First published in New Therapist, South Africa, Issue 26, July/August 2003
An interview with Penny Tompkins and James Lawley
Interview by John Soderlund, editor New Therapist.
JS: You have noted the extent to which even Carl Rogers was in the habit of changing and redirecting the attention of clients away from the material they present. How much of the effectiveness of the clean language approach do you believe has to do with the fact that the therapist, arguably more so in clean language than in any other single modality - simply and unquestioningly accepts and reflects the client's metaphorical world?
PT & JL: Accepting a client's metaphorical world just as it is and reflecting that back is an important aspect of Clean Language -- but it's only the beginning of a three-part syntax. The second part invites the client to attend to an aspect of their perception (ie. it specifies a context previously identified by the client). The third part sends the client on a quest for self-knowledge. The facilitator needs to accept and reflect, and they need to invite the client to attend to patterns of thinking, feeling and behaving which are currently outside their awareness. In this way the client constructs a map, a model or a metaphor of how they function -- and, importantly, how they would like to function.
Over the last seven years we have trained therapists from a variety of schools and we consistently observe that without the discipline of Clean Language, therapists unwittingly insert their own metaphors and assumptions into the interaction -- and this inevitably influences the client's perceptions.
In Symbolic Modelling the challenge to the therapist is to facilitate the client without introducing any of the therapist's own metaphors, concepts or opinions. This is an almost impossible task, but you can get quite close using Clean Language.
JS: David Grove suggests that in using clean language, "the I-ness of the therapist should appear to cease to exist". This sentiment and the assumptions underlying symbolic modeling seem to de-emphasise the therapeutic relationship more than is the case in almost any other therapeutic approach. And yet, some 50 years of research into the effective ingredients of psychotherapy suggest that the quality of the therapeutic relationship is the single-most powerful predictor of outcome. How central do you believe the therapeutic relationship is in the effectiveness of clean language therapy and how is it nurtured and maintained while the "I-ness" of the therapist is so diminished?
PT & JL: First we'd like to explain the relationship of Clean Language (originated by David Grove) to Symbolic Modelling. Clean Language itself is not a therapy. Clean Language is the means, client-generated metaphor is the medium, and modelling is the methodology. Used together they comprise Symbolic Modelling. And by 'modelling' we mean identifying 'what is' and how 'what is' works -- without attempting to change it.
It is also worth noting that while clients most often express themselves metaphorically in words, they can also use nonverbal symbolic representations. They can gesture, enact, draw, sculpt or use sound. Some people have even used Lego!
As your question notes, the therapeutic relationship is important. But there is more than one kind of therapeutic relationship. I believe the research you mention relates to contexts where the client and therapist dialogue: they pass information between them, the therapist empathizes with the client, the client responds, the therapist comments on what the client says, and so on. In Symbolic Modelling a different kind of therapeutic relationship is established. It's what David Grove calls "a trialogue". The client's primary relationship is with their own mindbody perceptions and metaphorical constructs. The therapist gets a glimpse of the client's inner world as a by-product of that relationship. The therapist asks questions about that world while attempting to keep out of it -- it's complex enough without adding another variable!
And then a strange thing happens. The client relaxes into their own process. They don't need to explain to the therapist, they don't need to learn the therapist's language or thought process and they don't have to respond to the therapist's behaviour, because the entire focus of the interaction is their relationship with their perceptions. Even very experienced clients report again and again, that this is a very liberating experience. Scary sometimes, but freeing nonetheless.
JS: You note that clients take some time to begin engaging their metaphorical worlds. Do you think it in any way advisable to explain some of the assumptions behind symbolic modeling to those who continually meta-comment, analyse or focus almost exclusively on daily events? Or would you prefer that they simply switch over to a more symbolic perception at their own pace?
PT & JL: Cognitive scientists have demonstrated that everyone uses metaphor most of the time. However, the vast majority of metaphors are not recognised as such by either the speaker or the listener. Many everyday words are in fact metaphorical in that they describe one kind of thing in terms of another. So it is not a question of whether a client uses metaphor, but rather when they become aware of their use of metaphor, and how much the therapist explains that beforehand.
If a new client is curious about Symbolic Modelling we will describe our way of working to them. However, we attempt to explain as little as possible about our philosophy, because we do not want the client to expect their process of change to unfold in any particular way. And we certainly don't want to introduce any of our metaphors for change. We want them to find out how they and their Metaphor Landscape evolve through the process of actually changing.
You mentioned clients who 'continually meta-comment'. We regard all behaviour as an external manifestation of the way the client organizes their experience in the moment. Or, as Maturana and Varela put it: "an outside view of the dance of internal relations of the organism." If it seems that their meta-commenting is related to them achieving or not achieving their stated outcome for therapy then we will direct their attention to those meta-comments. Naturally we do that with a clean question.
JS: My own experience of using clean questions with clients is that it is a real challenge to make them sound un-mechanical and less formulaic than, in fairness, they are. Clients also seem to notice the patterns of questions as somewhat repetitive and formulaic. Do you believe this to be a problem and, if so, how would you recommend to practitioners that they change it and address the bewilderment of clients?
PT & JL: Congratulations on using Clean Language! Continue to practice, it's the only way to become fluent. Here are some tips to make your questions sound more natural. The first thing to remember is that these questions may sound very different to you than they do to the client. It is vital to experience being asked clean questions in order to recognize the difference. Second, the questions have to become automatic so that your attention is not on the question but on how you deliver it, where it goes within the client's perceptual space, and to what aspect of their Metaphor Landscape you are inviting them to attend. Third, the more curiosity in your voice tone, the more the client will attend to their quest for an answer, rather than to the question itself. Fourth, the more the client is situated in their Landscape and the more they embody their metaphors, the more they will pay attention to their Landscape and the less to you.
JS: Working in a clean language way is a radical shift for most therapists because of the concentration required to clean up their language. How willing do you find your trainees are to shift to a cleaner language in their therapy?
PT & JL: Completely willing -- especially once they have experienced the benefits of being asked Clean Language questions themselves. After people have had their personal constructs, metaphors and perceptual space accepted in this way they become sensitized to other people reframing or rewording their language.
Becoming fluent in Clean Language doesn't preclude you using more directly influential language. It simply extends your flexibility -- then you have greater choice.
JS: As with any new or emerging therapeutic modality, those
to whom it is introduced tend to ask about the evidence of its
effectiveness. Has clean language been subject to any such assessment
and, if so, what do the numbers say?
PT & JL: The
short answer is, to our knowledge Symbolic Modelling and Clean Language
have not been subject to any quantitative assessment.
language encourages clients to go into what you in places have referred
to as a trance-like state. Do you believe this makes it unsuitable or
strongly recommended for any particular client populations?
PT & JL: The
first part of your question relates to trance. Every time a therapist
asks a question, makes a statement or tells a story that requires the
client to 'go somewhere' to make sense of what has just been said, they
go into a trance. Trance phenomena are, to a greater or lesser degree,
part of all therapy. It's just that in some modalities, such as
hypnotherapy, both the client and the therapist are aware that trance
is part of the process. In Symbolic Modelling trance phenomena happen
without any suggestion or intention from the therapist. Nevertheless,
to answer some questions the client has to go into a trance-like state
to access the answer -- but they are always conscious of the question
and their answer.
The second part is related to the suitability
or otherwise of trance for certain client populations. As we've never
met a person who didn't go into trance at some time, even if only for
fleeting moments, it's hard to answer the question. Rather than the
modality, we think it is more a function of the therapist's experience.
For example, if you have had little experience with clients who have
difficulty grasping consensual reality, we wouldn't recommend using
Symbolic Modelling (or anything else for that matter) without an
experienced supervisor present. On the other hand, if this is a client
population with whom you are competent, then Symbolic Modelling could
be very useful in aiding the client to become aware of the consequences
of the current belief structure, and, if they want, in establishing an
alternative inner world that more supports them to get what they want.
regard to recommended client populations, Symbolic Modelling grew out
of David Grove's work with healing the effects of trauma. Symbolic
Modelling has also been successful in assisting clients to make changes
in a whole range of everyday issues (not just in psychotherapy, but
also in other contexts such as education, business and health).
However, Symbolic Modelling seems to be particularly suited to those
less definable, intractable, higher level issues -- ones involving
binding and double binding patterns, dissociation, conflicting
intentions, issues of identity, and matters of life purpose and
JS: Clean language seems to
draw strongly from the narrative and postmodern schools of therapy. It
shares with Michael White and David Epston a stress on the de-centered
role of the therapist. It shares with the late Harry Goolishian a
"not-knowing" approach to the client's material. It shares some of the
core solution focused tenets of Bill O'Hanlon, Steve de Shazer and
Insoo Kim Berg. And it appears to steer well clear of the grand therapy
models to which a range of postmodern practitioners have developed
something of an aversion. Despite this, however, the work of the clean
language therapist is governed by nine central questions, a rigid
syntactical approach to client-therapist dialogue and a strict
adherence to mimicking both the content and tone of the client's words.
For these reasons, it has been accused of itself being a grand
technical scheme itself, albeit one which derives from a more
enlightened philosophical foundation. What is your response to this?
PT & JL: Yes,
Symbolic Modelling follows in the tradition of those postmodern schools
of therapy you mention. David Grove certainly deserves to be recognized
along side such great therapists.
Before answering your question
we'd like to clarify a few points. The rigidity of the syntax you
mention is there partly to keep therapists from adding extraneous
words. Every extra word has to be processed by the client and somewhat
alters the meaning of the question. One of the features of Clean
Language is that its simplicity frees the client to be only as complex
as they need. And therapists are NOT required to mimic the client, only
to adjust their voice to more closely match the emphasis and changes
evident in the client's speech. Not only does this maintain rapport, it
also informs the therapist's model of the client's way of being.
don't know if Clean Language is a 'grand technical scheme'. We do know,
however, that for over 15 years David Grove conducted clinical research
into the set of questions that most engage the client with their
perceptions AND least contaminate those perceptions with therapist
words and inferences. We identified nine questions as 'basic' to Clean
Language because they can be used in most any circumstance. These nine
questions get used about 80% of the time. There are a further 20
'specialist' questions which are occasionally used in response to
particular kinds of client information. The particular circumstances of
a client's Metaphor Landscape may sometimes call for the invention of a
tailored question, but we've found that most other questions tend, in
varying degree, to be less than clean.
Interestingly this is not
necessarily the case outside psychotherapy. We have been involved in a
project run by Caitlin Walker to teach Clean Language to police
officers who interview vulnerable witnesses. We found that when asking
questions about an actual incident, rather than a Metaphor Landscape,
the number of questions that can be defined as 'clean' grows
There is nothing sacred about the nine basic or
the 20 specialist questions. They are simply what we currently observe
works best. We are open to extending or reducing the set if that proves
more effective, efficient and elegant.
argue in your book an individual's system works perfectly at doing what
it does. The whole of the clean language approach is also insistent
that the therapist should not contaminate the client's metaphorical
landscape with anything other than clean reflections of the client's
material. Do you believe that there are other methods in which clients
might be encouraged to engage in the reorganization process themselves
- without the assistance of the therapist - effectively offering them
assistance in accessing their Metaphoric Landscape alone?
PT & JL: We
know of people who use both Clean Language and metaphor as a means of
self-development. Some ask themselves clean questions and write or draw
the answers. Others use their symbols as part of their meditation
practice. We know of couples who use metaphor to resolve conflict.
Having said that, some people seem to need to hear the questions asked
for them to have any effect.
JS: The clean
language approach is compelling in its use of the client's metaphorical
world as the starting point for a reorganization of the rest of that
client's world. Some therapists intrigued by the approach may object to
using such a highly structured approach for an entire session with a
client. What do you believe is the advisability of using aspects of the
clean language approach, interwoven with other approaches?
PT & JL: All
clients will at times describe their experience in metaphor. They
especially make use of metaphor when the issue is complex or difficult
to talk about, or when they describe their internal experience. That's
what metaphors are for.
Therapists often say that just
becoming more aware of the client's (and their own) metaphorical
language influences the way they conduct therapy. At the very least,
client's metaphors offer the therapist a golden opportunity to gain
insight into the client's internal perceptual world. Most words
indicate what the client is thinking and feeling, metaphors tell you how they are thinking and feeling it.
it may be obvious that art, drama, sand tray and NLP therapists can
easily incorporate aspects of Symbolic Modelling, we have found that
therapists from other modalities can also adapt Clean Language,
metaphor or modelling into their existing way of working. They report
that just asking the client a few clean questions about their metaphors
proves useful. This includes body workers whose clients describe
symbolic images and sensations while being touched; health
professionals whose patients use metaphors to explain their physical
symptoms ("A stabbing pain"); and teachers whose students use metaphors
for their learning difficulties ("My mind has just gone blank").
we encourage integration with other approaches we also recommend
practitioners learn how to conduct complete sessions with Symbolic
Modelling. In this way they develop the experience and flexibility to
facilitate a client to be as metaphorical as they like while exploring
the logic and organization of their mind-body-spirit as deeply as they
© Penny Tompkins and James Lawley, and New Therapist