[acb-hsp] sociological perspective on therapeutic alliance ...

[J.Rayl]

A sociological perspective on the therapeutic alliance: Ethnomethodology and conversation

analysis.

Author: Kozart, Michael F. 1 1 U California, Dept of Psychiatry&Biobehavioral Sciences,

Los Angeles, CA, US

Publication info: Psychotherapy: Theory, Research, Practice, Training 33. 3 (1996):

361-371.

https://login.libproxy.edmc.edu/login?url=http://search.proquest.com/docview/614319859?accountid=34899

Abstract: Traditional perspectives consider the therapeutic alliance as tied to specific

goals of positive patient change. Psychometric studies suggest that the therapeutic

alliance is better conceptualized as linked to aspects of the patient-therapist dyad

independent from therapeutic goals. A framework to capture the relational identity

of the alliance, proposed by E. Bordin (see record 94-105022-001), centers on patient-therapist

collaboration. Ethnomethodology, a sociological research paradigm, suggests that

"collaboration" in psychotherapy consists of methods that establish an impression

of common sense between patient and therapist. These methods involve verbal and para-verbal

cues, often subtle and implicit, that compose a subtext to the more explicit dialogue

about therapeutic goals. (PsycINFO Database Record (c) 2012 APA, all rights reserved)

Links: null

Full Text: Contents - Abstract

Introduction

Traditional Psychodynamic Perspectives

Bordin's Theory of the Alliance

Towards a Sociological Paradigm

The Science of Making Sense

Ethnomethodology Applied to the Therapeutic Alliance

Empirical Leads

Conclusion

Show less Abstract Traditional perspectives consider the therapeutic alliance as

tied to specific goals of positive patient change. Psychometric studies suggest that

the therapeutic alliance is better conceptualized as linked to aspects of the patient-therapist

dyad independent from therapeutic goals. A framework to capture the relational identity

of the alliance, proposed by Edward Bordin, centers on patient-therapist collaboration.

Ethnomethodology, a sociological research paradigm, suggests that "collaboration"

in psychotherapy consists of methods that establish an impression of common sense

between patient and therapist. These methods involve verbal and para-verbal cues,

often subtle and implicit, that comprise a sub-text to the more explicit dialogue

about therapeutic goals.

Introduction Multiple models have been developed to conceptualize the therapeutic

alliance ( Horvath & Luborsky, 1993 ). In these models the alliance has often been

conflated with specific goals of positive patient change, with the quality of the

alliance reflected in how far the therapy has progressed towards the achievement

of one or more change goals. In this article, however, a shift in emphasis will be

proposed, in which the work that leads to a therapeutic alliance does not represent

the primary work of therapeutic change, but rather the work of sustaining an interpersonal

focus on the patient's problems and various solutions to those problems. Accordingly,

it will be suggested that the impetus to sustain the therapeutic alliance is only

partially captured by the anticipation of direct therapeutic gain for the patient.

A broader definition of the motive force behind the alliance involves an imperative

to sustain a therapeutic encounter "as it is meant to be," reflecting individual

as well as socioculturally based norms of patient-therapist interaction. So conceived,

the therapeutic alliance represents the common sense character of any therapy situation,

enabling the patient-therapist dyad to be characterized by a normative purpose (e.g.,

the achievement of positive patient change), rather than the very purpose of establishing

the proper sense of the relationship.

Sociological perspectives on human interaction promise to enhance our understanding

of the therapeutic alliance, though they have rarely been applied in research on

the subject. One such perspective, known as ethnomethodology, will be considered

in this article. Ethnomethodology represents both a theoretical and methodological

approach to social behavior. Theoretically, it emphasizes social behavior as a product

of methods that enable individuals to achieve common sense meanings in and about

their daily situations. Orderly social life, according to ethnomethodology, consists

of a stream of practical behaviors that enable individuals to "see" the outlines

of familiar, commonplace meanings during and as a consequence of interactions with

others. Methodologically, ethnomethodology consists of study practices to identify

and interpret these interactions. Research in ethnomethodology suggests that methods

of everyday verbal exchange function ubiquitously in psychotherapy, though the specific

rules of interaction in therapy will likely vary according to the nature and severity

of the patient's clinical state, the ongoing course of the therapy, and the sociocultural

backgrounds of the participants. The main argument of this article is that these

methods in fact represent the operational definition of the therapeutic alliance.

Traditional Psychodynamic Perspectives While distinct models of the therapeutic alliance

have been proposed, reflecting different schools of psychotherapy (e.g., Rogerian

and Social Influence), most conceptualizations of the therapeutic alliance have their

origins in traditional psychodynamic notions of "reality" and "transference" in the

therapist-patient relationship ( Freud, 1912/1958 ). The association between the

alliance and the proportion of reality to transference in the patient-therapist dyad

has led to a view of the alliance as tied to specific goals of therapeutic change.

As Horvath and Luborsky (1993) note, Freud asserted that the therapist has to convey

a "serious interest" in, and "sympathetic understanding" of the patient to establish

a working relationship with the patient (p. 561). To the extent that the patient

grasps the therapist's actual intent to help, a reality based therapeutic alliance

emerges ( Sterba, 1934 ). On the other hand, the transference relationship involves

the patient's projection of attachment feelings onto the therapist independent from

what the therapist actually does. These feelings derive from prior experiences and

are stimulated by distorted perceptions of the therapy relationship. Positive transference

yields an alliance based on the patient's drive to satisfy unconscious, unrealistic

desires. Zetzel (1956) and later Bowlby (1988) claimed that patients initially develop

an alliance based on transference, though one based on reality assumes precedence

as the patient apprehends the therapist's intent to help. Greenson (1965) suggests

that a capacity for realistic bonding enables the patient to understand his/her own

neurotic transference tendencies and thereby resolve unconscious conflicts. The alliance

is thus seen to represent the "additive summation" of the patient's transference

and reality based attachments to the therapist ( Gelso & Carter, 1994 , p. 300).

The composition of the alliance shifts in accordance with the ebb and flow of the

patient's transference and overall progress towards unconscious conflict resolution.

In this manner, the work of building an alliance can be viewed as isomorphic with

the work of achieving positive patient change. Development of a reality based alliance

in fact represents the therapeutic goal. Doubts, however, have arisen as to whether

this accurately captures the full meaning of the alliance. Gelso and Carter (1994)

argue that a vital aspect of the therapeutic alliance is the mandate to work, and

that "no matter how much or how variable the overlap of the real relationship and

transference is with the working alliance, there is a way in which the working alliance

is independent of these other components" (p. 300).

In their writings, Gelso and Carter have conceptually split the alliance into components

that include the mandate to work or the working alliance, the transference alliance,

and the real relationship. Unfortunately these distinctions are problematic for researchers.

Reliable methods for the delineation of the transference alliance from the real relationship

have not been developed ( Bordin, 1994 , p. 279). Moreover, the distinction between

the working alliance and real relationship is confounding ( Greenberg, 1994 ). Finally,

Greenberg makes the point that Gelso and Carter's construct may apply only to therapies

based on the reflective application of psychodynamic theory. These criticisms suggest

that a reasonable alternative to Gelso and Carter's approach, avoiding any necessary

distinction between reality and transference, would be to centralize the notion of

the alliance solely around the mandate to work.

Rawn (1991) has done just that, defining the working alliance as a mechanism to maintain

a condition of constructive engagement between patient and therapist. Rawn argues

that the working alliance should be understood as "totally devoid" of transference,

though a more consistent reading of his work leads to the conclusion that "conflict,"

rather than "transference," is the operative term to be excluded from any definition

of the working alliance (p. 383). The working alliance emerges as a complex ego function

related to the intrapsychic balance of transference and reality, as well as resistance

and motivation to change. Alliance ruptures may result from an overabundance of transference

in the patient's mind, though when a proper balance of transference to reality is

preserved throughout the therapeutic dyad, actors maintain an ultimate objective

of uncovering therapeutic meanings. Thus in Rawn's scheme it does not make sense

to talk about separate transference and working alliances, nor to demarcate the working

alliance from the real relationship. In a related fashion, Rawn disengages the object

of therapy from the alliance. Alliance development does not reflect the patient's

resolution of unconscious conflicts, nor for that matter any other "mechanism" of

direct therapeutic change, but rather a situation of constructive engagement between

the patient and therapist. The alliance thus serves as a means to a therapeutic goal,

not the goal itself.

An anticipated problem in developing the therapeutic alliance concept around Rawn's

notion of the working alliance, according to Horvath (1994b ), is that the concept

will become so general as to lose its immediate relevance to clinical situations.

It will capture the quality of a warm, helping interaction, but have little to distinguish

itself from the everyday alliances that occur in all spheres of regular social life.

This worry, though, does not account for the influence that the overall context of

a therapy relationship must exert on the patient-therapist dyad. Though the therapeutic

alliance may be analagous to the interpersonal bonds present in many other everyday

interactions, it can be specifically distinguished by the unique role identities

of "patient" and "therapist" in clinical settings. What makes the therapeutic alliance

specific to psychotherapy is precisely the matrix of "warm, helping interaction"

behaviors that reflect the boundry concerns and role expectations of patients as

patients and therapists as therapists.

Bordin's Theory of the Alliance Edward Bordin devised a conceptual framework which

parallel's Rawn's notion of the working alliance ( Bordin, 1994 ). The model consists

of three elements: change goals, tasks designed to achieve goals, and the bond, which

describes the network of positive attachments that bind patients and therapists in

the accomplishment of tasks and goals. The principle unifying all components is collaboration.

The identification of goals and tasks involves a collaboration in which the patient

and therapist mutually agree on what the therapy should be all about. The "experience

of association in a shared activity" adds to the patient's sense that collaboration

in necessary to vanquish the common foe of his/her own pain and suffering ( Bordin,

1994 , p. 16). This establishes the patient-therapist bond, expressed in "felt terms

of liking, trusting, respect for each other, and a sense of common commitment and

shared understanding in the activity" ( Bordin 1994 , p. 16). It is possible that

these felt terms may reflect variable amounts of transference. However, like Rawn,

Bordin is not greatly concerned with the clear demarcation of transference and reality

in his definition of the alliance. Of greater interest is the way the bond weathers

the turbulence of strain that unavoidably enters into any therapy relationship. "The

collaborative process represents an arena in which the patient once more encounters

his self-defeating propensities rather than a place where he reports on the problematic

parts of his life outside of therapy" ( Bordin, 1994 , p. 18). Mild strains soon

develop as the patient wrestles with his/her resistance to change. These strains

emerge as negative feelings about the therapist, or questions about the ultimate

worth of the collaboration. Bonds that include greater amounts of transference will

more likely crumble under the impact of strain; bonds that include greater amounts

of "real relationship" will more likely cut through the strain and provide the leverage

to overcome the patient's self-defeating character. For Bordin, therefore, a research

priority should be the study of alliance strain, a subject that indirectly addresses

the degrees of transference and reality incorporated in the bond without the necessity

of specifying a clear demarcation between transference and reality.

Bordin's model has been widely embraced as "pantheoretical" since it does not assume

a particular viewpoint on the type of therapy necessary to achieve positive patient

change ( Horvath & Luborsky, 1993 ). It merely posits that successful therapies have

collaboration and hence bonding folded into the formulation and accomplishment of

tasks and goals. Empirical findings support this. A widely cited "meta-analysis"

of alliance studies by Horvath and Symonds (1991) found that elements common to most

if not all alliance constructs are "collaboration, mutuality, and engagement" (p.

147). Despite this finding, there remains ambiguity as to whether collaboration (and

the related terms of mutuality and engagement), can be distinguished from the specific

goals and tasks that distinguish different therapies. Can a concept like collaboration

stand on its own as a truly pantheoretical principle of psychotherapy, or must it

always be discussed in reference to the specific work performed in any given therapy?

Bordin draws a conscientious distinction between the bond and goal/task aspects of

the therapy, suggesting that the bond may continue to exist even when the patient

calls into question the worth or relevance of change goals and tasks. Results from

psychometric studies affirm this premise ( Bachelor, 1991 ; Horvath, 1994b ; Horvath

& Luborsky, 1993 ). Although no study suggests that a healthy bond can exist for

long in the absence of change goals and tasks that enable the patient to achieve

positive transformation, studies indicate that work entailed in bond formation is

fairly independent from work directly involved in change goals and tasks ( Horvath

& Marx, 1990 ). If we assume this to be true, than we must posit that priorities,

other than those directly associated with specific change goals and tasks, provide

the motivation and structure for patient-therapist bonding. A possible source for

these priorities consists in the therapy participants' initiative to enact their

mutual situation as it is meant to be enacted. A common sense characterization of

any "ideal" therapy involves the therapist acting to help the patient resolve significant

psychological problems, which by definition implies patient-therapist collaboration.

Little success would come of a therapy in which the therapist acted without reciprocal

effort by the patient to reveal internal conflicts, frustrations, anxieties, and

so on. Any solution to the patient's problems entails an active role by the patient

to process and apply lessons learned from therapeutic discussions. When we posit

that the structure of the bond is provided for by therapy participants' motivated

compliance with normative patterns of patient-therapist interaction, we assume the

necessary condition of collaboration.

We can thus conclude that Bordin's notion of collaboration, and hence bonding, is

indeed pantheoretical (i.e., independent from the specific goals and tasks of different

therapies), since it accounts for a universal, common sense definition of therapy.

We can extrapolate an additional hypothesis, namely that the very imperative to enact

a therapeutic alliance represents a motivated compliance with normative rules of

doing "proper psychotherapy."

Towards a Sociological Paradigm As a discipline devoted to the study of social order,

sociology is well placed to extend our discussion of the therapeutic alliance as

a mechanism to maintain the normative or common sense character of any psychotherapy

situation. Sociologists have traditionally assumed two general perspectives on the

interactional rules that give rise to orderly social occasions. The first, based

on the work of Talcott Parsons, begins with the premise that social behavior is guided

by internalized "norms as causal dispositions to action" ( Heritage, 1984 , p. 23).

To the extent that individuals perceive the importance of patterned collaborative

behavior in a particular field of social action, they can be expected to internalize

a disposition to express similiar behavior upon re-exposure to a homologous field

( Parsons, 1951 ). Parsons hypothesized that internalized norms are generally "opaque

to the actor," and only on "rare occasions do the actors become transparent to themselves

and grasp their own motivating forces" ( Heritage, 1984 , p. 21).

The second approach, based on the work of Harold Garfinkel, argues that Parsons'

framework objectifies the relationship between thought and action without accounting

for what actors know themselves to be doing. Garfinkel, who initiated ethnomethodology

as an alternative to the Parsonian system, argued that actors choose behaviors based

on a range of options, and after consideration of a range of circumstances. It is

the conscious awareness of how to produce normative conduct, rather than the automatic

operation of internalized norms, that characterizes the production of orderly behavior.

For Garfinkel, collaborative behavior evolves according to how actors consciously

conceive collaboration to be relevant to situations at hand. It is beyond the scope

of this discussion to fully develop the argument that renders the Parsonian account

of cognition and social action inadequate (see Heritage, 1984 , pp. 1-36). Suffice

it to say that the Parsonian system cancels out, or at least reduces to the status

of a mediating factor, the determining role played by conscious awareness. Ethnomethodology

presents a more realistic framework for understanding "why people do what they do,"

since it addresses the nexus that invariably exists between self-reflective cognition

and external situations that frame the thinking self.

A central concern of ethnomethodology is to explain mechanisms that sustain a normative

focus during and as a consequence of various communicative exchanges ( Garfinkel,

1967 ). Ethnomethodology argues that the imperative to maintain this focus does not

simply overlap with the goal or task of the focus per se, but rather with the imperative

to maintain a social encounter as it is meant to be. Also, the maintenance of normative

social behavior does not simply overlap with an epistemology of moral behavior, but

rather with a general set of rules for everyday rationality. To explore the implication

of these views for research on the therapeutic alliance, it is necessary now to refine

the theory of ethnomethodology and discuss basic principles of conversation analysis.

The Science of Making Sense For Garfinkel, social membership is principally defined

as an assumption of shared meanings about commonplace phenomena. The irony is that

in the "natural world" no two people apprehend the same phenomenon in precisely the

same manner, given natural differences in physical and psychological perspective,

and yet the assumption of common meaning prevails. An explanation for this is that

to conduct ordinary business with others, the actor or "member" must remain relatively

unconcerned with such differences of perspective. In this way the actor can focus

on "routine work" rather than the very work of establishing a basis of common understanding

with others. The practical work of everyday life consists in methods that affirm

or protect common sense assumptions. These methods imply a symmetry between what

the actor believes and what he/she does. As the actor believes situations to be predicated

on common sense, he/she functions within and about those situations according to

"business as usual," and this structures the social world according to ordinary expectations.

This is formulated by Garfinkel in his operational definition of ethnomethodology

as the study of "everyday activities as members' methods for making those same activities

visibly-rational-and-reportable-for all-practical-purposes, that is, accountable"

( Garfinkel, 1967 , p. 1). And the conclusion that Garfinkel and other ethnomethodologists

have reached is that "by his accounting practices, the member makes familiar, commonplace

activities of everyday life recognizable as familiar, commonplace activities" ( Garfinkel,

1967 , p. 32).

A major principle in ethnomethodology is that methods for a recognizable accounting

of everyday life must remain invisible. Individuals do not comprehend ordinary situations

as a subjective process of accounting, but rather as a process of seeing objectively.

To point out that knowledge of everyday life is mediated by a subjective process

is to indicate something out of the usual field of awareness. Yet this is not the

same as attributing normative behavior to the unconscious operation of causal dispositions.

What remains out of awareness is the existence of accounting methods as methods.

What remains persistently and seamlessly available to awareness is the problem of

making sense according to a presupposed "objective" framework of common sense.

When actors' assumptions of a known-in-common reality are threatened, the risk of

social disorder is great. Still, actors rarely fail to "discover" for themselves

the common sense in even the most disorderly of events. In a series of infamous "breaching"

experiments, Garfinkel had experimenters propose unintelligible explanations to subjects'

"reasonable" questions. The subjects went to great lengths to maintain the thread

of common sense between themselves and the experimenters. This involved expressions

of confusion and the need for clarification, as well as subtle cues to keep the dialogue

moving forward so that the eventual meaning of experimenters' statements would emerge.

When all efforts failed, subjects did not simply abandon beliefs in the shared status

of common sense knowledge, but rather proclaimed the experimenters to be purposefully

rude, insensitive, hostile, in short, disaffiliative. The orderly character of the

experimenter-subject interactions became chaotic, involving shouts and threats of

hostility. Garfinkel hypothesized that individuals constantly protect the assumed

reality of a known in common world, preferring to see others who might challenge

this assumption as anti-social rather than madly correct. For subjects, attributing

anti-social character traits to experimenters preserved a common sense feeling about

their interactions, though this "adjusted" common sense was quite distant from subjects'

initial, common sensible orientations to experimental situations. These situations,

designed to resemble standard scenarios, for example, a counseling session, quickly

lost their initial "normal" character and became something different. In the example,

the counseling session became an argument, in which the original point of counseling

gave way to the point of proving who is right. Once subjects were able to assume

that the situation was in fact an argument, it became common sensible for subjects

to argue, though this new state of affairs was clearly predicated on a different

objective than the initial common sense objective of receiving counseling.

Conversation analysis is the term that has been given to the ethnomethodological

study of everyday speech. It begins with the premise, as amply documented by Garfinkel

(1967) , Sacks (1984) and other linguists (see Heritage, 1984 , pp. 142-157), that

speech acts fail to specify all the information needed for their interpretation.

Speakers reference multiple information points, other than those palpably encoded

in the "here and now" speech act, to understand the meaning of the speech act. All

these points are organized in spatiotemporal and sociocultural contexts, and involve

preceding and subsequent stretches of speech, the relationship history of the speakers

themselves, and their personal and sociocultural backgrounds. Any communicative expression

remains inherently incomplete and subject for its meaning to an individually formulated,

contextualized, frame of reference. Because of this fact, no two individuals can

be certain they will emerge with identical understandings of any single utterance.

Yet, to preserve the sense of orderly conversation, speakers assume they can universally

know the meaning of each others' utterances, otherwise everyday speech would be riddled

with clauses like "I don't understand you," or "you need to be more clear." For this

reason, routine conversation is saturated with methods that establish an impression

of common sense within and about the speech situation. The methods function to keep

the interaction moving in a direction that makes sense for the participants, at the

same time "glossing" natural differences in perspective that threaten to impede the

"normal" work of the interaction. The methods remain largely seen but unnoticed,

enabling speakers to focus on conversation as a process of objective discovery rather

than subjective interpretation. The sense of objectivity reaffirms the trust that

speakers place in being bound to common objectives that can be universally known.

It is the capacity to use such methods spontaneously, with little need for direct

reflection, that defines the status of competency or fluency in everyday speech.

The methods consist in diverse para-verbal cues (intonations, pauses, behavioral

expressions), as well as occasional direct verbal references. The methods do not

overlap with the explicit content of the dialogue, except when the dialogue itself

involves the literal attempt to establish common sense, but rather constitute for

the most part an implicit infrastructure to each and every message. To return to

the counseling session example, subjects exercised methods to get experimenters to

clarify or affirm a common sensible purpose to the sessions. These would involve

diverse behaviors to signal a sense of "lost purpose" (e.g., expressions of confusion,

interruptions, exclamations). A problem in isolating such methods is that there exists

no standard lexicon for their categorization and use. Rather, as Garfinkel and other

researchers have shown, the methods are routinely abstracted from general, unwritten

codes of conventional discourse and then adapted to specific occasions. In this way

we can say that any conversation involves forces that are both specific to the localized

instance (context dependent), as well as socioculturally institutionalized across

whole classes of instances (context independent).

Conversation analysis involves two basic resources. The first consists of an elaborate

transcriptional system to isolate the specific verbal and para-verbal behavioral

methods of making common sense ( Atkinson & Heritage, 1984 ). The analysis generally

begins with a close inspection and transcription of a video-recorded conversation,

and proceeds to identify the behavioral sequences that "seem to matter" for the unfolding,

ordinary character of the conversation. Knowing "what seems to matter" invokes the

second resource, which is the researcher's own practical knowledge of methods that

speakers use to keep the interaction directed towards an ordinary purpose ( Sacks,

1984 ). Thus in conversation analysis, and ethnomethodology as a whole, rendering

any instance of natural social activity accountable to a social-science community

represents itself a next-order demonstration of the same phenomena under primary

study, namely the implicit use of methods to make familiar activities recognizable

as familiar activities.

Conversation analysis does not simply concern itself with the technical work of doing

conversation, but also with the feelings that interactants have towards one another.

Any actors' methods may be expected to follow a sequence of affective intensity in

line with his/her need to affirm the ordinary purpose of the situation at hand. If

early methods fail, speakers can be expected to display increasingly obvious needs

to "get things right," ultimately involving displays of emotional crisis (e.g., frustration,

confusion, or sometimes helpless resignment). Feelings of hostility can be part of

a common sensible dialogue, though these feelings emerge out of a sense of lost common

purpose. Conflict may be the last option for interactants who fail to establish common

cognitive bearings on issues of mutual importance. Alternatively, an interaction

that corresponds to participants' methods for keeping the interaction "on track"

will likely involve constructive engagement around a "proper" goal, and the same

felt terms of "liking, trusting, respect for each other, and a sense of common commitment

and shared understanding" that Bordin (1994 , p. 16) discussed in reference to the

therapeutic bond can be expected to emerge.

Ethnomethodology Applied to the Therapeutic Alliance The application of ethnomethodology

to psychotherapy is hardly new ( Labov & Fanshel, 1977 ), and this literature is

nestled in a larger body of work on medical discourse ( Freeman, 1987 ). This research

has examined how speech events in clinical settings unfold in an ordinary manner,

though the therapeutic alliance as a specific entity has not been examined. According

to our discussion, however, it is reasonable to propose that the therapeutic alliance

consists precisely in methods that sustain the common sensible purpose of therapy

as therapy. This implies that the alliance consists of behaviors that keep the interaction

moving to the ultimate goal of positive patient change, behaviors which enable therapy

participants to bypass invariable points of discrepancy or confusion that might otherwise

lead to a sense of lost purpose in the interaction. All this does not mean that for

a therapeutic alliance to exist, the patient and therapist must be locked in a state

of explicit agreement, but rather that they remain committed to a mutually agreed

upon, ordinary purpose above and beyond then-own beliefs about each other or the

nature of their work. This is equivalent to stating that the patient and therapist

define the purpose of their interaction as one of therapeutic gain for the patient.

The following three points can be made about an application of ethnomethodology to

the therapeutic alliance. First, to the extent that both participants account for

their mutual situation as recognizably familiar (i.e., ordinary), we can expect them

to engage in work that has meaning beyond the mere grounds of its existence. This

involves the effort to achieve therapeutic gain for the patient. To the extent that

this recognizably familiar sense fails to occur, we can expect the purpose of the

interaction to shift to the establishment of some proper sense in and about the situation

itself. Naturally this should not be the "official" point of the therapy, since that

is reserved for the goal of attaining positive patient change. Insofar as the participants

find themselves having to clarify and reclarify the meaning of their work together,

or having to wrestle with competing expectations over "proper" therapy, we can expect

disillusionment or disaffiliation to settle into the relationship.

Second, the accounting methods can be expected to involve subtle, non-verbal expressions,

as well as occasional overt references, that generate common cognitive bearings on

agendas and topics. The accounting methods do not always overlap with the explicit

dialogue, and often constitute an implicit system of cues. In this manner we can

understand how the work of the alliance remains independent from change goals and

tasks. What is said on the surface about goals and tasks will be supported by an

underlying structure of methods for maintaining the ordinary purpose in and about

the situation.

Third, because there is no codified manual for the methods that maintain common sense,

only general principles that must be tailored to specific occasions of use, the therapy

partners will naturally find themselves having to decipher each other's codes for

rendering their situation accountably familiar. If, for example, the patient fails

to confirm the therapist's expectation for how to make common sense, this should

not usher an immediate dissolution of the therapist's sense of the situation as common

sensible, bur rather instigate a search for a new interpretive frame in which to

read common sense into the situation at hand. It turns out that these are rich inferential

opportunities for the therapist, as Garfinkel's studies suggest:

Depending on the psychopathology or personality of the patient, different "deviations

from standard forms" will occur. As the therapist deciphers explanations for these

occurrences, that is, as he/she formulates a diagnostic impression, expectancies

will be generated as to the pattern of conversational behavior likely to occur in

the ensuing interaction, and this in turn will condition the therapist's own understanding

of the conversational norms for the particular interaction. As the patient improves

or changes during therapy, these norms, and the methods used to maintain them, can

be expected to evolve. In this sense, the therapeutic alliance (as the set of methods

that maintain the common sense character of the therapy situation), develops in relation

to the healing-work of the therapy, and yet remains fundamentally tied to aspects

of the patient-therapist relationship independent from specific change goals or tasks.

It needs to be remembered, however, that as with all social encounters, the use of

accounting methods in therapeutic encounters will be conditioned by the sociocultural

backgrounds of the participants. Thus, for the therapist, the task of diagnosis is

complicated by the need to discriminate how much of the patient's "breaching" behavior

represents an "internal" psychological process as opposed to an "external," socioculturally

conditioned pattern of conduct.

Empirical Leads What evidence can be mustered to recommend a definition of the therapeutic

alliance based on ethnomethodology? In fact, as already indicated, there is as yet

no literature either from sociology or psychotherapy that invokes ethnomethodology

in research on the alliance. Yet, a seris of studies has emerged from the Vanderbilt

psychotherapy research group, around William Henry and Hans Strupp, that adds further

impetus to our discussion.

Strupp and colleagues began studies on the patient-therapist relationship in the

1960s using measurement techniques driven by traditional psychodynamic formulations

of the therapeutic alliance ( Strupp, Fox, & Lessler, 1969 ). Later, Henry, Schacht,

and Strupp (1986) devised a method to code transcripts of patient-therapist sessions.

The coding targeted examples of disaffiliative and affiliative behavior. Loose descriptors

of disaffiliative or affiliative behavior were based on research by Benjamin (1974)

and Leary (1957) . The descriptors did not characterize the concrete behaviors that

create the sense of disaffiliation or affiliation, but rather provided examples of

the kinds of "senses" that constitute disaffiliation or affiliation. There was an

assumption that much of what the coders would be using to match aspects of the patient-therapist

dialogue with these "senses" would involve implicit or unstated understandings about

interpersonal behavioral acts ( Kiesler, 1979 ). It was noted that affiliative therapist

behaviors facilitated positive patient change, but even small amounts of dissaffiative

behavior by therapists unravelled or seriously jeopardized previous increments of

positive change ( Henry, Schacht, & Strupp, 1986 ). However, the history of the interpersonal

process in any therapeutic dyad influenced the sensitivity of acquired patient change

to therapist displays of disaffiliative behavior. For example, in a therapy relationship

with a long history of affiliative bonding, momentary negative interpersonal process

"might be 'filtered' through ... to produce less of a negative effect" ( Henry &

Strupp, 1994 , p. 74). All this established that the meaning of any therapist statement

depends on the quality of the relationship between therapist and patient (mostly

a function of nonverbal cues established over the course of therapy), and that a

healthy, affiliative relationship lends therapeutic value to a therapist's intervention

above and beyond its surface verbal content.

This research asserts a theme that figures prominently in ethnomethodology, namely

that the process of an interaction significantly influences its outcome. This process

does not simply overlap with the surface content of the verbal messages being exchanged,

but rather reflects communicative displays that are often implicit and para-verbal.

Henry and Strupp's work also demonstrates another theme embedded in ethnomethodology,

namely that "a little bit of bad process should go a long way" towards upsetting

the orderly work of an interaction ( Henry & Strupp, 1994 , p. 79). As we have noted,

ethnomethodolrogy situates the process of interaction in methods used to convey common

sense in about situations at hand. "Bad process" is captured by the way in which

behaviors undermine the common sense definitions of present situations. This lack

of confirmation can shift the very definition of a situation away from what it is

supposed to be, as illustrated in Garfinkel's "breaching" studies.

In an effort to develop a practical program of psychotherapy training based on their

results, Henry, Strupp, and colleagues subjected a pool of therapists to an intensive

program of "Time Limited Dynamic Therapy" directed to the cues that convey affiliative

and disaffiliative meaning ( Henry, Strupp, Butler, Schacht, & Binder, 1993 ). Far

from improving the quality of their interactions with patients, therapists who went

through the course became "significantly less optimistic, less supportive of patients'

confidence, more defensive, and more authoritarian" ( Henry & Strupp, 1994 , p. 67).

These therapists also demonstrated higher frequencies of disaffiliative behavior

with their patients. Henry and Strupp concluded that the behaviors they were trying

to identify in the course were not reliably tapped by training materials, and that

therapists demonstrated an insecurity in not knowing something that is nonetheless

crucial to know. This is well summarized:

Again, this aspect of Henry and Strupp's research is firmly accounted for by ethnomethodology.

According to Garfinkel, orderly interactive behavior should be ultimately reducible

to methods that enable a situation of common sense to arise. However, these methods

must remain "seen but unnoticed" for the "normal" work of the interaction to occur.

Participants are thus conditioned to overlook what it is that they actually do to

sustain the work of their interaction, and for that matter, researchers may be inclined

to use an implicit practical knowledge for explicating the sensible, recognizable

aspect of subjects' work, since the tendency to overlook the methods of production

for everyday behavior represents a common sense attitude. To enable these methods

to rise to the level of conscious awareness, a rigorous "out of the ordinary" technique

for a finegrained, moment-by-moment analysis of the interaction must be employed,

a technique which in fact makes the interaction appear differently to the researcher

than it naturally does to the participants. Strupp recognized the need for this kind

of technique for future, more practically oriented programs of psychotherapy training:

"we now believe that some type of fundamental training in the perception of moment-by-moment

interpersonal process should be an initial foundation for later training in different

theory-based therapies" ( Henry & Strupp, 1994 , p. 68).

We can use existing psychometric instruments as the starting point for ethnomethodological

surveys of the alliance. Multiple instruments have been devised to rate significant

features of alliance systems, enabling quantitative associations of these features

with other aspects of psychotherapy, such as outcome or patient pre-therapy characteristics.

Many of these instruments follow quite directly from different theories of the therapeutic

alliance, as for example the Working Alliance Inventory (WAI), which is a Likert

rating system derived from Bordin's work (Horvath, 1994 a ). Despite the careful

design of these instruments, little attention has been directed to the problem of

how informants should actually rate the different behaviors or utterances in a therapy

session. In the WAI, an implicit, unspecified common sense is assumed to guide the

informant in the decision as to whether a particular exchange represents a "high

or low" on a rating scale of task, goal, or bond strength. An ethnomethodological

study would seek to render this common sense transparent and identify as plainly

as possible the very methods that informants use to know what a given therapeutic

interaction actually represents. These methods can be expected to parallel the very

methods that subjects use to achieve common sense in and about the therapy situation,

especially when the informant is one of the subjects. Many of the rating scales,

like the WAI, are indeed available and utilized in a subject-response format.

Conclusion Psychotherapy researchers have frequently made the self-critical remark

that after years of work, no consensus has emerged around the precise identity of

the therapeutic alliance ( Gelso & Carter, 1994 ; Horvath, 1994 ). Proponents of

varying schools of psychotherapy have introduced different ideas to capture a genuine

sense of the alliance. In an effort to bring these diverse views into a concise framework,

Bordin forged a pantheoretical vision of the alliance based on the principle of patient-therapist

collaboration (expressed as the bond). Support for Bordin's model has come from a

meta-analysis of different alliance constructs that revealed collaboration, or mutual

engagement in the therapy process, to be central to most theory-specific models of

the alliance. However, Bordin's model failed to answer whether the bond should be

primarily expressed in terms of the technical work of the therapy (expressed as tasks

and goals), or according to nonspecific patient-therapist relationship variables.

Empirical data from psychometric studies of the alliance in fact affirms the latter

viewpoint, giving rise to a need to delineate the interactional variables that actually

comprise the bond.

This article suggests that such a delineation can be insightfully generated from

ethnomethodology, a sociological paradigm directed to the study of everyday social

order. What ethnomethodology promises, via its applied format of conversation analysis,

is a set of ideas and terms that focus the researcher's attention on subtle, and

often overlooked methods by which interactants jointly negotiate a sense of ordinary

purpose. Ethnomethodology not only presents a potential consensus position on the

definition of the therapeutic alliance, but promises to move alliance research beyond

another impasse, namely specification of the alliance building techniques that enable

therapists to achieve positive clinical outcomes. Ethnomethodology has already yielded

tools and concepts related to this process from studies of other social phenomena.

It is important to emphasize, however, that ethnomethodology offers no certain prescription

for an ultimate manual of alliance building techniques, since the methods that can

be expected to foster common sense in any therapeutic encounter will likely vary

according to the psychological and sociocultural experiences of the participants.

Nevertheless, this variation can be expected to occur within a larger framework of

common parameters. The very fact mat psychotherapy occurs according to institutionalized

rules of appropriate conduct reinforces the possibility that patients and therapists,

despite their individually unique backgrounds, possess shared methods to enact psychotherapy

as an ordinary, common sense act. It can be expected that an accumulation of studies

on the situated methods of making common sense in psychotherapy will eventually yield

general concepts that therapists will be able to apply, albeit with some modification,

to a wide variety of clinical situations.

REFERENCES

1 . ATKINSON, J., & HERITAGE, J. (1984). Structures of social action: Studies in

conversation analysis. Cambridge, England: Cambridge University Press.

2 . BACHELOR, A. (1991). Psychotherapy.

3 . BENJAMIN, L. S. (1974). Psychological Review.

4 . BORDIN, E., Horvath, A. O., & Greenberg, L. S. (1994). The working alliance:

Theory, research, and practice. New York: John Wiley.

5 . BOWLBY, J. (1988). A secure base. Clinical applications of attachment theory.

London: Routledge and Kegan Paul.

6 . FREEMAN, S. (1987). Text.

7 . FREUD, S., & Starchey, J. (1912/1958). The standard edition of the complete psychological

works of Sigmund Freud. London: Hogarth Press.

8 . GARFINKEL, H. (1967). Studies in ethnomethodology. Engle-wood Cliffs, NJ: Prentice

Hall.

9 . GELSO, C. J., & CARTER, J. A. (1994). Journal of Counseling Psychotherapy.

10 . GREENBERG, L. S. (1994). Journal of Counseling Psychology.

11 . GREENSON, R. R. (1965). Psychoanalysis Quarterly.

12 . HENRY, W. P., SCHACHT, T. E., & STRUPP, H. H. (1986). Journal of Consulting

and Clinical Psychology.

13 . HENRY, W. P., STRUPP, H. H., Horvath, A. O., & Greenberg, L. S. (1994). The

working alliance: Theory, research, and practice. New York: John Wiley.

14 . HENRY, W. P., STRUPP, H. H., BUTLER, S. F., SCHACHT, T. E., & BINDER, J. L.

(1993). Journal of Consulting and Clinical Psychology.

15 . HERITAGE, JOHN (1984). Garfinkel and ethnomethodology. Cambridge, England: Polity.

16 . HORVATH, A. O., Horvath, A. O., & Greenberg, L. S. (1994a). The working alliance:

Theory, research, and practice. New York: John Wiley.

17 . HORVATH, A. O., Horvath, A. O., & Greenberg, L. S. (1994b). The working alliance:

Theory, research, and practice. New York: John Wiley.

18 . HORVATH, A. O., & LUBORSKY, L. (1993). Journal of Consulting and Clinical Psychology.

19 . HORVATH, A. O., & MARX, R. W. (1990). Canadian Journal of Counselling.

20 . HORVATH, A. O., & SYMONDS, B. D. (1991). Journal of Counseling Psychology.

21 . KIESLER, D. J. (1979). Psychiatry.

22 . LABOV, W., & FANSHEL, D. (1977). Therapeutic discourse: Psychotherapy as conversation.

New York: Academic.

23 . LEARY, T. (1957). Interpersonal diagnosis of personality. New York: Ronald Press.

24 . PARSONS, T. (1951). The social system. New York: Free Press.

25 . RAWN, M. L. (1991). Psychoanalytic Review.

26 . SACKS, H., Atkinson, J. M., & Heritage, J. (1984). Structures of social action:

Studies in conversation analysis. Cambridge, England: Cambridge University Press.

27 . STERBA, R. F. (1934). International Journal of Psychoanalysis.

28 . STRUPP, H. H., FOX, R. E., & LESSLER, K. (1969). Patient's view their psychotherapy.

Baltimore, MD: Johns Hopkins Press.

29 . ZETZEL, E. (1956). International Journal of Psychoanalysis.

Show less

Address for Correspondence: Michael F. Kozart, M.D., Neuropsychiatric Hospital, Center

for the Health Sciences, 760 Westwood Plaza, Los Angeles, CA 90024.

© 1996 Division of Psychotherapy (29), American Psychological Association

Subject: Ethnology (major); Methodology (major); Psychotherapy (major); Sociology

(major); Therapeutic Alliance (major)

Classification: 3310: Psychotherapy&Psychotherapeutic Counseling

Population: Human

Identifier (keyword): ethnomethodology&role in therapeutic alliance

Title: A sociological perspective on the therapeutic alliance: Ethnomethodology and

conversation analysis.

Publication title: Psychotherapy: Theory, Research, Practice, Training

Volume: 33

Issue: 3

Pages: 361-371

Publication date: 1996

Format covered: Print

Publisher: Educational Publishing Foundation

Country of publication: United States

ISSN: 0033-3204

eISSN: 1939-1536

Peer reviewed: Yes

Document type: Journal, Journal Article, Peer Reviewed Journal

Number of references: 29

DOI: <a href="http://dx.doi.org/10.1037/0033-3204.33.3.361">http://dx.doi.org/10.1037/0033-3204.33.3.361</a>

Release date: 01 Jan 1997 (PsycINFO); ; 15 Dec 2006 (PsycARTICLES);

Correction date: 17 Jan 2011 (PsycINFO)

Accession number: 1997-02169-002

ProQuest document ID: 614319859

Document URL: https://login.libproxy.edmc.edu/login?url=http://search.proquest.com/docview/614319859?accountid=34899

Copyright: ©Division of Psychotherapy (29), American Psychological Association 1996

Database: PsycARTICLES

____________________________________________________________

Contact us at: http://www.proquest.com/go/contactsupport

Copyright © 2012 ProQuest LLC. All rights reserved.

Terms and Conditions: http://search.proquest.com/info/termsAndConditions

Jessie Rayl
thedogmom63 at frontier.com
www.facebook.com/Eaglewings10
www.pathtogrowth.org
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.acb.org/pipermail/acb-hsp/attachments/20121119/f803fc98/attachment-0001.html>