is, the analytic situation is real” (p. 277). In order to make use of transference in the treatment the analyst needed to open himself up to scrutiny by both himself and the patient in a way that had not been done before. Gill’s work complemented what the Kleinians had been writing since the 1950s and contributed significantly to the reconceptualization of countertransference as a useful element in the treatment, rather than as a reflection of shortcomings in the analyst. After publishing the two-volume Analysis of Transference (Gill, 1982; Gill and Hoffman, 1982) Gill went on to become an outspoken critic of ego psychology and the scientistic traditions in American psychoanalysis, in addition to becoming an advocate for interpersonalism.

Hoffman has been a similarly independent voice since his work with Gill although, more recently, he has become closely associated with the relational school. His work draws in particular on the interpersonalist Harold Searles, and on Hans Loewald and Heinz Racker. Hoffman developed what he called a ‘social-constructivist’ theory, later renamed ‘dialectical-constructivism’. With constructivism he sought to differentiate his work from models relying on positivistic beliefs that the analyst had privileged access to truth. Understanding could only emerge through the combined effort of the dyad, with the recognition that other truths may come from working with another analyst, or even with the same analyst under different conditions. Hoffman’s is an epistemological theory, directed at the question of how one acquires knowledge in psychoanalysis. This approach made use of a ‘radical’ critique of the ‘blank screen model’, i.e. the neutral and anonymous analyst. Hoffman’s radical form adheres to a view of the analyst as always limited by unconscious influences, which produce unnoticed biases, while the ‘conservative’ form assumes the analyst is able, at times, to overcome personal biases. He saw the conservative critique as ubiquitous, and reflected in the work of the Kleinians, of Sullivan and much of interpersonalism, of Kohut and other self psychologists, of Stolorow and his group, of some relational writers, and of many of the ego psychologists who sought to modernize that theory. For Hoffman, the analyst was never capable of providing an objective, unmediated, understanding of the analysand or of himself:

The personal participation of the analyst in the process is considered to have a continuous effect on what he or she understands about himself or herself and about the patient in the interaction ... the analyst’s understanding is always a function of his perspective at the moment ... Therefore, what the analyst seems to understand about his own experience and behavior as well as the patient’s is always suspect.

(Hoffman, 1998, p. 136)

The logical corollary of this is that, even when the analyst believes himself to be a blank screen, he continues to convey his biases to the analysand. The analyst must take a patient’s observations about him as seriously as he takes his own observations about the patient, i.e. as mixtures of reality and fantasy, since fantasy and reality can never be fully distinguished. He believes it can be useful to make the patient aware of the analyst’s struggle in sorting through his own thoughts in his quest to arrive at a better understanding. Yet he also feels it is important to maintain such a ‘personal perspective’ in a dialectic with a ‘clinical’ one to ensure that the patient, not the analyst, remains the focus of attention. This keeps the analyst attending to the ‘figure ground relationship’ that exists between those two perspectives.

Kernberg (1997) has objected that such a dialectic neglects what he terms the ‘third position’:

The analyst both clarifies the intersubjective field and adds a new dimension: an ‘outsider’s’ view of it, a reflection on what is experienced by patient and analyst, in addition to conveying his understanding of the patient’s subjective experience.

(p. 102)

This third position

is created by the analyst dissociating himself internally into an experiencing part that participates in the transference/counter-transference bind and an observing part that includes the analyst’s specific knowledge, technical tools, and sublimated affective investment in the patient.

(p. 103)

Important as this argument is, it seems to me that Hoffman leaves room for this on the clinical side of his dialectic. Kernberg’s third position has always been a part of analytic technique in its obligation that the analyst make all possible efforts to relinquish his attachment to personal perspectives and to look at himself objectively. Where they differ is that Kernberg seems to believe the analyst can reach such a self-objectivizing position, while Hoffman accepts the value of trying but also believes it is essential for the analyst to acknowledge that it can never be achieved.

The dialectical thinking Hoffman advocates has its greatest justification in a larger issue that need not be particular to an intersubjective approach. It is death that provides the ultimate example for constructivism in the limitations, finitude and meaninglessness it imposes on life. To fully face this condition of existence would make it impossible to go on, but to deny it would be unrealistic. Instead Hoffman calls for an ‘ironic attitude’ in which there are both awareness and avoidance. From this ontological outlook Hoffman builds his epistemological theory. One can imagine taking the same insight about existence and using it in other non-intersubjective ways, as Freud did, but few since Freud have taken on the task. Melanie Klein’s deep engagement with the death drive led in another direction, away from irony and toward the awareness of inborn aggression.

Along with Harold Searles’s earlier work, Hoffman brought the notion of analytic anonymity into question, claiming a degree of legitimacy for interventions expressing the analyst’s personal reflections. Clearly considered unacceptable in Freud’s theory, then briefly tested by Ferenczi in his experimentation with mutual analysis, self-disclosure had always been shunned by mainstream analysts. However, through a series of provocative clinical examples, Hoffman led the way in bringing about a serious reconsideration of it. He argued that, by allowing the analysand occasional access to the analyst’s personal thoughts and feelings, greater depth could be achieved in the ultimate aim of understanding the patient, and trust in the analyst strengthened. By demonstrating that the ‘rituals’ of analysis (e.g. anonymity) could be questioned, the analyst was also modeling the analysand’s potential to question his own self-definitions. Although self-disclosure certainly creates a de-mystification of the analyst, Hoffman has been clear that it is important that the analyst maintain a mostly unknown and idealizable presence. There is no absolute rule to follow with regard to remaining anonymous or revealing one’s thoughts in his approach. He feels that strict adherence to the rituals can be as exploitative as the lack thereof: there is always “more than one way to look at something and more than one right way to act” (Hoffman, 1999, p. 907). Every intervention must be decided in the moment, and will be based on an amalgam of analytic insight and adherence to the rituals along with spontaneous personal responsiveness.

Although it is well recognized that Freud made occasional lapses in the rule of anonymity, and that most analysts always have, Hoffman’s documentation of such events, and his assertion that they do not necessarily constitute analytic errors, produced considerable outcry from ego psychologists. Intersubjectivists are frequently criticized for such acts, and characterized as using self-disclosure routinely rather than as the rare event it appears to be – at least for Hoffman. Nevertheless, as Greenberg (2001) has observed, the clinical examples presented by Hoffman and many relationalists invariably present breaches in the traditional analytic frame, and this gives the impression that such breaches are the norm rather than the exception. Bollas (2001) has presented a different critique, stating that, although there is a place for self-disclosure by the analyst, it is implicit in standard technique. For Bollas, recognition of the analyst as an individual emerges through the unconscious content of all the analyst’s communications, and is picked up on by the patient’s unconscious. The core of analytic work is done in that unconscious to unconscious communication:

By remaining in the background and giving the patient the fundamental right of free speech, the psychoanalyst limits the contribution of his own personality to the analysis. Yet by surrendering himself to his own unconscious, he is deeply present as an intersubjective participant, communicated through the textures of silence, echoed words, affective affinities, questions, and interpretations. [To do more] destroys the analysand’s freedom to use the analytical object.

(Bollas, 2001, p. 102)

Benjamin with Hegel and Winnicott

Jessica Benjamin based her ideas on an integration of Hegel’s master–slave dialectic with Winnicott’s theories and recent infancy research. As in Hegel’s philosophy, her intersubjectivity prioritizes the ability to “recognize the other as an equivalent center of experience” (Benjamin, 1995, p. 28), i.e. a separate subjectivity.

The premise that there are two persons – an interpersonal dialectic – is a necessary but insufficient condition to constitute intersubjectivity... To get at the core of intersubjectivity, we must go beyond the mere knowledge that there is an interpersonal situation to the felt conviction of two-ness, in which the other person’s independent contribution is recognized.

(Benjamin, 1991, p. 530)

Feminist theory has also figured large in Benjamin’s work, and she has led the charge of making it (and gender studies) an important element of more contemporary relational theory (Benjamin, 1998). She draws on it to heighten the focus of psychoanalytic developmental theory on the interaction between mother and infant where she finds the origins of recognition and difference. The recognition of the other arises within an intersubjective developmental trajectory but, rather than emerging out of a Hegelian struggle, it emerges from a Winnicottian process of fantasized ‘destruction of the object’ which the object survives. The child develops a capacity to recognize the mother as a being with her own needs and desires that are independent of the child’s, rather than as an object who is there to meet each of the child’s needs. Benjamin maintains unresolved tension between two opposing trends: on the one hand, the fantasy of omnipotence and absolute independence with accompanied objectification and subjugation of the other, and, on the other, acceptance of the autonomy and power of others. Difference, not similarity, is what is important here. Objects are representations and can be manipulated to satisfy desires, but subjects exist in a reality that is always beyond grasp. Like Hoffman’s dialectic, Benjamin recognizes an alternation between fantasy and reality, the intrapsychic and the intersubjective. To emphasize one over the other would lose what is central to being human, the dynamic tension that gives each its meaning: “It is in contrast to the fantasy of destruction [i.e. the omnipotent fantasy] that the reality of survival is so satisfying and authentic” (Benjamin, 1995, p. 41). Through development discovering the difference that reality offers becomes a source of pleasure rather than a necessary ‘detour’ for pleasure or an assault on the ego. Benjamin’s intersubjectivity is meant to include classical intrapsychic theory, not to replace it. She emphasizes the need for both perspectives.

Given the similar attention to dialectical processes in the analytic setting there is much that Benjamin has in common with Hoffman. Their differences lie primarily in the theoretical bases for justifying and elaborating their ideas. Benjamin draws heavily from research on child development along with Winnicott. This brings her back repeatedly to the mother–infant relationship and how it changes over time. Hoffman finds his fundamental paradigm in the reality of mortality; Benjamin finds it in the transformations that occur in the child as it develops from egocentrism to intersubjectivity. This places them both solidly within an empirical, reality-based realm in the end in spite of the necessity of constructivism for Hoffman and fantasy for Benjamin. It also differentiates Benjamin from Ogden, who sees intersubjectivity as the original condition for the newborn.


Central to Ogden’s intersubjectivity is his notion of the ‘analytic third’. This is a concept he arrives at through examining Kleinian positions and projective identification, and Winnicott’s developmental progression from primary maternal preoccupation, to mirroring, to transitional relatedness and, finally, to object usage. In the progression from the primitive to the mature Ogden charts the movement from a primary condition of intersubjectivity to a later one of subjectivity, and is explicit in not assigning a hierarchy of value or health to one over another. The analytic third is the intersubjective space that is created in the transference–countertransference of any analysis. It is the relationship-as-subject that each contribute to and draw from, but that can only be accessed via reverie or bodily sensations. As Reis (1999) has pointed out, Ogden’s intersubjectivity goes beyond Hegel and Husserl, and brings the phenomenology of Merleau-Ponty’s ‘body subject’ into psychoanalysis. For both Ogden and Merleau-Ponty intersubjectivity is where we start from, prior to the emergence of individual subjectivity, and is based in the body, not in the mind. Most other theorists consider intersubjectivity as developing secondarily as a product of mind, and as an achievement after subjectivity and self-consciousness have developed. It is also important to note the creative quality of the analytic process that Ogden describes. Each analysis is unique to the analytic partners: “It is the analyst’s responsibility to reinvent psychoanalysis for each patient and continue to reinvent it throughout the course of the analysis” (Ogden, 2005, p. 6). Analysis relies on the ability of each participant to come alive within that relationship. He makes much use of Bion, particularly notions of alpha function, reverie and dreaming:

In participating in dreaming the patient’s undreamt and interrupted dreams, the analyst is not simply coming to understand the patient; he and the patient are together living the previously undreamable or yet-to-be-dreamt emotional experience in the transference–countertransference.

(Ogden, 2005, p. 8)

It is the engagement with unconscious process, via ‘dreaming,’ reverie, and the third, that is central to Ogden’s approach. None are directly accessible through willful, conscious effort, but are achievements one gleans indirectly from unbidden experiences and metaphoric elaborations. Writing, particularly poetry, becomes a part of his intersubjectivity in a series of papers on the relationship between writer, text and reader (1997). Each member of the writing–reading dyad must bring his own creativity to the experience, as in analysis, to give life to the text. In reading Ogden one feels it is not just that speech and listening in the analytic setting, like writing and reading, are saturated with potential to create new life, but that Ogden is presenting a philosophy of life in general for which psychoanalysis is a paradigm. In that regard, along with the association to Merleau-Ponty described by Reis, Ogden is also well placed alongside Ralph Waldo Emerson and the American Transcendentalists. This situates him opposite the pragmatism and empiricism of Sullivan’s interpersonalism and much of relational psychoanalysis. Among analysts Ogden’s intersubjectivity is closer to the Kleinians and the intersubjectivity of Christopher Bollas. Like Bollas his intersubjectivity is manifested through unconscious to unconscious communication, not through a particular dialogue about what is transpiring between analyst and patient.

Clinical implications

Intersubjectivity presents correctives to American psychoanalysis from three different perspectives. For some it represents a major revision to psychoanalytic metapsychology, and this is most clearly the case for those who use it to justify abandoning the drive concept and replacing it with a social idea of the psyche that is only a weakened, incomplete part of itself when viewed as an isolate. Another perspective regards intersubjectivity as providing a new model for clinical technique in which the ideal of the remote, austere analyst observing and interpreting the patient with objectivity is replaced by a creative, interactive, and unabashedly human analyst who is open to acknowledging his contributions to the unfolding process. A third point of view that I have added here concerns the social and political aspects of the psychoanalytic establishment in the USA. It regards intersubjectivity, specifically in its manifestation in the relational school, as a means for a disenfranchised professional group to form its own organizations and claim an equal, if not greater place, at the table of psychoanalytic scholarship and legitimacy.

If the influence of intersubjectivity has been greater with respect to clinical technique than with metapsychological innovation, how has it altered psychoanalytic practice in the USA? This is difficult to answer given the fact that all theoretical groups evolve from within at the same time that their evolution is propelled from without. One does get a sense from reading contemporary ego psychologists that their work has been significantly affected by intersubjectivity. There is an outpouring of interest in enactments, increased reference to the fallibility and subjectivity of the analyst, a less authoritarian stance in the use of analytic authority, and greater appreciation of the uniqueness of every analytic dyad and the creative aspects of the analyst’s work. One wonders if these changes are really nothing new at all, but merely practices that had been de-emphasized in clinical reports due to the prevailing cultural climate. Additionally, over the same 30 years ego psychology was evolving on its own in these directions as challenges to the refrigerator analyst from people like Leo Stone, Hans Loewald, Roy Schafer, James McLaughlin and many others began to take hold. The idea of a cool and remote analyst gradually faded as a mode of personal warmth and responsivity became acceptable. Over those same years Kleinian theory and the British independents gained increasing influence in the USA, along with our home-grown self psychology and interpersonalism. Intersubjectivity is clearly one part of a broad shift in psychoanalytic theory and practice. The corrective intersubjectivity has brought to psychoanalysis of shifting the analyst’s attention toward the manifestations of relationship is a corrective that much of psychoanalysis was only too ready to receive.

It is a mistake to assert there is a single clinical technique for either intersubjectivists or for relationalists. A recent paper by Carmeli and Blass (2010) describes the relational model of treatment as one “in which one helps the other through encouragement, suggestion, and offering a more benevolent environment” (p. 223). They find this approach theoretically “regressive” (p. 223) because it resorts to “an educational, corrective position” (p. 223). Perhaps this is true for some but it seems like a distortion if applied to the entire group. I have tried to show that there are a number of intersubjectivists, some of whom also consider themselves relationalists, for whom this description would not fit. Such a critique reflects issues Sass (1989) raised in his discussion of humanism and hermeneutics. If applied to the description by Carmeli and Blass, relationalists would be relying entirely on a humanistic approach while traditional technique would rely on a hermeneutic model. Yet Benjamin and Hoffman have each been quite clear in their advocacy for a mixed approach, one that balances humanism with hermeneutics.

That dialectic, which echoes throughout psychoanalysis, can also be considered in theoretical terms as object vs. drive or, in clinical terms, as attention being directed to the transference–countertransference vs. the unconscious. As the clinical focus has tilted away from the unconscious contents of the patient’s mind and toward the transference–countertransference process within the dyad, there is a move toward humanism and away from hermeneutics. When the balance goes beyond a certain point toward what is manifest in the here and now of the relationship, the ability to engage with the unconscious becomes more limited. It is this imbalance that is troubling to so many traditional analysts. Wallerstein (1998) describes his own position as ‘both and’ in this regard. Like Benjamin and Hoffman, he believes it important to include both perspectives rather than forcing a choice between one or the other. Doubts that intersubjectivists really believe in this both and-ism appear to stem from their clinical reports which typically stress the humanistic side. I would assume that inconsistency is due to what Green (2000, p. 10), referring to Freud and other innovators, called the need “to emphasize what is new in their conceptions”. Green, like Wallerstein, endorsed a dialectical perspective on the ‘drive–object equilibrium,’ emphasizing their mutual dependence.

Is it possible to find a balance that includes both sides of this dialectic? As an analyst who completed his training in the early 1990s just as these theorists were gaining a solid footing, the influence from intersubjectivity was considerable. These authors opened the door to reconsidering my clinical approach in ways I would not have without them: if they could “throw away the book” (Hoffman, 1998, p. 194), maybe I could too. Only once I felt licensed to be more creative in my technique and discovered clinical dividends from it was I able to begin to notice how it simultaneously led to losses in other aspects of the work. By encouraging analysts to be continuously considering the transference relationship, Gill’s approach could create a kind of interpersonal claustrophobia that lacked the space for thinking that came from sources outside the dyad, and outside the manifest content. The transference is not just about the here and now, and the consciousness-based dialogue between the players does not tell the entire story. I needed a second corrective, but this time from a more traditional approach, to bring me back to the unconscious.

Of the writers described in this essay, Ogden has foregrounded the importance of the unconscious more than the others. Even though he works primarily in the here and now of the transference, he listens for an Other he calls the intersubjective analytic third. The third provides a voice of difference that is a product of the analytic relationship at the same time that it speaks from a place outside the relationship that is unique to the individual. This has bearing on the conduct of the treatment, requiring silence from the analyst when there might otherwise have been dialogue. Bollas, quoted by Wallerstein (1998), put it this way at a panel presentation:

Is it not vitally important for us to know when we are being used as an object of thought for the elaboration of the analysand’s state of mind and when we are being coerced as an other, aimed into the interpersonal? Does not the former state call for our quiet and our sustaining the rights of thought itself?

(p. 1032)

We can wonder where we would be today in the evolution of psychoanalysis without the arrival of intersubjectivity and relational theory. For me they were crucial to my development as an analyst but, as with any theory, served as a model to include rather than as an end in itself. These writers suggest a range of new ways of working with the transference–countertransference and of understanding one’s role as an analyst, which I consider along with the other ways I work. Analysis has a liberating, emancipatory agenda. Over the course of many years, and from the work of a wide range of individual theorists, intersubjectivity has brought that revolutionary spirit to psychoanalytic theory. There is a place for that spirit, in its dialectic with tradition, in our clinical enterprise.


  • 1

    New York University Postdoctoral Program in Psychotherapy and Psychoanalysis.

  • 2

    Psychoanalytic Dialogues: A Journal of Relational Perspectives.

  • 3

    The International Association for Relational Psychoanalysis and Psychotherapy.

  • 4

    Donnel Stern, a relational analyst, provided this view. Benjamin sees herself as a relational analyst who works in the theoretical ‘category’ of intersubjectivity (personal communications).

  • 5

    For a more complete presentation of this topic, see Frie and Reis (2001) and Sass (1989).

  • 6

    The other analyst influenced significantly by Heidegger was his one-time student, Hans Loewald. Loewald is often cited by intersubjectivists as an important progenitor.

  • 7

    Ricoeur called psychoanalysis a ‘mixed discourse’ that is defined by an interpretive approach, based in the coherence of understanding and meaning, as well as by a mechanistic approach, based in explanation and force.

  • 8

    “The first is based on a glance that sees nothing: the King and then the police. The second is based on a glance which sees that the first sees nothing and deceives itself into thereby believing to be covered by what it hides: the Queen and then the Minister. The third is based on a glance which sees that the first two glances leave what must be hidden uncovered to whomever would seize it: the Minister, and finally Dupin” (Lacan, 2006, p. 10).

  • 9

    Hoffman refers to it not as intersubjectivity, but rather as the contemporary ‘paradigm shift’ in psychoanalysis, which he calls one of ‘constructivism.’

  • 10

    Philip Bromberg, Bernard Friedland, James Fosshage, Emmanuel Ghent and Stephen Mitchell.

  • 11

    Most prominent among them were Erich Fromm, Karen Horney and Clara Thompson.

  • 12

    As theorized by Charles Sanders Pierce, William James and John Dewey.

  • 13

    A theory of philosophy by Percy Williams Bridgman defined in Wikipedia as: ‘the process of defining a fuzzy concept so as to make the concept measurable in form of variables consisting of specific observations.’

  • 14

    “Since I began doing psycho-analysis and intensive psychotherapy nine years ago, I have found, time after time, that in the course of the work with every one of my patients who has progressed to, or very far toward, a thoroughgoing analytic cure, I have experienced romantic and erotic desires to marry, and fantasies of being married to, the patient. Such fantasies and emotions have appeared in me usually relatively late in the course of treatment, have been present not briefly but usually for a number of months, and have subsided only after my having experienced a variety of feelings – frustration, separation-anxiety, grief, and so forth – entirely akin to those which attended what I experienced as the resolution of my Oedipus complex late in my personal analysis – specifically, about five years ago” (Searles, 1959, p. 284).