Conversational Errors and Common Ground Activities in Psychotherapy—Insights from Conversation Analysis

Many patients leave psychotherapy although in need. What can professional practitioners and researchers assume what happened? Trying to receive a response from these patients we too often are left without an answer. In this paper I introduce to psychotherapy discourse some concepts taken from linguistics and Conversation Analysis (CA). The reason is that what psychotherapists of every kind do is “talk-in-interaction”. During such talk Typical Problematic Situations (TPS) appear which are well known in a macro-analytic perspective (if a patient comes late to the session, does not talk or blackmails the therapist with suicide threat). However, there are many TPS that can be detected by a micro-analytic perspective only. CA is a tool helping to idenfity this type of TPS. One relevant CA-concept is Common Ground, a psychological and linguistic concept which requires special activity from both participants in an interaction. Conversational “errors” that risk to tear the Common Ground often go unnoticed. Presenting segments of transcribed therapy sessions I want to direct attention to the details of how ‘errors’ in Common Ground activity happen, how they are noticed and dealt with by skillfull therapists or how they can become repaired. Among others I use transcription details of two suicidal patients. The transcripts are from the CEMPP-Project (Conversation analysis of Empathy in psychotherapy process), conducted at IPU, Berlin. Thanks to a grant by Köhler-Stiftung, Germany.


Person or Method-What Is the Problem?
The everyday saying, "it's not what you say but the way you say it" is too rarely scientifically observed in process research and clinical practice.Therapists are trained to think in terms of "intervention" technology reducing "disorders" and producing "outcome".However, prominent researchers warned: "As a rule, the study of psychotherapies has been favored over the study of psychotherapists-as if therapists, when properly trained, are more or less interchangeable ... We think that one reason for this relative paucity of research on psychotherapists is an implicit bias in thinking about therapy leading to the assumption that it is basically a set of methods, techniques, or procedures that are efficacious, in and of themselves, in curing or ameliorating psychological and psychiatric disorders...This bias views the personal element or the subjective equation in human experience and relations as a source of error in research to be minimized or controlled..." (Orlinsky & Ronnestad, 2005, p. 5) conflicting scientific debate (Elkin, Falconnier, Martinovich, & Mahoney, 2006;Kim, Wampold, & Bolt, 2006;Wampold & Bolt, 2006).However, is the method or the therapist the only choice?

Situationism-An Alternative Solution
There is another alternative: situationism that guides the work of conversation analysts, following certain methodological rules (Note 1) which can be summarized in guidelines: First, don't look primarily for external variables as e.g., social background, attachment style, motivation or type of personality.These abstractions produce generic explanations; however, in therapy we look for how these variables (and many others) are individually realized (or not) in situated interaction.
Second, make talk-in-interaction the center of analysis.This generates data close to the situational dynamics.These dynamics are steered by gaze, body movements and talk.Talk includes words, the embodied voice, rhythm used to achieve a definition of the situation.
Third, look for how a Common Ground (Enfield, 2006;Stalnaker, 2002) is established or not.Common Ground outlines the horizon we talk to, it is never a "given" but to be established in situations.
Fourth, talk-in-interaction has the double potential to repair (Dingemanse et al., 2015) an imbalanced Common Ground and to tear the Common Ground to pieces.
Fifth, direct your attention to how Common Ground activities are managed successfully or not.Without a Common Ground situationally maintained by interactional and talking activities every special technical procedure in psychotherapy heavily risks to fail.
A situationist approach furthers microanalytic observation of small scale events that have the power to deteriorate otherwise "good" relationships or vice versa, to steer flat relationships to deeper emotional experience.Negative events are called here "conversational error" because they are situated in conversation as the main and universal tool of all psychotherapy.What therapists do is "talk-in-interaction".In all psychotherapeutic endeavors there is a consistent amount of "drop-outs" close to 10%.Clients leave therapy although they need it.This may be due to false diagnosis or inappropriate institution (Werbart, Anderson, & Sandell, 2014).However, it might be useful to study the details of conversational "errors" that go unnoticed-that "errors" happen is not the problem.What is more important is whether they are noticed and repaired between familiy members (Corrin, 2010), especially mothers and their babies (Emde, 1981) or in psychotherapeutic interaction (Barnett, 1980;Castonguay & Hill, 2012).The problem is not the "error" but the lack of repair.Here I want to point to a level of conversation which linguists have termed "Common Ground".If "errors" on this level go unnoticed they have an effect to interrupt further progress.What follows is an explication of Common Ground.

"Common Ground" and Conversation Analysis (CA)
It is necessary to describe what Common Ground means.There are two offsprings.First, there is a psychological use: "She accomplishes this by choosing her wording based on their Common Ground so far, and by collaborating with her partner to reach the mutual belief that he has understood her" (Wilkes-Gibbs & Clark, 1992, p. 183).This is from a paper about "coordinating beliefs in conversation".How do people manage to make sure that they understand their words reasonably well?The answer is that we create a Common Ground by directing attention to a common perceptual object in our common environment.We point to it saying something like: "Nice book shelf you have there".I can see it and I can see that you see it and when You respond: "Yes, I have bought it directly from the factory" two important things happen: By "Yes" the perceptual object is transformed into a conversational one and then you link to the conversational object another experience or, even better, a series of experiences.So there are three transformational levels: perceptual-conversational-link.Later, when having studied some transcript material, I will add a fourth one.
Transforming the perceptual into a conversational object is an act of conversation.It informs both speakers that they have joint attention (Bangerter, 2004;Tomasello, 2003) and for at least one moment they have created a Common Ground.This opens a horizon to project further activities.Thus, although "Common Ground" sounds like a piece of territory it is meant as an activity.Common ground can disintegrate if this activity fails or is ended.
These activities are psychologically risky.If speaker A would utter his sentence and B would completely ignore it, neither respond with words nor a gaze, then sensitive persons can sometimes doubt if they have made the utterance?If an unpopular group member enters and his greetings are not responded by team members and they leave the room the recipient is punished by non-resonance.His attempts to build up a Common Ground are refused.Who ever made this painful experience knows how threatening it is and how much one's personal integrity depends on basic resonance.
Second, there are linguistic definitions.Linguist Nick Enfield gives a definition very close to the psychological one: "The pursuit and exploitation of mutual knowledge, shared expectations, and other types of common ground ... not only serves the mututal management of referential information, but has important consequences in the realm of social, interpersonal affiliation.The informational and social-affiliational functions of common ground are closely interlinked" (Enfield, 2006, p. 399).
If the conversational level is achieved a lot of further social psychological accomplishments become possible: to exploit mutual knowledge, shared expectations and interpersonal affiliation.Participants produce and reproduce Common Ground-why?Psychologists would point to generic factors like interpersonal attraction, sympathy, etc. From a linguistic point of view comes a further reason: "Common ground is a resource that speakers exploit in inviting and deriving pragmatic inference, as a way to cut costs of speech production by leaving much to be inferred by the listener" (Enfield, 2006, p. 401).
Conversation is very slow as compared to thinking.Conversation is linear, thinking is circular and has a multi-processing architecture.There is a speed difference between conversation and cognition.So, to share your thoughts completely with someone would become an aim forever unattainable.But if you share a Common Ground reliably you must not say everything because you can trust that the listener will infer correctly (Schmitz, 2014).This kind of trust is a conversational strategy to reduce complexity (Luhmann, 1979).Producing and reproducing a Common Ground, from a linguistic point of view, is indispensable for every talk-in-interaction because of the speed difference between conversation and cognition and, I think, this is a good argument to be added to the psychological ones.
If one has both aspects in mind one can more easily see how sensitive, how risky and how complicated Common Ground activities are.Applied to therapeutic conversation one must conclude that every single intervention has a good chance to fail if the Common Ground is not established or is in risk of decay.I know of no psychotherapy process research study paying attention to this level.
Common Ground is related to and different from other psychological concepts.Attachment quality, e.g., is acquired in early life and built into a person's patterns as "working model" (Stalker & Davies, 1998); attachment, although directed to others ("objects") is an individualistic conception while Common Ground is a conversational concept including, as I tried to show, processes like affiliation, joint attention, etc.In psychotherapy process Common Ground is related to the concept of "alliance" (Safran & Muran, 2000) between therapist and client.However, to come to a working alliance agreement is relatively late (not in the first session) and presupposes a lot of Common Ground activities.This is why conversation analysts pay much attention to how a conversation starts.
One observation seems to hold: there is a kind of hologram-principle, once formulated by William Blake, that there is a world in every grain of sand (in Bruder, 2003) (Note 2).However, this principle unfortunately is a non-symmetrical one: one single "false remark" sometimes has the power to destroy an otherwise well working therapeutic relationship; thus, lots of "good remarks" not necessarily indicate that a break of Common Ground could be ignored.Everything seems to depend upon the therapists' sensitivities to realize that a "Typical Problematic Situation" (TPS) is being build up.Can the therapist quickly reorganize one's mode of operation, attitude, thinking about the patient's disorder and utterance-or not?A general principle can be recommended: therapists should be trained to think freely about at least two types of response before delivering an utterance.Holding to such a principle gives the therapist even when "under fire" (Note 3) a grounded feeling-having a choice intensifies your sense to be an autonomous person making decisions which increases your self-worth.
Self-reports of expert clinicians have described macro-analytically difficult situations ranging from patients coming late to a session, not talking or blackmailing the therapist with suicide threats.Here I want to go through some TPS threatening to tear up the Common Ground that can be described only microanalytically.However, once analyzed they function as "eye opener".One can see and hear them.To gather a pool of TPS can in the future result in higher perspectives of typicality what makes TPS.I will present segments of transcripts (all translated from German).I will start with examples where the TPS was skillfully handled.

Agenda Transforming Utterances
The 28 th session of a psychoanalytic treatment with an obsessive-impulsive female patient treated by a female analyst starts as follows: The patient's first remarks-can they be heard as a narrative?A narrative (Jefferson, 1978) has a kind of standard format: beginning with a "pre-announcement" ("Wanna know what happened so terrific to me yesterday?")directing the listener's attention, a narrative is followed by "setting the stage", bringing the story slowly to a "climax" of tension, where the listener is supposed to utter his "stance" (Stevanovic & Peräkylä, 2015) which is followed by a decrease of tension.The story is often closed by a coda of evaluation ("moral of the story") exhibiting the conclusions the teller draws from his experience.
This tension curve of narration cannot be found here.The patient uses not a narrative, but a reporting mode.She adds one event after the next using "and then"-connections.No build-up of narrative tension can be found and, what is most important, she begins her report with an evaluative phrase that she got through the day somehow.
The slightly negative self-evaluation of her day (line 5) precedes the report while in a narrative it is placed at the end of the telling.Contrasting to the positive emotional content of the reported experiences is the negative evaluation at the start.
We see this mode of reporting in the prosodic analysis conducted with PRAAT (Boersma & Weenink, 2013) which shows a non-emotional mode of talk.I only represent here the German version of the patient's phrase (line 11-12) "after the weather wasn't so good then um" (for a full analysis see Buchholz & Reich, 2015).The overall report is in the same manner.Skillfully the therapist takes up the evaluation at the beginning, not the content of the report.Herewith the therapist produces an "agenda transforming utterance" (Stivers, 2007a) by installing a new evaluation: to have savoured her day.In the transcript (lines 30-31) we find that the patient laughs.Prosody changes, too.This PRAAT-graph shows the therapist's utterance (Line 29-29) and then the patient's response.We detect how the patient's prosody is changed.The turn-taking is indicated by the intensity line dropping to zero for a time-interval of 0.2 seconds after which the patient (line 30) agrees and laughs loudly.The flatness of F-zero and intensity disappeared indicating the return of assertion and emotional expression.Easily can be concluded that a TPS would have happened if the therapist would have focused her utterance to the single events reported.The flat emotional state would not have changed.A skillfull therapist helped to create a "moment of change" (Stern, 2004, p. 174).
What makes this a TPS?We have a decomposition of the multimodality of utterance with the patient's beginning.On a semantic level she reports enjoyable events, however her embodied telling creates a sharp contrast through her flat voice.The affect vanished.The therapist reacted somehow intuitively and managed to reintegrate enjoyability of events and affect; in following phrases the patient uses the word "savour" several times.One might be sure that this goes unnoticed by her.A kind of unconscious implementation ("re-emotionalizing") can be observed here.
My following example seems to be similar, however, there is another kind of implementation by the therapist's use of a new metaphor.I consider both types of skillfull handling a TPS as variants of "agenda transforming utterances" (Heritage & Clayman, 2010, p. 137) in psychotherapy.

Introducing a New Metaphor by the Therapist
In a first interview with his psychoanalytic psychotherapist a young student presents his compulsive symptom with the following words: He talks calmly in an expert manner, speaks about his "compulsion to control" not his emotions, but his actions: how he controls to have forgotten something when coming home to his girl friend from university.He speaks with "scratchy voice".His self-description of his symptomatic behavior happens in the first minute of the interview.At the end we find the following talk: The patient accuses himself that his control behavior has lost measure, without any relation to what he effectively could control.He expresses his sorrow how strong this compulsion is.The TPS is generated by his compulsion to "controll controlling".This is his project for therapy which includes a pragmatic paradox: "I urgently need your help-but on my condition".This condition is: controlling control.He fights symptoms by renewing symptomatic behavior.However, as long as he is encapsulated in this controlling-control-frame he cannot emotionally share his experience with the therapist.The patient's self-description of "control" has lost potential to establish an emotional Common Ground with the therapist.He is imprisoned in this powerful but restrictive frame of reference.He lacks the ability to reframe his activities from a different perspective.His ability to progress to establish Common Ground activities is limited.
Thus, the therapist's main task is to rearrange things in a way that enable him to emotionally participate.Here, at the end of the first interview the therapist contrasts his "project" of therapy ("this will keep us busy what you are looking for there") with the patient's therapy-as-controlling-control.Then, he replaces the verb "looking" by "seeking" and this is repeated twice.The therapist projects into the future what will engage both of them and while formulating (Antaki, 2008) he uses the new metaphor for what they will do together: instead of control he talks of "looking", then "seeking".This formulation (Note 4) changes the agenda completely: While controlling remains within the patient's frame the new project of "seeking" sets the patient free.Someone who "seeks" is not in need of any control, he can freely find out what he is seeking and he can emotionally participate with joy (if something is found) or disappointment.
The patient agreeingly takes up this metaphor and directly acknowledges the therapist's formulation by repeating "not only control it's seeking".Now the therapist continues the construction of an alternative agenda by adding a "looking around seekingly".The prosody looks like this: Figure 3.The prosodic interchange and turn-taking after therapist's project proposal Easily, we can see the therapist talking with "energy".His intensity is higher, his voice lips vibrate faster producing a higher tonality of speaking than the patient."Natural ear prosody" convinces listeners that he speaks in a mode of imposition; obviously he somehow realizes the implementation of a new metaphor.More than the first example this is a kind of successful "persuasive communication".The patient is portrayed as a seeking person, not knowing how to finish his exams, if his girl friend is the right person for him and how certain life events could be understood.The implementation of this new metaphor starts to make his symptomatic behavior understandable for himself.
Here I want to direct attention to the role of metaphor-change.When I spoke about Common Ground I described three transformational levels: perceptual-conversational-linking.Now appears a fourth level, a consensual exchange of a metaphor central for the patient's disorder.The therapist points (linguists call that "deictic gesture") to what the patient describes as controlling and by uttering his "seeing" he points to the metaphor as if it were a commonly shared new perceptual object.Both can see that.When the patient agrees he transforms the perceptual into a conversational object and links it with some reflection.Thus, we can observe how the metaphor establishes a fourth level of Common Ground activity.The new metaphor is then conversationally acknowledged.
From here the Common Ground activities recursively start with the same procedure at a higher level.A finite number of psychological and conversational instrumentation suffices to produce an infinite number of possibilities.
This 4-level model of Common Ground activity is confirmed in later sessions.Here is the beginning of the 7 th session.
The patient starts the session with his paradoxical and playful remark that he does not want to begin (perceptual).
The calm therapist received-token takes note of this paradoxical start (conceptual).The first utterance of the therapist affiliates with the patient asking if this were important for the patient (link).However, the patient denies that the therapist should begin.Taken on a lexical-semantic level only the situation herewith is contradictory.However, the therapist uses a metaphor of testing and now the patient agrees.This metaphor includes an oberver-position for the patient as tester and a being-tested-position for the therapist.The "unused part" of the testing-metaphor, then, is adressed in the use of another metaphor: wrestling match.Between a tester and the test subject is some fighting if no informed consent is agreed.All these topics are implicitly dealt with in a playful and humerous fashion.And the wrestling-match metaphor exactly describes what happens between the two of them.Maybe, this is something what Daniel Stern (Stern, 2004) had in mind when he presented his theory of a present moment.A present moment, then, could be described as a recursive turning point: if both participants establish a higher level of Common Ground restarting this activity by treating a metaphor as if it were a perceptual object describing the state of their common relationship correctly and consensually.Here is a table integrating the levels of Common Ground, the transformations and the required cognitive abilities of both participants described in advance: To my knowledge this special mode of operation has not yet been described in psychotherapy process research literature.One can see that the therapist does not technically intervene but he is establishing higher levels of Common Ground.Higher levels make deeper emotional exchange possible.

If Multi-Modality of Speech Decays
Conversation is an integral procedure where in most cases words and voice, affect and attunement, verbal and nonverbal expressions of face or gesture form a gestalt where each part fits to the other.In irony or play e.g., this integer gestalt-format of conversation can deliberately be brought into a state of disintegration, when e.g.someone's talk is caricatured.Conversation operates in a multi-mode fashion.In turn construction units it sometimes happens that this multi-modality decays.Someone ends a phrase that is not closed, another stops with the boundary tone given a high pitch which is normally a secure sign that the speaker wants to continue.But he interrupts himself and a long pause follows which is a sign that a "turn relevant place" is achieved where both speakers can select each other or themselves as next speaker (Sacks, Schegloff, & Jefferson, 1974).This is the case in my next example of a TPS.It illustrates a failure done by the female therapist treating a male patient because of his unability to maintain harmony with his wife.I show a segment of a session where a symmetric escalation is build up and I focus on the therapist's contribution to this.Deliberately, I choose a segment that makes the content for readers hardly understandable.The reason is that I want to direct attention to the organization of turn-taking.Watch, how the patient ends his phrasing and how the therapist comes in: Directly before the pause of 15 sec the patient ends his talk with a type of utterance ("errm") indicating that he wants to continue speaking.His ending with a high pitch boundary tone (I omit the prosodic graph) confirmes.Thus, this is no turn-relevant place (Schegloff, 2007).However, the relatively long pause of 15 sec, then, might give a listener the impression that the patient has finished.The decision if this is a turn-relevant place or not is transferred (Note 5) to the therapist.This phenomenon is rare in everyday interaction, however not so in therapy.The therapist's decision to take the turn or not has an unescapable chance to either being accused of not helping the patient to find the right word or by interrupting him permanently.This dilemma emerges from the decay of turn-taking device.It creates the TPS.
Here, the therapist after the pause takes up the patient's last word ("so::") and continues as if complementing the patient's utterance.The patient seems to agree with "yes", however long self-initiated interruptions follow and his phrase is left incomplete again, a long silence of 27 sec follows.
The same pattern is enacted a minute later in the same session: The patient ending with "but" indicates that he has not finished his utterance.This is confirmed by high pitch-prosody again.Yet his pause of 6 sec might be heared as transferring the decision to take the turn (or not) to the therapist.Again, the therapist decides to continue by taking up the patient's last word as if continuing the patient's incomplete sentence.This pattern happens six times in the session.Then we see the "interaction engine" (Levinson, 2006) of turn-taking procedures deeply irritated: Where the brackets are one cannot understand what was said.One hears only several breathing-ins indicating an attempt to start speaking and then stopping this impulse because one hears the other attempts to speak, too.Finally, the therapist takes the turn.Both fight for the right to speak.It is interesting that this segment is initiated by the patient talking about "rivalry".However, he narrates a scenario of rivalry from the weekend.I assume that this pattern makes visible what process researchers (Lippe, von der Monsen, Ronnestad, & Eilertsein, 2008) have termed "the pull of hostility".These authors talk of a "dance" that sometimes originates independent from dancers' intentions.Situationalism is supported by such a view.
Are there recommendations for a better solution?Obviously, the therapist here has "good intentions".She wants to support the patient to overcome what she might think is an inhibition.However, CA-guided analyses (Lerner, 2013;Schegloff, 2003) leads one to think differently about such situations.The well-intended help by the therapist is heard by the patient as impatience.The reason is that his self-interruption together with the high pitch boundary tone creates a type of pause (Frankel, Levitt, Murray, Greenberg, & Angus, 2006) indicating that another idea appeared in the patient's mind not yet prepared for a formulation.His parsing process, thus, has not come to an end.Understood in this way the patient is right would he complain of not being supported to overcome an inhibition but that he feels blocked to utter these other thoughts in his mind.In such situations therapists are guided better when following a maxime like "Just listen!" or "Give room for the unspoken".The idea of "interventionism" seems misguiding for such situations.

Suicidality
However, there are other TPS where "interventionism", better: therapeutic engagement, is demanded.A severly depressed woman in her end twenties begins her 30 th therapy session as follows: … 6 lines omitted… This TPS is composed of several interactional details.First, this patient's talk is full of self-interruptions and long pauses (see arrows).Many thoughts come to mind, however, she does not tell their full gestalt.Her turn constructions is not easy to follow.What might be relevant next?By the series of such utterances every single topic is systematically downgraded in relevance (Koerfer & Koehle, 2007), it is not a type of self-initiated self-repair (Schegloff, 2013).Nothing seems to be of a special relevance and the speaker herself seems to conclude that nothing at all is relevant to tell the listener.
Second, in the German version lines 4-5 ("and of course things always turn around the same way always") clearly express that she fears that the therapist might be bored by being forced to listen to "the same as always".She does not have the conversational power to re-build the Common Ground, it threatens to disintegrate.
Third, the patient intensifies her pressure outlining the existential dimension of protecting the core of herself.This ugrading relevance is enhanced by that she does "not know" adding "or who I am in this whole thing".This wave of up-and downgrading relevance has interactive effects: On the one hand she shows consideration of her therapist's mood which leads her to inhibit full story telling in order not to become boring, on the other hand she increases her demands for help.
Fourth, not-telling her full narration is replaced by intensifying symptomatic complaints; she suffers from a powerful and paradoxical inner image to drive along a wall that protects and encloses her; someone reproaches her, she has to protect herself against being destroyed and she is so worn out that she feels impulses to bring her life to an end.
Fifth, increasing her symptomatic complaints increases her therapist's feeling of being overwhelmed and not-knowing where to begin.In these 4 min of talk the therapist utters hardly anymore than go-ahead-tokens.
With emphasis the patient utters her sense of "desperation" (line 40).
Summarized, it is as if the patient would say "Urgently I need your help, however I let you know that nobody can help me, not even you!" The value of "situationism" can best be demonstrated if I take another example with the therapist better dealing with the complexities of such a situation.The patient had committed a suicide attempt what brought him to therapy.He had managed to deceive his family and his professional environment, he simply lied and did not talk about his serious depressive disorder.He speaks in the similar way as the female patient in the segment before (what I leave out here) and then says: In this segment the therapist again uses several tools to repair the Common Ground: He takes the many details not as "single problems" to be solved in a one-after-the-next fashion but as the patient's attempt to communicate a dark state-of-mind "with sinister ideas that come to your mind" (see arrow).By this phrase he transforms into a conversational object to what the patient tries to point to ("deictically").
Level one and two of Common Ground activities are established.He links, then, with the patient's self-accusations using an "I-positioning" while speaking as if he were the patient (line 368-369) and this helps to re-normalize the patient's experience as he implicitly learns that there is someone who blatantly knows what he is talking about.Finally, the therapist introduces a complex metaphor of mother and child.These procedures seem to restore the Common Ground again.The segments end with a first pause of nearly 12 seconds.
However, so simply a seriously depressed patient's complaints cannot be cured.The patient comes up with a lot of similar complaints and two minutes later the following sequence is enacted: Changing to the topic of caring for his own children the patient seems to indicate why he is unable: because he feels as a child as the therapist uttered a few seconds before.Feeling a child himself his children become an unbearable burden.Accusingly he appeals to himself that he is the father but cannot comply with his role and is frightened by ideas of his children becoming drug-addicted.
We find the therapist using the metaphor of a child in a way the patient can easily accept (428)(429)(430).This is later confirmed by the patient telling that he sleeps in his children's room to feel their closeness.The therapist affiliates with the patient's helplessness by verbally doing what he is talking about-taking the "child's" hand.This is more than positioning, it is therapeutic agency.
This segment, finally (line 437-449), results in what Lerner (Lerner, 2013) has called "other completion".Both participants begin to move into a micro-universe of distributed knowledge mutually completing their sentences.This type of collaboratively co-constructed utterances is described by Hutchins in the following way: "In the most frequently studied type of collaboratively constructed utterance, one speaker begins an utterance in a way that projects possible completions.Another speaker then contributes utterance elements that are incorporated into a jointly produced utterance.The acceptance by participants of a collaboratively constructed utterance is strong evidence for the establishment of Common Ground understanding" (Hutchins & Nomura, 2011, p. 29).
It is easy to see that what is described here as a Collaboratively Constructed Utterance (CCU) is what we in everyday language term as "speaking with one mouth".Therapist and patient have found a way to repair the endangered Common Ground.The therapist contributed in active agency.

Discussion and Conclusion
It proves useful in psychotherapy process research to study not only method or therapist or therapist-patient-matching.These studies have ambiguous results.I proposed here to use conversation analysis as a hard empirical program of research (Schegloff, 2007) from linguistics and the social sciences to make use of a concept of "situationism".Some methodological rules were outlined.
To study "situations" which are well known to clinicians but hardly studied in micro-analytic detail can enrich professional practice and process research.Voice is such an informative organ.Some think, voice contains more individual information than finger prints do.However, to include these levels of analysis, of which I gave some examples, does not make things easier.Analyses become extremely complex and need more expertise.This holds for video-records, too.To include all these levels of analysis is very demanding and yields data whose complexity and relevance is not easy to understand.
The analysis of my conversational data in psychotherapeutic talk shows how important Common Ground is.
Based on these data I showed that Common Ground is not a kind of psychological sympathy-relationship alone, but Common Ground has a linguistic dimension.Common ground cares to adapt the speed difference between (slow and linear) conversation) and (fast and circular) cognition.If something of relevance is to happen in a two-person relationship (not only in therapies) such a strategy of reducing the speed difference is required.Dealing skillfully with Common Ground produces trust in a psychological sense, reproduces the Common Ground and reduces the enormous complexity of conversation.However, Common Ground is no once-and-for-ever stable environment, it must be produced and reproduced with agency.I showed that there is a closeness between psychological and linguistic conceptions of Common Ground and developed a recursive theory of Common Ground activities ranging from perceptual to conversational levels where linking with other conversational reportable experiences becomes possible and then, sometimes, the metaphorical level is achieved.If a metaphor is conversationally acknowledged and ratified it is treated by conversationalists like a perceptual object (e.g., as a description of the common relationship) and the procedure beginning with descriptors of perceptual objects can recursively start again.By this way one can gain an understanding how a process of relationship-development can become microanalytically described.Future research, empirical and conceptual, will have to test these conceptualisations for other data.

Figure 1 .
Figure 1.Prosody of the patient's affectively flat conversation

Figure 2 .
Figure 2. Prosodic analysis of the therapist's replacing of the patient's metaphor and the patient's turn-taking

Figure 4 .
Figure 4. Schema of Common Ground activities; conversation and cognition Therapeutic practitioners can be helped to better understand what happens and what it is precisely what makes them helpless and, if situations are compared as here presented in two cases of suicidality, how to deal better with such situations.Researchers could value the precise description of situations which are overlooked if student raters are used in studies to evaluate process and/or outcome in sessions.Obviously, as in childhood research, it requires a long training to see what can be seen in a well made transcript.At International Psychoanalytic University (IPU) in Berlin (Germany) we have begun to cooperate with linguists in order to come closer to details of voice.