Psychotherapy Integration: a Postmodern Critique – Jeremy D. Safran and Stanley B. Messer

Pluralism is “an antidote to parochialism and the attitude that absolute certainty is obtainable”, and contextualism is “the hypothesis that an event cannot be studied as an isolated element, but only within its setting,” we are reminded by Safran and Messer. Pluralism and contextualism are two of the main characteristics of the postmodern attitude, which comes up against the pretences of absolute and neutral knowledge that are often considered to be the hallmark of modernity. Whether these pretences belong to modernity as such–as an inevitable aftermath of the Enlightenment–or they are better understood as symptoms of the unachievement of the Enlightenment project, as some maintain, it is a controversial matter. An injection of pluralism and contextualism is in any case a remedy for those pretences. As for all medicines, on the other hand, its beneficial effect is dose dependent. While a proper dose of pluralism and contextualism is surely beneficial, an excessive dose could produce some side effects. One of these is certainly relativism, which is a risk for postmodernism as the uncritical objectivation of knowledge was for modernity. The Authors are well aware of such risk, and point to dialogue as to an antidote to it: “The reason that dialogue is critical is because it provides a means of moving beyond our preconceptions toward a better understanding of the things themselves”. To dialogue means to communicate on the common ground of the logos, that is on the ground of the universal reason, and it is made possible by a movement beyond all preconceptions. But what remains of pluralism and contextualism, once our preconceptions are removed? Aren’t they the very substance of our preconceptions? And wasn’t a postmodern tenet that there is nothing like “the things themselves”, as there are only “things within their settings”? This contradiction is only apparent, if we consider that postmodernism is not, and cannot be, an ideology to replace modernism (it would be self-contradictory, if it had this pretence). Postmodernism, at its best, is a healthy reaction against rationalism, i.e. the tyranny of reason. As such, it is not a substitute for the logos, but a correction for its misuse and abuse. As there is no choice between one and many, by the same token there is no choice between modernity and postmodernism. What is needed, in both cases, is a dialogical-dialectical perspective, as Safran and Messer suggest throughout their article. 

Psychotherapy Integration: a Postmodern Critique 


This article critiques major trends in the psychotherapy integration movement from the postmodern perspectives of contextualism and pluralism. A contextualist position asserts that psychotherapeutic concepts and interventions can be understood only within the linguistic, theoretical, and ideological frameworks in which they are embedded. Therefore, they take on new meanings when extracted from their original context and are incorporated into an eclectic therapy. Pluralism holds that there is no single theoretical, epistemological or methodological approach that is preeminent and no one, correct integrative system towards which the field of psychotherapy is evolving. In light of this critique, we argue that the goal of the integration movement should be to maintain an ongoing dialogue among proponents of different theories and world views thereby allowing for the clarification of differences as well as the judicious integration of alternative perspectives and techniques. The article also spells out the implications of contextualism and pluralism for psychotherapy theory, practice, and research.

The last two decades have witnessed the beginning of an important shift away from the prevailing climate of factionalism and parochialism among the psychotherapies towards one of dialogue and rapprochement (Arkowitz, 1992; Bergin & Garfield, 1994; Norcross & Goldfried, 1992; Stricker & Gold, 1993). Integrative links have been forged, for example, among psychodynamic, behavioural, and family system therapies (e.g., Wachtel & McKinney, 1992), and among experiential, cognitive, and interpersonal approaches (e.g., Safran & Segal, 1990). Commonalties across the different therapies have been distilled into single therapies (e.g., Garfield, 1992; Prochaska, 1995) and techniques from several sources have been employed eclectically in connection with the differing needs of individual clients (Beutler & Hodgson, 1993; Lazarus, 1992). A poll which surveyed clinical psychologists, marriage and family therapists, psychiatrists, and social workers has documented that from 59% to 72% endorse eclecticism as their preferred approach (Jensen, Bergin, & Greaves, 1990).

In this article we adopt a postmodern perspective to critique common approaches to psychotherapy integration and to highlight the more radical implications of the integration movement for shaping our attitudes towards psychotherapy. A recurring theme in postmodern discourse, deriving originally from Hegel (1910), is that self-identity emerges only through the construction of “the other”. The unfortunate effect of this construction is that “the self” always gets defined in contrast to “the other”, who is thereby deprived of genuine standing. This functions to validate and maintain the privilege of the self or of the dominant group. Foucault (1967), for example, argues that during the 18th century the insane (the “irrational”) were placed in the category of “the other” as part of the process of protecting and enshrining the rationalistic values of the enlightenment. An important function of postmodern critique is to challenge constructions of reality which have the effect of marginalizing “the other”.

One way to view the recent trend towards psychotherapy integration is as a response to confrontation with “the other”. In the conventional discourse that has taken place among therapeutic traditions, each approach has defined itself in contrast to the other. For example, psychoanalysis is defined in contrast to behaviour therapy by its emphasis on the unconscious, and behaviour therapy is defined in contrast to psychoanalytic therapy by its emphasis on social influence. As in the case of racial, ethnic or cultural differences, perceived positive qualities of one’s own group take on ritualistic significance whereas other traditions are assigned a negative, caricatured quality. The other is thus appropriated and used to define and enshrine the values of the self (Sampson, 1993).

From a postmodern perspective, one of the most important functions that the psychotherapy integration movement can serve is to help theorists and practitioners move beyond the attitude of superiority, contempt, and aversion which frequently arises from the confrontation of adjoining therapeutic “cultures” towards a sense of surprise and eagerness to learn, which is also a natural human response to difference (Feyerabend, 1987). One can compare the task of the psychotherapy integrationist with that of the cultural anthropologist. Shweder (1991) refers to the “astonishment of anthropology” in describing the core value which orients the cultural anthropologist’s stance:

Astonishment and the assortment of feelings that it brings with it — surprise, curiosity, excitement, enthusiasm, sympathy –are probably the affects most distinctive of the anthropological response to the difference and strangeness of “others.” Anthropologists encounter witchcraft trials, suttee, ancestral spirit attack, fire walking, body mutilation, the dream time, and how do they react? With astonishment. While others respond with horror, outrage, condescension, or lack of interest, the anthropologists flip into their world-revising mode.

Such a consciously inculcated stance of astonishment is one of the most valuable attitudes that can emerge from the psychotherapy integration movement. To the extent that confronting alternate therapeutic paradigms and techniques flips us into a “world-revising mode,” versus the more common stance of outrage and condescension, there is the possibility of its leading to a dialogue which can truly deepen our understanding of the human change process. The importance of dialogue of this type is a recurring theme throughout the article and later we will explore the central role that it plays in the scientific enterprise.

In the following, we critically examine the three most frequently employed strategies for psychotherapy integration–technical eclecticism, common factors, and theoretical

integration–in light of two defining characteristics of the postmodern attitude: contextualism and pluralism. We also explore the obstacles to integration that emerge at metatheoretical and epistemological levels of discourse. The article concludes with the implications of contextualism and pluralism for psychotherapy theory, practice, and research. 

Contextualism is the hypothesis that an event cannot be studied as an isolated element, but only within its setting. Every event is said to have quality and texture. Quality is the total meaning of the phenomenon, and texture refers to the parts that make it up (Pepper, 1942). Quality entails a fusion of the textural details, e.g., “Lemon, sugar, and water are the details of the taste, but the quality of lemonade is such a persistent fusion of these that it is very difficult to analyze out its components” (Pepper, 1942, p.243, after William James). The postmodern notion that there is more than one correct theory or perspective by which to view any phenomenon is known as pluralism. It is an antidote to parochialism and the attitude that absolute certainty is attainable. Seeing how other theories get a grip on the world can lead to enhanced understanding and improvement of the theoretical ground on which one stands (Nozick, 1981). While contextualism notes that context often determines which of many possible interpretations or meanings we give to an event, pluralism acknowledges that there are multiple perceptions of truth, each one influenced by the context out of which the perceiver arises in making his or her judgments.

 Technical Eclecticism

 There has been discussion in the psychotherapy integration literature as to whether integrative efforts should have a more applied or a more theoretical emphasis (Garfield, 1994). Technical eclecticism holds that theoretical integration involves fusing theories that are irreconcilable, and that techniques should be combined pragmatically on the basis of observed or presumed clinical efficacy ( Lazarus, 1996; Lazarus, Beutler, & Norcross, 1992). Lazarus’s multimodal therapy is a good example of this approach. Techniques from gestalt, cognitive, behavioural, psychodynamic, and family systems therapy all may be applied in one individual’s therapy.

One of the problems with this form of eclecticism is that it often proceeds as if a therapeutic technique is a disembodied procedure that can be readily transported from one context to another, much like a medical technique, without consideration of its new psychotherapeutic context ( Lazarus & Messer, 1991). The problem can be illustrated by reference to the hermeneutic circle which stresses the contextual nature of knowledge (Messer, Sass, & Woolfolk, 1988). Within this view, a fact can be evaluated only in relation to the larger structure of theory or argument of which it is a part, even while the larger structure is dependent on its individual parts. Thus a therapeutic procedure such as an interpretation or empathic response does not stand on its own, independent of the framework of meaning created by the entire therapeutic system.

This part-whole interdependence can be illustrated in various ways. For example, a client whose treatment has been primarily cognitive-behavioural may experience a therapist’s shift to empathic/reflective responding as a withholding of needed psychological expertise. Conversely, a client whose treatment has been client-centered or psychoanalytic may experience a shift to cognitive-behavioural interventions as controlling. Although such interventions have the potential to be effective, their meaning and impact should be explored in their new context. (e.g., see Frank, 1993, Messer, 1992).

In a second type of technical eclecticism, different therapies or techniques are prescribed as optimal for different kinds of problems or clients, rather than combined in one client’s treatment. This is known as prescriptive matching (Beutler & Clarkin, 1990; Beutler & Harwood, 1995), differential therapeutics (Frances, Clarkin, & Perry, 1984), or selective eclecticism (Messer, 1992). In asking the question: “Which therapy is best for which type of client?”, selective eclecticism is a movement towards greater contextualization of therapy.

The prescriptive matching approach, however, ignores the fact that two clients with the same diagnosis often have very different case formulations (Collins & Messer, 1991; Persons, 1991). Moreover, clients change both within one session and over the course of therapy. This requires the skilled clinician to constantly modify interventions in a context-sensitive fashion in attunement with a changing process diagnosis, rather than applying a therapy module in response to a static diagnosis or formulation (Rice & Greenberg, 1984; Safran, Greenberg, & Rice, 1992). The failure to conduct psychotherapy research in a sufficiently context-sensitive manner is probably one of the factors underlying the difficulty demonstrating a consistent pattern of therapist by client interactions (Beutler, 1991; Omer & Dar, 1992). It is thus important for psychotherapy researchers to conceptualize relevant variables in more process oriented, phase specific terms that take ongoing context into account. 

Common Change Principles as Integration

A second form of psychotherapy integration consists of the discernment of common principles of change across different therapies (e.g., Frank & Frank, 1991; Goldfried, 1980; Weinberger, 1995). For example, a common principle in many forms of psychotherapy consists of helping clients to become aware of and challenge their self-criticism. A closer look at the ways in which this is accomplished in different therapies, however, reveals important distinctions. In the scientific and rationalistic spirit of cognitive therapy, clients are encouraged to challenge self-criticism by treating their negative thoughts as hypotheses to be tested through examining relevant evidence, or by considering alternative perspectives. In gestalt therapy, by contrast, self-criticism is challenged by means of eliciting an emotional experience through what is known as “the empty chair” exercise. In this approach, clients’ self-criticism is expressed while sitting in one chair, and then confronted by their emotional reaction to it while sitting in a second chair.

Although both of these techniques share the common principle of “challenging self-criticism,” important differences emerge when we take into account the theoretical context in which interventions are employed (Goldfried & Safran, 1986). The hypothesis-testing intervention in cognitive therapy takes place within a theoretical framework which views self-criticism as maladaptive thinking to be recognized, controlled, and eliminated (Messer & Winokur, 1984). It is embedded in a modernist world view, which values rationality, objectivity, and pragmatism (Woolfolk & Richardson, 1984). Gestalt therapy, by contrast, regards self-criticism as an aspect of the self which must be recognized and then integrated with other parts of the self. In this therapy, the values of emotional experiencing, subjectivity, and the complexity of personality are paramount.

Since different therapies convey different over arching values or messages (Beutler, Crago & Arizmendi, 1986; Kelly & Strupp, 1992), any intervention must be understood as part of a general process through which such values are transmitted to the client. In the attempt to extract common principles one can lose sight of important features of the overall therapeutic system and the process through which it works. As Wittgenstein (1953) once remarked, it is a mistake to try to get to the essence of an artichoke by divesting it of its leaves.

We are not arguing that there is never any value to extracting common principles. The utility of the specific common principle that is articulated, however, depends upon the function that it serves in the phase of dialogue between systems of therapy. In early stages of integration, the articulation of common principles can play an important role in facilitating dialogue where none previously existed. In this way it can help to reduce the sense of “otherness”. As the dialogue progresses, however, it becomes more critical to explore similarities and differences between orientations from a more nuanced perspective. Anthropologists refer to this type of contextualized exploration as “thick description” (Geertz, 1973). Thick description provides a corrective to older forms of anthropological investigation which are more likely to assimilate aspects of new cultures into existing knowledge structures (Schwartz, White & Lutz, 1992). Geertz has argued that it is only by understanding each culture in its uniqueness that we can learn something new about the human condition. Similarly, the exploration of other therapeutic systems in a refined, contextualized fashion can lead to new understanding of both other systems and our own.

Thus differences among therapies in their higher level theoretical constructs should not be ignored. Theories have a “trickle down” effect on clinical practice. To return to our earlier example, challenging self-criticism may convey a different message in the approach of a therapist who subscribes to a theory that self-critical thoughts are distortions to be eliminated, than it will in the approach of a therapist who views them as reflecting a part of the self containing the seeds of important strengths. 

Theoretical Integration 

In this form of integration, different theories are combined in the attempt to produce a superior, over arching conceptual framework. Wachtel’s (1977) joining of psychoanalytic and behavioural theories within an interpersonal psychodynamic framework, and Safran and Segal’s (1990) wedding of cognitive, experiential, and interpersonal approaches within a single theory of therapy are good examples of this genre. Such super ordinate integrative theories are said to lead to new forms of therapy that capitalize on the strengths of each of its elements.

While the integration of pure form theories into one that is super ordinate may bring certain advantages, the integrative theory could lose some of the practical wisdom that has evolved over time in its component therapeutic systems. In the same sense that interventions cannot be understood outside the context of the theory in which they are embedded, a theory of therapy cannot be fully comprehended without reference to the details of its clinical implementation. As Geertz (1983) suggests, in order to truly understand a culture there must be “a continuous dialectical tacking between the most local of local details and the most global of global structures in such a way as to bring them into simultaneous view” (p.69). Similarly, a proper appreciation of a therapeutic approach requires a tacking back and forth between theory and the specifics of its implementation.

Organicism Versus Pluralism 

There are other potential problems with theoretical integration, to which a postmodern outlook alerts us. The task is sometimes approached as if there were one correct integration waiting in the wings to be discovered. Labeled “organicist” by the philosopher Stephen Pepper (1942), this perspective (or “world hypothesis” as he calls it) presumes that by organizing data at a higher level, the appearance of conflict between ideas or findings is resolved by their incorporation into an organic whole. Organicism posits that in the world we encounter fragments of experience — such as the observations of a school of therapy. These appear with certain contradictions, gaps, or opposition from other fragments of experience — such as the observations of other theories of therapy. The various fragments have a tendency to be resolved by incorporation into an organic whole that, all the while, was implicit in the fragments and that transcends them. In this view, progress in theoretical integration is achieved by including more and more of the fragments into a single, integrated, and unified whole.

There is an alternative view to organicism, namely that psychology, by its very nature, is pluralistic: “Paradigms, theories, models (or whatever one’s label for conceptual ordering devices) can never prove preemptive or preclusive of alternate organizations” (Koch, 1981, p. 268). The pluralistic perspective holds that all theories are necessarily limited and that the best way of approaching the truth is through the ongoing confrontation of multiple, competing theories with data and with each other.

Integration as Translation

Theoretical integration typically involves some element of reconceptualization or translation from one framework into another. For example, in an attempt to place the insights of psychoanalytic theory on a firmer scientific footing, Dollard and Miller (1950) translated psychoanalytic concepts into learning theory. Contemporary examples include drawing on concepts from cognitive psychology to refine psychoanalytic theory, such as efforts to account for the phenomenon of transference in terms of schema theory (Safran, & Segal, 1990; Singer & Singer, 1992; Westen, 1988), and attempts to reformulate the psychoanalytic theory of the unconscious by means of cognitive theory (Erdelyi, 1985). Within a contextualist view, however, language and theory are inextricably intertwined, which forces us to consider carefully what has been added by the translation. Psychological meanings only make sense by virtue of their interrelations to other terms within their conceptual setting. Thus, for example, while attempts to translate a concept from one theory into the terms of another may result in ease of empirical testability, some of the concept’s richness and subtlety could be lost.

Translation can also lead to the reductionistic fallacy which holds that theory A (regarded as nonscientific) is more adequately explained in terms of theory B (regarded as scientific). For example, it is a mistake to assume, a priori, that the principles of Chinese medicine can be better explained in terms of the principles of Western medicine. As Sampson (1993) argues:

To examine a culture’s own system of understanding requires us to become familiar with the culture in its terms, rather our own. This requires a dialogic rather than a monologic approach. We must carry on a dialogue with the other culture. In this dialogue our framework and theirs meet. Out of that meeting a newly cast understanding of both them and us is likely to emerge.

Metatheoretical Integration

In comparing the visions of reality contained within psychoanalytic, behavioural, and humanistic therapies, Messer and Winokur (1984) have illustrated the difficulties of integration at the metatheoretical level. They argued that psychoanalytic therapy is guided primarily by a tragic view of reality in which people are subject to forces not of their knowing and which can be only partially ameliorated. Behaviour therapy, by contrast, falls more within the comic vision, where conflicts are viewed as external and more readily resolvable. Empirical findings on the process of these two therapies are consistent with this description (Goldfried, 1991). The humanistic therapies, by contrast, are characterized by the romantic vision which prizes individuality, spontaneity, and unlimited possibilities in life.

Fundamental differences in world view are not readily integrated because they are mutually exclusive in many respects and are typically held as unquestionable presuppositions. Nor can they be resolved by reference to the data. What Kuhn (1970) has said about the incommensurability of different paradigms applies here: there is no set of rules to tell us how rational agreement can be reached or that would settle all conflicts between paradigms or world views. It is tempting to think that the relative value of different therapeutic systems can be resolved definitively through psychotherapy research. However, the evaluation of therapeutic outcome is inextricably tied to values and shades of meaning (Messer & Warren, 1990). This is unlike the situation in engineering where a bridge will collapse if the correct method of building it is not employed, or in medicine where a child will die if an incorrect procedure is applied to repair a heart valve.

For example, if an individual comes to accept her shyness and finds meaning in it, can we consider it a good outcome, or does there have to be a substantial reduction in her shyness? Gandhi (1957) maintained that his own shyness had become one of his greatest assets, since it forced him to think before he spoke. If an individual loses his phobic symptoms upon joining a cult, should this be considered a good outcome? Rilke, one of the great poets of the twentieth century, chose to cultivate his pain and solitude in order to deepen his art. Would Wittgenstein’s life have been “better” if he had been happy in the conventional sense? Of course there are some outcomes on which most, if not all, clinicians will agree. For example, few clinicians would argue that reducing suicidal behaviour is not a desirable outcome in the treatment of a severely depressed patient. Differences will, however, emerge when it comes to other types of outcome with the same patient. For example, the existentially oriented therapist is likely to be more concerned with helping him or her to live authentically than the cognitive therapist.

Metatheoretical systems are best thought of as multiple lenses, each of which can bring into sharper focus different phenomena and different aspects of the same phenomenon. For example, while tragic and comic visions cannot easily be integrated they can each be usefully brought to bear in different clinical contexts, and in highlighting different dimensions of one person’s experience. This can be conceptualized as a type of dialectical thinking that allows one to take into account the paradoxes and contradictions that are inherent in life.

A long term psychoanalytic therapist may be suspicious of the good outcomes reported by short term behaviour therapists, seeing these as superficial and unenduring. From a short-term behavioural perspective, the psychoanalytic emphasis on structural change may be viewed as presumptuous insofar as the therapist claims to know what changes clients need to make. Dialogue about this type of issue can lead to questions such as: How ambitious should the therapist be regarding change? How should the therapist and client negotiate differences in desired outcome? When should a reemergence of a problem be considered a relapse, and when should it be considered a new problem? What types of change should health insurance pay for?

This is not to say that research is irrelevant or that clinicians should feel free to define outcome as they will. Rather, different kinds of outcomes emphasized by different therapies must be viewed within the context of the values and visions of life each holds to be true, and this multiplicity of values is merely a reflection of the complex nature of life. Psychotherapy integration does not solve this problem, but serves to highlight it. A postmodern perspective directs us to confront this complexity rather than to gloss over it or ignore it. It encourages us to engage in ongoing dialogue with colleagues who hold different world views. It also encourages dialogue with clients about the tasks and goals of therapy. This type of negotiation constitutes an important part of the process of establishing a therapeutic alliance (Bordin, 1979).

The recent shift in behaviourally oriented theory towards an emphasis on self-acceptance rather than self-control (Jacobson, 1994), provides an example of the type of metatheoretical elaboration that can result from dialogue among different theoretical traditions. Although it has not been uncommon for behaviour therapists to borrow techniques and concepts from other traditions, they are usually assimilated into a fundamental world view which emphasizes the importance of self-control. By explicitly proposing that change be viewed as self-acceptance, an outlook typically associated with the experiential tradition, Jacobson is challenging the underlying paradigm through which change is understood. The resulting shift does not necessarily have to radically change the specific techniques that are employed, but the different ends to which they are put may affect their ultimate impact.

Messer (1992) has referred to this kind of importation of concepts as “assimilative integration” (pp. 151-155). It is the incorporation of attitudes, perspectives, or techniques from one therapy into another in a way that is cognizant of how context shapes the meaning of foreign elements. This mode of integration favors a firm grounding in any one system of psychotherapy, but with a willingness to incorporate or assimilate perspectives or practices from other schools (see also Stricker, & Gold, 1996). This is an evolutionary process in which the contact with difference leads to a de facto, even if unacknowledged, integration. However, to carry on such a dialogue with the other in a meaningful fashion, one must be knowledgeable about and firmly rooted in at least one tradition, and know where one stands. 

Integration at the Epistemological Level

Different therapeutic traditions tend to be associated with different epistemological stances, and this also creates an obstacle to integration. A survey by Morrow-Bradley and Elliott (1986) found that, in general, practicing therapists find little of value in psychotherapy research, and that psychodynamically oriented therapists are less likely to make use of psychotherapy research findings than are their behavioural peers. The behavioural tradition subscribes to the epistemological stance of logical empiricism (Scriven, 1969) and its associated methodology of experimental research. The empirical/experimental method of truth-seeking, which psychologists have adopted from the natural sciences, relies heavily on observation, laboratory studies, elementism, and objectivism (Kimble, 1984; Krasner & Houts, 1984). It stems from the philosophy of scientific modernism which includes the belief that nature has an existence independent of the observer and is accessible to the operations of the human mind (Schrodiner, 1967). Findings are presumed to be context-free and lead to universal, nomothetic laws.

Psychoanalysis, by contrast, has traditionally been associated with an epistemological stance which is more hermeneutic in nature (Messer, et al., 1988). Under Brentano’s influence, Freud distinguished psychology from the natural sciences and instead developed a “descriptive science based on the direct observation of psychological life, with a focus on its meaning” (Wertz, 1993). Psychoanalysis was thus originally understood to be a descriptive and interpretive science rather than an experimental one.

Proponents of psychoanalysis have, to some degree, accommodated themselves to the cannons of experimental research. But, as Hornstein (1993) has stated, “American psychologists did to psychoanalysis what they did to every verstehen-based psychology that arrived on the boat from Europe–they ignored its underlying assumptions, skimmed off what they could use, and repackaged the remaining content in the sparkling language of positivist science” (p. 586). Even while this synthesis of psychoanalysis and experimental method took place, there was never any extensive debate about the fundamentals of scientific practice (Hornstein, 1993). This may account, at least in part, for the failure of experimental research to have had a substantial impact on the practice of psychoanalytic therapy.

For some time now, there has been a call for methodological pluralism in psychology (Polkinghorne, 1984), which we endorse as an important feature of postmodernism. Cook (1985), for example, recommends agreement from independent epistemological perspectives as the best foundation for approximating truth. Similarly, Bevan (1991) warns us to be wary of rule-bound methodology. “Use any method with a full understanding of what it does for you but also what constraints it may place on you…Be mindful of the potential value of methodological pluralism” (p. 479). Such methods may include traditional experimental research, case analysis (both quantitative and qualitative), skilled reflection (Hoshmand & Polkinghorne 1992), phenomenological description, anthropological field studies, action research, and narrative approaches.

Calls for methodological pluralism, however, come up against strong emotional barriers. Hudson (1972), in a book with the ironic title, The Cult of the Fact, suggests that experimentalists (the “tough-minded”) tend to think of nonexperimentalists (the “soft-minded”) as sloppy, even morally remiss, in their unwillingness to treat hard data seriously. Nonexperimentalists, on their part, tend to view experimentalists as mechanistic, dehumanizing, and simpleminded. Part of what is at stake here is the question of what constitutes “science.” A number of philosophers of science from Kuhn (1970) onwards have demonstrated that the process through which science evolves is very different from the picture portrayed in the “standard view” of science (Manicas & Second, 1983). Science has an irreducibly social and interpretive character. Data are only one element in a rhetorical process through which members of a scientific community attempt to persuade one another (Weimer, 1979).

The rules and standards of scientific practice are worked out by members of a scientific community and are modified over time. Many contemporary philosophers and sociologists of science assert that the demarcation criteria between “science” and “non science” are not as clear-cut as they were once thought to be. They argue that the logical empiricist view of science is a reconstruction according to certain criteria of rationality rather than an accurate portrait of the way science really works (Bernstein, 1983; Feyerabend, 1975; Houts, 1989; Kuhn, 1970; Safran & Muran, 1994; Weimer, 1979). The “research-practice split” is thus, in part, fueled by the same type of marginalization of the “other” associated with the contest between different therapeutic orientations. 

Beyond Relativism

The appreciation of the relative merits of different psychotherapies within a pluralist outlook, and the willingness to engage in informed debate about philosophical and epistemological issues can lead to the conclusion that all are equal and “anything goes”. That is, one can confuse openness to other approaches with a kind of intellectual anarchy or wishy-washiness. A relativistic position is said to characterize our culture in this postmodern era in general. It has led critics (e.g., Bloom, 1987) to argue that our culture lacks fundamental moral and political convictions and, in the current relativistic climate, there is “no enemy other than the man who is not open to everything” (p. 27).

Finding a stance that is both pluralistic and nonrelativistic is a central concern for many contemporary philosophers, and a new understanding of the nature of science is emerging (Bernstein, 1983; Gadamer, 1980; Habermas, 1979; Rorty, 1982). A central theme in this understanding is the importance of dialogue among members of the scientific community. This emphasis on dialogue should not be confused with sentimentalism. The point is an epistemological one. A central theme in the contemporary philosophy of science is that our understanding of things is inevitably shaped by our preconceptions. There are no theory-free observations (Hanson, 1958).

The reason that dialogue is critical is because it provides a means of moving beyond our preconceptions towards a better understanding of the things themselves (Gadamer, 1980). Through the process of recognizing our preconceptions and engaging in dialogue with that which is alien, the possibility of seeing beyond our preconceptions emerges. True dialogue involves seeking to listen to and understand what the other is saying, and a willingness to test our opinions through such encounters. Rather than a facile acceptance of alternative positions, true dialogue involves an active engagement in the process of truth seeking. This recognition of the importance of dialogue emerges out of historical and sociological analyses of the way science actually operates rather than the way it should operate. Scientific practice involves deliberation among members of the scientific community, interpretation of existing research, and application of agreed upon criteria for making judgments and debate about which criteria are relevant. The absence of absolute foundations is not equivalent to arbitrariness. Bernstein (1993) refers to the underlying philosophical position as one of “engaged fallabilistic pluralism.” This means “… taking our own fallibility seriously – resolving that however much we are committed to our own styles of thinking, we are willing to listen to others without denying or suppressing the otherness of the other”.

A parallel can be drawn between scientific practice and the process of making judicial decisions (Bernstein, 1983; Polanyi, 1958). Principles of judicial arbitration evolve over time through rational deliberation and precedent. Evidence plays a critical role, but this evidence is always subject to interpretation. Each case must be dealt with in its particularities. Rather than applying universal principles, general rules of argument are given more or less weight depending on the specific nature and circumstances of the case. These contextual features of common law do not make judicial decisions “irrational” or “nihilistic”, but they do make it impossible to adequately model them through universally applicable algorithms.

We are thus advocating ongoing dialogue at all levels of analysis — empirical, theoretical, metatheoretical and epistemological — and not an uncritical acceptance of all therapeutic orientations and techniques. The challenge that psychotherapy theorists and researchers face as we enter the twenty-first century is one of learning to live with an irreducible ambiguity, without ignoring it and without wallowing in it (Bernstein, 1993).

 Implications for Theory, Practice and Research

What are the implications of pluralism and contextualism for psychotherapy theory, practice and research? At a theoretical level we have highlighted, in accordance with pluralism, the importance of maintaining a continuing dialogue among multiple perspectives. Rather than aspiring to one superordinate theory, such a dialogue leads over time to a degree of assimilation of ideas and techniques from one theory or therapy into another.

One might argue that there is a contradiction between stressing appreciation for the “otherness” of the other, all the while critiquing the different forms of integration and advocating some assimilation of them. Critical analysis, however, is part of the dialogue. Although there is an inherent tension between appreciation of difference versus a critique or assimilation of differences, we have argued for a dialectical process between them, and not a facile or wholesale acceptance or rejection of difference.

In line with a contextualist viewpoint, theoretical dialogue must be grounded in the specifics of clinical practice. Just as study of a culture requires tacking back and forth between theory and observational detail, a theory of therapy has to be embodied in the particularities of practice. It is not enough, for example, to discuss the differences between transference and stimulus generalization in theoretical terms. Comparison on the theoretical level must be grounded in clinical material. Cultivating an attitude of astonishment among psychotherapy researchers and clinicians can play a critical role in creating a climate conducive to presenting videotapes and audiotapes of actual clinical material in public forums, thereby facilitating clinically grounded dialogue across theoretical orientations.

An implication of pluralism for practice and training in psychotherapy is that we should be fluent in more than one therapy language and mode of practice (Andrews, Norcross, & Halgin, 1992; Messer, 1987). In the same way that one has to spend time in other cultures in order to truly understand them, one has to immerse oneself in other therapeutic orientations in order to be able to appreciate their strengths and recognize their limitations. Clinical psychology programs are too often conducted within one theoretical perspective, which does not allow students to be “multilingual” and “multicultural” in relation to the multiplicity of existing therapeutic languages and cultures. Aside from book knowledge, the best ways of learning about other approaches is to be supervised in their practice or to experience them as a client.

While pluralism emphasizes our attaining knowledge of several approaches, contextualism highlights the need for clinicians to evaluate a technique they incorporate from a different orientation in the ongoing context of therapy. A technique takes on the colouring of its surround and it must be assimilated in such a fashion that it fits comfortably within the theoretical and clinical framework into which it is imported. One must attend carefully to the effect on clients of such a change in the therapist’s manner, perspective, or technique.

Regarding the implications of postmodernism for research, it is important to find ways to take into account the context and complexity of clinical phenomena. A finding from a randomized clinical trial that a treatment approach is effective with singly-diagnosed clients, does not speak sufficiently to the practicing clinician who has to work with complicated (often dual-diagnosed) clients whose nuances of personality and psychopathology are not readily captured by their diagnosis or the research protocols (Fensterheim & Raw, 1996; Goldfried & Wolfe; Safran & Muran, 1994; 1996). Group designs which study subject variability are unable to mine the context-rich information that can be extracted from the study of intrasubject variability. Although it is difficult to generalize from such single-subject research, this can be accomplished by multiple replications or by combining intensive and extensive analysis (Barlow, 1981; Greenberg, 1986; Kazdin, 1982; Messer & McCann, in press; Safran, Rice, & Greenberg,1988).

Thus, research comparing different treatment modalities at a global level (e.g., cognitive therapy versus interpersonal therapy), or examining client by treatment interactions, should be augmented by research that investigates specific interventions that are effective in specific contexts and the processes that underlie such change. For example, Safran and colleagues (Safran, Crocker, McMain & Murray, 1990; Safran, Muran & Samstag, 1994; Safran & Muran, 1996) have developed an empirically based model of the processes that lead to the resolution of ruptures in the therapeutic alliance. This model specifies which specific therapist interventions will be effective in the context of specific client processes along the pathway to resolution.

Messer and his students have studied the effect of therapists’ competence and their adherence to a psychodynamic focus on the ongoing progress of individual clients. Raters had access to the flow of clinical material thus allowing context to affect their ratings (Messer, Tishby, & Spillman, 1992; Tishby & Messer, 1995). Collins and Messer (1991) adapted Plan Formation methodology (Curtis, Silberschatz, Sampson & Weiss, 1994) to study how case formulations are influenced by the context of a rater’s favored theory.

A fruitful strategy for promoting the development of integrative knowledge can consist of identifying important therapeutic contexts or markers (Rice & Greenberg, 1984) that may be responded to differently by therapists with different orientations (Safran & Inck, 1995). For example, how do different traditions respond to instances of patient self-criticism or to defensive manoeuvres? Are there markers that are favored by, or unique to, specific orientations? By working with these smaller units of analysis (i.e., intervention A in context B) there is an opportunity to get beyond name brand theories, allowing the results to become more accessible and relevant across traditions. It is also closer to a level that is meaningful to clinicians and therefore can be used to guide practice in a complementary way to randomized clinical trials.

Thus, research programs consistent with the spirit of integration need not necessarily evaluate the effectiveness of integrative treatment programs per se. When researchers dialogue with one another, within a spirit of pluralism, around the kind of process research just described, they can more readily absorb results stemming from other viewpoints because it gets around their emotional attachment to a brand name therapy.

Another implication of pluralism for research is the importance of being open minded about methods other than those that are experimental or correlational. Each method has its assets and shortcomings but too often we sacrifice richer, contextual meaning for exactness and narrowly focused certitude. Some combination of quantitative and qualitative methods employed within the same research paradigm, for example, may lead to a better understanding of the complexities of psychotherapy than either approach alone. 


In summary, the development of an open and engaged stance towards integration among theorists can lead to more fruitful cross-theoretical dialogue rather than the advocacy of a premature, unified paradigm (see Mahoney, 1993; Stricker, 1994). The greatest value of the psychotherapy integration movement lies in the creative and growth oriented confrontation with and dialogue about difference, and it is in this process that the payoff lies.

Our call for a more contextually based, pluralistic approach towards psychotherapy integration may seem to some to invite unnecessary complications into a field which is already complex enough. To be sure, there are times when the strategy of simplification through ignoring context or alternative perspectives is the most appropriate way to proceed. Ultimately, it may be best to pursue an ongoing dialectic between the strategy of simplification and that of thick description (cf. Elliott & Anderson, 1994).

The search for a single, unified therapeutic model and laments about the preparadigmatic and unscientific state of psychotherapy theory stem from a misunderstanding of the nature of science. In the natural sciences it is recognized that multiple, contradictory theories are necessary to capture different aspects of the underlying phenomenon, and that a given theory captures some of these aspects at the expense of others (Nozick, 1981). Moreover, contemporary philosophers of science state that science evolves through methodological pluralism rather than a uniform set of procedures and criteria.

Over a century ago, John Stuart Mill (Cohen, 1961), a strong advocate of empirical methods in scientific procedure, argued that a plurality of views is critical for the following reasons:

1. A view which one rejects may be true nevertheless, and to reject it assumes one’s own infallibility.

2. A problematic view may contain some portion of the truth since the prevailing view is never the whole truth. It is only by collision with contrary opinions that the remainder of the truth has a chance of being recognized.

3. A point of view that is wholly true, but not subjected to challenge, will be held as a prejudice rather than on a rational basis.

4. Someone holding a particular point of view without considering alternative perspectives will not really understand the meaning of the view he or she holds.

5. Decisive evidence against a perspective only can be articulated once an alternative perspective is advanced. This results from the fact that evidence in the absence of theory is meaningless.

Both psychotherapy integration and science flourish in an atmosphere of confronting and discussing difference rather than shunning it. Once an integrative system becomes codified, creativity and openness whither. One can become an adherent of an integrative system in the same way that one becomes a cognitive therapist, a Freudian, or a Jungian. A theoretical system is always in danger of becoming a fossilized remnant of what was once a vital insight, even in the hands of the person who developed it. It was presumably for this reason that Jung once remarked (in Progoff, 1953): “I am not a Jungian and I never could be.”