NLP Practitioner

Meta-Model Distinctions – Deletions

When we communicate with others we assume that what we are talking about is actually understood by the people we are conversing with. We base this assumption on a number of factors;

  • Shared Experience – We are all human beings, living in a world that works in a particular way. Our experience of the world is similar enough to our neighbours that certain things can be taken for granted.
  • Shared Knowledge – Sometimes we will be in a conversation with someone and we know that we don’t have to be overly specific as we both have a certain level of knowledge about what we are talking about.
  • Shared Context – If we are with someone at breakfast, proffering a drink would imply tea, coffee or juice, not beer, wine or spirits. The current context is taken into account whenever we engage with another person.

It is because of this level of shared information that when we interact with our friends and colleagues we can leave out a lot of the detail as we assume that these people understand what we are talking about. Even when talking to strangers we assume that there is a level of common ground and so we often only go in to detail on those things that we believe the other person is unfamiliar with.

The assumption of knowledge leads to Deletions in our communication and it is this deletion of information that often causes confusion and misunderstanding and only by putting back the missing detail can we truly understand what people mean.

The original Meta Model contained three basic types of deletion:

  • Simple Deletion – The removal of a single piece of information. (What)
  • Unspecified Verb – A vague description for a series of actions. (How)
  • Loss of Referential Index – The ‘who’ is missing or are generalised without clarity. (Who)

In addition to the basic types two additional deletions have been added, they originally came from the Milton Model of ambiguous language. They can be considered complex or compound deletions as they are usually made up of a collection of the basic deletion types, and because they occur regularly in our language they have been given a special status within the Deletions of the Meta Model, these are:

  • Lost Performative – Opinion expressed as a fact.
  • Comparisons – An unspecified comparison without evidence.

Simple Deletion

As the name implies this is simply the deletion of a single piece of information from a statement, often replacing the information with ‘it’ or ‘that’.

“It’s not true what they’re saying.” – What isn’t true?

“Can you pass me that book.” – Which book?

“I’m not sure it’s ready.” – What isn’t ready?

The simple deletion is so easy and pervasive that is can be difficult to spot them all within a sentence. Also it can be really annoying if you start to question every single simple deletion as sometimes what is being left out is obvious because of the context.

For example, if someone has passed you a cup of coffee, asking the question “Is it hot?” contains a deletion, but there is implied knowledge, changing this to “Is the cup of coffee you have just passed me hot?” may be more complete, but is more than a little excessive.

Unfortunately, this is a trap that many inexperienced practitioners fall into, where they begin to try and restore all of the simple deletions into a conversation. It’s good practice to listen out for them when they happen, but to constantly as for clarification can be just plain irritating.

Understanding when to add back the information from a simple deletion comes with experience, however it can be useful just to clarify occasionally, during some of the more convoluted conversations one has.

Unspecified Verb

The Unspecified Verb deletion deals with what actions are taking place. It is not uncommon to remove a series of distinct activities and lump them together under one or two words.

Again shared knowledge and context can be used to get to some of the details of what is going on. For example, after returning from a Baseball game, someone says ‘They played well today’, then the verb ‘played’ is fairly well understood, however exactly what that means (did they pitch & field well, get a good number of home runs, were generally good sports etc.) is lost.

When we are talking to clients it is important to clear up what is really going on by getting to the detail of the Unspecified Verb, if someone was ‘hurtful’, what did they actually do to cause hurt?

“Mary was really mean to me.” – In what way was she mean?

“They killed it.” – How did they ‘kill’ it, specifically? What does that really mean?

Unspecified Verbs pop up all over our language and because of their nature they can mean different things to different people, even those with a very similar cultural and environmental context, in the last example the word killed has a number of meanings and the statement could be negative (as in dead) or positive (as in a winner).

The goal of the Meta Model is to gain clarity on a current situation and this can only really be done if all of the information is fully understood, assuming that you know what someone means without seeking clarification is a mistake on the part of the practitioner, no matter how much they know about their client.

Lack of Referential Index

In language terms we what through the Simple Deletion, we’ve found the activity with the clarification of the Unspecified Verb and now we need to identify the subjects, the who by clearing up the Lack of Referential Index, or more simply ‘who are we talking about?’.

Have you ever been in a conversation which sounds a little like this?

“He said ‘no’, and then she said ‘well that’s not right’ and he said ‘I know, but that’s what She said’, and she said ‘I don’t believe it, that’s not what She said to me.'”

What the…! Although this seems like an extreme case of Referential Index Loss it is not uncommon to hear conversations like this as we ‘know’ who ‘he’ and ‘she’ is, don’t we? Ah, and there’s the rub. Often in a conversation of this complexity it is easy to get lost, as the listener, and to attribute actions to the wrong person.

Not only do we generate Referential Index Loss through shared knowledge and context but we also create them implicitly when we assume that our listener ‘knows’ who we are talking about, so you come across ‘He’ and ’She’ when it is assumed that you know who they are, without ever being introduced to the subject in the first place.

In addition a client will create arbitrary collections and in these cases we will come across ‘They’, ’Them’, ‘Us’ and ‘We’ within the language;

“They know everything we do” – They who? Identify them.

“They’re on to us” – Who are they? Who are us?

“We are clear about the next step” – Who are we? Who fall outside the group.



As NLP became more established two additional Deletions were added to the Meta Model; Lost Performative, which will be covered later and Comparisons.

Comparisons are a form of Unspecified Verb type where the verb is a comparison with some unknown quality or quantity.

Typically a Comparison deletion will use words like ‘better’, ‘worse’, ‘easier’, ‘faster’ etc. but they will be missing the comparator. The key to a Comparison within the Meta Model isn’t that the comparison exists but the fact that it is complete.

He’s a worse runner. [Comparison Deletion]
He’s a worse runner than Usain Bolt. [Complete Comparison]

The point here isn’t that comparisons are a bad thing, they’re not, they motivate us to be better. The issue is that incomplete comparisons are bad, they don’t give us anything to measure ourselves against.

“You must be better.” – Better than whom?
“This is easy.” – In comparison to what?
“He’s much quieter now.” – Than when/what?

Often a client will have unrealistic comparisons that they don’t voice, which produces a feeling of inadequacy. In these situations the NLP Practitioner will need to elicit the comparatives from the client, and then use a more appropriate measure.

Internal comparisons can be both the most motivating and debilitating forms of internal dialog. When we create an effective connection between where we are and where we want to be it can motivate us to achieve great things, if on the other hand you take the comparison and make in unrealistic then only problems will ensue.

Lost Performative

Just as Comparisons are an extended form of Unspecified Verb, Lost Performatives are an enhanced version of Lack of Referential Index where the person doing the judging is deleted from the statement. So “You’re not a very good swimmer.” begs the question, “Who says so?”, and that’s the point of a lost performative, a judgement is being made, but unless the person involved is appropriate then the state has no meaning.

Lost Performative statements are very common amongst clients with low self-esteem. “I’m no good”, “I can’t do anything”, “I’m useless”, “I’m a bad person” all make judgement statement without explicitly identifying who is making the claim, often simply asking “According to whom?” (or “Who says so?”) can open up the conversation and identify the faulty thinking in the client.

Of course many of these statements will come from an internal source, they are self-referential and could easily be prefaced with “I think…”, or “In my opinion…”, but we rarely do this as when we do the statement often lose a great deal of their power.

In addition to the internal referenced Lost Performatives we will also hear them in language that we overhear, often they are prefaced with “Obviously”, “Clearly” and the most common of all “Surely”, all of which puts the onus on the listener to contradict what everyone knows because it’s so certain.

With these language patterns the speaker is almost daring the listener to contradict them, and yet there is no substance to the comparison it is purely an opinion served up as Guaranteed, Genuine, 100%, No Money Back Hogwash!

Blame it on the Parents

Unfortunately, many of the internalised judgements start in childhood with well meaning parents making judgment calls on their children. This leads to an ownership problem, because they are opinions coming from an authority figure it is too easy to take the statements as fact and then make them our own.

With careful thought these ideas can be challenged and then broken which often solves many problems with a client.

Adding Ownership and Metrics

To remove a Lost Performative from a statement (if you lose something that is Lost, do you find it?) you need to identify who the owner is and what the metric (actual comparison) is. To do this is relatively simple:

  • Ask “Who?” – Who has made this judgement? What is their authority on the matter? What are their credentials?
  • Ask “How? – How are they forming the comparison? Against what/who are they comparing? What are they using to measure the two sides.

Once the these pieces of information have been returned to the statements it can be easier for the client to evaluate the reliability and accuracy of the comparison, they can then decide whether to keep or discard the concepts as they see fit.

Deletions – A Word of Caution

For the vast majority of NLP Practitioners the Meta-Model is a revelation and deletions, because of their simplicity and frequency, are often the most easily spotted and targeted by the novice.

A litany of “Who, specifically?”, “How, specifically?”, and “According to whom?” are at best repetitive and at their worst, really annoying for the client.

Working with the Meta Model, as with all of NLP should be a subtle activity and overuse of any of the components will render the whole thing useless and the client will feel that they are little more than a ‘lab rat’ running through a maze created by the practitioner.

Carefully select the deletions to focus on and you will be a much more successful practitioner and your clients will appreciate that you really do understand them.

Leave a Comment