Clean Language was developed by David Grove in the 1980s as a result of his work on clinical methods for resolving clients' traumatic memories. As Lawley & Tompkins describe it, "He realised many clients naturally described their symptoms in metaphor, and found that when he enquired about these using their exact words, their perception of the trauma began to change."
Clean Language also is the basis of Symbolic Modelling, a stand-alone method and process for psychotherapy and coaching, which was developed by James Lawley and Penny Tompkins.
Clean Language combines four general elements of communication in a very specific way, that is syntax, wording, vocal qualities and nonverbals.
Note: we refer to the person asking the questions as the 'facilitator' and the person receiving the questions as the 'client'. This habit comes from the therapeutic roots of the Clean Language process. Depending on the context, these labels could be coach/coachee, interviewer/interviewee, doctor/patient,
Clean Language questions are cleansed as far as possible of anything that comes from the questioner's "maps" -- metaphors, assumptions, paradigms or sensations -- that could direct the questionee's attention away from increased awareness of his/her own metaphorical representation of experience.
Clean Language offers a template for questions that are as free as possible of the questioner's inferences, presuppositions, mind-reading, second guessing, inferences, references and metaphors. Clean questions incorporate all or some of the speaker's specific phrasing and might also include other auditory components of the speaker's communication such as sighs, pitch, tonality, etc. The questioner might also draw attention to any non-verbal signals that coincide with the client's auditory output, i.e., a fist being raised simultaneously with a sigh, that might also represent elements of the client's metaphorical representation of experience.
Besides words of the client, also his/her nonverbals are repeated or referenced in the question, as far as the questioner noticed them and if they might be of symbolic significance.
Note: Clean Language facilitators do not follow popular generalised assumptions about the meaning of 'body language' (e.g. assuming that crossed arms mean the person is 'closed'), preferring to ask and find out what such behaviour means to the client.
In this example B is the 'client' and A the 'facilitator':
B: "I feel strange."
Not-clean response
A says one of: "Have you got a headache?", "Are you ill?", "You're probably catching a cold", "You must be hung-over!", "Stop complaining! Take a pill..." etc.
Cleaner response
A asks one of: "Where do you feel strange?", "What kind of strange?", "Strange like what?", "Is there anything else about that 'feel strange'?", "What happens just before you feel strange?"
While there is a set of 9+3 basic Clean Language questions that get used about 80% of the time, the concept of being 'clean' resides not in the questions themselves (which are merely the medium) but in the intention of the facilitator.