NLP Training – Meta Model

Written by Terry Elston

Picture explaining the Metal Model | NLP World.

The meta-model in NLP or neuro-linguistic programming (or meta-model of therapy) is a set of questions designed to specify information, challenge and expand the limits to a person’s model of the world. 

It responds to the distortions, generalizations, and deletions in the speaker’s language. The meta model forms the basis of Neuro-linguistic programming as developed by then assistant professor of linguistics, John Grinder and Richard Bandler. Grinder and Bandler “explained how people create faulty mental maps of reality, failing to test their linguistic / cognitive models against the experience of their senses.”

The meta model draws on transformational grammar and general semantics, the idea that language is a translation of mental states into words, and that in this translation, there is an unconscious process of deletion (not everything thought is said), distortion (assumptions and structural inaccuracies) and generalization (a shift towards absolute statements). Likewise in hearing, not everything said is acknowledged as heard.

These language patterns were based on the work of family therapist Virginia Satir, gestalt therapist Fritz Perls and linguistic patterns from Transformational syntax. It is claimed that the Meta-model “yields a fuller representation of the client’s model of the world – the linguistic Deep Structure from which the client’s initial verbal expressions or Surface Structure, were derived” by offering challenges to its limits, the distortions, generalisations or deletions in the speaker’s language.The reverse set of the meta-model is the Milton-model; a collection of artfully vague language patterns elicited from the work of Milton Erickson.

The following examples were derived from therapeutic contexts. The developers state that these patterns can be identified in all human communication. 

Definition of the meta-model:

“People respond to events based on their internal pictures, sounds and feelings. They also collect these experiences into groups or categories that are labeled with words. The meta-model is a method for helping someone go from the information-poor word maps back to the specific sensory-based experiences they are based on. It is here in the information-rich specific experiences that useful changes can be made that will result in changes in behaviour.”

Terry Elston “Knowing NLP”

Following are EXAMPLES of HOW to use your language to recover lost sensory experiences from the speaker:


1. Mind Reading: Claiming to know someone’s internal state. Ex: “You don’t like me.” “How do you know I don’t like you?” Recovers Source of the Info.

2. Lost Performative: Value judgments where the person doing the judging is left out. Ex. “It’s bad to be inconsistent.” “Who says it’s bad?” “According to whom?” “How do you know it’s bad.” Gathers evidence. Recovers source of the belief, the Performative, strategy for the belief.

3. Complex Equivalence: Where two experiences are interpreted as being synonymous. Ex: “She’s always shouting at me, she doesn’t like me.” “How does her yelling mean that she..?” “Have you ever shouted at someone you liked?” Recovers Complex Equivalence. Counter Example.

5. Presuppositions: Ex: “If my husband knew how much I suffered, he wouldn’t do that.”
There are 3 Presuppositions in this sentence: (1) I suffer, (2) My husband acts in some way, and (3) My husband doesn’t know I suffer. (1) “How do you choose to suffer?” (2) “How is he (re)acting? (3) “How do you know he doesn’t know?” Specify the choice & the verb, & what he does.
Recover the Internal Rep., and the Complex Equivalence


6. Universal Quantifiers: Universal Generalizations such as all, every, never, everyone, no one, etc. Ex: “She never listens to me.” Find Counter Examples. “Never?” “What would happen if she did?” Recovers Counter Examples, Effects, Outcomes.

7. Modal Operators: a. Modal Operators of Necessity: As in should, shouldn’t, must, must not, have to, need to it is necessary. Ex: “I have to take care of her.”
b. Modal Operators of Possibility: (Or Impossibility.) As in can/can’t, will/won’t, may/may not, possible/impossible. Ex: “I can’t tell him the truth.” a. “What would happen if you did?” (“What would happen if you didn’t?” Also, “Or?”

b. “What prevents you?” (“What would happen if you did?”)
Recovers Effects, Outcome.

Recovers Causes


8. Nominalizations: Process words which have been frozen in time, making them nouns. Ex: “There is no communication here.” “Who’s not communicating what to whom?” “How would you like to communicate?” Turns it back into a process, recovers deletion, and Ref. Index.

9. Unspecified Verbs: Ex: “He rejected me.” “How, specifically?” Specifies the verb.

10. Simple Deletions: a. Simple Deletions: Ex: “I am uncomfortable.”
b. Lack of Referential Index: Fails to specify a person or thing. Ex: “They don’t listen to me.”

c. Comparative Deletions: As in good, better, best, worst, more, less, most, least.

Ex: “She’s a better person.” a. “About what/whom?” b. “Who, specifically, doesn’t listen to you?”
c. “Better than whom?” “Better at what?” “Compared to whom, what? Recovers Deletion.

Recovers Ref. Index.

Recovers Comparative Deletion.

OK, the following is a take on the Meta Model from another source: You can read this to allow yourself to think more about language or just stick with the above for clarity!



Presupposition, refers to an assumption whereby the truth is taken for granted.

Crucially, negation of an expression does not change its presuppositions: I want to do it again and I don’t want to do it again both mean that the subject has done it already one or more times; My wife is pregnant and My wife is not pregnant both mean that the subject has a wife. In this respect, presupposition is distinguished from entailment and implication. For example, The president was assassinated entails that The president is dead, but if the expression is negated, the entailment is not necessarily true.

Example 1:

* “My wife is pregnant.”
* Presupposition: You have a wife.

Example 2:

* “You are doing that ‘thing’ on me again?”
* Presupposition: I have done ‘it’ already, at least once.
* Challenge: “What kind of ‘thing’? How are you so sure I already did it, whatever it is?”

Example 3:

* “My husband is as lazy as my son.” (you can also see this as a comparative deletion.)
* Presuppositions: You have a husband; you have a son; you believe your husband is lazy.
* Challenge: “How lazy is your son? Am I to assume that you believe your husband and son are lazy?” “Compared to what or whom?”


Cause-effect, the inappropriate use of causal thinking (x means y, x makes me y, or x makes y happen)

Causality always implies at least some relationship of dependency between the cause and the effect. For example, deeming something a cause may imply that, all other things being equal, if the cause occurs the effect does as well, or at least that the probability of the effect occurring increases.

Example 1:

* “That news makes me angry.”
* Presupposition: He/she is angry
* Challenge: “If it weren’t for that news, you would not be angry?”
* Challenge: “How does that news cause you to choose to be angry?” (when asking this question, you need a lot or Rapport.)

Example 2: (complex equivalence)

* “Being late means she does not love me.”
* Challenge: “How do you know?” “Were you ever late for someone you cared for?”

Mind Read

Mind-reading violation occurs when someone claims to think they know what another is thinking without verification.


* “If he doesn’t start paying his share of the bills, she is going to leave him.”
* Challenge: “How do you know this? Has she told you that she intends to leave him if he doesn’t?”


Nominalisation occurs when a verb is transformed into a noun. A dynamic process (i.e. a verb) is transformed into something static (i.e. a noun). It’s like taking a snapshot of a moving object, you don’t see the movement any more, just the (static) object.

In English, some verbs and adjectives can be used directly as nouns, for example, change and good. Others require a suffix:

* applicability (from applicable)
* carelessness (from careless)
* difficulty (from difficult)
* failure (from fail)
* intensity (from intense)
* investigation (from investigate)
* movement (from move)
* reaction (from react)
* refusal (from refuse)
* swimming (from swim)
* nominalisation (from nominalise)


* “The communication [from ‘communicate’] in this company is poor.”
* Challenge: “Who is not communicating with whom?” Or (if you know who) How could we communicate more effectively?”

* “They need my decision [from ‘decide’] by Monday.”
* Challenge: “What are you deciding?” (takes it back into the process) If you know what they are deciding upon – “What have you decided already, before we spoke?”

Note: there are 2 simple tests that can be used to determine if a word or expression is a nominalisation:

* the wheelbarrow test: if you can put it into a wheelbarrow, it is NOT a nominalisation. E.g. A drink is a noun, but it is not a nominalisation… as it is tangible, it can be put into a wheelbarrow and carried around. Quality control fails the wheelbarrow test and is a nominalisation.
* If the word continuous can be put in front of the noun and still make sense. E.g. improvement becomes continuous improvement, hence improvement is a nominalisation. (The fact that continuous can be added indicates that there is a dynamic aspect to this static element).

Complex Equivalence

Complex equivalence draws an unrelated conclusion from an event to create a logic that “does not follow”


* “And now my secretary quit. I’ll be bankrupt by the end of the year!”
* Challenge: “Are you telling me your fortune depended on your secretary’s employment?”


* See also: hasty generalization, Glittering generality


All known human languages make use of quantifiers, even languages without a fully fledged number system. For example, in English:

* Every glass in my recent order was chipped.
* All of those people standing across the river are racist.
* All of the people I talked to didn’t have a clue who the candidates were.
* Everyone in the waiting room had at least one complaint against Dr. 
* There was somebody in his class that was able to correctly answer every one of the questions I submitted.
* All blue eyed people are smart.

Universal quantifiers occurs when someone attempts to characterize an entire set (all people, every X, everyone, everything, …). This NLP meta model question can be used when someone is generalising too broadly.


* “My co-workers are all lazy.”
* Challenge: “All of them?” or “Which co-workers, specifically?”

Modal operators

Modal operators are intuitively characterised by expressing a modal attitude, such as necessity (have to, must, should) or possibility (can, might, may) towards the proposition which it is applied to. (see also: wishful thinking)


* “I can’t put myself together.”
* Challenge: “What would happen if you did/didn’t?”

* “I must put myself together.”
* “Challenge: “What would happen if you didn’t?”


Simple Deletions

In a simple deletion an important element in a statement is missing. For example:

Go and do it. That is important. I feel bad.

Key words to look out for are it and that.

The appropriate response would be to ask what, where or when exactly? “Go and do what exactly?”

“I feel really bad” – “About what, whom”. “My mother told me” – “Told you what?” This may sound very simple, but the number of times even trained people step over and accept simple deletions defies belief.

Unspecified Verbs

In an unspecified verb it is not clear how the action creates or created the result. For example:

I created a great impression on them.

The appropriate response is to ask how exactly does taking “x” action lead to “y” result. “How exactly did you create a great impression (and note the unspecified referential index “them”) on who exactly?”


Unspecified Comparatives or null comparative is a comparative in which the starting point for comparison is not stated. These comparisons are frequently found in advertising. For example, in typical assertions such as “our burgers have more flavour”, “our picture tube is sharper” or “50% more”, there is no mention of what it is they are comparing to. In some cases it is easy to infer what the missing element in a null comparative is. In other cases the speaker or writer may have been deliberately vague in this regard, for example “Glasgow’s miles better”. The null comparative


* “That was not the best plan.”
* Challenge: “What were some of the other plans?”

* “Our picture tube is sharper”
* Challenge: “Sharper, compared to what?”

Unspecified referential index

Unspecified referential index, refers to the use of personal pronoun when the context is unknown, or can not easily be understood based on the preceding sentences. For example uncontextualised use of they, them, you, …


* “They say I should go into business, but I don’t know if I have the
* Challenge: “Who is it that says you should go into business?” and/or, “What do you mean by confidence?”

* “Yeah, I have tried alcohol before. It makes you say stupid things.”
* Challenge: “Wait, does it make everyone say stupid things?”

* “I hate watching the Tottenham in the playoffs. We always lose and it
   makes me depressed.”
* Challenge: “By ‘we’, do you mean that you are part of the football team?”

Lost Performative

Lost Performative, makes reference to an action but the person who performed the action is unspecified.


* “Her book was highly acclaimed.”
* Challenge: “Highly acclaimed by whom?” or “How do you know that?”


John Grinder did his doctoral thesis on Noam Chomsky’s Transformational Grammar.

It can also be traced to the nominalistic tradition of William of Ockham.

An effort unrelated by origin but going in the same direction of improving clarity of communication is the constructed language Loglan (and its close cousin, Lojban).

See also

* Problem of universals
* Neuro-linguistic programming
* Richard Bandler – co-creator of this meta model
* John Grinder – co-creator of this meta model
* Fritz Perls
* Virginia Satir
* Transformational Grammar
* List of NLP topics
* Modeling (NLP)
* Cognitive distortions