An overview of the Meta Model and explanation of the five distortion categories

March 16, 2016

Distorting perceptions through language

Human beings rely on language to share experience. For the most part, the sentences we form are unconscious productions. We often limit ourselves through the language we use, and distort immediate perception from our deeper reality.

The NLP Meta Model, is an effective model for dealing with distortions in everyday language, to help people have richer experience in language and an enhanced internal personal experience. In this article, I will present the history of the Meta Model, an overview of the three categories distortion, deletion and generalisation and describe five meta model patterns in the distortion category.

A history of the Meta Model

The NLP adventure began way back in the early 1970’s when John Grinder offered assistance to Richard Bandler and Frank Pucelik in providing a ‘code’ for their (Bandler and Pucelik’s) unconscious assimilation of the work of Fritz Perls. Bandler and Puclik were very good at the ‘how’ in replicating the work of Perls, they (like Perls) did not have a conscious representation of ‘what’ specific patterns were systematically applied by Fritz Perls.

Grinder being Grinder, insisted on learning the HOW first, before providing any representation of the ‘WHAT’.

To that extent Grinder, modelled Bandler and Pucelik in their Gestalt work to get an unconscious representation of the patterning before making the patterning explicit through coding.  Once Grinder was proficient behaviourally, Grinder and Bandler went to work on creating a code.  Those who have studied NLP, will be aware that Grinder, Bandler and Pucelik continued their modelling work with Virginia Satir, noted an overlap with Fritz Perls and created NLP’s first model ‘The Meta Model’

Grinder and Bostic (2002) refer to the Meta Model as a ‘common function model’ i.e it has different sources (Perls and Satir)  but the patterning functionally in the same application.  The meta model is a syntactically based language model with thirteen verbal patterns to challenge the ‘structure’ (form) present in a person’s language.

Each meta model pattern has a series of challenges or questions which when answered by the client forces the client to expand or modify the limitations within their mental maps providing the client with greater access to the choices they have naturally within.

The Universal Modelling Process

In NLP we refer to two mental maps, i/ the sensory based map which is the internal representation of what we see, hear, taste and smell moment to moment and ii/ the linguistic representation of the sensory based map. In early NLP, the sensory based map was known as the 4Tuple and the linguistic map known as ‘deep structure’.

The verbal expression of the linguistic map was known as ‘surface structure’ Grinder and Bandler (1975) used the term ‘Universal Modelling Process’ to describe how we form our mental maps.

The Universal Model Process has within it three elements

1. Distortion – The process of representing parts of the model differently than how they were originally represented e.g process to event

2. Generalisation –  The way a specific experience (or number of experiences) is mapped to represent the complete category of which it is a member

3. Deletion – Where portions of the mental map are omitted and do not appear in the verbal expression

Grinder and Bandler borrowed the term ‘deep structure’ and ‘surface structure’ from transformational grammar.  Deep Structure is the full linguistic representation which on which distortions, deletions and generalisations occurs creating a limited map of the world, where the client does not have full access to their internal resources.

The impoverished map of the world is evident in the verbal expression (surface structure), the NLP Practitioner listens to how the client is distorting, deleting and generalising their world and provides a meta model challenge to enable the client to connect to the full deep structure where more resources are present

Meta Model Distortions

Mind Reading: – Claiming to know the thoughts, feelings, intentions, meanings, motivations, or other internal processes of another person – with no basis in reasonable, logical grounds for interpretation or direct, sensory observation.


‘I know you don’t like me?’
‘I know you are curious’
‘I know you are sad?’

There is an obvious distortion in the sentences above as the speaker has no evidence in deep structure for the statement. To challenge the coach asks the speaker to specify their evidence for the hallucination

Challenge: ‘How do you know I am …”

Lost Performative

Value judgement where the performer of the judgement is not stated. The speaker expresses the statement as a generalisation true for the world. (Good/bad etc)


“It’s bad to be inconsistent’
‘It’s rude to be loud’
‘It’s good to be timely’

The distortion, has no evidence, it’s frequently based on family or cultural filter. To break the distortion the NLP Practitioner asks the speaker to specify evidence for their statement

Challenge: According to whom? “Who says so?“ “How do you know?”

Cause Effect Pattern:

The implication or direct claim that one thing causes, or is caused by, another when there is no well-formed logical support or demonstrable, sensory-based evidence to support a causal connection.


‘You make me sad?’
‘People talking at once causes me anxiety?

Cause ”“ effect can only exist in mechanical systems, there is no cause effect in biological system. There is nothing anyone or anything can do to affect your internal state, the distortion here is, it’s our response to the external stimuli, not the stimuli itself. The meta Model challenge is designed for the speaker to access resources and have a different response, and thus choice.

Challenge: “How specifically does people talking at once ‘cause’ you to choose anxiety ?”
Or offer a counter example

Complex Equivalence Pattern

Statements where complex situations, ideas, objects or their meanings are equated as synonymous.


‘You always yell at me means you don’t like me’
He’s a man, he is lucky

The statements above are highly distorted the speaker is creating meaning by linking two unrelated components. The meta model question is designed to break the equivalence and recover choice.

“How specifically does my yelling mean I don’t like you?”
“Has someone ever yelled at someone you liked?


Presuppositions are the linguistic equivalent of assumptions.To make sense of a sentence the listener and/or speaker accept the presuppositions/assumptions as being true. Presuppositions are the distorted assumptions inherent in natural language.


“If My husband knew how much I suffered he wouldn’t act that way”

Presuppositions in the sentence

She suffers
Her husband acts in some way
Her husband does not know she suffers

You can use any Meta Model pattern to challenge the distortions that are present assumptions in language .

How do you choose to suffer? Recovers choice.
How specifically is he acting? Specifies the verb.
How do you know he doesn’t know? Challenges the mind read

In the next article on the Meta Model, we will visit generalisations and deletions. In the meantime keep your ears open to distortions in the language people use around you. Use the Meta Model questions to help the people access more choice and of course be mindful of how you distort your own perception through your language.

Latest insights from our experts

Blog Derren Brown, placebo, pills and psychology April 22, 2024
Videos The Impeccable Inception – the birth of NLP March 29, 2024
Videos The Impeccable Inception – the birth of NLP March 29, 2024
Videos The Impeccable Inception – the birth of NLP March 29, 2024
Videos The Impeccable Inception – the birth of NLP March 29, 2024
Videos The Impeccable Inception – the birth of NLP March 29, 2024
Videos The Impeccable Inception – the birth of NLP March 29, 2024
Videos The Impeccable Inception – the birth of NLP March 29, 2024
Videos The Impeccable Inception – the birth of NLP March 29, 2024