Modelling in the symbolic domain

The art of Clean Language
Print Friendly, PDF & Email

Published in Rapport 80, pp. 30-31, Autumn 2023

“And what would you like to have happen?” is the Clean Language question with which we usually start coaching or therapy. “I would like the inflammation of my gums to stop” was the translated reply. The answer came from a member of a group of Ukrainian psychologists we have been supporting since the Russian invasion in February 2022.

Since we are not doctors nor dentists we cannot provide medical advice, but we could work with the client’s relationship with her symptoms. It turned out that she considered the inflammation a reaction to her situation. 

Our client had been evacuated to live in the west of Ukraine with the mother and son of a soldier on the front line. Two days before our session, the mother was notified that her husband had been killed in the war and she asked for help telling her son.

Our client wanted “to tear apart with my teeth all these abusers who kill our people”. Her immediate problem was that she did not know how to find the words to help this young mother tell the child his father had died.

Of course we could have suggested any number of supportive words, but that’s not how we work. Our process is strongly predicated on the presupposition that “people already have all of the resources they need to act effectively”.[1]

David Grove

It’s 40 years since David Grove first came up with the idea of Clean Language. He applied it as a therapeutic method for helping people suffering from traumatic childhood and adult memories, and with military veterans returning from war with PTSD.[2]

He noticed that when his clients described their experience, they would often stop just before the worst moments and say “It was like …” and give a metaphor. 

Instead of asking them to recount the details of the trauma – which was a common method at the time – he got curious and began asking questions about the metaphor they were using, rather than the traumatic event itself.

His genius was to realise that metaphors did not like being asked ordinary questions. He spent years finding the few questions that would give life to people’s personal metaphors. Sometimes he even referred to himself as a midwife!  And that’s how Clean Language was born.

David discovered that by staying clean and not ‘contaminating’ his client’s experience with his own interpretations and suggestions the metaphor changed organically, healing the suffering in a way unique to each and every person. Sadly David died unexpectedly in 2008.

The field of Clean

Today, a whole field under the umbrella of ‘Clean’ exists. Clean approaches are used in coaching, education, organisational change and group facilitation, to name but a few. Perhaps the fastest growing application is as a research methodology. There have already been several articles published in high-quality academic journals as well as a groundbreaking book, Clean Language Interviewing: Principles and Applications for Researchers and Practitioners.[3] Almost every week we receive notifications of another academic or other kind of project which has made use of Clean Language.

The proliferation of Clean Language into more and more applications can be traced to our modelling of David Grove in the 1990s. 

Modelling excellence

In the 1970s Richard Bandler, John Grinder and others modelled exceptional therapists giving rise to the field called Neuro-Linguistic Programming (NLP). We adopted a similar approach to study and codify Grove’s innovations. While we thought it would take a year to model David, it took us 4 years! In part because we only had access to him every six months and because in each intervening period his method had evolved. Eventually we realised we needed a model of the essence of what he was doing that included his past innovations and could encompass likely future changes to his way of working. In 2000 we published the results of our modelling in Metaphors in Mind: Transformation through Symbolic Modelling.

Symbolic Modelling

Symbolic Modelling didn’t simply recapitulate what Grove was doing. By including the latest developments in Cognitive Linguistics, self-organising systems theory and evolutionary dynamics we gave the approach a solid theoretical underpinning and a wide remit.

On any introduction to Clean Language you will be given 8-12 of the ‘basic’ Clean Language questions. After thirty years these exact same questions still form the core of all we do. And yet our level of mastery has multiplied many times. How are we and others able to facilitate clients to experience such transformative changes with so few questions? How can something so apparently restrictive facilitate clients to be so creative? How does less become more?

Working cleanly involves more than just asking Clean Language questions. For a start we called our approach Symbolic Modelling for a reason – because first and foremost it is a modelling methodology. In our approach, ‘change’ is not a step in a technique, it is not even the aim of the process. In Symbolic Modelling change happens spontaneously as an organic by-product of facilitating the client to self-model the structure and organisation of their own unique inner world.[4]


David Grove discovered that people’s inner worlds, their mind-body maps, are fundamentally metaphoric in nature and have their own coherent logic. This was a revolutionary idea in the early 1980’s – and still is today! It is rarely appreciated that people commonly use six metaphors a minute, even in everyday conversations.[5] Mostly these metaphors are not recognised as such, and those that are easily recognisable are often highly idiosyncratic. Here are few from sessions with the Ukrainian group:

Initial description


Feeling the movement of life

The ignition key in a car

The impulse before searching

Sparks of Bengali fires

When my body functions

Sweet honey

A feeling of independence

Dark red, like raspberry

My attention 

A small white rabbit sleeping in a hole in the forest 

Perhaps because of his Maori heritage, David Grove was highly attuned to his client’s metaphors and he took them as a perfect description of their inner world. He developed Clean Language as a way to facilitate the client to engage directly with their symbolic domain – a process David called “a trialogue”.

Metaphors are vital to “The study of the structure of subjective experience”; one of the earliest definitions of NLP. Grove discovered that personal metaphors do not respond well to leading questions – those that attempt to reframe, change the client’s experience or otherwise suggest an answer. Whether intentionally or not,  these questions introduce the coach or therapist’s beliefs, values and presuppositions into the client’s inner world.

Approaches based on leading questions and suggestions can be very effective. However, something special happens if, as a therapist or coach, you don’t use them for an entire session. In other words you only use Clean Language.

Another component of Symbolic Modelling is a particular kind of outcome orientation that we brought to the field. David Grove was at heart a problem solver, a healer, and some would say a shaman. Until we arrived on the scene little attention had been paid to using Clean Language to develop embodied desired outcomes. Outcome orientation involves more than SMART goals or NLP’s well-formed outcome conditions. We see it as a process that continually orientates the session to the client’s desired outcome. It is a way of using the client’s desires, wants, would like’s and needs as dynamic reference points which act as beacons or signposts for a process that doesn’t know where it is going to end up.

We arrived at this approach in two ways. We modelled expert therapists who worked ‘bottom up’ – David Grove, Steve De Shazer, Robert Dilts, Steve Andreas (and ourselves) – and we borrowed from the work of Robert Fritz. This led to our Problem-Remedy-Outcome model and ‘vectoring’. A vector is a cluster of questions which together head in a process direction determined by the logic of the client’s information and in particular their desired outcomes.

Together, Clean Language, metaphor, modelling and outcome orientation are four components which enable a facilitator to model symbolically and facilitate clients cleanly. However, there is plenty going on in the background that may not be obvious to an untrained observer which we’ll cover in future articles. 

Finding words in a spiral galaxy

Our Ukrainian client with the inflamed gums was now deeply emotional. She said “I don’t know how to find the words” to help the mother tell her son his father had died. This statement implies that the words are not the problem, the problem is the client does not know how to find them. It might seem paradoxical, but we take clients’ metaphors (like this one) literally.

As the session progressed the client accessed the state of “a warm and trembling love that wants to support” the bereaved mother and child. Through a series of Clean Language questions, this resource developed into a metaphor of “two oval spots” located in her solar plexus. Spontaneously the spots morphed into two connected galaxies. We asked:

So when it’s like those two galaxies are connected, there in your solar plexus, and that’s a warmth and a trembling of a love that wants to support there, what happens to words in a memory?

After a pause she answered that the words she needed to find “can be put together” like “planets in a galaxy that go one-by-one … in a spiral [gestures depict the metaphor].”

And so when those words are put together like planets in a galaxy, then what happens?

“The words help the child to experience … a kind of connection” to his father and “to have an opportunity to experience joy and happiness even though there is also grief.” 

The client didn’t need to find the words there and then. Instead she had access to a resource metaphor which meant the appropriate words would come when she needed them.

At the end of the session we enquired, “And what happens in your gums?” 

After awhile the client reported, “Right now I don’t feel any negative feelings or sensations.”

And finally …

Clean Language questions allow clients’ untarnished experience to be revealed – first to them, and then to you. The art is in recognising that a metaphor holds a corresponding organisation and structure to what it represents. As Clean Language questions are asked and a metaphor landscape is developed, psychoactivity occurs – the metaphor spontaneously changes – and these changes will be reflected, isomorphically, in the client’s experience and behaviour.


1. Dilts, R. (1998). Presuppositions.

2. Grove, D. J., & Panzer, B. I. (1989). Resolving Traumatic Memories: Metaphors and symbols in psychotherapy.

3. Cairns-Lee H., Lawley J. & Tosey P. (Editors). (2022). Clean Language Interviewing: Principles and Applications for Researchers and Practitioners.

4. Lawley, J. & Tompkins, P. (2011). Symbolic Modelling: Emergent Change though Metaphor and Clean Language. Chapter 4 of L. M. Hall & S. R. Charvet (Eds.), Innovations in NLP.

5. Tosey, P., Sullivan, W. & Meyer, M. (2013). Clean Sources: Six Metaphors a Minute?

6.  Bandler, R., Grinder, J., Dilts, R. & DeLozier, J. (1980). NLP Volume 1: The Study of the Structure of Subjective Experience.

7. Tompkins, P. & Lawley, J. (Feb 2006). Coaching for P.R.O.’s: The Problem, Remedy, Outcome model. Coach the Coach.

8. Full transcript available at:

Part 2 of this three-part series is available at:

Introducing Clean Language

Part 3 is available at:

Resources, embodied metaphor and Clean Language

Print Friendly, PDF & Email
body * { color: inherit !important; }