NCP Logo

The National Council of Psychotherapists

Est. 1971

Methodical Meta and Magical Milton Models:

Language and the Nature of Therapuetic Transformation.’


Patrick Jemmer

Alowvelkí Consulting Limited


This article investigates the interleaving of two diametrically opposed yet complementary techniques of language manipulation, the Milton and Meta models of Neuro-linguistic Programming, which taken together can provide an almost magical toolkit for elicitation and change in a therapeutic setting. 


O’Connor and Seymour (1990) provide simple yet useful definitions for the Milton and Meta models, as follows:


‘[The meta model is] A model that identifies language patterns that obscure meaning in a communication through the processes of distortion, deletion and generalization, and specific questions to clarify and challenge imprecise language to connect it back to sensory experience and the deep structure.’


‘[The Milton model is] The inverse of the Meta Model, using artfully vague language patterns to pace another person’s experience and access unconscious resources.’

The particular usefulness of these definitions is the way in which the fundamental relationship between the two models is stressed, rather than any possible distinctions. It is therefore important to investigate the nature and use of both models together rather than in isolation: this is done by considering the application of both models to a particular problem, and can be best described in terms of ‘chunking’ or breaking the problem into parts. The Meta model is used for chunking down, that is, going from a higher to a lower logical level: this allows the focus of attention to shift from the more general to the more specific, and thus aids a move towards the definition of a specific problem. In linguistic terms, the Meta model moves from the surface structure that is the words spoken, to the deep structure that is the internal representation of the underlying concepts and things. The Milton model is the reverse of this, and can be a resource in solution-finding, by chunking upwards from the part to the whole, from the specific to the general.  A fundamental tenet of Neuro-linguistic programming is that ‘the map is not the territory,’ and that by changing a client’s internal representation or map, their whole world of experience can be changed. The process of chunking allows a therapist to achieve this by first pacing her client, interrogating the client’s map and finding and installing solutions appropriate to that client. These ‘magical’ effects can often be facilitated by the use of hypnotic trance.


In terms of actual therapy, in contrast to theory, Battino and South (1999) categorize Meta model interventions into three categories. First, it is necessary to gather information and clarify; then one must find out what limits the client is imposing on his map of the world; finally the therapist concentrates on ‘fuzziness’ in the client’s logic as applied to his map. In fact, it becomes clear that the most effective and least painful method for applying the Meta model is to tackle distortions in the map first, followed by generalizations, and then deletions. These ideas are illustrated below.


For example, a client may come to see a therapist and present in the following way: ‘I don’t know why I’ve wasted my time in coming to see you. Nobody listens to me. They all hate me. I’m so unhappy.’ The therapist can use the Meta model to challenge the client’s statements and so recover the true meaning of these as represented in the client’s map. To do this she uses specifying questions such as ‘And what exactly…?’ As a highly telescoped sample of a therapy session, the therapist is likely to begin by challenging the generalization of the universal quantifier and ask ‘So no one at all listens to you, then?’ She could then move on to investigate distortions such as mind reading by asking ‘How exactly do you know that these other people hate you?’ Finally, she might turn to deletions, by querying simple deletions ‘And what specifically are you unhappy about?’ These Meta model queries (asked with compassion and congruence on the part of the therapist!) might elicit responses that indicate that in fact the client believes his partner does not listen to him, is dissatisfied in his relationship, and has issues relating to his sexuality: the point is that these underlying meanings would be very difficult and longwinded to discern from his surface structure (if that were possible at all), without the use of the Meta model. It would be easy for both client and therapist to ‘tie themselves in knots’ and become quickly frustrated is less structured and incisive methods were to be used. Having determined at least one important part of the client’s map that it would be useful to change or enhance (for example to allow a fuller exploration of relationships, or in developing self-esteem), the Milton model is a useful device to use together with trance, to facilitate this process.


As mentioned previously, the Meta and Milton models are defined in terms of each other and are complementary strategies for communication and problem-solving. Thus if the Meta model can be used as a scalpel to dissect a problem to its bare bones, then the Milton model with its ‘artfully vague language’ provides the very ‘magical’ incantations used to conjure up solutions, as described by Bandler and Grinder (1975,1977). Once again we encounter the ‘transderivational search’ in which the client ‘goes inside’ to seek meaning for himself: this is naturally accompanied by a trance-like state. And whereas Hall (1998) tells us that ‘meta-modeling brings a person out of trance,’ the Milton model can be used to enhance the search, and can so also be seen as a tool for entrancing. The way in which this is accomplished is to chunk upwards from the specifics of the client’s recovered deep structure of the external world, to general ideas and principles. In the Milton model generalizing questions like ‘And how do you know…?’ are used to accomplish this. However, the idea is now to generate surface structures different from the original, problematic internal map representations, and so offer new choices that increase flexibility and freedom. The ideas of distortion, generalization and deletion are used again: however, in applying the Milton model, rather than trying to eliminate these categories of linguistic ambiguity, a therapist actively introduces them into his conversation or trance induction.


In terms of trance induction and as a methodology for accessing resourceful problem-solving states the Miltonian devices of distorting statements, making statements general rather than specific, and deleting portions of statements, are used to confuse the client’s conscious processing systems. Eventually, the failure of these to cope with the barrage of linguistic information causes them to ‘give up,’ allowing direct (or indirect) communication with the out-of-conscious processing system. This can be accompanied by other techniques developed to perfection by Erickson over his entire career, such as quoting, using ‘hidden’ commands, employing specialised voice dynamics and inserting ‘tag’ questions. Some of these devices are illustrated in the paragraph below, which indirectly discusses some of these techniques in a Miltonian way, and which might be used to entrance trancers who claim not to understand the Milton model.


So, now, then, having found a ‘glitch’ in the client’s map, the therapist can use appropriate tools to enhance choices for the client, or even inappropriate ones if they work, can’t she? And after all, what’s appropriate but just that that someone can appropriate for themselves. So what a good time you can have to avoid a hitch and mend a glitch since a stitch in time saves… for a better future. But but me not buts since as we’ve said we’re going on this journey… begins with a single step… into trance. And Battino and South (1999) tell the story that once upon a time someone asked Milton Erickson whether he was aware of the way in which he used words, but it was not the weather he was interested in, nor whether Milton was a witch or a poet or whatever, and he replied to him that he was and that he wanted to emphasize the importance of that awareness to him. And as reading you are this, under-standing gaining we must be. As we stand under the apple tree, everyone else and me. And as me and you can see, or maybe not, but in either case it’s all becoming clear that Erickson was a wizard with words, and Milton the poet said that the mind is its own place and in itself can make a heav’n of hell [Paradise Lost I 254-255] and with this… and that… now at last, as the night has passed, we’re back to our friend [Aeneid IV] the Meta model. And that’s poetry, no; or is it magic?


In conclusion, then, the Meta and Milton models are intrinsically linked and use the same techniques to achieve different goals in the same therapeutic game. The Meta model, based on the Chomskyan doctrine of Transformational grammar is used to ‘clean up’ a client’s statements and help him recover the true deep structure; this helps in identifying the true nature of his problem. The Meta model reverses this process to move from the specifics of the problem to new and better solutions by cleverly used choice of both directive and meaningless language. This is the true nature of beguilement.



·        Bandler, R. and Grinder, J., The Structure of Magic I. A Book about Language and Therapy. Palo Alto CA: Science and Behavior Books (1975).

·        Bandler, R. and Grinder, J., Frogs into Princes. London: Eden Grove Editions (1996).

·        Battino, R., and South, T. L., Ericksonian Approaches: A Comprehensive Manual. Carmarthen UK: Crown House Publishing Limited (1999).

·        Grinder, J. and  Bandler, R., The Structure of Magic II. A Book about Communication and Change. Palo Alto CA: Science and Behavior Books (1976).

·        Hall, L. M., The Secrets of Magic. Carmarthen UK: Crown House Publishing Limited (1998).

·        Hall, L. M. and Belnap, B. P., The Sourcebook of Magic. Carmarthen UK: Crown House Publishing Limited (2000).

·        O’Connor, J. and Seymour, J., Introducing Neuro-linguistic Programming. London UK: Thorsons (1995).





Article Menu - Members Directory - Navigation Page - Home