Part One: Models and Modelling

First published November 1998

Article in PDF Format: RemodellingNLPPart1

“The purpose of the model is to enable the user to do a better job in handling the enormous complexities of life. By using models, we see and test how things work and can even predict how things will go in the future. The effectiveness of a model can be judged by how well it works, as well as how consistent it is as a mechanical or philosophical system. People are very closely identified with their models, since they also form the basis for behaviour. Men have fought and died in the name of different models of nature”.
Edward T. Hall, Beyond Culture, Doubleday 1976 (p 13-14)



The article in the last issue of Rapport by Martin Roberts has prompted me to share some of my developments and concerns both in regard to modelling and NLP more generally. I wholeheartedly agree with Martin about the limits of the NLP modelling method and the danger in exaggerated claims. It is very important to know the limits of any approach. Knowing the limits directs our attention to where improvements can be made and new developments are required. For a number of years I have held back from writing about my developments for publication in Rapport. I did not want to fall into the trap of making premature claims and unfounded assertions. After thorough testing and successful application over many years I am confident of the effectiveness and efficiency of my work in applying a modelling methodology and the results of applying it to NLP. I hope that this and subsequent articles will continue the healthy discussions stimulated by Martin’s article.

My experiences in secondary school and university were influential in developing my early understandings of models and modelling. At school and at university each year we were told to forget the previous year’s teaching as we would be learning a better, more recent and more accurate model in physics and chemistry. If I had been concerned with learning the ‘right’ model I would have found the experience very frustrating, instead it was freeing. To be a chemist is to be the builder of models. The application of the models is, however, seen as different – a technician applies them. I have always considered this unfortunate as a practical developer greatly benefits from

At the end of the first year at university I had decided to change courses to psychology. In psychology there appeared to be a broad range of models to choose from rather than a series of developments. However the result in my opinion was diminished development. This situation exists in any system which is eclectic and closed to change.

From 1979 – 1989 I worked in projects in social services which were concerned with developing new, more effective support systems for young people and families. Over a ten year period I had the opportunity for experimentation and innovation in these developmental projects working with family therapy models, group-work models, social education and the dynamics of inter-agency negotiation and joint projects. I first heard about NLP in 1979. I was at that time trying to integrate the diverse areas I was exploring and it was claimed that NLP used a modelling approach to identify the underlying patterns of the areas in which I was working. Rather than reinventing the wheel it made sense to study NLP.
My experience of formal training in NLP was both exciting and disappointing. The excitement was in the skills training. The disappointment was the lack of a modelling methodology. As a cook, I want to make the most of ingredients, so I need cooking skills and cooking principles and not just recipes!

I was fortunate to work with a client group and in contexts which were very different from the middle-class, co-operative clients with whom I practised on NLP courses. Fortunate, because I really needed to understand the necessary structure of the techniques and skills of NLP in order to adapt them, for example, adapting them from an office setting to make them applicable for counselling gang members on street corners. I was therefore forced to explore their underlying structure – I needed to model them. The limits of the strategies model were quickly apparent in the joint education projects where we soon went beyond the simple memory and spelling strategies, developing problem solving strategies and linking learning to identity. Through the great variety of areas of application I learned to have a great respect both for the benefits of NLP and also its limitations.

Not knowing the methodology that was used to develop NLP, I established my own calling it Developmental Behavioural Modelling (DBM) and intended it as a stopgap until I learnt the ‘real’ one. To learn the ‘real’ one I arranged to train with Richard Bandler. In 1990 Richard Bandler asked me to train for him in Switzerland and certified me as an NLP Trainer. He certified me as a Master Trainer later that year in the USA.

In 1993, Martin Roberts and I brought Richard to train in England. By this time I had realised that if a more formal modelling methodology in NLP existed it was not formally available. I was as keen as ever to pursue precision and effectiveness in therapy and change, so the only solution was for me to commit further to my own ‘fledgling’ approach, DBM. This meant that I have gone in a different direction from that chosen by Richard. I am pleased that he and Paul McKenna met at our training and have successfully worked together since. Since 1993 I have applied my DBM methodology to NLP.

The value of NLP

For me the great contribution of NLP has been to make more explicit HOW to do what works. The early books emphasised modelling. But what kind of modelling?

Robert Dilts is an excellent example of an NLP modeller. He is the creative source of many pioneering models about NLP. In his latest book “Modelling with NLP” he says:

“Behaviour modelling involves observing and mapping the successful processes which underlie an exceptional performance of some type. It is the process of taking a complex event or series of events and breaking it into small enough chunks so that it can be recapitulated in some way. The purpose of behaviour modelling is to create a pragmatic map or ‘model’ of that behaviour which can be used to reproduce or simulate some aspect of that performance by anyone who is motivated to do so. The goal of the behaviour modelling process is to identify the essential elements of thought and action required to produce the desired response or outcome. As opposed to providing purely correlative or statistical data, a ‘model’ of a particular behaviour must provide a description of what is necessary to actually achieve a similar result”.

It is useful to note the sequence described here. Firstly the breaking down, secondly the mapping and finally, the reproducing. It is also useful to note that the motivation is separated from the behavioural model and how to put the bits together again, putting Humpty Dumpty together again, is unclear.


Models and Modelling

Bandler and Grinder say this about models.

“No two human beings have exactly the same experiences. The model that we create to guide us in the world is based in part upon our experiences. Each of us may, then, create a different model of the world we share and thus come to live in a somewhat different reality…………

………We want to make two points here. First, there is a necessary difference between the world and any particular model or representation of the world. Second, the models of the world that each of us creates will themselves be different. There are a number of ways in which this can be demonstrated”. Magic I Page 7

From this we can identify three elements in subjective models, the product – the model, the process – the subjective modelling processes, and the principles – the modelling principles operating.

In Frogs into Princes, Bandler and Grinder clarify modelling and themselves as modellers.

“We call ourselves modellers. What we essentially do is to pay very little attention to what people say they do and a great deal of attention to what they do. And then we build ourselves a model of what they do. We are not psychologists, and we’re also not theologians or theoreticians. We have no idea about the “real” nature of things, and we’re not particularly interested in what’s “true”. The function of modelling is to arrive at descriptions which are useful. So, if we happen to mention something that you know from a scientific study, or from statistics, is inaccurate, realise that a different level of experience is being offered you here. We’re not offering you something that’s true, just things that are useful.

We know that our modelling has been successful when we can systematically get the same behavioural outcome as the person we have modelled. And when we can teach somebody else to be able to get the same outcomes in a systematic way, that’s an even stronger test”. Page 7

Once again we can identify the three elements in relation to the modeller, the product – the model created, the process – the formal (or informal) modelling processes, and the principles – the modelling principles operating. These may be different from those in models.

Replicating successful behaviour – product and process.

There is of course a big difference between a product and the way that product is produced. So, for example, knowing the language patterns produced by Milton Erickson, the product, does not give any understanding of how he produced them, the process, nor why he did so, the principles he used. Nor does it explain the relationship between these three. What life experience does he have that lets him remember or calculate an intervention that is likely to be successful.

“We often think that when we have completed our study of one we know all about two because ’two’ is ‘one and one’. We forget that we have still to make a study of ‘and’”.
A. S. Eddington

What is involved here is not just different skills but a different level of skills with supporting life experience. Erickson’s language pattern is technology, his approach is a methodology. Both are useful but having a technology does not add to our understanding of how and why he did what he did. It is very important to understand these limits in the scope of the NLP modelling method.

If we are unaware we are not in a position to develop. All closed systems result in a lack of self critical development and eventually become obsolete in the face of new developments. Another consequence is an increase in myths and mistakes.

A great inspiration for me was watching and listening to Bandler perform change work when he was doing what he called “Torpedo Therapy”. This was what I had expected from NLP – an example of dynamic modelling and re-modelling that truly met the client at their model – not a technique in sight! (See “Magic in Action” for some examples).

It was with this inspiration and a desire to replicate this successful modelling that I pursued the practical application and further development of Developmental Behavioural Modelling.

The Value of DBM

“To know how to criticise is good, to know how to create is better’” H Poincare

We are all modellers in a naturalistic way. We all have our models and our own way of building them. What if we understood more about these models and how we build them and developed new, more effective ways to change them? This would involve product, process and principle. This has been the aim of DBM.

Models can be limiting as well as enhancing. Models should be the result of understanding and not the beginning. The use of NLP is too often to use models instead of building an understanding. Richard Bandler used to frequently say that NLP is 95% information gathering and 5% change work. All too often the reverse is what is practised, with a rush to use a technique.

DBM follows the philosophical position of constructivism – we all construct subjective models of the world. This has two immediate areas of interest for me. Firstly how effective are our models and secondly how do we build and change them? How we gather information is central to this – how we build a unique model of each unique client. This is similar to what Milton Erickson claimed to do. In his case he claimed to build a theory for each client.

DBM is a practical integration of a formal modelling methodology and practical application and is designed to work with everyday behaviour. Formally it is an integration of behavioural technology, much of it based on the core NLP skills (procedures), a holistic modelling methodology (processes), and an epistemology (principles) based on the work of Gregory Bateson.

The basic evaluation model in DBM “E-valuates” any model as to its Effectiveness,
Efficiency and Elegance. We use a tool called a Life Grid to identify levels of processing and areas of life experience. Life experience is a central element and crucial to include in any modelling approach. The lack of its inclusion in NLP justifies Martin Roberts concerns in his last article.
I also extended the scope of study from the WHAT level of NLP, to the processes (how) and principles (why) involved. The Epistemology Grid is a modelling tool to identify the appropriate level being investigated. At each of the levels there are more detailed what, how and why. This is a Fractal Modelling approach.

DBM LevelFractal LevelBrief Description
Epistemology Principles WhyWhyWhy the principles work
HowHow the principles are sequenced
WhatWhat principles are involved
WhyWhy the processes work in the order they do
HowHow the processes are sequenced
WhatWhat processes are operating
Technology Procedures WhatWhyWhy in a particular order – NLP change model
HowHow to do it – Main focus of NLP skills training
WhatWhat to do that works

Levels of Modelling

Detailed distinctions are important for precision modelling. That includes modelling itself. In DBM we identify a number of levels to modelling. All of them are useful.
1. Naming. Names are irresistible. Identifying something often begins by naming it. It can also be a trap as the name is not the thing, the result can be the illusion of understanding.
2. Listing. Collecting things, grouping of things.
3. Classifying. Relating the list. Very popular result of testing, as in IQ or personality. Again there is the possibility of illusory understanding.
4. Sequencing. Tracking changes over time. Attending to the dynamic qualities.
5. Mapping. Formalising a sequence. Useful in giving a simultaneous representation but at the cost of reducing the temporal dynamic.
6. Processing. Identifying the key functioning of the sequence.
7. Replicating. Copying particular sequences.
8. Patterning. Identifying a common map across examples.
9. Modelling. Replicating product, process and principle.
10.Recursioning. Going beyond by applying model to self.
11. Modelling Modelling. Identify the product, process and principles of the modelling.

NLP aims to replicate successful behaviour (level 7). The strategies model supports this by mapping the sequence of senses used in a skill (level 5). If we follow Bandler and Grinder’s advice to understand the client’s model – the structure of subjective experience and concentrate on what they are doing, that would be (level 4), if we want to build a working model that would be (level 9). In DBM we are also interested in how the client constructs and changes their model of the world – the ongoing processing and patterning of subjective experience (levels 10 and 11). This requires modelling what, how and why modelling works, and the working of modelling, these are tasks beyond the scope of the NLP method.

NLP aims to model how things work. DBM aims not only to model how things work
but how to work things.


Models-Myths-Metaphors-Mistakes, (4Ms Model)

One of the first things to identify with a client or a formal model is what exactly the model is, separate from any mistakes, myths and metaphors involved. In an old bush it is good to clear the rubbish, to cut out the dead and dying wood in order to identify the healthy growth. With that in mind I will begin with identifying a number of myths and mistakes that have grown in NLP over the years. It is difficult to improve NLP when it is a collection of disparate bits from a variety of other models further confused by numerous mistakes and unsupported claims. Having cut these out we will begin in part 2 with the central task of re-modelling NLP.

Below are some examples of myths and mistakes I have identified in NLP over the last ten years. You might be aware of some of them, some you may be committed to. My concern at this point is less in the detail (although at other times I am very concerned about detail) than with the lack of inclusion in many NLP training courses of skills in critical self reflection and practice of the attitude of curiosity and discovery.

I will explain a number of the myths and leave it to you to practise with the rest what I am preaching – positive exploring through healthy questioning. I have included twenty-one myths as a metaphor for the coming of age. Ideally, a maturing of belief and understanding where the illusions and simplifications of childhood are transcended. I will outline the others in future parts.

Myths and Mistakes in NLP
1. 93% of communication is non-verbal.
2. Eskimos have seventy words for snow.
3. Language is always literal.
4. Beliefs are difficult to change.
5. “Why” should not be used.
6. “Try” should not be used.
7. Flexibility and variability are the same.
8. Excellence and Excelling are the same.
9. Gestalt: The whole is greater than the sum of its parts.
10. NLP Ecology Check.
11. NLP is client centred.
12. NLP is Holistic.
13. In NLP there is no split between mind and body.
14. Personal “Parts” exist.
15. Beliefs and identity are neurological levels.
16. The brain is the same as the mind.
17. Modelling method is a methodology.
18. We have perceptual Filters.
19. Sensory acuity is all about clean sensory channels.
20. NLP is Systemic.
21. Systems thinking is the same thing as thinking systemically.

1. 93% of communication is non-verbal.

This is one with which I was never comfortable. If it was true then it should be easy to communicate the precise meaning including the 93% without the lowly 7% of verbal communication – try it! Feeling that it is wrong and knowing why is very different. My friend and colleague, Chris Norris, was able to help me when he attended an Advanced Master Practitioner training which I ran a number of years ago. He knew the name of the source, Mehrabian. Mehrabian (1971) wrote a very interesting book about how we communicate and perceive feelings. On page 77 he states the generalisation from their findings on liking to feelings more generally:

“Generalising, we can say that people’s implicit behaviour has more bearing than their words on communicating feelings or attitudes to others. So we have rewritten our equation for any feeling. Total feeling = 7% verbal feeling + 38% vocal feeling + 55% facial feeling”

He later goes on in page 79 to caution:

“Incidentally, we should be careful to note that these assertions about the disproportionate contribution of implicit, relative to verbal, cues is limited to feelings (pleasure, arousal, dominance) and like-dislike. Obviously, implicit expressions are not always more important than words. In fact, implicit cues are ineffective for communicating most referents denoted by words (for example, “I’ll see you tomorrow afternoon at 2:00 pm”, “I was wearing my new velour suit yesterday”, or “x + y = z”)”.

There are three points of note here. The first is obviously that when the percentages
quoted are applied to all communication and not just how we perceive whether we are
liked we are making a gross error. The second is that the details of them are often
changed from vocal and facial to non-verbal movements. The third is the use of these
percentages to sell non-verbal communication skills training. This is a dishonest sales
practice whether done knowingly or unknowingly.

There are three points of note here. The first is obviously that when the percentages quoted are applied to all communication and not just how we perceive whether we are liked we are making a gross error. The second is that the details of them are often changed from vocal and facial to non-verbal movements. The third is the use of these percentages to sell non-verbal communication skills training. This is a dishonest sales practice whether done knowingly or unknowingly.

2. Eskimos have seventy words for snow.

On page 15 of Frogs, Bandler and Grinder state that “Eskimos have some seventy words for snow”. Pullum explores this myth in his book ”The great Eskimo Vocabulary Hoax”. He cites the work of anthropologist Laura Martin of Cleveland State University who spent some of her research time during the 1980s attempting to slay the constantly changing, self-regenerating myth of Eskimo snow terminology. She traced the myth to the original source of Franz Boas’ introduction the Handbook of North American Indians (1911).

“And all Boas says there, in the context of a low-key and slightly ill-explained discussion of independent versus derived terms for things in different languages, is that just as English uses separate roots for a variety of forms of water (liquid, lake,river, brook, rain, dew, wave, foam) that might be formed by derivational morphology from a single root meaning ‘water’ in some other language, so Eskimo uses the apparently distinct roots aput ‘snow on the ground’, gana ‘falling snow’, piqsirpoq ‘drifting snow’, and qimuqsuq ‘a snow drift’. Boas’ point is simply that English expressed these notions by phrases involving the root snow, but things could have been otherwise, just as the words for lake, river, etc. could have been formed derivationally or periphrastically on the root water”. P161

“Don’t be a coward like me. Stand up and tell the speaker this: C W SchultzLorentzen’s Dictionary of the West Greenlandic Eskimo Language (1927) gives just two possibly relevant roots: qanik, meaning ‘snow in the air’ or ‘snowflake’, and aput, meaning ‘snow on the ground’. Then add that you would be interested to know if the speaker can cite any more”. P167

As a modeller I too am interested in what people do especially when perpetuating inaccurate models. My concern for a modelling field is the lack investigation and accurate modelling. It is one level of error to believe the Emperor is wearing fine clothes if you haven’t seen the Emperor, it is another when you have.

10. Ecology Check

The question at the end of techniques is not a test of ecology but of the client’s congruence with the changes. It is a test of “affect”. On its own it could never be a test of “effect”, what the results would be on the client as a whole system. It is a useful thing to do but deceptively mis-leading to give it a grander title than is accurate.

18 Perceptual Filters

It is often stated in NLP that we have Perceptual Filters. These are responsible for filtering the incoming information about the world. As a metaphoric description this is an interesting one, as a process description it is the exact opposite of what happens. Our nervous system works on thresholds and news of a difference. We learn to become sensitive to particular stimuli and so actively select what we let through – the exact opposite of a filter whose active response is to hold things back. What could hold a perception back and what happens to all the rubbish in the filter?

9. Gestalt: The whole is greater than the sum of its parts.

This is one that can have such misleading connotations as ‘big is better’. In my work with gangs, dysfunctional teams and organisations it is clear that the opposite is the case – the whole is less than the sum of the parts. I’m a little embarrassed to think how many years it took me to work this out (however I am somewhat consoled by the fact that no one else I have met has realised it until I tell them!). The accurate conclusion is that the whole is different from the sum of its parts.

I hope that this is enough to stimulate for now. The others will be clarified in future articles.

In this article I have outlined some of the beginnings of ‘re-modelling’ NLP to make it more effective, efficient and elegant. In future articles I will go into more depth on each of the main elements of practitioner and master practitioner level NLP. I hope that this will encourage an open and developmental approach to NLP through constructive criticism. Only if this happens can NLP continue as a field of study rather than becoming a dead technology. By demonstrating openness to learning and development in our own thinking and behaviour we offer our clients a deeper pattern of successful behaviour, one that will benefit them beyond the range of any particular technique.

Since the introduction of sub-modalities in the early 80’s there has been little in the way of major developments of the NLP model. There have been many developments beyond NLP; chaos theory, fractal structures, fuzzy logic to name a few. As Bateson said in the introduction to Magic I “But, indeed, much that was so difficult to say in 1955 is strikingly easier to say in 1975″p xi. We have moved on a lot since 1975.

Part two of this series will be “Re-Modelling the Meta and Milton models” and include a simple model of conversational belief change that is also a basis for constructive questioning of NLP.


Bandler, Richard & Grinder, John, The Structure of Magic vol.1, Science and Behaviour Books, Inc 1975
Bandler, Richard and Grinder, John, Frogs into Princes, Real People Press 1979
Bandler, Richard, Magic in Action, Meta Publications 1984
Dilts, Robert, Changing Belief Systems with NLP, Meta Publications 1990
Dilts, Robert, Modelling with NLP, Meta Publications 1978
Hall, Edward T. , Beyond Culture, Doubleday 1976 Mehrabian, Silent Messages, Implicit Communication of Emotions and Attitude
Second Edition 1971, 1981
Pullum, Geoffrey K, The Great Eskimo Vocabulary Hoax, The University of Chicago
Press 1991

DBM is a registered trademark

Personal Profile

John is the creator of DBM and Systemic Counselling and Consultancy. He assisted and trained for Richard Bandler in Europe, Australia and USA from 1990-93. He was certified as a Master Trainer of NLP in 1990. He frequently presents his developments at the ANLP conferences. He is the Director of Sensory Systems Training.