So, you think your counselling practices are collaborative?

Print Friendly, PDF & Email


Return to Journal Articles

Anne Hollingworth, PsyD, Clinical and Forensic Psychologist, Private Practice, New South Wales.



As noted by Andrew, Farhall, Ong & Waddell (2009) in recent times the delivery of mental health services in Australia features collaboration as a matter of both policy (e.g. National Mental Health Strategy, 2002) and action (e.g. Mental Health Council of Australia, 2000). Although the concept tends to lack a clear and consistent definition, Andrew et al. (2009) contend that collaboration consists of effective partnerships between patients, carers and mental health professionals. The concept can also be applied to how relationships are conducted across all sub-systems in a human service delivery system (i.e. professional–professional as well as professional–client). Henneman, Lee & Cohen (1995) conducted an analysis of the concept structure of collaboration in the literature and proposed the following essential, defining attributes without all of which collaboration cannot be said to be present: commitment to a shared venture; willing participation; team approach; shared planning/decision-making; shared contribution of expertise; a non-hierarchical relationship in which power is shared and based on knowledge rather than role or title [emphasis added]. This list of attributes implies successful collaboration is a result of reciprocity and mutuality between the interacting parties. However, it appears from the research of Andrew et al. (2009) into the collaboration between mental health professionals and family caregivers that each side attributed responsibility for collaboration to the other party. The professional’s collaborative practice (i.e. professional listening, professional confidence, sharing and supportive relationship) was the only dimension to contribute uniquely to the caregiver’s perception of collaboration.

Assuming that it would be fair to extend this concept analysis of collaboration by Henneman et al. (1995) to the professional-client relationships of clinicians who claim that collaboration is the cornerstone of their practice, I wish to focus on the last essential feature of collaboration (in italics above), with particular reference to the speaking practices of mental health professionals which enact (or do not enact) to share power and knowledge in non-hierarchical relationships. Therapists of a social constructionist bent (such as Narrative Therapists) contend that their speaking practices minimise the therapist’s ‘expert’ role and provide clients with the conversational opportunity to be ‘the expert’ on their own autobiographies. Assumptions that inform Narrative Therapy include that people have expertise on their own lives and that the therapist should ensure an atmosphere of curiosity, respect and transparency (cf. Morgan, 2000, for discussion). Following this line of reasoning, collaborative speaking practices should open up a conversational space in which the client becomes the final arbiter on the meaning of personally-lived experience.

Do mental health professionals (such as counsellors, psychotherapists, social workers, psychologists and psychiatrists) receive training in communication skills which will help them to develop collaborative relationships with their clients? In fact, do we know what sort of speaking practices would be effective in opening up the conversational space in which a non-hierarchical relationship in which power is shared and based on knowledge rather than role or title could thrive? Conversely, do we know what sort of speaking practices tend to shut down the conversational space in which collaborative professional–client working relationships might arise and flourish?

I want to begin by considering the speaking practices which are commonly taught as basic counselling skills. I am defining counselling as a speaking heuristic which is not aligned to any particular psychotherapeutic model although it may be common to many. Counsellors, psychotherapists and community workers from various helping professions typically receive training in basic interviewing and counselling skills (often referred to as micro-skills) prior to learning the specific intervention techniques of their discipline or treatment approach.

What Speaking Practices are Taught as Basic Counselling Skills?

The Skilled Helper (Egan, 1994) is a good example of a text which was widely used for many years to teach basic counselling skills. Egan offers a three-stage model of helping. The first stage deals with identifying problems and unused opportunities. Egan supplies some prototypical sentence patterns for eliciting, summarising and interpreting information. Speaking practices at later stages of the helping process are also described, such as suitable ways of phrasing such things as challenges (e.g. to distorted perceptions) and invitations (e.g. to review personal resources, set goals or to commit to an action plan). Developing accurate, basic empathy has long been stressed as the most fundamental skill in basic counselling texts (cf. Egan, 1990; Nelson-Jones, 1997)

Notions such as accurate empathy and distorted perceptions imply that speakers and hearers are jointly seeking to find and share ‘true’ meanings. In a successful collaborative search for meaning we might therefore expect a high degree of conversational cohesion. Various features of discourse help create cohesion, including conjunctions, such as ‘so’, ‘but’, ‘or’ which display the connection between the underlying propositional content of two clauses, that is, the clause in which the conjunction appears and a prior clause (Schiffrin 1987, p. 9). The link is established because the cohesive element (i.e. conjunction) in the later clause refers to and/or presupposes information from a prior clause. For example, ‘so’ typically introduces a clause which indicates result, as in: The job was boring so I left. During a conversational turn if one speaker completes the proposition implicit in the other speaker’s immediately prior utterance, and if that proposition is affirmed by the other speaker in the next turn, there will be conversational cohesion, and it will seem to the participants that a collaborative meaning-making venture has occurred (cf. Schiffrin, 1987).

If the proposition-completion made by the speaker is acceptable to the hearer, then the following turn should include some affirmation and a further elaboration (Yes and what’s more . . .). Whereas, if the proposition-completion is unacceptable the client’s next turn should include a rejection or revision, (No /Yes but . . .). On this basis, we could say that a proposition-completion accepted by a client reflects accurate empathy on the part of the therapist. (Note that no privileged truth status for any proposition is implied. The assumption is that all propositions are interpretations. If the client accepts the therapist’s interpretation as valid that usually counts as accurate empathy, in terms of the understandings provided by basic counselling texts). This process of conversational cooperation is illustrated in the following excerpt from a case example of a solution-focused approach for enhancing a person’s sense of competence:

C: In the past few weeks I’ve felt so terrible that I wished I was dead, but I could never do it. I think maybe one percent of me feels hopeful that life might get better in the end. It’s almost like a little voice inside me says ‘ It won’t always be like this’

T: So, there’s a part of you that feels hopeful that one day you’ll have the kind of life you’ll like

C: Yeah, but there’s a big part . . . sometimes that part’s bigger and I go to counseling. There was someone else . . . I ended going to a psychiatrist but I only went for about seven or eight visits. I explained to him all about my parents and everything and he ended up saying, ‘I don’t understand, there doesn’t seem to be enough data in what you’ve said to explain why you are the way you are’. So I just gave up

T: It’s interesting . . . when you describe the therapy you’ve had it sounds as like . . . the tone of what you’re saying is ‘I didn’t give it enough time, I should really hang in there’ and so on

C: Yeah

T: But at the same time, it sounds like you’ve made some valid decisions about what’s felt okay to you, what’s been helpful and what hasn’t you’ve made some decisions along the way that either certain people or methods were not that helpful to you

C: So I had a valid reason to stop?

T: Yeah  (Durrant & Kowalski, 1993, p. 118)

Note that in the excerpt above although the counsellor’s responses were grammatically formed as statements it would be fair to assume that these ‘statements’ were delivered with a rising intonation which allowed them to be heard as questions. The client certainly responded each time as though answering a question. In linguistic terminology, statements delivered with a rising intonation are prosodic questions.

Discourse analysis of some counselling texts indicates that prefacing a response with ‘So’ is almost exclusively the prerogative of the conversational partner trained in accurate empathy, i.e. the therapist of whatever ilk. A few examples are given below. (Note all propose themselves as teaching texts, except for Epston’s transcripts):


Source. Approach. Avowed purpose.

Speaker turns per excerpt

% of therapist’s speaking turns introduced by ‘so’

% of client’s speaking turns introduced by ‘so’

Durrant & Kowalski (1993, pp. 114–123). Competency-based approach. Enhancing a client’s view of self as competent.


23% (11/48)

2% (1/49)

Beck (1995, pp. 86–7). CBT. Identifying the problematic situation.




Egan (1975/82/86/90). Problem Management approach. Empathy and probing, pp. 133–4; developing an external perspective, p. 193, three short excerpts.

 3-4 av.

50% (4/8 av.)


Book (1998, p. 114). Brief psychodynamic therapy. How to clarify behaviour patterns.


16% (1/6)


Morgan (2000, p. 81). Narrative therapy. Example of a re-membering conversation.


40% (2/5)


Epston, D (nd), Narrative therapy. Three sessions of therapy (undated, unpublished, deidentified transcripts) 8020 words.

 33% of airtime

< 0.033%

(1 turn /3 sessions)


Since accurate empathy is held to be fundamental to counselling skills, and since in basic teaching texts conversational cohesion indexes empathy, and since So is a feature associated with conversational cohesion, we might expect So to be a common opening gambit in therapists’ utterances regardless of the ‘brand’ of therapy they espouse. Therapists of different brands would use So to indicate logical relationships between their responses and client’s prior utterances, although the type of therapist response would vary with the treatment modality. For example, we might predict that proposition-completion by narrative therapists would reflect the therapist’s curiosity about information only the client has the knowledge/expertise to provide, in which case So would introduce a question (congruently grammatically-coded as a ‘Wh- question’, the question form used to ask for specific detail). In contrast, therapists of orientations which stress the importance of origins of problems (e.g. psychodynamic approaches) or the cognitive distortions which maintain a problem (e.g. CBT) might be expected to complete the same client proposition by a reading which presupposes and foregrounds the client’s negative affect, cognition or self-defeating behaviour, most typically in the grammatical form of a statement (= a prosodic question). Therefore, we would expect So to be clocked at similar rates across all types of counselling discourse.

However, in a cursory survey of the published therapeutic discourse of narrative therapists Michael White and David Epston (cf. Epston & White, 1992; White & Epston, 1990; White 1988/1989, 1995) therapist utterances beginning with So were not prominent, although they did occur. Likewise, in the transcripts of three therapy sessions (anonymous, unpublished, undated) supplied by Epston—which I analysed for other research purposes—I found only one therapist utterance beginning with So. Furthermore, both White and Epston informed me (personal communications, D. Epston 1999 and 2002; personal communication, M. White, 2001) that they were attempting to minimise starting a conversational turn in therapy with So and advised their trainees to do the same. Why might the founders of Narrative Therapy feel uncomfortable about a speaking practice which is ubiquitous in counselling discourse as well as in teaching texts?

In a nutshell, my answer is that ‘So + paraphrase’ can be used as a coercive discourse strategy whereas Narrative therapists pride themselves on working collaboratively with clients. Therapy is an inherently unequal encounter. If, at the outset, argumentative resources were not asymmetrically distributed between therapist and client the therapist would not be in a position to help open up new ways of talking about problems. Some ‘talking cures’ use discourse strategies which ensure the therapist keeps firm control of the unfolding problems-talk, and, as a consequence, the direction of the client’s quest for meaning in lived experience. Narrative practice aims to re-distribute the argumentative resources so the therapist becomes the secondary author in this re-storying process (cf. White, 1995; Morgan, 2000). Famously, Narrative therapists use questions to ‘externalise’ a problem which has usually been unhelpfully identified previously as an inalienable feature of the client’s identity (White 1988/1989). In contrast to a chain of Narrative questions, a chain of So + paraphrase utterances is the conversational equivalent of the ‘iron hand in the velvet glove’—enabling the therapist to tightly constrain the discourse without seeming to.

In order to demonstrate how I reached this conclusion by applying insights from theoretical linguistics to the analysis of therapeutic discourse, I must first explain how using prosodic questions to paraphrase what other people say constrains their options for responding. Then it will be necessary to explain how So can be used as a linguistic resource for distributing meaning-making rights asymmetrically in a conversation.

Prosodic Questions Severely Constrain the Other Person’s Discoursal Options

In counselling textbooks, the speaking practice of paraphrasing what clients say as a statement is part of reflective listening. Paraphrasing is widely recommended in the counselling literature as a means of enacting empathy; however, it also operates to distribute speaking rights. According to systemic functional linguistics Gerot and Wignell (1994), “a speaker in uttering selects a speech role for himself and simultaneously and thereby allocates a speech role to the addressee” (p. 22). For example, if a speaker provides you with information s/he is implicitly inviting you to consider it, if s/he asks you a question s/he is implicitly demanding that you be given that information. In communication, the participants adopt a speech role (giving or demanding) in order to exchange commodities (information or goods and services). Here are the combinations of options with the language forms which typically realise them:



Good & services






Goods & services






The listener has the following response options:

EITHER                                     OR













(Adapted from Gerot & Wignell, 1994, p. 23)

However, when for reasons of politeness, tact or manipulation a speaker wants to downplay the fact that she or he is making pronouncements or giving orders, grammatical metaphor can be used to mask the communicative intent. Metaphor is a literary device for talking about something as if it is something else e.g. You are my sunshine, and by extension acquires the special features and powers of that something else i.e., You light up my world. Similarly, grammatical metaphor involves one grammatical form standing in for another, to realise a different speech role than the speech role(s) prototypically associated with that grammatical form. A question (which typically realises a demand for information) may be presented as if it were a statement (which typically realises an offer of information).

For example, You feel uncomfortable with the idea of discussing our common interests before we discuss possible solutions, is a question masquerading as a statement, and often spoken with a rising intonation. In linguistic terms, such an utterance is called a prosodic question. The grammatical form of the sentence makes it a statement, but intonation (i.e. its prosody) makes it heard as a question. The privilege of occurrence of such a counsellor utterance also prompts the listener to hear it as a question, regardless of prosody, because in counselling discourse an utterance like, You feel uncomfortable because . . . , occurs at speaking turns where a question would be a more likely response in everyday conversation.

Questions are important resources for achieving politeness via conversational implicature. Because questions intrinsically provide the response option of refusal, in choosing to code a proposition or proposal as a question the speaker implies his or her inability or unwillingness to compel the other’s compliance. A speaker can signal more or less respect for the listener’s personal sovereignty through the choice of question–form used to realise a demand for information, goods or services. The form of a question constrains the response of the listener in so far as it implicates certain possibilities for responses and not others.

For example, Wh- questions limit the responses to general categories within which the respondent may make a personal selection e.g. Where were you yesterday? (locations); Why did you do that? (reason); How did you do that? (manner); What was happening at that time? (circumstance). Whereas ‘Yes–No’ questions reduce the choice of responses to two. In the implicational hierarchy of questions (cf. Woodbury, 1984) prosodic questions are among the most coercive and can severely constrain the emergence of new information in discourse. A prosodic question is a Yes–No question in disguise. In addition, i) the statement form implies the speaker believes the truth of the proposition ostensibly offered for negotiation; ii) the statement form focuses the listener’s attention on given information, implying new information is unwanted; iii) the question intonation implies the speaker is respecting the listener’s right or refusal, thereby disarming resistance to the proposition or proposal. It requires high-level communicative competence to identify and explicitly challenge conversational implicature. Prosodic questions are hard to dispute.

Reflective listening is not the default setting for English conversation (if it were there would be no need for training courses in counselling) and therefore may have novelty value for encouraging self-disclosure and exploration. Verbatim paraphrases are useful for checking a message was accurately received. But, aside from ice-breaking and confirming accurate reception of the other’s message, is ‘So + prosodic question’ a speaking practice compatible with collaborative practice—particularly the value of respect for persons and the notion of power-sharing? What attitude should practitioners who claim to work collaboratively with their clients strike to speaking practices which are not what they seem to be: questions which pretend to be statements, and ‘demands’ which purport to be ‘offers’? Does one speaker have the right to consistently name another’s experience while pretending to be merely reflecting the other’s words?

‘So’ is a Linguistic Resource for Staging a Text AND for Distributing Speaking Rights

As discussed previously, the discourse marker So indicates cause-effect relationships, hence a conversational response introduced by So indicates that the proposition to follow has been warranted by what preceded it (Schiffrin, 1987). If after listening to a client’s outpouring a counsellor says, So you feel unhappy with the way power is shared in the marriage?, So makes the interpretation a warrantable inference from what the other person has just said, while at the same time, as explained in the previous section, the rising intonation of the prosodic question allows the speaker to signal a cooperative stance to the listener (cf. Schiffrin, 1990). Clearly, it would be uncooperative for a speaker to make unwarranted inferences about cause-effect in others’ behaviour. Consider the implications about how speaking rights and argumentative resources are to be distributed henceforth, should the client accept the counsellor’s interpretation about power-sharing in the marriage as a warrantable inference. A speaker allowed to repeatedly make inferences about the meaning of the other speaker’s prior narrative becomes the arbiter of meaning—and also to some extent the arbiter of when a search for meaning has concluded. Furthermore, that person acquires this speaking right by stealth, not as a result of any explicit, consensual process.

To make this point clearer, we should consider at what stage in a story about problematic life events inferences are warrantable. Storytelling can be divided into a range of narrative types according to how they are structured, according to Plum (1988). The stories people tell about problematised events and their concerns are most often structured as the genre Plum (1993) calls ‘Narrative’, which constructs lived experience as a quest. Eggins and Slade (1995) say that “successful Narratives have a telos. They give listeners a sense that they are moving towards an end, towards a resolution of some conflict” (p. 236). According to the pioneering work of Labov & Waletsky (1967, pp. 37–39) a canonical narrative of personal experience revolves around a perturbation to the normal flow of events which must be resolved and is typically staged as follows:

Stage                          Purpose


Summarise the story to encapsulate its point


Orient the listener in respect to place, time, situation


Describe sequences of events rising to a climax—the main part of a narrative


Reveal the attitude of the narrator towards the narrative


Show how the protagonist’s actions resolve the crisis


To make some point about the text (e.g. the moral of the story) and also to return the perspectival stance to the time of speaking.

The discourse marker So would typically arise at the Evaluation stage of a narrative, and/or at the Coda. A counsellor who makes a statement introduced by So at the conclusion of a chapter in a client’s story about life events may think s/he is merely summarising, but this is a delusional belief. Not only is the therapist imposing an interpretation on events, s/he is asserting the right to make interpretations by the very act of executing this stage of the narrative. The counsellor’s statement introduced with So which sums up the (presumed) attitude of the narrator towards his/her narrative implicitly declares that the story has reached the Evaluation stage. At the Evaluation stage of a narrative the search for meaning is over. Unless the narrator (= client) wishes to explicitly reject the counsellor’s interpretation, no further exploration will be made of alternative interpretations, as the conversational space for the topic will have been closed down.

If ‘So . . .’ is Not a Collaborative Speaking Practice Why Stick to it?

I hope by this point to have demonstrated to the reader three main ideas, namely:

• So is a ubiquitous feature of counsellors’ discourse. Counsellors start many of their

conversational conversations with So, whereas clients rarely do.

• So + prosodic question is a conversational gambit which constrains the discoursal options of the

other (in this case, the clients of counsellors)

• So + prosodic question is a speaking practice which, especially if used to excess, enables a

speaker to usurp the authorship of another’s story.

However, three other ideas are needed to counter-balance overgeneralisation, namely:

1. Not all therapist utterances consisting of So + prosodic question constitute bids to be recognised as the rightful arbiter of meaning. On the contrary, genuine attempts to clarify meaning could be realised in this form.

2. Not all therapist utterances beginning with So are equally coercive.

For example, there is a significant difference between So + prosodic question and So + Wh- question. As previously discussed, according to the implicational hierarchy of questions in English, the prosodic question is a form which implies that the speaker (e.g. the therapist) believes the propositional content of their utterance. Furthermore, since So is a discourse marker used to introduce warrantable inferences, the speaker’s position as an oracle is further strengthened by this grammatical form (e.g. “So, there’s a part of you that feels hopeful that one day you’ll have the kind of life you’ll like”. Durrant & Kowalski, 1993, p. 83). In this case the therapist is the one entitled to discover and announce the meaning of the client’s lived experience. (N.B. I am not arguing that insights offered to the client by a therapist speaking in this way are never useful. This is discussed further below).

In contrast, So + wh- questions constrain the client’s options in terms oftopic’, but response options are less constrained (i.e. not simply Yes or No). Consider a So + Wh- question such as “So what would Baby Born call that quality you have in not giving up when things get hard?” (from Morgan, 2000, p. 81). The propositional content of the question (i.e. that the client has a quality of not giving up when things get hard) is smuggled in as a pre-supposition and thus not offered for negotiation. It is Baby Born’s view about the proposition rather than the truth of the proposition that is being queried. In addition, the discourse marker So asserts that the proposition is a warrantable conclusion given the client’s preceding utterances. The therapist’s question steers the client towards exploring an emerging theme the therapist thinks could be useful. Wh- questions ask for further specification on a topic. Such specification (incrementally) provides the supporting detail and argumentation which will enable the client to draw his/her own conclusion in response to this line of ‘Socratic’ questioning. The therapist chooses the topic to investigate but the client retains the right to discover and proclaim meanings.

3. Not all therapist interpretations offered as So + prosodic question will destroy the therapeutic alliance, because clients perceive the therapeutic relationship variously. Unequal encounters are not inherently bad. Human benefit can result from the authoritative interventions of benevolent ‘experts’. In a study examining client perceptions Bachelor (1995) found that facilitative relationships are not invariable: some clients value a friendly relationship (respect, empathy, attentive listening) above all; others want expert input (improved self-understanding gained through therapist clarification of the client’s material); and others want a collaboration in which both parties contribute ideas and take responsibility for outcomes. According to the work of Duncan et al. (2004) the therapist with a flexible relational stance who will adapt his/her theory of change to the client’s theory of change is likely to be the most efficacious change-agent.

From this point of view, for the client who values (or at least expects) a therapist to offer insights, good results might be achieved with a timely offering of the therapist’s conclusions in the So + prosodic question format. However, if offering interpretations in this way is the counsellor’s default setting, then the hierarchy will be constantly invoked. The speaking practice may or may not be therapeutically effective, but it cannot fairly claim to be collaborative.

The role of a psychotherapist who aims to be collaborative is not to determine and proclaim the meaning of a person’s lived experience, but to foster that person’s sense of agency in making sense of his/her own life. Speaking practices which help people reflect upon and re-author their problematical life narratives will promote professional-client collaboration in counselling and psychotherapy. Arguably, So + prosodic question is a speaking practice which can (especially if used to excess) assert status differences, reinforce the power hierarchy and thus throw up barriers to client empowerment. If everyone is so keen to be collaborative these days, why are counsellors still talking to their clients like this?

One powerful reason may be because this way of speaking permits therapists to maintain dominance in the conversation without sacrificing affiliation in the relationship. Many people who seek therapy are not confident about their role-rights in a conversation—especially those who have been maltreated in childhood. When, from a young age, a person has been subject to the arbitrary authority of people who wield socially-legitimised power (e.g. parents, teachers, foster-carers) that person may come to the (unspoken) conclusion that reality is what the other people say it is. The constant imposition of interpretations by the therapist will fulfill such a person’s belief in being unfit to judge the meaning of their own lived experience, as well as the tacit assumption of being under-entitled relative to others in relationships (conversations).

Righteous indignation will not necessarily be the response to the therapist’s ‘So + prosodic question’ utterances even in people who do not feel under-entitled in relationships, because grammatical metaphor and conversational implicature operate to render invisible the inherent discoursal constraints of the sentence format. Hence the therapist as much as the client may be blind to the unintended consequences of the speaking practice—which are that dominant narratives have more opportunity to prevail, and neglected narrative themes have less opportunity to enter the conversation. Therapist and client may be equally docile in their oppression by received ideas.

Summary and Conclusions

I have argued that by repeatedly prefacing responses to a client’s conversational contributions with So + prosodic questions that counsellors and other mental health clinicians risk usurping the client’s right to be the arbiter of personal meaning. Because prosodic questions are hard to dispute they are coercive. Because the discourse function of So is to introduce a warrantable inference, it may be interpreted as ushering in the Evaluation stage of a story. At the Evaluation or Coda stage a Narrative is complete, whereas unstoried elements in clients’ self-reports might lead to a fruitful Resolution if discovered and explored in an on-going conversation. By and large, helping professionals claim to value respectful ways of speaking and client self-determination. In order to be consistent with these values and goals interpretations offered in the form of a So + prosodic question should form a minor part of the speaking practices of mental health practitioners who sincerely wish to work collaboratively with their clients in counselling and psychotherapy (or in community development/advocacy). However, starting every conversational turn with So may be a speech habit which helping professionals uncritically acquire during their basic training then consolidate by years of practice.

If theory is when you have ideas and ideology is when ideas have you, what ideology do So + prosodic questions serve? Is it possible that through the unreflective use of So + prosodic questions counsellors, psychotherapists and community workers unwittingly help to keep alive outmoded, received ideas about helping relationships which rely on maximising the authority and expert status of the mental health professional? Like it or not, many of us acquired the ‘so’ habit early in our careers. Old habits die hard. The question I am still left pondering is whether this means engaging in collaborative, respectful speaking practices is more a matter of skill or more a matter of virtue?

Limitations and Further Directions

The analysis of one specific, speaking practice (no matter how widespread that practice may be among practitioners) cannot address all aspects of therapist behaviour which may, or may not, enact a respectful, collaborative stance. My analysis of micro-aspects of therapeutic discourse was narrowly-focused for a specific purpose. Within the area of sociolinguistics called Conversation Analysis there is a large literature which addresses helping conversations more broadly. Some scholarly works, such as the landmark study of therapeutic conversations by Labov and Fanshel (1977), are certainly not light reading. For readers who would like to pursue further insights into how Conversation Analysis can illuminate therapeutic discourse, authors who are more accessible include Heritage, Robinson, Elliott, Beckett and Wilkes (2007), Jager (2016), Lock and Strong (2012).


Andrew, M., Farhall, J., Ong, B., & Waddell, F. (2009). Perceptions of mental health professionals and family caregivers about their collaborative relationships: a factor analytic study. Australian Psychologist, 44(2), 94–104.

Bachelor, A. (1995). Clients’ perception of the therapeutic alliance: a qualitative analysis. Journal of Counseling Psychology, 42(3), 323–337. http://dx.doi.org/10.1037/0022-0167.42.3.323

Beck, J. (1995). Cognitive therapy: basics and beyond. London: Guildford.

Book, H. (1998). How to practice brief psychodynamic psychotherapy. Washington: APA.

Duncan B.J., Miller S.D. & Sparks J.A. (2004). The heroic client: a revolutionary way to improve effectiveness through client-directed outcome-informed therapy. San Francisco, CA: Jossey-Bass.

Durrant, M., and Kowalski, K. (1993). Enhancing views of competence. In S. Friedman, (Ed.), The new language of change. New York, NY: Guildford Press.

Egan, G. (1990). The skilled helper: a systemic approach to effective helping (4th ed.). Pacific Grove, CA: Brooks Cole.

Eggins, S., & Slade, D. (1997). Analysing casual conversation. London: Cassell.

Epston, D., & White, M. (1992). Experience, contradiction, narrative and imagination: Selected papers of David Epston and Michael White, 1989–1991. Adelaide, SA: Dulwich Centre Publications.

Gerot, L. & Wignell, P. (1994). Making sense of functional grammar. NSW: Antipodean Enterprises.

Henneman, E.A., Lee, J.L., & Cohen, J.I. (1995). Collaboration: a conceptual analysis. Journal of Advanced Nursing, 21,103–109.

Heritage, J., Robinson, J. D., Elliott, M. N., Beckett, M., & Wilkes, M. (2007). Reducing patients’ unmet concerns in primary care: the difference one word can make. Journal of General Internal Medicine, 22, 1429–1433. DOI 10.1007/s11606-007-0279-0

Jager, M. (2016). Therapists’ continuations following I don’t know—responses of adolescents in psychotherapy. Patient Education and Counseling, 99, 1778–1784.

Labov, W., & Fanshel, D. (1977). Therapeutic discourse: psychotherapy as discourse. New York: Academic Press.

Labov, W., & Waletsky, J. (1967). Narrative analysis: oral versions of person experience in J. Helm (Ed.) Essays on the verbal and visual arts (pp. 12–44) Seattle: University of Washington Press.

Lock, A., & Strong, T., (Eds). (2012). Discursive perspectives on therapeutic practice. New York: Oxford University Press.

Mental Health Council of Australia (2000). Carers of people with mental illness project: final report. Deakin, ACT: Mental Health Council of Australia.

Morgan, A., (2000). What is narrative therapy: an easy to read introduction. Adelaide, SA: Dulwich Centre Publications.

National Mental Health Strategy (2002). National practice standards for the mental health workforce. Canberra: Commonwealth Department of Health and Ageing.

Nelson-Jones, R. (1997). Practical counselling and helping skills (4th ed.). London: Cassell.

Plum, G. (1993). Text and contextual conditioning in spoken English: a genre-based approach (Unpublished PhD thesis). University of Sydney, Sydney, Australia.

Schiffrin, D. (1987). Discourse markers. Cambridge: Cambridge University Press.

White, M. (1988/1989). The externalizing of the problem and the re-authoring of lives and relationships. Selected papers. Adelaide, SA: Dulwich Centre Publications.

White, M. (1995) Re-authoring lives: interviews and essays. Adelaide, SA: Dulwich Centre Publications.

White, M. (1997). Narratives of therapists’ lives. Adelaide, SA: Dulwich Centre Publications.

White, M., & Epston, D. (1990). Narrative means to therapeutic ends. New York: Norton.

Woodbury, H. (1984). The strategic use of questions in court. Semiotica, 48(3/4),197–228. https://doi.org/10.1515/semi.1984.48.3-4.197



Return to Journal Articles