Conversation Analysis – A Powerful Tool for Psychoanalytic Practice and Psychotherapy Research

Introduction

In recent years many efforts were undertaken to understand what psychoanalysis is and what psychoanalysts do. For former generations it sufficed to define psychoanalysis as the science created by Freud. What psychoanalysts do could be described by the word “interpretation”. Several developments darkened this idyllic picture:

1. Many studies conducted by experienced psychoanalytic psychotherapy researchers (see for an overview Bachrach et al,. 1991) tried to demonstrate systematic differences between “psychoanalysis proper” and “psychodynamic psychotherapy”. This difference could not convincingly be established (Sandell, 2012). In high frequency psychoanalyses not only interpretations were used but analysts also provided advice and emotional support (Bush & Meehan 2011); in once-a-week psychodynamic psychotherapies also unconscious material was interpreted in addition to advice and social support.

   1. Correspondence concerning this article should be addressed to Prof. Michael Buchholz and Prof. Horst Kächele, International Psychoanalytic University (IPU), Stromstr. 2-3. 10555 Berlin/FRG, Germany. E-mail: ,

   Page 4

2. To define the “common ground” of the multitude of psychoanalytic schools turned out to be increasingly difficult. The Rome Conference papers profoundly demonstrated that every author - pretending to know what the “common ground” was - just figured out his personal position. Titles were “Common ground: The centrality of the oedipus complex” (Feldman 1990), “The search for the common ground” (Schafer, 1990), “Empathy. A common ground” (Feiner et al., 1994), “Countertransference: The emerging common ground” (Gabbard, 1995), “The illusion of common ground” (Green 2005), and, finally, ironic titles like “Common (under)ground of psychoanalysis: The question of a Weltanschauung revisited” (Figueira, 1990).

3. David Tuckett (1993) pointed out that there is even no agreement what “clinical facts” are; “facts” appear in the horizon of a given theory. Therefore Tuckett lamented of having Winnicottian, Kohutian, Bionian, Kleinian and classical facts and today one could add a lot more, e.g., Bollasian or interpersonal/relational “facts”.

4. Studying psychoanalysts-at-work (Tuckett, 2012) allows to regain the domain of conversation. This rediscovery will open the door to a world of a rich caleidoscopic shimmer in all colours of the human rainbow.

For example an EPF working group identified 6 types of interventions psychoanalysts have described (Tuckett et al., 2008, p. 140):

1. Maintaining the basic setting

2. Adding an element to facilitate unconscious process

3. Questions, clarifications, reformulations, aimed at making matters conscious

4. Designating here and now emotional and phantasy meaning of the situation with the analyst

5. Constructions directed at providing elaborated meaning

6. Sudden and apparently glaring reactions not easy to relate to a certain analyst's normal method

This group recommends shifting attention more on what the analyst does. Such a comparison of what psychoanalysts do, what kind of inter-action psychoanalysis is comes in view. Such a comparative approach unavoidably discovers how different analysts work – beyond every idealistic description and self-description. Dana Birksted-Breen warns about the risks of this endeavor:

The novelty of this comparative method lies in a dual shift in perspective: from the patient to the analyst and from emphasizing a single ‘truth’ to valuing difference. Instead of the traditional discussion of a ‘case’ in which the leader and participants suggest a better or deeper understanding of the clinical material, there is a radical shift of focus towards understanding the analyst herself, what she is doing in the psychoanalytic encounter, and for what reason. The attempt to elucidate the implicit model of the presenter/analyst shifts attention away from evaluation and towards understanding (Tuckett et al., 2008, p. 2)

Having presented a clinical report the German analyst Helmut Hinz echoes in the same direction: “To recognize difference is the beginning of recognizing the richness of the universe of reality” (Hinz, in Tuckett et al., 2008, p. 109).

What psychoanalysis is begins to be defined in terms of what psychoanalysts do. To apply CA-methods to psychoanalysis endorses that line of thinking. CA contributes to precisely describe how it is done. This seems to be a remedy for this confusing situation and it is in good psychoanalytic tradition. Anna O. termed what Breuer and she had done together as “talking cure”. Today, this valuable metaphor is much more to the point than ever. In psychoanalysis no “variables” are “applied”; simply two people talk together although in a quite specific way. When directing attention to conversation we have a chance to catch sight of what “clinical facts” are.

The aim of our paper is to inform about some of the research results and outline further possibilities. This seems politically important. Influential parties in evidence-based medicine have constructed manuals instead of old fashioned textbooks of therapy in which a therapist is someone producing preformulated text blocks with respect to diagnoses. Psychotherapy in general is in danger to be reduced to something like a trivial machine without own reflection. Against this trend it might be important that psychoanalysts remind psychotherapy in general as a highly specialized and creative form of dialogue and directing research to analyze the special ingredients of psychoanalytic conversation as opposed to other therapies.

Conversation – A Core Concept

The formula of “talking cure” opens sight for the core questions of psychotherapy research: What are the qualities of a “healing conversation” (Symington, 2006)? What are the details? How to describe conversational practices with the aim of “healing”? And: How come that conversation influences cognition? If we glance into the consulting rooms of therapeutic neighbors like music therapy or art therapy or even cognitive behavior therapy we are endowed with a superior argument: more than 90 percent of activities there are not the special ingredients which a given therapeutic school proclaims to be effective. No, it's just - talking! And, of course, at the same time psychoanalysts could stick to the claim “Not Just Talking” (Pain, 2009); conversation analysts see this as the gift to therapy and they did begin to analyze “Talk as therapy” (Pawelczyk, 2011). In this seeming contradiction of “Just talking” vs. “Not just Talking” we see how psychoanalysis is formed by a central paradox: Therapy = Talk and Therapy ≠ Talk. Both equations are valid at the same time. Psychoanalytic patients and their analysts do talk and, according to Freuds didactic formula nothing happens but an “exchange of words”. Yet, still more happens. As “talking cure” psychoanalysis operates in a medium which transcends the logic of secondary process thinking, it comes close to primary process thinking where these logical contradictions are tolerated and creatively processed. But how is it done? And the question then arises as to the difference between psychoanalysis and other therapies?

To prefer “talking cure” over “chimney sweeping” - the other term Anna O. unforgettably coined for her psychoanalytic experience - has good reasons. Talking is the most neutral description for what happens in a treatment room; although, of course, not every talk is treatment or psychoanalysis. Of course, Freud did not exclude gesture and tears, mimic

and prosody, nor pause and silence. Explaining how to introduce the basic rule he used the term “conversation”. In such a conversation it happens that certain kinds of experience are condensed in a metaphor like “chimney sweeping”. Others, often heard, are e.g., “diving deeply” or the (religious) metaphor of “confession” and many others. Anna O's “chimney sweeping” was later replaced by Freud by “purification”.

These metaphors encompass a smaller or greater part of conversation. They never contain all aspects. Metaphors for conversation become part of conversation when they are occasionally used as frames for the ongoing conversation. As such they are part of conversation and, at the same time, create a reflective stance for conversation. Thus, conversation or “talking” is the superior concept one has to begin with.

Why Not Speak of Communication, But of Conversation?

Early communication research used a technological model where one speaker is analogized with a transmitter, sending a coded message to a receiver who has to decode the message . This technological provides an instruction for learning “communicative behavior”, to teach people to “say what you really mean”. It is widely used in mass communication, social psychology of groups and in many other areas like coaching, communication trainings etc.

The technological model ignores many highly relevant dimensions of human affairs (Peräkylä, 2004). One of them is that a listener pays attention to a speaker! What might be the reason for this active contribution by the listener? Before he knows that a “message” is to be “decoded” the listener listens! How is it done that they don't start talking together at the same moment? One must simply remember that babies do this and one without further instruction grasps that there must be a history in it. And more: A listener pushes back of his mind any irrelevant “noise”. Before he can “decode” he has to differentiate the “noise” from the rest. This active participation of the listener is completely overlooked in the technological model, but it is highly relevant for conversation. Participants in conversation have an implicit knowledge of this dimension. One can observe how speakers end their contribution with what CA-researchers call a “tag”, a little sound like “eh!”, in German some use “gell”, in Switzerland “o:dr” (“oder?”) is used; in British conversation one encounters “isn't it?”. Obviously these activities are directed to activate the listeners attention and in some cases to provoke the listener to give a prosodic sound that he still listens: “hm:m”. Another problem is how a listener might be sure to have “decoded the message correctly”? He cannot have a measure for correct translation before the beginning – which would be required in the technological model. Otherwise a “decoding” of the message would not work.

In conversational activities there is more than the conveyance of information. The technological model is a “folk theory” of talking. Reddy (1979) termed it the “Conduit metaphor”. Conversation is more. It consists of, at least, three components that have to be identified from the outset: A listener has to differentiate being addressed from the content and he has to use this difference in order to produce understanding. This is the full circle. Conversation does not happen from one to the other (this might be the case in military commands), it demands active participation on both sides.

Conceptual obstacles of a technological model are easily to overcome if we do speak of “conversation” in a tradition that goes back to Wittgenstein. Language is less a transporter system for “information” but a form of human practice. Thus, CA-researchers reject a technological blueprint for human and naturalistic affairs as it denies the sociable, gregarious dimension, the convivial element in conversation. Originally, conversation was not a scientific concept. It was meant for the coming together, for the social gathering of well educated people in a friendly manner. The place where they met was the salon (Miller, 2006, p. 71) like the salon of Rachel Varnhagen in Berlin or the Ephrussi family's salon in Paris or those in St. Petersburg or Moscow so marvelously described by Tolstoj. What these people did was called “doing conversation”. Their conversation followed certain rules all designed to support one aim: conversation should not stop, it should go on. Thus, tone had to avoid themes like religion or politics which might result in quarrels or serious controversies. Topics should be changed before it came to a decision because a decision in a question of relevance might someone feel excluded. Exclusion was considered a serious violation of conversational rules. Another rule was to care for the inclusion of everyone, even women. Those who were not excluded where considered members in conversational participation. Persons excluded should be re-“membered” into the conversational circle. So they could be re-“minded”. Etymology reflects the idea that to be re-minded meant to become healthy again by being remembered, by being re-included in the community.

In contrast, a scientific conversation had completely different aims. Scientific questions urge to be answered, they have a penchant for closure in order to turn oneself to other questions. If you want to know if an answer is “true” or “false” you have to debate it to the end, you cannot change topics or consider about someone being excluded. From this double origin of conversation the famous Freudian inseparable bond of “cure and research” herited (, p. 256). His formulation includes both in order to make perceptible by scientifically guided self-reflection the healing effects of conversation. Healing is effected by re-including.

Freud's use of the term “conversation” displays several connotations. To analyze conversation and to analyze a “psyche” has a huge semantic overlap of meaning; everything you can analyze must come through the narrow passage of conversation, including all elusive phenomena as slips of the tongue, a glance, a gesture or a curious word used. Everything that can be heard and seen can become part of conversation analysis. Affect displays are included in this conception of conversation (Krause & Merten, 1999). Or vice versa: if it is not heard nor seen it cannot be noticed and then, not analyzed until it appears on the common screen of conversation.

Psychoanalysts are interested in certain topics in their patients’ life which they think are repressed and refrain from appearing on the conversational screen. So we have a first (not the first) description of what analysts sometimes do: they actualize hidden dimensions excluded from conversation (Smith, 1991). They do this actualization in the medium of – conversation and cannot avoid this paradox. It follows, that even interpretations are embedded in conversation; they are part of conversational history. Much can be said to further analyze the How, When and Why of different kinds of actualizing activities. But for the moment we want to turn to another line of thinking.

The Origins of CA

Conversation analysis is the name for a research practice deriving from several traditions in sociology and sociolinguistics. It was Erving Goffman (1964, 1983) who directed attention to the phenomena of everyday verbal and nonverbal exchange between participants in interaction. He considered the difference between backstage and front stage in a very similar way as psychoanalysts, he saw how important “frames” of interaction are and could make plausible that these frames guide institutional conversation in a way perceptible for everyone, but largely ignored. In families, preschoolers are sometimes asked “Paul, do you know how much is two and two?” At school, then, the format of the question is changed: “How much is two and two – Paul?” The personal reference changes its place from the first utterance to the last. This format communicates meaning which is not in the words. It fits to classroom conversation (Gardner, 2013) which is different from family intimacy. Harold Garfinkel, another social scientist and founder of what he called “ethno-methodology” worked at similar problems and found that every subcultural ethnic community uses “methods” how they organize their conversations in order to make sure that there is “order at every point”. Even in excited quarrels one can find “orderliness”, but in times before Garfinkel (1967) neither participants nor observers could see it.

It was Harvey Sacks (1978, 1980), a gifted young student of conversation, who coined the term of conversation analysis (CA). His lectures were later published by Gail Jefferson (1992); Sacks untimely died in an accident in 1975. Gail Jefferson and Emanuel A. Schegloff belong to the first who understood the enormous relevance of studying the details of conversation in everyday talk, in institutions as patient-doctor interactions, in classrooms or at court, and influenced others to pay special attention to the details of conversation among children. In all these different institutional places it could be shown how conversational routines as questions and answers, referencing to present or absent persons or practices like storytelling make influential differences. Even the expression of emotion, the regulation of affiliation, the production of laughter or gaze direction was analyzed and it could be demonstrated that in conversation not only words are exchanged but bodies participate in a way confirming the basic premise of “order at every point”. Jefferson ( and ) could demonstrate this orderliness in such emotional moving interactions as “trouble talk” or, more positive, in how laughter is “organized”. Others have felt touched by the potential of CA to analyze psychotherapeutic processes in general more broadly. Madill et al. (2001) defined CA as “the study of talk in interaction and of how, through talk, people accomplish actions and make sense of the world around them” (p. 414). Forrester and Reason admit that CA might be “an exercise of disciplinary demystification” (p. 44) but they cannot see how CA deals with the “ontological status of the unconscious” (p. 53). Both papers focus on sequence organization of therapeutic talk and hope to gain a more detailed view of therapeutic practices.

There is still the other line of thinking in CA, the analysis of category bound activities (; Sacks, 1980; Schegloff, 2007). Let's have a short look at the case illustration used by Forrester and Reason (, p. 53). The patient has reported a dream. They, as observers, see “from a psychoanalytic-psychotherapeutic perspective”, that “the client offers the therapist a gift” (p. 48).

To interprete a dream(-telling) as a “gift” constructs a metaphorical equation – by observers. Nothing is shown how participants, therapist and client, treat this episode. To

criticize the therapist for not acknowledging the dream as a gift might be relevant from a clinical point of view. Clinical thinking is full of constructing metaphors of this kind. This is, what clinicians helpful do, it is one of the practices often used in doing psychoanalysis.

CA analyzes such practice as category activity. How do participants create and introduce into conversation new categories such as metaphors (or others) used to self-describe what they are just talking about? Are there relevant places where such categorical constructions can be introduced? Which conversational problem of the participants is solved when one of them does this? How come that Anna O. suddenly used the metaphor of “talking cure” for the kind of conversation she had with Joseph Breuer? Metaphors are powerful conversational tools to condense conversational experience of both participants and sometimes they create a new level of reflection, sometimes they fail and next speakers refuse them or do not agree. Here is a chance to integrate metaphor analysis (Ahrens, 2012; Angus, 1996; Buchholz, 2007; Cameron & Maslen, 2010; Carveth, 1984; Quinn, 2010) and CA.

A link between conversation and cognition (Molder & Potter, 2005) can be drawn. Leading cognitive linguist George Lakoff (1987) used “What categories reveal about the mind” as subtitle for his famous book and he drew on categories of very different kinds, logical and, of course, metaphorical. We have analyzed these links methodologically (Buchholz 2006) and used them empirically in our research about sexual offenders (Buchholz, Lamott & Mörtl, 2008; Mörtl et al., 2010). In our view, it is not that CA overlooks the “ontological status” of the unconscious; CA enriches psychoanalytic conceptions of the unconscious by focusing on other activities (sequences and categories) more than traditional conceptions of the unconscious do. Participants unconsciously use tacit skills for three purposes:

1. to orient their contribution to a common activity focus (“What are we doing here together?”) – in conversation is a capability for cooperation. In the future one could describe the special kinds of cooperation patients display or refuse.

2. to commonly organize – without conscious planning – their turn-by-turn organization; they “know” certain formats of talk, e.g., when a storyteller makes a pause but wants to continue or, alternatively, the listener may take the turn and pose a question; they “know” when it is interdicted to tell a joke (Sacks, 1978) or how to gossip (Bergmann, 1987) and violate moral rules of conduct (Hakulinen & Sorjonen, 2011; Bergmann, 1999; Pomerantz, 2012). This kind of tacit knowledge contributes to the overall organization of talking, in psychoanalytic observation it is widely ignored. Common organization means that there is a sense of sharing something together.

3. to further process their common conversational tasks and aims embedded in a context of interactional organization. For example, when you hear a “thank you” you follow the rules of polite complementarity, e.g., “you are welcome”, “my pleasure” etc.). But therapists, when they hear at the end of an interview a “thank you” while shaking hands don't respond this way. Therapists nod with their heads and keep silent sticking to their professional attitudes documenting that the interview was not an everyday kind of talk. Therapists use small violations of an everday-rule-of-conduct as professional markers for demonstrating the difference between everyday-types of conversation and the consulting room. This is an example for a general human sense of using self-produced context to .

In recent years a Finnish conversation analyst, Anssi Peräkylä, trained as a psychoanalyst, began to research psychoanalytic conversation. Psychoanalysis is talking, but, of course, not every talk is psychoanalysis (Peräkylä et al., 2008). So what details make the difference? To find this out is one aim of CA-research in psychoanalysis. It will result in a variety of knowledge of how psychoanalysts talk. Thus, CA-research is a powerful tool to contribute to important psychoanalytic questions, e.g., what clinical facts are. If you go back in the history to CA-research activities one is surprised to find how close this comes to psychoanalytic interests. This closeness has several dimensions:

1. The kind of data is not numbers, but talk-in-action (Heritage & Clayman, 2010)

2. The mode of careful observation is not global (“overall diagnosis” as in DSM or ICD) but directed to the details

3. Conclusions go from gathering the details to the more global conceptions, or, as William Blake once said: “There is a world in every grain of sand”.

Here are some examples to exemplify this closeness between CA-research and psychoanalysis.

Small Examples

These examples are not taken from the consultation room. They are a selection of early CA-research in order to show how these researchers are engaged in topics which Freud paid careful attention to. They are elusive phenomena, often overlooked. Jefferson in 1996 published an article “On the Poetics of Ordinary Talk” which led her close to Freudian positions. She started with an observation: The poet's job is it to arrange sounds and categories. She provided a few examples and then went further to a clinical question: “What is the difference between what the psychotic does and what the poet does?” (p. 4) She gathered examples of Freudian slips. Here are three of them:

1. A record from the US-Crandall Show where the host is reading out a commercial for suits named “Bond’ Blue Chips”. His exact wording but was:

   “~B[ig, [be)eautiful] from America's [l]argest [cl[othier. [B)oh-Bond's. Blondes, my goodness. Wuh that's a Freudian Slip” (quoted from Jefferson, 1996, p. 6)

   Jefferson sees the double sound-row of the letters b and l and comments there is “something moving towards ‘blonds boo’”. This is a classical Freudian interpretation based on an exact analysis of what the speaker said. Freud saw a slip of the tongue composed of a conscious intention to say something interfering with another unconscious intention to say something different. The condition for the slip is this combination of conscious and unconscious intentions.

   Here is a second example from Jefferson:

2. A plane made a rough landing and the stewardess a few seconds later announces:

   Stew: On behalf of the who(l)e f(r)ight - f(l)ight c(r)ew I'd like to thank you for flying Air California (p. 7)

   Page 11

   And Jefferson comments like a Freudian analyst using her countertransference: “‘Fright’, and how! Freudian Slips! Lots of nudging and grinning among us passengers. But then I thought, no, it's one of those sound-selection things. As with dreep drop, blonds boo, flight foyed, here we are on the way to ‘fright cloo’” (p. 7)

3. Jefferson informs about another type of slip based on category-formed errors. An example is when a male speaker, Larry, says:

   “Hi. I'm Carol's sister - uh brother” (p. 10)

This example is one of “categorization”: one's self is wrongly categorized – and then corrected. Examples of this kind are well known in psychoanalysis. We mention them to counteract the prejudice that CA is anti-mentalistic or simply behaviorally oriented. CA is engaged in similar topics as psychoanalysis. If Freud had had the technology of audio-and video recording he would have used it with respect to his patients’ needs for anonymity. As in his days this technology was not available he had no other possibility than to ensure his readers that he had a “phonographic memory” as in the introduction to his 1916-17 “Introductory Lessons” (Freud, 1916-17).

Senses and Embodiment

The astonishing equivalence between CA and psychoanalysis goes further. CA analyzes talk in a manner including the body and senses and with an excellent attention to the details. Details are elusive and cannot be reproduced nor experimentally manipulated, but they are those hearable and viewable elements in the Freudian tradition that are considered able to show that there is unconscious activity. Schegloff (2007) explains his interest in how people acknowledge each other with a personal anecdote and an important observation:

I often find myself walking on campus and encountering someone coming the other way who was an undergraduate student in my class. And we have this odd game of not-quite-mutual gaze. They look at me half expectantly, and as my eyes start coming to them they look away, figuring that there is no way I would recognize them as they recognize me, and to be caught looking at me like that would be … what? Intrusive? Presumptuous? Mocking? And if the pas de deux goes their way, we pass each other without ever meeting one another's gaze and with no mutual acknowledgement; and if it goes my way, I trap them, and recognize them – sometimes by name which blows their mind – and we greet each other, and it's very nice. This is the way the logic plays out when the very issue is whether there is to be any interaction at all in the first place.

Page 12

On the telephone, the parties are already in the interaction, so it plays out a bit differently, but the same logic is involved. I know him, but does he know me? and does he know that I know him? and does he know that I know that he knows me? (Schegloff, 2007, pp. 132-133).

This inclusion of personal experience follows Freud in showing how his personal experience “from everyday” steered his scientific interests deeply. Important here that the interpersonal “logic” of this “pas de deux” plays out in another area: On the telephone when people have no visual contact to each other! Here are some of Schegloff's examples:

Example (6) TG, 1 (: #42)

1. ((ring))

2. Ava: H'llo:?

3. Bea: –> hHi:,

4. Ava: Hi:?

Example (7) NB, #114 (: #44)

1. ((ring))

2. Cla: Hello::,

3. Agn: –> Hi:::,

4. Cla: Oh: hi:: ‘ow are you Agne::s,

Example (8) HG 2

1. 1 ((ring))

2. 2 Nan: H'llo::?

3. 3 Hyl: –> Hi:,

4. 4 Nan: HI::.“ (Schegloff, 2007, p. 133)

These examples clearly demonstrate the same pattern of “pas de deux” on telephone calls as when meeting on the university campus. From the senses of seeing to the sense we encounter the same transmodal pattern.

Schegloff explains:

In their first turn, callers do a greeting that in the first instance claims to have recognized the answerer as the person they meant to reach, and which also provides a voice sample to the answerer from which callers, in effect, propose and require that the answerer recognize them. In these three instances, it is about as small a voice sample as it could be; some callers are a bit more generous and say ‘hello’, providing the answerer with two syllables from which to recognize. In these three instances, and in a great many more, it works. With no hearable delay, answerers return the greeting in the next turn, which serves not only to reciprocate the greeting, but to claim that answerers have reciprocated the recognition as well. The operative word here is ‘claim’; in Example (7) Clara shows that she has recognized the caller (her sister) by addressing her by name; in Example (6) and (8), no such demonstration is provided. (p. 133)

These examples display interesting dimensions of CA-research for psychoanalytic topics and theorizing. Freud's programmatic declaration that the essential part of the Ego is the body today is worked through in other fields under the heading of “embodiment” research (Pfeifer & Leuzinger-Bohleber, 1986). From Schegloff's fresh analysis one can get an impression of how this sensual “transference” from “seeing-each-other” to “calling-each-other” might work. The same pattern of “pas de deux” is executed and it is this kind of pattern we look for in psychoanalytic practice.

We mention just one example (Emde, 1988) here from the rich body of evidence in infant research. The baby is given into his open mouth a nubbly dummy without the possibility to see it before. The baby feels the dummy is different. After the dummy is removed the baby sees several pictures of dummies on a screen. The baby stares longest when it sees the picture of a nubbly dummy. She makes a kind of “conclusion” from one sensual experience to the other, from the mouth to visual perception. This mental “gestalt” (Bernfeld 1934) confirms what Freud had in mind when he spoke of “transference” in his first versions – from one sensual modality to another. Or, in Schegloff's example, from one encounter on a University Campus to another when starting a telephone call. Or, as Emde (1988) formulated, from relationships to “relationships on relationships”. Infant researchers show in this kind of experimenting an observation and the baby's “method” to proceed from the observable to the unobservable. We see how mind comes into sensual experience, works in sensual transference and then up to relationships on relationships.

Nobody would conclude that Schegloff's observation of “pas de deux” in gaze-to-gaze-movement and then in acoustic verbal exchanges while starting a telephone call would be

“regressive” and “baby-like”. No, one can learn that these processes are elementary and constitutive for the beginning of new interactions.

Resistance in Conversation

There is an important debate what basic psychoanalytical concepts mean. Spurling (2008), in reviewing some famous case histories asks himself why he in his consulting room does not encounter the phenomenon of “regression” which is to appear regularly according to authors like Winnicott he re-reads? He concludes that in regression there is much contribution from the therapist's side. At the same conclusion arrive other psychoanalyst (Minolli, 2004; Rousillion, 2010; Wainrib, 2012), when reconsidering the phenomena of narcissism. Kächele et al. (2009) insist that psychoanalytic concepts have value only when they can be precisely documented in conversational practice in the consulting room. Convincingly they documented how by tape-recording of analytical session complex levels of psychoanalytic treatment can contribute to theoretical debates. Clinical facts can be documented when analysts give up their resistance against tape-recording of sessions. There is a small, but growing body of CA-literature on resistance. Gerhardt & Stinson (1995) have shown how ambivalently the concept of resistance can be used as a basic operation for a patient's striving for autonomy or as opposing a method. Others (Fleisher Feldman, 1995; ) have responded to this exposed alternative but the conclusion was that there is no “objectivity” in defining resistance; everything depends on the analyst's emotional and epistemological position.

“Resistance” has been borrowed from psychoanalysis generating productive conversational research in related fields as advice giving, e.g., in medical conversation, genetic counseling, telephone help services and the like. How do professionals in these areas deal with resistance? Hepburn and Potter (2011) describe three practices:

1. Resisted advice is packed in a more idiomatic form

2. Sometimes a “tag” is added so that the listeners is defined as someone who already knows the relevant version

3. The counselor dampens down the requirement for a response by “continuing past the transition place” (p. 217).

To give advice has a normative dimension and it is asymmetric. The advice giver defines himself as someone who claims to know how things are done “right”. These interactional dimensions are handled by the three practices described. The difference to the psychoanalytic notion of resistance is not huge. To overcome resistance also in psychoanalysis means to make the patient accept the momentary asymmetry and to accept that the analyst might be “right” – and the analyst tries to balance this asymmetry by verbal activities of the kind described. Resistance can be aggravated by the conversation format of how a patient's problem is formulated (Morris, 2005). In question-driven therapies resistance is aggravated by this kind of conducting therapy (MacMartin, 2008). Patients find covert modes to resist advice while they openly consent. Hepburn and Potter (2011, p. 221) describe an “endemic epistemic asymmetry” – the advice giver in telephone emergency calls never has a full picture and the caller can withdrew to informations not yet or never given. The advice giver never gains control whether his advice is followed or not. This might operate in psychoanalytic conversations, too. Vehviläinen (2008), contributing with the first article on CA about resistance in psychoanalysis, describes the “interpretative trajectory”:

The analyst does preparatory work to create the relevance, and an interactional ‘slot’, for the interpretation, thereby co-constructing it with the client. Connections and contradictions in the associated materials, pointed by the analyst, provide puzzles: noteworthy, enigmatic issues calling for exploration and explanation. In a stepwise manner, the analyst treats some aspects of the client's talk as worth exploring. The interpretative statements that typically follow attend to these puzzles, providing explanations. They draw on the materials the client has provided, but reorganize them or add something new. This is, then, the core interactional practice of showing the client something that she or he ‘has not been aware of’… (pp. 121-122)

This trajectory is not a theoretical description but substantiated by a series of transcribed data. Thus we have two basic operative procedures in dealing with resistance: First, the analyst uses the client's material but arranges it in a new way. Second, the analyst adds something new, a new perspective or a new combination with other materials the client had delivered in earlier moments. We propose to term these practices under the headline of “changing the frame”-procedures with a notable aspect: the new frame has to be made relevant for the client in order to be accepted. Here, Vehviläinen is right in speaking of analysts solving puzzles. Vehviläinen (p. 137) points out that the analyst's task to figure out the puzzles in a client's talk must provoke “trouble”, attempts to justify oneself, to rationalize or to defend oneself in other manners. These practices show how psychoanalysis makes clients accountable for actions where accountability has been denied. This involves risky aspects of being blamed. But, of course, it is inavoidable. So, how this is done becomes a question of enormous practical relevance. We are at the beginning here. Freud was right, when he saw a cornerstone of psychoanalysis in the concept of resistance and claiming that the whole psychoanalytic theory is built on this concept.

From today's point of view it might be worth considering that theory is a necessary protective resource for this special kind of conversation psychoanalysts try to conduct every day. In his 6-days-a-week practice Freud (1913, p. 460) observed what he called the “Monday crust”, the patient reappearing a little bit hardened as compared to the Saturday session. But the humorous metaphor of “Monday crust” clearly is more than an observation. It serves as a means to continue with a practice Freud called in other writings as an “impossible profession”.

Talk in Professional Psychoanalytic Practice

We hear and see what can be heard and seen and listened to and in everyday interaction we turn our attention away from things which co-interactants commonly exclude – these things are “categorized” as “not-relevant to what-is-going-on”. In contrast, as

psychoanalysts we try to re-mind what has been excluded from conversation, we pay special attention to what was made strange and try to re-member these elements. We strive to catch for the invisible – but we can catch it only when we are attentive to what is simply “documented” in what can be heard and seen. The “behavioral” dimension of conversation is just the access to the so important invisible dimension which we term the Unconscious. Here CA-research and psychoanalytic intuitions converge. In this paragraph we follow the work of Peräkylä for some convincing demonstrations.

Traditional domains of CA are medical communication, in court, in classrooms – summarized as “institutional communication”. In institutions of that kind on can find conversational routines which make deviances meaningful. Obviously, psychotherapeutic communication cannot be subsumed under the headline of “institutional conversation” in general. The variations from analytic dyad to analytic dyad are immense.

But, there are of course routine parts as e.g., the way how to end a session. Most analysts use a standard formula like “Our time is over” or the like to end a session. In this standard routine lies a chance for deviance. When there was a touching session with tears or otherwise emotionally moving so that analyst and patient share a common experience sensitive analysts end a session in deviance from the standard formula. This deviance then is uploaded with an unsaid meaning as the patient can “co-read” the analyst's being emotionally moved and his sharing of emotional experience during the session – but the analyst must not explicitly “inform” his patient about his participation. Sensitive analysts do this because intuitively they feel that to use a standard routine of “good bye” can be “heard as” a kind of cold neutrality. In anticipation of this possibly hurting “being heard as” the good-bye formula is changed. And, as it were, incidentally let the patient participate for a moment of the own emotional state. Without saying or telling, nor talking nor informing, it is no “communication”. It just violates a standard routine. Psychoanalytic conversation might be found in this “intermediate” region where something is “said” and “not said” in the same move. Everyday logic cannot think this “at the same time”. In everyday logic A=A and not A≠B. But in our conversations sometimes it happens that we meet another logic which comes close to what Freud (1900) in “The Interpretation of Dreams” had termed the primary process. The new discovery is that these characteristics of primary process appear not beyond or “behind” the conversational surface. They appear on the surface and we can direct our attention to them.

When analyzing the use of metaphors in psychoanalytic dialogue (Buchholz, 1996) there is a similar observation. Metaphors operate on the base that A=B and A≠B. “His mind is a mill” means, of course, that his mind is not a mill. Stählin (1914) showed in order to understand the meaning of a metaphor it is important to develop a “consciousness of double meaning”. If one hears the sentence “He is gone crazy of love for her” and you expect to find the man in a psychiatric ward you have not understood the metaphor. To understand metaphors means to renounce the sentence of excluded third. Metaphors show that there is an excluded “Third” between formal structured logic. Metaphors (Ogden, 1997; Borbely, 2008; Cacciari, 2008; Aragno, 2009; Buchholz, 1993) demonstrate this primary process logic on the surface of our talking. This kind of intuitively “catching” emotional hints is observed by Peräkylä (2011) in other examples of psychoanalytic conversation, too.

Peräkylä (2013, p. 552) critizices this “going beyond” of what is said and done.

While ordinarily in interaction, language and other signs are understood as means for displaying and recognizing the speaker's communicative intentions…, in psychotherapy there is an endemic orientation in the therapist, and usually in the patient, to examine the patient's talk beyond its intended meanings.

Psychoanalysts’ practices can be described as using formulations. Formulations (Antaki, 2008) are utterances which indicate in what way a speaker came to understand what in the other speaker's turn has been said before. Using such formulations psychoanalysts do not exactly reproduce the words the patient used. They give meaning by slightly shifting utterances. They use turn-initial particles like “It seems as if you were …” or “It sounds like…”. The practical implication of such formulations is that they sound like saying the same and operate as “just-understanding”. But they do more. Vehviläinen (2003) sees that in such formulations a new content is packed in making a different (unconscious?) meaning for the hearer easier acceptable.

Psychoanalysts expand meanings by repeating a patient's words and add additional formulations: “You say this made you angry, but no feelings of being disappointed before?” or they make utterances like “You say it is hard. I think you mean painful”. Rae (2008) calls this a “lexical substitution” aiming to find the correct word for an emotional state and its intensity.

Peräkylä (2011) made an interesting observation. The analyst's formulation is spoken in a “first position”, the patient's response illustrates a move from a “second position”. When analyzing the material of 58 audio-recorded sessions of psychoanalyses conducted by two psychoanalysts with 3 patients Peräkylä (2011) finds that analysts modify slightly what they have said in their first-position move. Sometimes they intensify the words for emotional experience or they additionally point to layers of experience not addressed to in the first-move-utterance before. These reformulations are not made explicitly with an accent on “this is right now!”, rather they come in a way to open a discrete opportunity for the patient to change how she understood the first interpretation given. Analysts contribute to the process by respecting the patient's autonomy and at the same time intuitively demonstrate their adaptation to the patient's response to the first interpretation given. Can this kind of micro-analyzing clear the difference between therapy and other forms of talk? The answer is yes and a short review can show this.

CA in Psychoanalytic Therapy Research

One distinguishing feature is the amount questions from the therapist. Putting questions to someone has a multitude of conversational dimension of which we mention here only one: it responds to the silent question who holds the initiative? In question-answer therapies the conversation stops if no questions are posed or this threat is permanently present. In contrast, psychoanalysis has aimed to give the patient the power of initiative and thus to pose the analyst in a listening position. After silencing of some length

therapists utter a “hm-hm”, obviously an attempt to bring themselves in a listener's position (see first example below, line 5).

Pittenger et al. (1960) in their research on “The First Five Minutes” established some principles of therapeutic talk (“recurrence”, “immanent reference”, “adjustment” (p. 229-244) which demonstrate the firm wish to empirically base their intuition of difference in therapeutic talk on research data. Scheflen (1972) included in his analysis the bodily movements of hands and gestures and positioning which synchronized with conversational intent. His material was a detailed analysis of a family therapy session which influenced family therapy research of authors like Wynne (1984). Labov and Fanshel (1977) analyzed the first 15 minutes of session 25 observing four levels: a) the elements of text including all pauses; b) the elements of paralinguistic cues (p. 42) with which loudness, laughter, breath and rate was included; c) the level of “expansion” by bringing together verbal and paralinguistic data together with elements of the situation; d) the level of interaction by which participants identify their actions mutually. Their methodological instrument was speech-act theory. They come to formulate 4 basic types of conversational action: metacommunication, representation, request and challenge (p. 77-110). Turner (1976) directed his attention to the question when a session does begin? What makes the difference between “informal” talk before the “official” work session starts? His material were group therapy sessions and he finds that in the informal first parts therapist omit those “second moves” to which they are urged after “official” session parts had started. This distinction is a complicated matter in handling psychoanalytic sessions. We have gathered three starts of psychoanalytic sessions from fully trained psychoanalysts, examples which demonstrate how important it is to draw the line between “informal” and “official” session. In all three examples one participant starts talking with a special demand:

First Example

1. T (male): Mrs R., t'day (.) I would let my cell phone online and if i:t were my family I wou::ld respond the call becau:z (..) my father is in the hospital

2. (1,5)

3. P (female): My stepgrandfather too. I mea::n Irene's father. A:lso (.) came today to the hospital

4. (3)

5. T: m:mh

6. (7)

7. T: How are you doing?

Second Example

1. P (male): well, at first uhm; next Tuesday (.) we could drop our session (-) or shift it.(1) Be::cause (-) I would have to attend another meeting

2. ((P and T are laughing))

3. T (female): ä::uhm (1,5) ä::uhm we could shift it to Thursday

4. P: Yeah:: okay::uhm it should work ((stands up, walks to his bag and takes his blackberry out))

5. T: this is the 25th.

6. P: also at seven pm or? That (.) sounds good (-) I ll put it down (-) to be on the safe side (--) perfect!

7. (8)

8. ((P and T are laughing))

9. P: tzen I ll switch off (.) my machine(--) you really look good, relaxed

10. T: thanks. ((laughs))

11. P: sun burned ((laughs))

Third Example

1. T (male): Uhm::hh (..) I have a small info first (..), uhm

2. (1)

3. P (male): Yes::

4. T: It will ring any moment (.) someone will deliver at the door (-) I ll first have to open it

5. (.)

6. P: ((sighs))

7. T: that you (-) will be (.) prepared

8. P: ((sighs, loud inspiration,))

9. (60)

10. P: Yes: =I just had a nice experience and a surprise (reports a pleasurable situation with his wife)

These examples have something in common and they are different. The three psychoanalysts follow the technical rule to begin a session with information about a change of the formal frame of the session if this should become necessary. These things

happen in psychoanalytic sessions and it is, following Tuckett, not necessary to deal with it whether it is appropriate or not. CA has not the task to evaluate a procedure but to describe the steps by which conversation is organized.

In the first example the therapist informs his client that he wants to learn what happens with his father in a hospital and that the session might probably be interrupted. The client's answer is of a very ambivalent kind. Even after repeated careful listening to the tape it was not possible if this is a response of rivalry-in-suffering (My stepgrandfather too) or is it an utterance categorized as empathic with the analyst's situation. The ambivalence of how to categorize the client's response makes the analyst (line 5) attempt to bring himself into a listener's position by uttering “m:mh” which produces no client's response but a continuation of silence. After seven seconds the analyst tries an everyday start in order to overcome the line between the informal and the official part of the session.

In the second example it is the client who wants to change the date of a session. With “well, at first uhm” he informs the analyst about his drawing a line between the two parts of the session, it is his categorizing activity. But here the analyst does not come to participate in drawing that line. The common laughter (in line 2 and 8) is a co-production of resistance against differentiating “informal” from “formal” parts of the session. The last chance to differentiate these two parts of a session is lost after the internal pause (--) in line 9. This is a turn-transition point, the client switches of his blackberry, informs the analyst about this activity and ends with a pause. If the analyst does not take the turn here this can be heard by the patient as a “silent continuer” - as if a “go on” is uttered. The client goes on with informal talk and makes his comments about the after-holiday fashion of his analyst. The analyst responds in a conventional everyday manner (“thanks”) and instantiates the patients hearing that the line between informal talk and official work is not yet established. This might further the patient's courage to talk about his analyst or it might lead to an aggravation of this kind of resistance.

The third example starts in a similar way like the first. The analyst does not have such a justifying account as if he expects the session to become interrupted because of a father in hospital. But he justifies his pre-session information in a similar fashion (line 7: “that you (-) will be (.) prepared”). The patient moans several times, the therapist does not actively try to restart conversation. After a “felt long” pause of 60 seconds it is the patient who takes his initiative.

These examples might serve as an illustration as to how CA can help to clearly analyze the enormous difficulties to cope with in psychoanalytic practice. It is not only sequentiality of turn-taking, it is category-bound activity (to differentiate the informal from the formal part of the session) as well as mutually reading intentions of both participants (including unconscious intentions) that has to be skillfully managed in professional practice. In psychoanalytic practice more happens than just “giving interpretations”.

This early discovery originated from a project on “discursive structures in psychoanalytic therapy” (Flader et al. 1982) showing special properties of psychoanalytic conversation: a) a specialization of mental participation in dialogue; b) deviant participation structures simply as to the amount of verbal activity by the therapist; c) therapists exhibit unusual conversation practices in long silences and non-responding to questions; d) both

participants need time for establishing their special kind of cooperation; e) finally, the question of asymmetric anonymity.

These authors used the concept of “defense” for explaining certain conversational features instead of analyzing how this acts in conversation itself. One task they described might be remembered because of its practical value: Patients at the beginning of treatment have to be socialized into that special kind of conversation tolerating deviances from everyday expectations as not answering, long pauses and the analyst's attempt to keep secure his anonymity. There is training in specialized conversational practices than only gaining “insight”. A patient cannot, as in most conversational opportunities of other kind, select the analyst as a next speaker. The patient might put a question to the analyst and make the experience not to receive an answer. Spence et al. (1994), while “monitoring the analytic surface” observed a related phenomenon. They were interested in just one type of utterance when a patient directly addresses his analyst by sentences like “I think of yesterday when you said…” This connection of “I” and “you” in close approximation in one sentence can be easily detected. They find that there are sessions with a lot of such sentences (“related hours”) and sessions without any (“isolated hours”). The important conversational feature is: in “related hours” the analyst does not only speak more, but earlier in the session.

The phenomenon once found it is not difficult to explain. When the analyst is directly addressed this exerts a certain conversational pressure to respond. This kind of research detects a phenomenon of how the analyst is steered by his patient's utterance format. And this phenomenon must be conversationally unconscious. Up to now it had not been described in textbooks and it is not a documented part of treatment technique – but it operates in psychoanalytic conversation.

Grabhorn et al. (2005) analyze speech of an anorectic patient during a 12-week treatment in a psychotherapeutic ward. At the beginning of treatment they find high levels of resistance indicated by the SASB-method (Benjamin, 1974); in the middle of treatment resistance gradually disappears and a therapeutic alliance can be built up. At the end they find more autonomy. Spence et al.'s (1994) findings are more fine grained: in the beginning of treatment this patient used the addressing of the therapist very often, clinically this indicated low levels of autonomy and autoplastic adaption to treatment conditions. When in the middle phase of treatment the conversation phenomenon disappeared this was clinically interpreted as a necessary self-reorganizational retreat; the reappearance of addressing the therapist was considered as a new step in autonomy.

CA – Opening a New Way between the Scylla of Hermeneutics and the Charybdis of Science

So far we have reviewed relevant CA-research and included some excerpts of our own study in order to show that the distinction of metacommunication, representation, request and challenge, as described by Labov and Fanshel (1977) would not suffice to fully describe the complex phenomena here. More mental or cognitive dimensions like category-bound activity, sequentiality or unconsciously luring of the analyst's response cannot be described within the framework of Labov & Fanshel's early work. CA underlines that patients need confirmation and ratification by the other subject. This reminds one of Tomasellos (2008) differentiation between types of conversation as

inform, request and sharing. According to this author the ability to share (information, perspective, knowledge) is a human condition and not found in apes.

Addressed in this line of research is a very relevant aspect of process research: how conversational and cognitive processing relate to each other. Hepburn and Potter (2011) regret in the period of developing one's own research strategies, “conversation analytic work has not attended to cognitive matters” (p. 219). Today CA-research addresses the psychological more directly. Analyzing category-bound activity during conversation characterizes the interface of CA and psychoanalysis.

This is a very delicate domain as it is traditionally assumed that talk follows other rules than mental and cognitive processes, that thinking is much faster than talking and that both have very different capabilities. Psychoanalysis is somewhat “mentalistic” while conversation is seen as outward directed and concerned with the observable only. If things are left in this way conversation and psyche could never be brought together. This is obviously a false conception as people at least in therapies strive for letting another person know what they feel is in their psyche. So a way must be found to bring both sides together without giving up methodological rigorism and precision of observation.

Peräyklä (2011), experienced in both practices of analysis, proposes as a third way to overcome the dichotomy of “inner” and “outer” experience and to see this difference as one produced by cultural and conversational practice. He defines psychoanalysis

as a practice in which the client and the analyst explore their inner experiences, and step by step either recognise dimensions of affect and cognition that appear for them as ones that have always been there but have not been perceived with clarity before, or achieve new dimensions of affect and cognition that are real but have not been possible to be experienced before the psychoanalytic process. In short, feelings, thoughts, hopes, desires appear as real phenomena in psychoanalytic practice – not merely as artefacts or projections produced by linguistic and interactive processes in the consultation room or elsewhere (p. 237).

He comes to propose a new sight. The “inner” world and its difference to the outer world are demarcated by a symbolic and interactive border. Here operates a category-bound activity of a very important kind. This border is in itself constituted by conversational and interactive practice defining an individual's personal, private, “inner” sphere and it is a culturally widespread distinction. In case of a projection this inner sphere is outside of the individual, in case of introjections the outside world is inside.

Self and conversation can be considered as the two sides of a coin: the system of mutual affective regulation. Interactional and more psychological (“inner”) processes of

regulation can be brought together. The details of this “come together” are tasks for the future. Here we offer a perspective to analyze process dimensions of psychoanalytic talk from both sides: from conversation and psychoanalysis. The psychoanalyst's responses to a patient are describable as related to contexts which are self produced for the moment, emerging and then fading away replaced by new contexts in and by conversation. Thus, a central problem of psychoanalysis can be brought closer to a solution: How is the relationship between a general theory and a practice which operates helpful only when the individual dimension of the unconscious is approached and can be recognized, addressed, touched and seen? This dilemma of general theory and individual truth is contained in the ongoing and never ending controversy about the question: is psychoanalysis a science or is it hermeneutics? (Thomä & Kächele, 1975).

This controversy goes back to the German philosopher's Wilhelm Dilthey distinction between these types of scientific endeavor and it lasts on to our days (Boesky 2008). One could conclude that the two models of “hermeneutic” and “science” are in itself insufficient. A psychoanalyst does not “interprete a text” and he does not “apply” a general theory to an individual patient. Merton Gill remarked that to apply a theory in psychoanalytic consultation equals like having earwax in the third ear. Both, hermeneutic and science, differently appreciate the subjective dimension of human existence. While hermeneutic positions favor this dimension they are rebuked for this as they never could achieve in a mature position in academic controversies; while scientific positions claim to work scientifically they are rebuked from the other side to ignore individual suffering, meanings, dimensions. If human beings were fully determined by the laws of nature there would be no place for subjectivity and individual decisions; if human beings were fully “free” in every respect one never could substantiate human sciences. This controversy became part of repetition compulsion in psychoanalytic generation building. Same topics are treated again and again in every new generation and clearing up operations in theoretical questions that have been solved are widely left out. The most serious consequence was that psychoanalysis did not evolve a research paradigm in its own right. Both positions suffer from a fundamental inability to conceptualize a basic dimension of humanity, which is the social dimension expressed in conversation and interaction. It is in this social dimension only that we are constituted as individual persons with a personal history brought into new interactive encounters with others who contribute to change our conceptions of who and how we are. It is time to include a (micro-)social dimension into psychoanalytic theorizing and give it a more central place. This enables psychoanalysis to give answers to what “clinical facts” are. Psychoanalysis and its clinical facts are locally produced, naturally organized, reflectively accountable, ongoing in practical achievement, always, only, exactly and entirely, by participants’ work in and during sessions - this is the fundamental phenomenon.

Psychoanalytic process research seeks to provide detailed analyses of the assemblages of practices which are partially based on tacit knowledge through which the work of accomplishing local “social order” of an individual psychoanalysis is achieved. It is as Winnicott wrote that there is no baby without its mother. And there is no psychoanalytic patient without an analyst. We can turn this statement around: there is no analyst without a patient. Both must come together to produce what we consider as psychoanalysis. This production is a social phenomenon including hidden dimensions of conversation, of tacit knowledge and individual skillfulness on both sides. We can approach the riddle of “clinical facts” when we begin to consider the dyadic nature of the psychoanalytic endeavor. And CA is a powerful tool for psychoanalysis. The “common ground” of

psychoanalysis might neither be Oedipus theory nor early envy, neither repetition compulsion nor death instinct, neither phantasy nor reality and what other conflictual themes might come to mind. This all is on the level of theory “applied”.

Our theoretical conceptions allow a huge range of individual variability of technique embedded in an enormous amount of individual variability of patients and analysts. Debates might receive a new and fresh drive when psychoanalytic treatments are conceptualized as social patterns of relating and affiliating, of formulations and reformulations, of discovery and interpretation. Including more strictly the social dimension might help to join a research program that has begun to show that some of our most urgent problems can be brought closer to a helpful solution. If hermeneutics and science are two sides of a coin the whole debate always saw one side covered in the dark, either hermeneutics or science, either subjective dimensions or the more general law, either individual meanings or the more general theory. To bring this coin to stand up in order to make both sides visible will demand to give this coin the knack: to make it turn as fast as possible by conversation and other social influences. There is no standstill.