The NLP Meta Model is one of the most well-known set of NLP Language Patterns. This model utilizes strategic questions to help point your brain in more useful directions. The Meta Model also helps with removing distortions, deletions, and generalizations in the way we speak.
0:16 History of the Meta Model
1:18 The Goal of the Meta Model
2:14 The Meta Model Compass
3:30 Presuppositions
5:15 Mind Reading
6:07 Lost Performative
7:01 Modal Operators
9:58 Universal Qualifiers
10:46 Cause & Effect
11:33 Nominalization
12:16 Predicates
13:10 Complex Equivalence
14:33 Comparative Deletions
16:04 Non-referring Nouns & Unspecified Verbs
18:33 Frame of Reference
19:03 The Framing Tool
20:02 Inoculation
23:24 Changing People's Minds
History of the Meta Model
Back in the 1970s, Richard Bandler, one of the co-founders of NLP, and Frank Pucelik learned Gestalt therapy from modeling the founder himself, Fritz Perls.
They began teaching it other students at the University of California but lacked the knowledge of the specific patterns being used.
John Grinder, the other co-founder of NLP, helped dissect the specific patterns they were using by first modeling each of them, to get an unconscious representation of the patterns, and then codified it.
They continued their modeling work with Virginia Satir, a world-renowned Family therapist, and noted an overlap with Fritz Perl’s work.
In 1975, The Structure of Magic was published and the Meta Model was introduced for the first time. It was originally intended to be used by therapists, but it can be applied in a wide range of applications including business, personal development, and coaching/consulting.
The Goal of The Meta Model
The Meta Model is simply that, just a model. But for our purposes, we will be using the Meta Model to help others point their consciousness in more useful directions.
When it originally appeared in the Structure of Magic, it was intended to be used by therapists to help their clients.
We do that by treating the map (or statement) as a whole map and asking questions that challenge the main idea behind an utterance or statement.
First, there is the sensory-based map which forms an internal representation of everything we see, hear, feel, taste, etc. from a moment-to-moment basis.
Secondly, there is the linguistic map which is a symbolic interpretation of the sensory-based map.
In the Structure of Magic, Bandler and Grinder outline what they call the “Universal Model Process”, which describes how we form our linguistic maps.
There are 3 distinct elements:
- Deletion - A process which removes portions of the sensory-based mental map and does not appear in the verbal expression.
- Distortion - The process of representing parts of the model differently than how they were originally represented in the sensory-based map.
- Generalization - The way a specific experience is mapped to represent the entire category of which it is a part of.
In the Structure of Magic, Bandler and Grinder make a distinction between “surface structure” and “deep structure”, two terms that they borrowed from transformational grammar.
The surface structure refers to the words or utterances that correspond to the internal representation.
This is where we notice the deletions, distortions, and generalizations that gives us insight into their model of the world.
Traditionally, therapists used the Meta Model to help their clients recover the deep structure in their language or "restore the well-formedness conditions of the surface structure."
The deep structure is the pure experience which the surface structure is based upon. It exists at an unconscious level.
By using the Meta Model challenge questions, we’re able to help someone connect to the deep structure where more resources are present.
The Meta Model Compass
I want you to start thinking of the Meta Model in terms of a compass. This idea was originally formulated by Michael Breen of NLPTimes.
On the north end of the compass, we have generalizations, conclusions, abstractions, and summaries.
This is a high-level overview of someone’s thought.
On the southern end of the compass, we have the sensory-specific information.
This is the kind of information that could be picked up on a video camera, other than taste and smell.
In other words, just the facts.
Every statement that someone makes is relatively specific or relatively abstract in relation to the context that they are speaking about.
By the time you hear what someone has said, billions of processes have already taken place in their brains.
What you hear is a conclusion based on many neurological processes.
On the eastern end of the compass, we have what’s called “inside the map.”
“Inside of the map” is the type of information that a journalist would look for: who, what, when, where, and how.
We can usually infer this information from what someone has said.
On the west side, we have the patterns that relate to what’s outside of someone’s map or model.
These are the linguistic elements from inside the map or model that someone cannot perceive.
We will be using this compass as we go through each pattern to demonstrate how each pattern relates to the compass.
As an added note, there are also patterns that fall on the “backbone” of the compass which is neither inside nor outside of the map or model.
Presuppositions -NLP Meta Model
Every statement that is made is a claim to knowledge.
All of the language patterns in the Meta Model act as presuppositions in the statement that is made.
As a quick exercise, I’m going to give you a sentence and I want you to presuppose what has to be there in order for the statement to make sense.
Here it is:
“The cat sat on the mat.”
Think about it for a sec.
For starters, we have to presuppose there’s an entity called a “cat”.
We have to presuppose there’s an object or entity called “mat”.
Lastly, we have to presuppose there’s a relationship between the two entities called “sat”.
Because presuppositions exist within every single utterance, we’re going to place it at the top of the Meta Model.
In the Structure of Magic, Bandler and Grinder outlined 29 syntactic environments, which are places within an utterance or sentence, where presuppositions can live.
The problem with this approach is that it makes learning the Meta Model seem a lot more complicated than it really is.
For our purposes, we won’t be going over each one, but focus on the general idea of presuppositions.
Mind Reading - NLP Meta Model
A Mind Read is a claim that’s made without stating how you know.
In order to challenge this language pattern, we must first assume that in order for someone to make a particular claim, they have to have a way of making that claim.
Also, if someone is making a claim about something that is less than useful, isn’t helpful, or it’s getting in the way, we can ask someone for the foundation of that claim and make a change in how someone thinks.
Here’s how you would challenge a mind read: how do you know?
Asking them how they know allows us to see where they’re drawing that information from.
Lost Performative - NLP Meta Model
The Lost Performative is the first pattern we look at from outside of the map or model.
It has the same overarching influence as the Mind Read but from the outside.
Challenging this pattern allows you to take a statement from a free-floating reality back to a statement about a specific time, a specific place, and a specific reason.
More often than not, we tend to drop who said it, when they said it, and under what conditions.
When we fail to define the scope of a particular statement, our nervous system treats it as a universal, meaning all situations, and at all times.
Here are a few questions for challenging a Lost Performative:
- Who says?
- According to whom?
- Where did you get that from?
Modal Operators - NLP Meta Model
Modal operators describe the mode of operating in a sentence.
They show us the boundary of the map or model.
You can also learn what motivates someone by the modal operators they use.
There are 3 types we concern ourselves with:
- Possibility - can, could, may, might
- Necessity - should, ought to, must, need to
- Desirability - love, like to
There are a number of ways you can challenge modal operators.
First, there’s the conventional approach.
For example, someone says, “I can’t do something.”
You can respond, “What would happen if you could?”
This kind of question provides us information from outside of the map.
You can also respond with, “What stops you?”
This provides us with information from inside of the map.
Here are some other challenge questions you can ask:
- How do you know that you can’t? (Mind Reading)
- According to whom? (Lost Performative)
By the way, when someone says that they can’t do something, they’re literally saying “I can engage in the act of not doing x.”
Here’s a quick story:
There was a VP of communications from a big company that was afraid of giving talks in public.
One day, he met with an NLP Trainer and told him, “I can’t give talks in front of other people.”
The NLP Trainer asks, “How do you know?”
The VP says, “I gave a couple of talks and I didn’t like how it turned out.”
The NLP Trainer responds, “Ok, so you’ve given a couple of talks and you didn’t like the outcome. You ever gave a talk and it was neither here or there?”
“Yes, plenty of times,” responded the VP.
The NLP Trainer says, “We’re no longer talking if you can or can’t. The question is of quality now.”
If you hear Modal Operators around limitations, use counterexamples or challenge the inferred or implied universality.
We’ll talk more about universals in the next section.
Universal Qualifiers - NLP Meta Model
Universal Qualifiers define the scope of the map.
Some examples include “every”, “all”, and “only”.
There are also times when universal qualifiers aren’t being used explicitly.
Unless someone offers a form of qualification, there is an implied or inferred universality to what they’re saying.
In order for a statement with a universal qualifier to be valid, there cannot exist a single counterexample.
To challenge a universal qualifier (implied or explicit), come up with a counterexample that could be true, based on what they say.
Cause & Effect - NLP Meta Model
Cause & Effect gives us the structure for how the model works.
The basic structure for cause and effect is “if x, then y”.
On the “x” side, we have the sum total of evidence that points to “y” as the outcome.
It could be one thing, or it could be several things, depending upon the belief.
For more information on challenging the cause and/or effect, check out the Sleight of Mouth language patterns.
Nominalization - NLP Meta Model
Nominalization allow us to take complex activities and put it together into one thing so we can think about it.
If you’re unsure if a word is a nominalization, use the wheelbarrow test.
Any noun that you can’t put into a wheelbarrow is a nominalization.
Here are some examples of nominalizations:
- education - educate
- conclusion - conclude
- demonstration - demonstrate
If a nominalization is being problematic, it can prevent someone from taking effective action.
To solve this, put the nominalization back into verb form.
Predicates - NLP Meta Model
Predicates are words that tell something about the subject.
We can use these words to influence what kind of representations someone makes in their mind.
There are 2 main types of predicates: Time/Space and Sensory.
Time/Space predicates allow you to place things in regards to submodalities.
Here are a few examples of each:
- Time - before, look back, happened
- Space - here, there, inside
Sensory predicates are words that imply sensory information.
In NLP, there are 5 major representation systems: Visual, Auditory, Kinesthetic, Olfactory and Gustatory.
Each type of sensory predicate corresponds to one of the representation systems.
Here are a few examples of sensory predicates:
- Visual - review, out of sight, show me, colorful
- Auditory - Outspoken, say, Shout, unheard of
- Kinesthetic - Heartwarming, Firmly, Solid, Gripping
- Olfactory & Gustatory- stinking, fragrant, sweet, stale
Complex Equivalence - NLP Meta Model
A complex equivalence occurs when a person equates a particular, qualitative word and their experience of the world around them.
Here’s an example: “I just missed my appointment. I’m such a disappointment”.
In this example, “missing the appointment” is the same thing as “being a disappointment”.
To challenge a complex equivalence, you simply ask them how they equate the qualitative or descriptive word with their experience of the world around them.
Usually, they will give you a list of things that let them know how those two things are equal.
In that list, there will be some sensory-specific language, but you want to look out for nominalizations.
This is partly because Nominalizations and Complex Equivalences are on the same level of the Meta Model Compass.
Nominalizations describes the bits and Complex Equivalences show how they all fit to mean one thing.
In addition, the Sleight of Mouth language patterns are also great for challenging complex equivalences.
Comparative Deletions - NLP Meta Model
Every evaluation that’s made, from the biggest abstractions to the sensory-specific, comes from making comparisons to other things.
Traditionally, this pattern was used for statements that contained phrases like “better than”, “worse than”, best, etc.
However, you can still use this pattern even if there’s no explicit mention of a comparison.
If there’s no standard of comparison mentioned in the map, you can simply ask about it.
Here’s an example:
Person A: This is the worst sandwich I ever tasted.
Person B: Compared to what? A sandwich from a five-star restaurant?
By asking for a standard of comparison, you’re able to sort out the relative difficulty for that person.
Lack of Referential Index - NLP Meta Model
A referential index refers to the subject of the sentence.
Lack of referential index is a language pattern where the “who” or “what” the speaker is referring to isn’t specified.
Examples include he, she, it, and they.
To challenge a lack of referential index, ask “who?” or “what, specifically?” to gain clarity on what the speaker is referring to.
Non-Referring Nouns & Unspecified Verbs - NLP Meta Model
We will present the last two patterns together because they are interrelated with one another.
Both of them live at the bottom of the Meta Model, in relation to the compass.
The subject-verb relationship is fundamental to the acts of cognition. We have an inherent need to know what something is and what it’s doing.
It’s also the first linguistic structure that children learn to generate.
Non-referring nouns are also called Unspecified Nouns.
When we challenge this language pattern, we’re able to find out more information about who or what is being talked about.
To challenge a non-referring noun, we use the same questions as those in lack of referential index.
The words are the same, but the function is different.
In the case of unspecified verbs, it’s important to note that every verb is relatively unspecified.
To challenge an unspecified verb, simply ask “How, specifically?” to get more information.
Keep in mind that just because you collect more information about a verb, doesn’t provide more understanding about what’s going on.
Here’s an example: Let’s say you’re with a guy and he tells you, “My friend hit me.”
You ask, “How?”
He responds, “She hit me hard.”
You ask, “With what?”
He responds, “With an idea.”
You ask, “why?”
He responds, “Because she was trying to help me.”
It’s not until we ask our third question that we finally realize that the guy is speaking figuratively.
This is why it’s important to start at the top of the model and then work your way down.
Otherwise, you’ll end up collecting a bunch of useless information.
Frame of Reference
In order to change someone’s mind, you have to first consider their frame of reference.
When you’re talking with someone, they’re making references within their own worldview to a particular set of representations, references, and/or ideas.
We call those references or ideas “frames” or “frames of reference”.
If you’re trying to persuade someone, everything you say is filtered through their frame of reference.
In order to change someone’s mind, you need tools to help them shift that frame of reference.
The Framing Tool
The framing tool is another model created by NLP Master Trainer Michael Breen, that allows you to determine what would be the best frame of reference to help someone reframe what they’re thinking.
The framing tool is the Meta Model compass with the added dimension of time.
In the center of the framing tool lies the utterance or statement that was made.
On the north end of the framing tool, we have purposes, intentions, and worldviews.
These are higher levels of abstraction which is directly correlated to the north end of the Meta Model compass.
On the southern end of the framing tool, we have sensory-specific details like evidence and examples.
On the western end of the framing tool, we have the influences or states that led to the utterance or statement.
These are things that must be true or must be valid, in order for the statement to be valid.
On the eastern end of the compass, we have the consequences, or outcomes as a result of the statement or issue.
By working with the framing tool, we can start asking better questions around what frame of reference someone is using.
From there, we can help them reframe what they’re thinking about using the Meta Model language patterns.
Inoculation
As we stated earlier, people will filter what you’re saying through their own understandings, expectations, hopes, and fears.
Unless you find a way to contextualize and make sense out of what’s going on in those filters, you set yourself up for failure or difficulty.
I would highly suggest that before you begin any meeting, consultation, or interaction, that you do an “inoculation”.
To put it another way, you want to bring up anything that could be a potential problem, whether it’s a “thought virus”, an “elephant in the room”, or the like, and create a way for people to deal with it before it becomes a problem later on.
There was an NLP Trainer that was hired to do some work for a big financial firm.
One day, he met with two of the senior executives and they had very concerned looks on their faces.
The executives told him that the group that he would be training had been very unkind to previous consultants.
They said that they “chewed them up and spat them out.”
Later that day, the NLP trainer went to his hotel and took some time to think about what had to be going on in their heads in order to make those behaviors the right one.
He knew that the group had a minimum of 10-15 years of experience.
Some had even ran businesses making $100 million to $500 million per year.
He figured that if someone were to patronize them or lord over them, that could be one reason why they’d be hostile.
He thought of a few more possibilities and went to sleep.
The next day, he met with the managing director and was solemnly walked to the room.
He was then introduced by the managing director and was left alone.
The NLP Trainer started off by saying, “Before we begin, there’s a number of things I’d like to acknowledge. Most of you have far more experience and I’m not here to tell you how to do your job better. There are unique competencies here in this room and each of you is here for a reason. You’re already successful. I’m here to bring you some information you might not have heard of. I also have some things that can help anyone. I’m not asking for your immediate compliance nor believe everything I say. I invite your benevolent skepticism. But, there’s one thing you need to know before we start. I’ve already been paid for this work and for the next 3 days, we can either work enjoyably and have a wonderful experience, or be miserable. It’s your choice.”
As he was saying this, he noticed how they were reacting.
Some of them were smiling, others were nodding their head in agreement.
When he mentioned the part about already being paid, they broke into laughter.
After his speech, he went through his agenda, and they were off to the races.
At the end of the day, the Managing Director pulled him aside and asked, “What did you do differently?”
He responded, “I simply didn’t step on their toes.”
We can also inoculate a situation by using presuppositions on how it will be addressed.
Anything that can’t be put on the table will become a foundation for passive-aggression, withholding participation, etc.
In another story, there was an NLP Trainer who was working with a company that had just released 15 percent of the workforce.
To add insult to injury, the process was also poorly managed.
The NLP Trainer had also heard through the grapevine that a lot of people were still concerned and grieving from the downsizing.
In the introduction to the meeting, the NLP trainer said, “Although 15% of the workforce had been released, some people are still in the process of grieving, all that means is in the work we’re gonna be doing today, we need to be aware that this process is still going on, and we can use this process to help achieve an even better result.”
As he said this, a number of people took a deep breath and relaxed.
The issue was finally out in the open and rather than it being problematic, he told them how it was going to be used to make something better happen.
Another way of dealing with inoculation is if you suspect there’s a certain attitude about a situation, you can create a caricature out of it.
Then, you create an alternative for whatever that attitude is.
For instance, let’s say you’re about to hold a meeting and the general consensus is that people aren’t too thrilled about it.
At the start of the meeting, you can address this by saying, “I know some people are thinking ‘oh no, do we have to go through another meeting?’’ and maybe if it was like other meetings you’ve been to before, that would be the right thing to say. But in this meeting, if you’re willing to go through the process, and give your attention in the way that I’m asking for, I’d like to suggest to you that by the end of the day, when we’re looking back over what’s been accomplished, you’re gonna be quite surprised over how much progress has been made.”
By caricaturing the attitude, you’re signaling to others how it’s going to be dealt with, but you must give them another option or choice.
Changing People’s Minds
Every question that you ask will point someone’s consciousness in one direction or another.
Also, the way that you ask a question will have a radical impact on how the other person understands the question, and where they go to access the information.
You need to think about what category of answers you want and use presuppositions to elicit them out of other people.
Consider these 2 questions:
- If your income doubles, what would you do with it?
- What would doubling your income do for you?
In the first question, you probably thought about the different things you could buy if your income doubled.
In the second question, you probably thought about more abstract concepts.
On a final note, we do not ask questions merely to collect data.
Before asking any question, we must first determine whether the frame of reference we’re using is the correct one.
The correct frame of reference is usually determined by the client. It could also depend on what we’ve been asked to do, as well as our belief on the problem.
We then utilize the framing tool to find a particular frame of reference that’s soluble.
Sir, I just went through your article on ”meta models”.
NLP has come a long way since 2000 when i first had a taste of it.
Thank you for the revisions and updates i enjoyed.
Glad I can help!
Very cool post, Jacob: Table of contents, vídeo, well-written, great images e wrote, congrats!
What I’ve get: I’ve understood the north, south, east, and west functions of the Compass Model, and also I’ve clearly understood each pattern. It is also clear that the model was designed for navigation, orientation and knowing the next step.
Now, I have some questions:
1 – How would I use the compass model in order to do fewer questions or more useful questions? I haven’t understood that very well.
According to the NLPTimes sales page, the Compass Model was designed also to avoid stimulus response questions and many unseful questions.
If a client says “I can’t do that”, how do I navigate the compass model in order to make useful questions? Should I identify this phrase as a ‘modal operator’ and then choose the right (’cause & effect’) or left (‘universal quantifiers’) option in the compass?
2 – Considering the above scenario:
For example, why ’cause & effect’ and ‘universal quantifiers’ are at the same level in the compass? What being at the same level means?
Does it mean that if I identify in the client’s speech a ’cause & effect’ pattern then should I ask a question that challenges ‘universal quantifiers’?
3 – About ‘inside’ and ‘outside of the client’s map:
I didn’t understand why ‘lost performatives’ and the other patterns on the left of the Compass were positioned there.
If a client says “All the men are not trustworthy”.
This is a big generalization using ‘universal quantifiers’.
Shouldn’t this pattern be positioned on the right side since it is part of the internal representations of the client?
Shouldn’t questions about accuracy (who, what, why, how …) or questions about the client imagining possibilities positioned in the ‘outside’ (left) of the Compass Model since these are things that the client has not perceived yet?
Hi Max,
1. If a client says I can’t do that, I would generally recommend starting at the top of the model.
Asking the Mind Reading question is normally a safe bet.
2. Same level means that they are on the same level of abstraction.
As stated in the beginning of the post, the north and south end of the compass represents how abstract or sensory-specific a given statement is.
In regards to Cause and Effect, I recommend looking at the Sleight of Mouth language patterns.
Universal qualifiers are normally best challenged by using a counterexample.
3. Lost performative is outside of the map because the information that’s missing cannot be inferred from the statement itself.
Lost performative is also a type of distortion pattern where a person misrepresents external reality.
In order to correct this distortion, it requires pulling in information from outside of the map so that it can be adjusted accordingly.
You are correct in stating that a universal qualifier is a type of generalization.
As with the lost performative, generalization involves a misrepresentation of reality and requires information from outside of the map in order to bring it back into alignment.
I’m not sure how questions about accuracy would be outside of the map.
As stated in the post, inside of the map means information that can be inferred based on what someone has said.
We can usually infer undisputed facts based on what someone has said.
On a final note, I don’t believe the Meta Model compass is infallible.
If you haven’t already, I recommend checking out the Meta Model Revisited, which provides a much more comprehensive overview on how the Meta Model is ordered.
To the best of my knowledge, the Meta Model compass is not an official ordering.
It is fairly useful, but it is lacking a linguistic foundation, which is addressed in great detail in the Meta Model revisited.
Hee’s the link to the post: https://pcialpha.com/nlp/nlp-meta-model-revisited/
Nice video for reference…. Could I ask where you got the meta model framing tool chart?
Excellent chart. I know it is in some study materials, yet cannot seem to find it online.
Fun material, difficult to monetize into coaching income. Good Luck.
Thank you
Hi Robert!
The framing chart came from Michael Breen of NLPTimes.
Fantastic article on the Meta Model guys, loved the presentation and the sheer amount of ground you covered!
Congrats and keep up the great work.
Thanks Mike, appreciate the kind words!