1. Top of page
  2. Abstract
  3. Method
  4. Analysis
  5. Discussion
  6. References

The present article focuses on the study of circular questioning “in situ,” that is, in the context of the discursive arena of an actual first systemic family therapy session. Two typical circular questions are selected, and discourse analysis heavily drawing from the discursive action model is deployed with the aim to highlight their function in the context of therapist and family members' problem talk. The analysis demonstrates the gradual building of two respective patterns, which both exhibit signs of shifts toward the instillment of a systemic epistemology in relation to problem talk: a shift from homogeneity to heterogeneity in family members' voices, which legitimizes the existence of different viewpoints within a system, and a shift from the construction of an accusation toward its deconstruction, which challenges the linear perception of causality underlying the accusation, thus introducing a more circular perspective. Implications are discussed in relation to contemporary, constructionist systemic family therapy practice, and reference is made to the methodological “promises” and challenges of the deployment of discourse analysis for the scrutiny of systemic family therapy therapeutic techniques and tools under a discursively informed light.

Since its advent in the early 1980s by the Milan Associates (Palazzoli, Boscolo, Cecchin & Prata, 1980), circular questioning (CQ), the well-known “interviewing tool,” has been closely related with the “debated” (e.g., Anderson, 1986) practice of the Milan model and its developments. Initially designed to serve as an exploratory tool eliciting “news of difference,” in a Batesonian sense (Bateson, 1979), it aimed at facilitating the therapist's effort to conduct the session by remaining “loyal” to the three well-known principles of “hypothesizing,” “circularity,” “neutrality,” by denoting an adherence to a systemic epistemology (Palazzoli et al., 1980). Following the shift from the epistemology of first-order to that of second-order cybernetic and constructivist approaches, reflected in respective developments (Bertrando, 2007; Boscolo, Cecchin, Hoffman & Penn, 1987), CQ gradually acquired the additional quality of an interventive tool, argued as itself ample to generate change (Tomm, 1987a).

Despite its links with a currently out-fashioned era for the field of systemic family therapy, namely the first-order cybernetic one, CQ still “survives” largely in the European scene (Bertrando, 2007; Linares, 2001; Perosa & Perosa, 2010; Pote, Stratton, Cottrell, Shapiro & Boston, 2003). Nevertheless, research on CQ has been and still remains particularly sparse. Although a number of writers have worked on demonstrating, classifying, and expanding the use of circular questions and their function in the therapeutic process (e.g., Brown, 1997; Freuridas, Nelson & Rosenthal, 1986; Tomm, 1985, 1987a,b, 1988), this body of work has a purely theoretical character, even in cases where transcripts of actual sessions are presented (e.g., Penn, 1982). Only a few studies have set out to test the hypotheses that the relevant literature states about CQ, methodologically located mainly in the positivist tradition (e.g., Dozier, Hicks, Cornille & Peterson, 1998; Ryan & Carr, 2001; Scheel & Conoley, 1998). While presenting with interesting insights, these studies locate the study of CQ in a context “outside” the context of its actual use; Scheel and Conoley's (1998) data, for example, are derived from interviews conducted subsequently to the actual linguistic interaction between the therapist and the family members. With this article, we will attempt an alternative perspective by relocating the focus of CQ's study “in situ,” that is, in the actual context of the therapist-family members' discursive transaction over the course of a therapeutic session. To do so, “research wise” we will deploy a discourse analytic perspective.

Discourse analysis (DA) is a theoretical and methodological orientation currently running throughout the whole body of the social sciences, reflecting the basic premises of the wider intellectual movement of social constructionism, widely known as the “turn to language” (Burr, 1995; Gergen, 1999). Our use of the term here points to certain developments in the field of social psychology, originating in the 1980s (Potter & Wetherell, 1987).

Discourse analysis has been founded on the rejection of the traditional notion of language as a neutral medium, used to represent the physical and the psychological realm. It focuses on language in use (discourse) and argues for the inseparable nature of discourse and social action. Different traditions share a family resemblance, as they all rest on the above assumptions, but draw different connections between discourse, society, and the individual (Wetherell, 2001). Discourse analytic approaches that focus their interest on psychological issues are usually clustered under the term, “discursive psychology.” Instead of seeing discourse as the site where internal states are represented, discursive psychology approaches discourse as the site, where psychological states are constructed and, thus, introduces discourse as the prime object of study (Edwards, 1997). On a methodological level, a major implication of this is that what counts as data are, usually, naturally occurring instances of talk in interaction (Taylor, 2001).

Although discourse analysis can be used to investigate any type of psychotherapeutic process, most discourse analytic studies have so far focused on social constructionist therapies, due to, as argued, a theoretical and epistemological compatibility (Avdi & Georgaca, 2007; De Haene, 2010; Georgaca & Avdi, 2009). Discourse analytic findings are more meaningful to clinicians who see the formation and the solution of psychological problems as taking place in a discursive arena. Systemic family therapy has approached the emergence and treatment of psychological problems in interactional terms long before social constructionist ideas officially entered the field (see Watzlawick, Beavin-Bavelas & Jackson, 1967). Nowadays, a range of contemporary approaches endorse social constructionism and turn to language to offer alternative conceptualizations of clinical practice (e.g., Anderson, 1997; Anderson & Goolishian, 1988; Bertrando, 2007; Seikkula & Olson, 2003; White & Epston, 1990). As a result, family therapy research exhibits a growing tendency to make use of discourse analysis or conversation analysis in explorations of the therapeutic process (e.g., Avdi, 2005; Couture & Strong, 2004; Couture & Sutherland, 2006; Harvie, Strong, Taylor, Todd & Young, 2008; Kogan, 1998; O'Reilly, 2007; Sutherland & Couture, 2007; Sutherland & Strong, 2011). By drawing attention to features of talk, which usually remain unacknowledged, discourse analysis can enhance therapists' reflexivity (see also, Couture & Strong, 2004) by sensitizing them to the microdynamics (e.g., Couture & Sutherland, 2006) or the macrodynamics (e.g., Avdi, 2005) of the therapeutic conversation and to their own role in this complex discursive space.

Following a similar rationale, in this article we will attempt to investigate CQ exclusively in a discursive arena, as part and parcel of an ongoing circular discursive interaction between the therapist and the family members in the context of the evolving therapeutic dialogue. Our focus in the following analysis will be on two typical circular questions, “What is the problem in your opinion?” and “How do you explain the other's behavior?” and their deployment in the context of a first family therapy session. Customarily used as the opening question of a first session, the “problem definition” question addresses the links between the problem and the relationships in the family in the present, whereas the “explanation” question addresses respective links in the past (Penn, 1982). Our aim here is to highlight the ways in which these two questions function while the therapist and the family members engage in a dialogue about the family's problem.


  1. Top of page
  2. Abstract
  3. Method
  4. Analysis
  5. Discussion
  6. References

All the data presented and analyzed below are derived from a single therapeutic session, which took place in 2007, in the context of a training program in systemic family psychotherapy for mental health professionals located in Greece. Analysis took place following the termination of the particular therapy, from which this session is derived. Twenty-two audiotaped sessions available in this context were considered for their appropriateness for the present study. This particular session was selected on the basis of three criteria, denoting a preference for first sessions, securing the absence of prior interactions between the therapist and the family and, thus, strengthening analytic claims, a high level of expertize in the Milan model on behalf of the therapist, and lastly, a paradigmatic phrasing of circular questions. The therapist involved in this selected first session is an instructor at the training program and has many years of experience in the Milan model. A therapeutic team consisting of trainees, with both authors present, was watching the session behind a one-way mirror. In the beginning of the session, written informed consent was obtained from all family members regarding the trainees' presence behind the mirror, the audiotaping of the session, and the potential use of the audiotapes for research purposes. In the text that follows, all names have been replaced with pseudonyms, and care has been taken to preserve anonymity. Additional permission for proceeding with publication of the study was also requested and granted by the Institute running the training program.

The family who participated in the session included a father and his two sons, George (24 years of age) and Nick (22 years of age). Initial contact was made by the father, who expressed concerns about his younger son. According to the father, during the last year, Nick's behavior had been so strange and so difficult to cope with that he had begun to worry about Nick's state of mind. In the extracts presented below, “TH,” “F,” “G,” and “N” are used to indicate the speaker, corresponding respectively to the therapist, the father, George, and Nick, whereas line numbers indicate the location of the extract over the course of the session.

The whole session, which was conducted in Greek and lasted for one and a half hour, was audiotaped and subsequently transcribed verbatim. The transcription system selected here is close to the conventions of a standard written text, following the requirements of the analytic method employed and also our aim to ensure the readability of the transcript. Once the analytic process was completed, the most illustrative extracts were translated in English. In the presented extracts, the symbol “…” is used to indicate a noticeable pause in talk, whereas the symbol “(…)” indicates segments of the transcript that do not add to the analysis and/or threaten the preservation of anonymity and are, therefore, omitted.

The analysis draws heavily from the “discursive action model” (DAM) introduced by Edwards and Potter (1992). DAM is related with multiple trends in the history of language study (Horton-Salway, 2001), including linguistic philosophy (Austin, 1962; Wittgenstein, 1953), ethnomethodology (Garfinkel, 1967), conversation analysis (Atkinson & Heritage, 1984), and poststructuralism (Shapiro, 1988). In line with the fundamental principles that underly all models of discursive psychology and retaining the emphasis it places on context, DAM approaches psychological phenomena related to emotional and cognitive states as constructed in people's accounts in everyday interactions and expects the same person's account to vary depending on the function it performs in different contexts (Edwards & Potter, 1992). DAM offers an alternative perspective on psychological life by highlighting the actions the discursive constructs of psychological states perform within the interactional context in which they appear (Edwards & Potter, 1992). Its analytic focus on the local interactional context of talk accounts for our choice to deploy it here, as we aimed at illuminating the function of circular questions in the context of the therapeutic dialogue.

Discursive action model distinguishes three key features of people's discursive practices: “action,” “fact and interest,” and “accountability” (Edwards & Potter, 1992). Firstly, DAM proposes that talk is action oriented; it is designed to perform specific actions embedded in activity sequences such as invitation refusals, blamings, or defendings. It is, thus, argued that speakers always construct their accounts, descriptions, and formulations in a particular rhetorical context. The focus of analysis is on how psychological phenomena like remembering and attribution are put together in talk and what actions these constructs are designed to accomplish (Edwards & Potter, 1992). Secondly, DAM suggests that, given that reports are located in a rhetorical context, they are designed in ways that protect them from being undermined as motivated by stake or interest. DAM points to a variety of discursive techniques, like “vivid description,” “extreme case formulations,” “lists and contrasts,” etc., that speakers use to construct their reports as factual (Edwards & Potter, 1992). Lastly, DAM addresses the management of accountability in talk in interaction. It identifies two distinct levels of accountability: on the one hand, reports have to deal with issues of accountability raised in the reported events; on the other, as reports perform particular actions in specific interactional contexts, they have to bear the accountability that accompanies these actions. For DAM, one of the analytical tasks is to examine the ways in which these two distinct and often intertwined levels of accountability are managed in talk (Edwards & Potter, 1992). It is important to note that this mode of analysis cannot be seen as an alternative form of individual psychology. The key features of talk identified above are not connected to attributes of the individual speakers that do the talking. They are rather seen as characteristics of the discursive practices speakers resort to, yet are not in sovereign control of (Edwards & Potter, 1992).

The analysis presented here involved both researchers separately going over the data repeatedly, especially working on segments of the transcript entailing circular questions. Both the structure and the content of discourse were analytically attended to, and an attempt was made to ground interpretation on a turn-to-turn basis. Attending to DAM “directives” was translated into a practice of an action-oriented interpretation of the identified patterns. This was based on the adoption of a rhetorical view of the interactional context, as well as on the pursuing of a specific consideration of issues of interest and accountability as corresponding rhetorical concerns. It is important to note, however, that DAM does not entail specific suggestions as to how to proceed with analysis. Like all discourse analytic approaches, it rather points to broadly defined features, which need analytic attendance (Wetherell, 2001; Wood & Kroger, 2000). Following that, in our analysis, we focused on the exact ways in which the participants' reports are put together. Instead of examining their features in respect to their referential properties, we have examined them in respect to the particular, rhetorical actions they seem to be designed to accomplish. An example of this can be found in the analysis of extract 2 (p. 15, first paragraph), where we analyze the actions that seem to be performed by the therapist's account. Furthermore, we also attempted to attend to discursive features related with fact and interest. The analysis of extract 1 (p. 12, second paragraph) presents such an example, by explicating the ways in which the father's report is constructed as factual. Finally, a way of analytically attending to the two intertwined levels of accountability that DAM brings forward is exemplified in the analysis of the first extract (p. 13, last paragraph): on the one level, the father's report presents his younger son as accountable for certain reported behaviors; on the other, as a direct consequence, the father himself is automatically accountable for laying blame on his son. Related accountability issues are further pursued in the analysis of extract 4 (p. 20, first paragraph).

The analysis presented below incorporates a form of triangulation (see Silverman, 2006) in that it is the outcome of the comparison and combination of the two researchers' independent analytic attempts. However, triangulation here is not intended to support claims to objectivity, but to enhance the quality of our work. To that aim, we have also included extensive transcribed extracts, as well as a note on reflexivity, thus attempting to adhere to criteria for quality control in DA studies (Wood & Kroger, 2000).


  1. Top of page
  2. Abstract
  3. Method
  4. Analysis
  5. Discussion
  6. References

What is the problem in your opinion?

The norm. Extract 1 appears in the beginning of the session, right after the participants have introduced themselves, have discussed the particularities of the setting, and have signed the relevant consent forms.

Extract 1

10 TH: I am aware that you have discussed with Mrs. T. on the phone, since this is our procedure, but

11 would you like to tell me what brings you here today? You might not all agree on that.

12 F: To begin with, it is not necessary to describe the facts we have experienced. I think you are

13 informed, or would you like us to …?

14 TH: Whatever you wish. I am asking you what brings you here now. What is the problem right now.

15 F: In a nut-shell, my wife, the mother of my children, died about 2 years ago in a way that I would

16 say deeply marked our whole life, she committed, that is, suicide, as a matter of fact, it was (…). The

17 kids witnessed their mother in this condition. I do believe that this has definitely affected all of us. For

18 the past 2 years we have been trying to stand on our feet. Sometimes we manage, sometimes we

19 don't; all this results in certain tensions and disputes, there is no harmony in our house, no

20 coordination, no consistency and all this may come up, let's say, through certain phrases like when my

21 older son said “in my life I feel alone.” My younger son said “daddy I feel like we are roommates.” In

22 other words, we feel that our family is not functioning, we all feel that our family is not functioning,

23 maybe there are not enough psychological resources to keep it running after what has happened to us.

24 In the meantime, during that period, we had to face two more deaths, my father's and my mother's.

25 They were quite old, but …

26 G: They came before

27 F: Yes, in 3 years, that is, we had three deaths.

28 TH: What was the order?

29 F: First my mother, after about one and a half year, roughly speaking, it was my wife, and after one

30 and a half year it was …

31 TH: Your father

32 F: Yes, it was my father

33 G: After 2 years and 3 months it was mum

34 F: That is, this is the order of the deaths in these 3 years; and I think it is only natural for anybody

35 to experience some sort of oddities because of these facts. I am very straight with my children, I feel,

36 lets say, that I have to be honest that my younger son is the way he is, he is short-tempered, he is

37 irritable; I may be the one to blame when I provoke him sometimes with some of my wrong actions

38 and we want, lets say, to set things right (…)

In the beginning of the extract, the therapist opens up the conversation with a typical circular question (10–11). She asks about the problem that brings the family to therapy introducing right from the start the idea that the family members' views about their difficulties may not be identical (“You might not all agree on that”). The father is the one who responds to the therapist's question. The upshot of the answer he offers to the therapist is explicitly stated in line 22: the problem is that “our family is not functioning.” The distinctive organization of this answer, though, links the father's response to a much more sophisticated action than simply offering this piece of information.

It can be noticed that the father constructs his report as factual using several of the techniques, which Edwards and Potter (1992) characterize as “techniques of fact construction”: he talks about “facts” (12), he incorporates his opinion about the problem in a narrative sequence in which the problem acquires a causal explanation and appears plausible (15–23), and he makes use of vivid description, direct quotation, and independent witnesses' consensus (20, 21). However, the father seems to be working toward an additional task, much more delicate than simply constructing his report as factual; he constructs his report in a way, which emphasizes that he is presenting not just his own opinion about the problem but an opinion that all members of the family share in common. That is, the father answers the therapist's question on behalf of all three members of the family. For the most part, the father's account is offered on a “we” instead of an “I” basis (18–24). At the point where he describes the problem, he chooses to prove its existence (“all this may come up through certain phrases”) by directly quoting his two sons (20, 21). To sum up, instead of using a neutral, empiricist expression, of the kind that is typically used in fact construction, like “our family is not functioning,” he uses an expression that underlines a subjective, yet collective feeling (“we feel that our family is not functioning”). The reformulation that appears immediately after (“we all feel that our family is not functioning”) indicates the great significance attached here to the collective character of the report (22).

George, the oldest of the two sons, interrupts the father's report twice (26, 33) to provide a more precise description of something that his father portrays in a vague manner, rather than to correct him. The father, on the other hand, does not treat George's utterances as interruptions or objections, but seems to recognize their supportive character, as he incorporates the detailed information suggested by his son in his own report (27, 34). It seems here that George and his father both treat the account the father offers to the therapist as a collective account that represents the views of all family members.

The father uses the first person to refer to his younger son, Nick, whom he had identified as the “patient” at this initial contact (34–38). The picture he draws of Nick by his own eyes has a distinctive negative character (“short-tempered” and “irritable”). His account shifts from the previous rather flowing style to a more elaborate one, exhibiting extra caution, as he seems to attend to the following issues: he naturalizes Nick's condition (“it is only natural for anybody to experience some sort of oddities because of these facts”), he emphasizes that he is not accusing Nick, he does not hold Nick responsible for his strange behavior (“I may be the one to blame when I provoke him sometimes with my wrong actions”), and he justifies the fact that he is speaking of Nick (“I feel that I have to be honest”) in his presence (“I am very straight with my children”). His account seems to orient to the multifaceted accountability issues that his talk about Nick brings to the forth. It can be seen as forming a carefully organized defense, on the one hand against the possibility that Nick is held accountable for the reported behavior, on the other hand against the possibility that the father himself is held accountable for making accusations against Nick. Simultaneously, Nick's potential objection to this report is managed, resulting to the protection of its collective character. Nick himself seems to collaborate to this, as he does not take a turn at this point to object and, thus, disrupt his father's report.

Extract 1 illustrates that in the beginning of the session, the family appears to respond not only to the question “what brings you here today?” but also to the remark “You might not all agree on that.” All three members of the family, each in their own distinct way, seem to cooperate at this point in their denial of the possibility to which the therapist's remark refers. This elaborate effort points to the norm to which the participants' talk seems to orient in the beginning of the session: the total agreement of all family members on factual, incontestable problems.

Reversal. Extract 2 appears short after extract 1, as soon as the father draws his report to a close.

Extract 2

54 TH: And what would you say is the problem now in your opinion? Your current problem?

55 F: You mean for me or for my whole family; what do I think of this problem?

56 TH: Yes, what do you think of it.

57 F: Well, first of all that we are not bonded as a family. Now and then, there are times when there are

58 conflicts among us, especially, I would say between the children, which saddens me a lot that there are

59 big conflicts. I would like to see my child, my children, to be happy, to move on, to have ambitions

60 and goals. I don't see these things happening and I can understand that something is wrong.

61 TH: Do you have Nick in mind when you say this?

62 F: Yes, Nick, who is the younger of the two.

When the father's report comes to an end, the therapist repeats the question she posed earlier in a slightly different manner (54). Here, she explicitly asks for an account of the problem from the father's own perspective (“And what would you say is the problem now in your opinion? Your current problem?”). By reformulating her original question, the therapist seems to perform two actions; she demonstrates that her original question has not yet been answered and points to the type of answer she seeks, namely a personal view on the problem. This time, her formulation does not present the existence of multiple, personal views on the family's problems as a possibility, like in her original question, but takes it as given, thus potentially suggesting that disagreement between the family members in relation to problems is expected as the norm.

The father asks for clarification (55) by offering two different interpretations and asking which of the two is the correct one (“You mean for me or for my whole family; what do I think of this problem?”). The therapist's last utterance (“Your current problem?”) is ambiguously stated; it fits with either one of the two interpretations that the father proposes. This, however, is not the case with her preceding question (“And what would you say is the problem now in your opinion?”); it matches only one of the interpretations proposed by the father. It can be argued that it is not the ambiguity of the therapist's question that causes communicational trouble here. Trouble possibly arises from the fact that this question constructs the existence of divergent perspectives with regard to the family's problems as expected and natural and, thereby, reverses the norm the participants so far orient to.

Once the therapist makes clear what she means (56), the father presents a new account of the problem, this time from his own, personal perspective (57–60). He predominantly uses the first person (“I would say,” “which saddens me,” “I would like to see,” “I don't see these things,” and “I can understand”) and does not make any effort to justify or naturalize Nick's behavior this time. On the contrary, he uses an expression that points to pathology (“something is wrong”). As this new account does not have a collective character, the other participants' consensus does not seem to be an issue anymore.

The new norm. Later in the session, the therapist investigates George's and Nick's ideas about the family's problems. Extract 3 appears, when Nick's account is suddenly interrupted by the father.

Extract 3

143 TH: Well, I wonder whether what you have said about anybody being able to stand on their feet is

144 actually the way you account for this current situation.

145 N: I believe that yes it accounts for it.

146 TH: It accounts for the way you are because each one of you is trying to stand on your feet, to move

147 on with your lives.

148 N: Yes

149 F: You will allow me to make an interruption here, since I believe that things are not exactly so, that

150 everyone can stand on their feet. When a house is kept up and provided for by the father, it is the

151 father who does the washing up, the cooking, the shopping, the ironing and I believe that I

152 wouldn't say that anybody feels alone or standing on their feet. All this has to do with the material

153 aspect; the emotional aspect, if what I have heard has to do with it, well, there seems to be there a

154 great deal of truth.

155 TH: Mmmm

156 F: But I wouldn't say that one feels alone and paves one's own way when one expects to be fed by

157 somebody else, to have one's shopping, cooking and ironing done by somebody else, to have the

158 house expenses covered by somebody else.

159 TH: In other words, it seems that on a material level you are a family.

160 F: On a material level, yes.

161 TH: But somehow you are all implying that on an emotional level your relation is …

162 F: I don't know… my children… we have come here and let me hear their opinion, if there is one. On

163 a material level, I think, there must be no complaint whatsoever, because, I do believe, that I do much

164 more than what I should.

The father interrupts the dialogue between Nick and the therapist and states his opposition to Nick's ideas about their problems (149–154, 156–158) by means of a personal, subjective account (“I believe,” “I wouldn't say”). As such, this account is in need of validation and is, therefore, accompanied by logical arguments that support it (“When a house… on their feet,” “I wouldn't say that… covered by somebody else”).

The therapist takes a turn going over the main points of the father's talk twice (159, 161). Her first statement paraphrases one the two basic premises embedded in the father's argument (“In other words it seems that on a material level you are a family”), and he agrees with it straightforwardly (160). Her second statement refers to the other, yet in a different manner: here, the father's idea is constructed as an idea that all members of the family share in common (“But somehow you are all implying that on an emotional level your relation is…”). This time the father does not agree with the therapist (162). His response has the characteristic “dispreferred turn shape” typically used in disagreements (see Hutchby & Wooffitt, 1998); it is delayed by pauses in talk, qualified (“I don't know”), and accounted for (“my children… we have come here and let me hear their opinion, if there is one”). The reason he gives seems to be that the appropriate thing for him to do “here” is not to assume but to “hear” his children's “opinion.” What should not be assumed seems to be the uniformity of opinions, which the therapist's remark implies. Therefore, the father goes on elaborating on the distance between his own and Nick's opinion (163–164). In strong contrast to the beginning, it seems that later in this session, it is the existence and acknowledgment of divergent opinions about the family's problems between participants, which now seems to have acquired the status of a norm that organizes the participants' talk. This seems to be the case in father's talk, when he turns to such an idea to justify his conduct.

The three extracts presented above keep track of a change, which seems to take place, that is, the reversal of an implicit norm, which underlies the organization of the participants' talk. In the beginning of the session, the uniformity of the participants' views concerning their problems seems to have a powerful, normative character, whereas later in the session it is the exact opposite, the divergence of relevant opinions, what seems to have the status of a norm. The “problem definition” question seems to play an important role with regard to this change. However, the present analysis sees talk in interaction as inherently open-ended and unpredictable; it connects the “problem definition” question to the change discussed above and argues not for the causal, in a lineal sense, but for the dynamic character of this link.

How do you explain the other's behavior?

The accusation. The following extract appears later in the session, at a point, where the relationship between the two brothers is being discussed.

Extract 4

330 N: Let me mention something, what my father said about my girlfriend. He has asked me that she

331 should not sleep in our house when he is around and I have generally respected that, except for two

332 times, when it was really cold outside and we couldn't go anyplace else … my father asked me that

333 whenever he goes out to visit his acquaintance … that I shouldn't bring my girlfriend over here, that I

334 shouldn't spend the night with her in my house … that's it.

335 TH: Hmm, hmm, and what did you think about that?

336 N: It simply doesn't feel right

337 TH: It doesn't feel right …

338 N: It doesn't feel right

339 TH: Would you like to elaborate more on this?

340 N: …. (Sigh).

341 TH: You mean to say that your father judges things …

342 N: From his point of view, yes. He doesn't consider what … would make me happy, to be, that is,

343 with my girlfriend in my house, and not to be going out anywhere but my house. With my girlfriend.

344 TH: He doesn't understand you, in other words, he doesn't understand how much you need this.

345 N: Yes, yes.

346 TH: And how do you account for this behaviour, why does he do that?

347 N: … I don't know… I don't know what he might have in his mind

348 TH: On what grounds, according to your father, shouldn't this be happening?

349 N: I don't know… He may have grown up with some other models, maybe … I don't know

350 TH: What about you … do you know why your father behaves like this on this matter? What …

351 thought lies beneath these actions?

352 G: I don't think that this is exactly how he sees this matter, no, I can't understand it.

353 TH: You can't understand, that is, on what grounds he acts like this…

354 N: May I say something else too? With this sort of behavior, without even saying a single word, my

355 father has made my girlfriend not to want to come again to our house. No matter what I have told her

356 to reassure her that there is no problem, my father's behavior has shown that under no circumstance

357 does he want this girl in our house.

In the beginning of the extract, Nick takes a turn to “mention something,” namely an accusation against his father: “my father asked me… that's it.” (332–334). Nick's father is here presented as accountable for a behavior that “doesn't feel right” (336). Before stating his complaint, though, Nick inserts a small introduction (“He has asked me… anyplace else,” 330–332) presenting a situation, in which he has “generally respected” his father's wish instead of objecting to it. In this introduction, accountability is reversed: Nick is here accountable for not complying with his father's wish and therefore gives a good reason for this (“when it was really cold outside and we couldn't go anyplace else”). In Nick's talk, the father's presence/absence from the house is connected with the discrepancy between the situation described in the introduction and the one depicted in the following accusation. Using alternatively the expressions “our house” and “my house,” to point to when father “is around” and “when he goes out to visit his acquaintance,” respectively, Nick constructs a whole set of interests, rights, and obligations. In this way, the same request on behalf of the father appears as justified in one case and unjustified in the other. The introduction Nick inserts in his talk seems to have an important function in the construction of the accusation. By contrasting the two conditions, Nick indicates what he is actually accusing his father of, namely of behaving in an unpleasant manner with no good reason whatsoever.

After discussing with Nick about his accusation for a while, the therapist poses a typical circular question, she asks him to explain his father's behavior (346). The long pauses in Nick's response signify a problematic topic in talk (347). Nick does not offer the explanation that the therapist asked for (“I don't know”). Nevertheless, his response is indicative of the way he interprets the therapist's question (“I don't know what he might have in his mind.” 347), which is then validated by the therapist's reformulation of her question (“On what grounds, according to your father, shouldn't this be happening?” 348). Like before, Nick responds with long pauses (349) and makes a hesitant effort to account for his father's behavior (“He may have grown up with some other models, maybe”). The type of account he offers with uncertainty does not fit with the interpretation of the question that Nick himself, as well as the therapist, suggested. Nick seems to recognize this, as he closes his turn by repeating “I don't know.”

After laying blame on his father for a certain behavior, Nick seems to treat the therapist's request for an explanation of this behavior as problematic. Had he not carried out so, this would possibly have undermined the current action performed in his talk, namely the act of blaming. The therapist's question invites an explanation and, thereby, a possible justification of the father's behavior. Nick's accusation, on the other hand, is predominantly based on the unjustified character of this behavior, which is carefully constructed in his talk. As long as the therapist's question remains unanswered, the accusation continues to be effective, which allows Nick to carry on with the act of blaming by elaborating on the unpleasant consequences of his father's behavior in his next turn (354–357). As he keeps talking, the accusation grows bigger covering all the more reprehensible issues.

The explanation. The conversation about Nick's accusation goes on for a while. George makes an attempt to answer the therapist's question stating that he sees his father's actions as protective. Extract 5 appears right after the father has agreed with George.

Extract 5

396 TH: How come …errrrr … what I think listening to you is how come if anybody of you shows a sign 397 of care towards the other, this eventually ends up in…

398 F: Yes

399 TH: It ends up in a fight, doesn't it? Well, I don't know, I can understand now your point of view, but

400 something like this …

401 N: May I say something?

402 TH: Yes

403 N: Errrr, to begin with, I think that it's only natural for a couple to be at the crossroads after some

404 months; to experience some sort of tension. (…) But then again, if there was some sort of …

405 errr…genuine interest from my father's part, I think, that he wouldn't have come to me to tell me

406 to break up with her, he would have tried to help us sort things out rather than just break up.

407 TH: If this wasn't genuine interest, then what sort of interest was it, do you say?

408 N: This wasn't any kind of interest at all.

409 TH: Mmm, then how do you account for your father's action, let's say?

410 N: I do believe that … it's nobody's business what I do in my personal life. (…)

411 TH: What is it, do you think, errrr, that makes your father behave like this, that makes him prevent

412 you from having a relationship, that makes him urge you to find a job … what is it that makes him…

413 interfere with …

414 N: The fact that he wants what is best for me, but it's the way that he does it that's the real issue here.

415 And I surely should have mentioned some other things that took place in our house too, like some

416 really annoying details. That is, he came to me and made, ok, a well-intentioned remark, because I

417 had used a towel for the beach to dry myself in the bath… in the bathroom and then he was annoyed,

418 which really got on my nerves, since for me it was only a towel, and I don't use separate types of

419 towels.

420 TH: Is there anything though that your dad does right? Because it appears to me that there is, there 421 are some things for you that are right, from the ones your dad does?

422 N: Oh yes, definitely. Many things

423 TH: Hmmm, would you like to mention a couple of them?

424 N: The fact that he is in the house as much as he can, as much as his energy allows him to, the fact

425 that he supports both of us as much as he can, and that he wants us both, my brother and myself, to be

426 ok.

The therapist poses a question, which incorporates the explanation that George and the father put forward (396–397), that is, she refers to the father's behavior as “a sign of care.” Right after she starts formulating her question (“How come”), she makes a pause, hesitates (“errrrr”), and repairs herself (“what I think listening to you is how come”). Her self-repair is possibly intended to correct a substantial error in her talk: her original question treats George's explanation as factual, when not all of the participants have agreed on that. By repairing herself and presenting the explanation as a subjective opinion that she personally espouses (“what I think listening to you”), the therapist orients to this issue and marks a sensitive topic in talk.

Nick takes a turn and challenges the idea that he is in need of care (403–404), as well as the idea that his father's intention is to care for him (404–406), thus countering the explanation the therapist's question suggested. Accepting Nick's opposition to the available explanation, the therapist returns to her original question and persistently asks for his own account of his father's behavior (407, 409, 411–413). Nick repeatedly resists the therapist's request (408, 410).

Once Nick agrees to the explanation proposed by the other participants (“The fact that he wants what is best for me,” 414), he then proceeds to introduce a new topic in talk (“but it's the way that he does it that's the real issue here”), thus demonstrating that the topic discussed so far no longer stands. To continue with the act of blaming, Nick has to produce a whole new accusation (“And I really should have mentioned… annoying details.”). However, as the so far established account of the father's behavior becomes a part of this attempt (“ok, a well-intentioned remark”), the construction of a new accusation becomes problematic. By inserting a justification of his own behavior in his report (“since for me… separate types of towels”), Nick demonstrates that accountability is here debatable and no longer clear-cut. The presence of a firm explanation of the father's behavior seems to have the power both to deconstruct a particular accusation against him, but also to undermine the very construction of an accusation against him altogether. This transformation seems further reflected in the therapist's subsequent question, which shifts the topic of the conversation from blameworthy to agreeable things that the father does (420–421), as well as in Nick's straightforward response to this shift of topic (422, 424–426).

In the extracts presented above, the therapist poses the “explanation” question once somebody is being accused. Without objecting to the accusation, as it is a question and not a statement, this question invites an answer, which presents a justification of the accused behavior and, thereby, seems to play an important role in the accusation's collapse. It can be noticed that, once a justification of the accused behavior is established, the causal structure of the reported events is automatically modified: the father's behavior is no longer a single action with certain consequences, but part of a longer causal sequence, itself a reaction to preceding facts.


  1. Top of page
  2. Abstract
  3. Method
  4. Analysis
  5. Discussion
  6. References

The analysis presented above investigates the function of two typical circular questions with respect to the discursive construction of a family's problems in the first session. It indicates that the questions of interest trigger respective shifts in the way the problems that bring the family to therapy are talked about. The “problem definition” question seems to set off a move from a norm of absence of difference to a norm of introduction of difference in family members' talk about their problems. In an analogous manner, the “explanation” question seems to set off the deconstruction of an accusation.

Such shifts can be seen as facilitating the introduction of a systemic epistemology in the discourse about the family's problems in relation both to the principle of the celebration of different viewpoints within a system and of the preference for circular causality as the organizing principle of patterned interaction between parts of a system (see Watzlawick et al., 1967). By contributing to the replacement of the norm of homogeneity with the norm of heterogeneity, the “problem definition” question seems to validate and legitimize the existence of multiple viewpoints linked to different positions within the family system. Respectively, by disrupting a blaming process, the “explanation” question appears to attack the linear perception of causality introduced by the arbitrary punctuation that an accusation of a family member's behavior entails. A number of recent, discourse-oriented studies have addressed the management of multivoicedness and blamings in the interaction between the therapist and the family (Kurri & Wahlstrom, 2005; Stancombe & White, 1997, 2005). The present article brings circular questions into this dialogue by drawing attention to the subtle ways in which they impinge on such issues in therapeutic talk.

The role of the studied questions in relation to the identified changes in problem-centered talk poses an interesting project to pursue research wise, given both the acknowledgment of the dynamic and flexible nature of talk in interaction, but also its structural properties: questions have the power to constrain the co-participants' actions by setting up the limits within which appropriate answers can be located (Hutchby & Wooffitt, 1998). Should we also acknowledge the asymmetrical character of psychotherapeutic interaction, with participants occupying discursive positions of differential power (see e.g., Guilfoyle, 2003), one can argue that the studied circular questions have the power to initiate and give direction to the identified discursive shifts to a great extent. On the other hand, the methodological approach adopted here emphasizes the co-constructive practices entailed in the production of meaning and their dialogical nature (Horton-Salway, 2001). Our choice to deploy it has hopefully highlighted that the aforementioned shifts constitute joint accomplishments of therapist's and family members' talk in interaction, resulting in a co-constructed discursive endeavor of creation and re-creation of meanings over the course of the therapeutic dialogue, entailing a “systemic quality,” gradually built up even in family members' emerging discourse. Thus, by approaching meaning as the “product” of a “systemic mind,” in tune with the premises underlying Post-Milan developments (e.g., Bertrando, 2007), this article does not point to what the therapist's deployment of the particular circular questions does to the family but to how these operate in the discursive event that the therapist and the family jointly generate. Moreover, it is not unproblematically assumed that the identified pattern is relevant for other instances of participants' lives or other therapeutic sessions.

In line with the theoretical and epistemological foundations of discourse analysis, we would like to acknowledge the relative, situated, and constitutive character of our discourse in this article. Like in the case of the data above, the presented analysis is itself a discursive, action-oriented construction tied to our “identity.” It reflects our interests and values, informed by our dual category membership: we are both systemic family therapists, trained in the Milan/Post-Milan approach and also have an affiliation with discourse analysis, with an ongoing interest and effort to contribute to their meeting (Tseliou & Eisler, 2007). The first category membership possibly accounts both for our preference for the particular “bottom-up type” of DA, whose affiliation with the pragmatic systemic theory of communication (Watzlawick et al., 1967) is perhaps hard to miss, and also for the particular reading of our data. On the other hand, following social constructionist premises, our claims, although situated in nature, could be seen as having a wider relevance: the identified features are not attached to particular individuals but to people's discursive practices and to the shared social knowledge and rhetorical competence all therapists, families, and analysts draw upon.

There is one final point we would like to make here. The type of DA deployed here is not equipped to address the level of interpretation addressing nonlinguistic features of communication, ordinarily deployed by a therapeutic team in action. Furthermore, it may come across as “handicapped” in analytically capturing fundamental aspects of human experience, like, for example, emotions, thus giving a cold, mechanical impression of denial or ignorance of those (Potter & Whetherell, 1987). This danger is all the more pertinent in cases like here, where deeply personal and emotionally laden instances of talk are being addressed. The discursive approach we adopt in this article neither denies nor disregards “psychological reality”; without making any ontological claims, it turns to discourse as the site where it is constructed and managed (Potter & Whetherell, 1987). Perhaps future developments will methodologically incorporate current theorizing in the field (Wetherell, 2012) or insightful theoretical suggestions (e.g., Baxter, 2004) pointing to alternative discursive theories, like, for example, the Bakhtinian dialogic theory (e.g., Bakhtin, 1981) and their methodological applications in current innovative attempts (e.g., Seikkula, Laitila & Rober, 2011). The limited available space here restrains us from further pursuing this discussion.

Our aim in this article has been dual. By introducing a discourse analytic perspective to the study of CQ, on the one hand, we aimed at providing a detailed examination of the role of circular questions in the discursive therapeutic space and, thus, at linking this long-established therapeutic tool to the concerns of postmodern systemic clinical practice. On the other hand, we aimed at providing an example of DA's potential in two ways. DA may constitute a means for the “scrutiny” of therapist's interventions in the context of therapeutic dialogue, thus offering valuable insights for a reflexive appraisal of the whole endeavor. Perhaps its deployment for the analysis of a therapy, while the latter is still in progress and with DA findings fed back to the therapist/therapeutic team, could better attend to such an aim. On another level, DA could add to the study of systemic family therapy “techniques,” “tools,” or even “dialogic principles,” like others have also claimed in relation to its potential for psychotherapy research (e.g., Avdi & Georgaca, 2007; Strong, Busch & Couture, 2008). Our suggestion is that such a venture, although still at its infancy (blinded Tseliou, in preparation), could, nevertheless, enrich their understanding by supplementing the theoretical arguments and the empirical observations that support them with a language-based consideration of the possibilities and the complexities of their use. The methodological challenge lies ahead.


  1. Top of page
  2. Abstract
  3. Method
  4. Analysis
  5. Discussion
  6. References
  • Anderson, C. M. (1986). The all-too-short trip from positive to negative connotation. Journal of Marital and Family Therapy, 12, 351354.
  • Anderson, H. (1997). Conversation, language and possibilities. New York, NY: Basic Books.
  • Anderson, H., & Goolishian, H. (1988). Human systems as linguistic systems: Preliminary and evolving ideas about the implications for clinical theory. Family Process, 27, 371393.
  • Atkinson, J. M., & Heritage, J. C. (Eds.) (1984). Structures of social action: Studies in conversation analysis. Cambridge: University Press.
  • Austin, J. L. (1962). How to do things with words. Oxford: Clarendon Press.
  • Avdi, E. (2005). Discursively negotiating a pathological identity in the clinical dialogue: Discourse analysis of a family therapy. Psychology and Psychotherapy, 78, 119.
  • Avdi, E., & Georgaca, E. (2007). Discourse analysis and psychotherapy: A critical review. European Journal of Psychotherapy and Counselling, 9(2), 157176.
  • Bakhtin, M. M. (1981). The dialogic imagination: Four essays (C. Emerson & M. Holquist, Trans.). Austin: University of Texas Press.
  • Bateson, G. (1979). Mind and nature: A necessary unity. Glascow: Fontana/Collins.
  • Baxter, L. A. (2004). A tale of two voices: Relational dialectics theory. Journal of Family Communication, 4(3-4), 181192.
  • Bertrando, P. (2007). The dialogical therapist. London: Karnac.
  • Boscolo, L., Cecchin, G., Hoffman, L., & Penn, P. (1987). Milan systemic family therapy: Conversations in theory and practice. New York, NY: Basic Books.
  • Brown, J. (1997). Circular questioning: An introductory guide. Australian and New Zealand Journal of Family Therapy, 18(2), 109114.
    Direct Link:
  • Burr, V. (1995). An introduction to social constructionism. London: Routledge.
  • Couture, S. J., & Strong, T. (2004). Turning differences into possibilities: Using discourse analysis to investigate change in therapy with adolescents and their families. Counselling and Psychotherapy Research, 4(1), 91101.
  • Couture, S. J., & Sutherland, O. (2006). Giving advice on advice-giving: A conversation analysis of Karl Tomm's practice. Journal of Marital and Family Therapy, 32(3), 329344.
  • De Haene, L. (2010). Beyond division: Convergences between postmodern qualitative research and family therapy. Journal of Marital and Family Therapy, 36(1), 112.
  • Dozier, R. M., Hicks, W. M., Cornille, T. A., & Peterson, G. W. (1998). The effect of Tomm's questioning styles on therapeutic alliance: An analog study. Family Process, 37, 189200.
  • Edwards, D. (1997). Discourse and cognition. London: Sage.
  • Edwards, D., & Potter, J. (1992). Discursive psychology. London: Sage.
  • Freuridas, C., Nelson, T., & Rosenthal, D. (1986). The evolution of circular questions: Training family therapists. Journal of Marital and Family Therapy, 23(1), 2740.
  • Garfinkel, H. (1967). Studies in ethnomethodology. Englewood Cliffs, NJ: Prentice Hall.
  • Georgaca, E., & Avdi, E. (2009). Evaluating the talking cure: The contribution of narrative, discourse and conversation analysis to psychotherapy assessment. Qualitative Research in Psychology, 6(3), 233247.
  • Gergen, K. (1999). An invitation to social construction. London: Sage.
  • Guilfoyle, M. (2003). Dialogue and power: A critical analysis of power in dialogical therapy. Family Process, 42, 331343.
  • Harvie, G., Strong, T., Taylor, L., Todd, N., & Young, F. (2008). Turning points in therapy: Discursive analyses and therapist comments. Journal of Systemic Therapies, 27(2), 3349.
  • Horton-Salway, M. (2001). The construction of M.E.: The discursive action model. In M. Wetherell, S. Taylor & S. J. Yates (Eds.), Discourse as data (pp. 147188). London: Sage in association with The Open University.
  • Hutchby, I., & Wooffitt, R. (1998). Conversation analysis: Principles, practices and applications. Oxford: Polity Press in association with Blackwell.
  • Kogan, S. (1998). The politics of making meaning: Discourse analysis of a ‘postmodern’ interview. Journal of Family Therapy, 20, 229251.
  • Kurri, K., & Wahlstrom, J. (2005). Placement of responsibility and moral reasoning in couple therapy. Journal of Family Therapy, 27, 352369.
  • Linares, J. L. (2001). Does history end with postmodernism? Toward an ultramodern family therapy Family Process, 40, 401412.
  • O'Reilly, M. (2007). Who's a naughty boy then? Accountability, family therapy, and the “naughty” child The Family Journal: Counseling and therapy for couples and families, 15(3), 234243.
  • Palazzoli, S. M., Boscolo, L., Cecchin, G., & Prata, G. (1980). Hypothesizing-circularity-neutrality: Three guidelines for the conductor of the session. Family Process, 19(1), 312.
  • Penn, P. (1982). Circular questioning. Family Process, 21(3), 267280.
  • Perosa, L. M., & Perosa, S. L. (2010). Assessing competences in couples and family therapy/counseling: A call to the profession. Journal of Marital and Family Therapy, 36(2), 126143.
  • Pote, H., Stratton, P., Cottrell, D., Shapiro, D., & Boston, P. (2003). Systemic family therapy can be manualized: Research process and findings. Journal of Family Therapy, 25, 236262.
  • Potter, J., & Wetherell, M. (1987). Discourse and social psychology: Beyond attitudes and behavior. London: Sage.
  • Ryan, D., & Carr, A. (2001). A study of the differential effects of Tomm's questioning styles on therapeutic alliance. Family Process, 40(1), 6777.
  • Scheel, M. J., & Conoley, C. W. (1998). Circular questioning and neutrality: An investigation of the process relationship. Contemporary Family Therapy, 20(2), 221235.
  • Seikkula, J., Laitila, A., & Rober, P. (2011). Making sense of multi-actor dialogues in family therapy and network meetings. Journal of Marital and Family Therapy, 38(4), 667687.
  • Seikkula, J., & Olson, M. (2003). The open dialogue approach to acute psychosis: Its poetics and micropolitics. Family Process, 42(3), 403418.
  • Shapiro, M. (1988). The politics of representation: Writing practices in biography, photography and policy analysis. Madison, WI: University of Wisconsin Press.
  • Silverman, D. (2006). Interpreting qualitative data: Methods for analysing talk, texts and interaction. London: Sage.
  • Stancombe, J., & White, S. (1997). Notes on the tenacity of therapeutic presuppositions in process research: Examining the artfulness of blaming in family therapy. Journal of Family Therapy, 19, 2141.
  • Stancombe, J., & White, S. (2005). Cause and responsibility: Towards an interactional understanding of blaming and ‘neutrality’ in family therapy. Journal of Family Therapy, 27, 330351.
  • Strong, T., Busch, R., & Couture, S. (2008). Conversational evidence in therapeutic dialogue. Journal of Marital and Family Therapy, 34(3), 388405.
  • Sutherland, O., & Couture, S. (2007). The discursive performance of alliance in family therapy: A conversation analytic perspective. The Australian and New Zealand Journal of Family Therapy, 28(4), 210217.
  • Sutherland, O., & Strong, T. (2011). Therapeutic collaboration: A conversation analysis of constructionist therapy. Journal of Family Therapy, 33, 256278.
  • Taylor, S. (2001). Locating and conducting discourse analytic research. In M. Wetherell, S. Taylor & S. J. Yates (Eds.), Discourse as data (pp. 548). London: Sage in association with The Open University.
  • Tomm, K. (1985). Circular interviewing: A multifaceted clinical tool. In D. Campbell & R. Draper (Eds.), Applications of systemic family therapy (pp. 3345). New York, NY: Grune and Stratton.
  • Tomm, K. (1987a). Interventive interviewing: Part I Strategizing as a fourth guideline for conducting therapeutic sessions. Family Process, 23, 313.
  • Tomm, K. (1987b). Interventive interviewing: Part II Reflexive questioning as a means to enable self-healing. Family Process, 26, 167183.
  • Tomm, K. (1988). Interventive interviewing: Part III Intending to ask lineal, circular, strategic or reflexive questions. Family Process, 27, 115.
  • Tseliou, E., & Eisler, I. (2007). ‘You’ and ‘I’, ‘Us’ and ‘Them’: a systemic-discursive approach to the study of ethnic stereotypes in the context of British-Greek heterosexual couple relationships. Family Process, 46, 515522.
  • Tseliou, E. (in preparation). A critical methodological review of Discourse and Conversation analytic studies of family therapy.
  • Watzlawick, P., Beavin-Bavelas, J., & Jackson, D. D. (1967). Pragmatics of human communication: A study of interactional patterns, pathologies and paradoxes. New York, NY: Norton.
  • Wetherell, M. (2001). Debates in discourse research. In M. Wetherell, S. Taylor & S. J. Yates (Eds.), Discourse theory and practice (pp. 380399). London: Sage in association with The Open University.
  • Wetherell, M. (2012). Affect and emotion: A new social science understanding. London: Sage.
  • White, M., & Epston, D. (1990). Narrative means to therapeutic ends. New York, NY: Norton.
  • Wittgenstein, L. (1953). Philosophical investigations. Oxford: Blackwell.
  • Wood, L. A., & Kroger, R. O. (2000). Doing discourse analysis: Methods for studying action in talk and text. Thousand Oaks, CA: Sage.