A further response in the ongoing dialogue between Shawn Carson and Steve Andreas, regarding the HNLP Meta Pattern, (including comments from Connirae Andreas) (round 5b)


         In this post I’ll respond to Shawn’s claim that “all human change is based on the HNLP Meta Pattern” and “all NLP patterns follow the Meta Pattern.”

         Any general formulation such as the Meta Pattern will always be incomplete, and have omissions and implicit presuppositions. That is what makes a general formulation useful. However that incompleteness also provides opportunities for misunderstanding.

         First I’ll point out what I think are some of the significant omissions in the Meta Pattern, and then provide examples of interventions that don’t fit the four steps of the Meta Pattern.

         Finally, at the end, I’ll offer an alternative way of understanding the different structures of change, a reorganization of the different patterns of reframing.


(Below are links to the Previous Exchanges.)

Steve’s Original Article 

Shawn’s Response

Steve’s Response to Shawn

Shawn’s Response to Steve

Steve’s partial response to Shawn


Shawn wrote:      

         Finally the Meta Pattern…

In my blog post on the Swish (which itself was in response to Steve’s original blog post), I explain that the second step of the HNLP (Humanistic Neuro Linguistic Psychology) ‘Meta Pattern’ (not the Meta Model!) dissociates the client from their problem state. And because all human change is based on the Meta Pattern, all NLP patterns follow the Meta Pattern, and therefore all contain a ‘break-state’. Steve argues this is not true, citing a number of examples, the most intriguing of which is the Compulsion Blowout that clearly doesn’t contain a break-state…

…or does it?

Origination of the Meta Pattern

The Meta Pattern was observed and formulated by John Overdurf. Note that it was not ‘created’ by John, any more than relativity was created by Einstein; it was ‘discovered’.

Steve says that he learned the Meta Pattern from Bandler. This may be the case, but I doubt it for a couple of reasons. Firstly John Overdurf told me he developed it and strictly attributes to others where appropriate; so John Overdurf did not learn it from Dr. Bandler (despite being a Master Trainer under Dr. Bandler). Therefore if Dr. Bandler was previously teaching it, he must have stopped teaching it to his Master Trainers before John Overdurf trained with him.

Secondly, John Overdurf discussed the Meta Pattern with John Grinder and Grinder gave no intimation that it was something he had taught before. So either Grinder wasn’t present when Bandler taught it, or he had forgotten.


Connirae Andreas joins the discussion:

         Shawn, I’ve heard a lot of good things about John Overdorf, so I believe what you are saying about his integrity, and I also respect that you’ve learned a lot from him. This note is about Steve’s saying we learned a “meta-pattern” from Bandler’s group. I’d like to tell you my experience of this.

         This was a long time ago, so some of my memories are a bit vague. The part I remember clearly was that Steve and I sometimes taught “a” meta-pattern in our Master Practitioner Trainings. I remember teaching it as part of some of the advanced language patterns sections. Early on Steve and I taught that together, and then I was usually the one who taught that section. We taught the “meta pattern” as part of understanding how to do “conversational change.”

         The “meta-pattern” that we taught had 4 basic steps: Stage 1: The “client” starts out being associated into the problem experience, e.g. “So you are problem-ing…” Stage 2: Dissociate the person from the problem in the present, and place the problem in the past. “And this has been a terrible problem, hasn’t it?” Stage 3: Elicit the “solution” or resource as dissociated future possibility “What would it be like,… when you have all the resources you need for things to go in a much different way (voice tone indicating pleasure)…” Stage 4: Associate the person into the resource, in the present. “…Now!”

         So the essential aspects of understanding this meta-pattern are a) time frame: being aware of the time frame and shifting it as appropriate, and b) shifting association or dissociation, plus c) shifting experience from problem to solution. (This example is just for illustration – there would be many ways to do it which could be much more detailed and thorough.)

         This meta pattern is quite similar to the one you describe John Overdorf mapping out, although there are a few differences. I had thought we learned this meta-pattern from the Bandler group explicitly. However, it’s possible that this formulation was what we came up with in trying to understand Bandler’s demonstrations of conversational change. Steve and I went over examples together, and also Cathy Modrall (who was studying with Bandler at that time) shared some additional examples with us, and we were all trying to figure out the structure. I don’t know if Bandler ever taught it formally, but he was systematic and I think it was quite clear he was following that structure.

         We were teaching it as one key way to organize presuppositions and language patterns to make change possible. There were other ways that I used conversational change that didn’t involve that sequence.

         And also, as it sounds like John O. did, we presented this as a fundamental sequence through which change happens through the NLP formats. And how when one understands this underlying structure, one can organize presuppositions and language patterns to create change.

Re: John O. being in Bandler’s trainer training and not being taught this. That doesn’t surprise me at all. Bandler never was very systematic about training. He wasn’t interested in teaching people a systematic or “complete” NLP experience. He was always more interested in teaching whatever he felt like teaching—or exploring. This often annoyed some of his colleagues, who were trying to offer a more “complete” or “systematic” training.

Steve and I often took on the role of making things more thorough, complete, and systematic, so that people would learn what (at that time) was “the whole thing.” As I mentioned earlier, we didn’t learn this meta-pattern from Bandler in any training, either.

It’s also not a surprise that Grinder had never heard of the meta-pattern. When Bandler started demonstrating conversational change, that was already quite some time after he and Grinder had stopped teaching together. Even earlier when they showed up to teach in our practitioner trainings in ‘79 and ‘80, each wasn’t aware of what the other had taught. The conversational change and “meta-pattern” was happening after the time that the “Big 4” (Richard, John, Leslie & Judith), went their separate ways. From then on they really didn’t know what each other was doing. If I understand the timing right, John O’s trainer training was after the “Big 4” had split up.

This “meta pattern” was evident in the change personal history pattern early on. But we didn’t teach it as a meta-pattern until the ‘80’s sometime. I’d have to look up more specific dates. It was when I was doing conversational change, and somewhat before that.

After writing the above, I noticed we’d published a version of this “meta-pattern” in our book, Change Your Mind—And Keep the Change (1987).

“All NLP involves accessing and resequencing experiences in time. One very broad general formulation of change work is that you start with a problem state, and then identify and access an appropriate resource state. Finally you install the resource state so that it’s triggered in response to the same cues that previously had been cues for the problem state.” (p. 28)

And on Page 29 of Change Your Mind there is a more detailed 5-step version of this “meta-pattern”:


“1. If I have a problem I can’t solve, I am associated into the problem state, in the present. I may have no awareness of the desired state.

“2. The first step toward change is for me to think of the desired state, dissociated, as a possibility in the future. So I see myself acting resourcefully in the future.

“3. Next I dissociate from the present problem state, and associate into the future resourceful self. Now I am in the future, with resources.

“4. From this future vantage point, I can see the old problem behavior in the completed past.

“5. Now I can collapse the future “now” with the present “now” so that I experience the resources in the present, and the problem as over in the past.”


         As Shawn mentions, some of the above steps need to happen before other steps. However some can happen in a different sequence. For example, one can dissociate from the problem state before thinking of the resource, OR one can think of the resource first, before dissociating from the problem state. We taught with examples both ways.


Shawn wrote:

Thirdly, Steve doesn’t understand the Meta Pattern (I’ll explain why in a moment) so if Steve learned it from Bandler, then Steve has also forgotten it. Now obviously Bandler (and Grinder) taught applications of the Meta Pattern (because all NLP patterns follow the Meta Pattern), for example ‘The Structure of Magic’ is a clear application of the Meta Pattern. So I suspect Steve may be seeing the Meta Pattern in something Bandler taught and thinking that was the Meta Pattern.


Steve Andreas comments:

Though not so named, the four steps in the Meta Pattern are explicit in Bandler and Grinder’s early presentation of the “change personal history” pattern in the 1979 book, Frogs into Princes (pp. 82-86) which I believe is some years before Overdurf got involved in NLP. That was the basis for my saying that I learned it long ago. The Meta Pattern was a fairly crude early outline that was useful but oversimplified—something that typically happens in the early stages of any field. Subsequent discoveries and developments have revealed a number of important distinctions and exceptions to the Meta Pattern, some of which I’ll discuss below.


Shawn continues

Meta Pattern Overview

Overdurf noticed that all human change follows four basic steps. These steps are associating into the ‘P/S’ (think of P/S as the ‘problem-state’ for now), dissociating from the P/S, associating into the R/S (think ‘resource-state’ for now), then ‘collapse’ (think ‘collapsing anchors’ for now). Overdurf called these four steps the ‘Meta Pattern’ of change. Note that Steve said I was over-generalizing by saying all NLP patterns follow the Meta Pattern, here I’m over-generalizing even more by saying all human change follows the meta-pattern!

Let’s represent this pictorially in the case of the Collapsing Anchors pattern:



Steve writes:

I agree that the “collapsing anchors” pattern (which has the same steps as “Change Personal History,” mentioned earlier) is a reasonably good fit for the Meta Pattern.

         However, even here the fit is a clumsy enough that I don’t think the Meta Pattern alone will be a practical guide for practitioners. For example, let’s look at step 4 “collapse anchors.” If you want to collapse anchors, you need to have 2 states to collapse. However, in Shawn’s 4 steps, there isn’t a step to reelicit the problem state so that you have something to anchor and then collapse with the resource state.

         Also, there is a significant ambiguity about what to “collapse”; the resource state could be collapsed with the problem state, or with the trigger for the problem state, and the results of each will be different. Either the problem state or the trigger has to be re-elicited before it can be “collapsed” with the resource state, but there is no such re-elicitation step in the Meta Pattern.

         The Change Personal History pattern itself has more precision in it. You elicit the problem state, do a break state (dissociation) then ask, “How would you like to feel in this situation?” You access and anchor this resource, then say, “Take this resource, into that situation, and notice how things are different.” (collapse)

         I have never liked using the word “collapse” as a metaphor to describe combining two states. What, exactly “collapses”? When two states are elicited in a brief time frame, they can combine in a variety of ways. For simplicity, I’ll only mention two of the most common useful ones:

  1. Integration, in which two states simultaneously blend into a third state containing elements or aspects of both.
  2. Chaining, in which the two states remain unchanged, but one state I linked to the other sequentially.

         Although I have been using the word “state” in this dialogue it is helpful to realize that the word is ambiguous. Some people use the word “state” to mean emotional feeling alone, while others would include all the different sensory elements (VAK) of an experience in a moment in time—but without specifying any of them. So it’s good to keep in mind that the word “state” is a sort of “place holder” for a package of ignorance.

         Shawn’s Step 3, “associate resource state,” specifies association; however in the swish pattern the resource is dissociated, not associated, in order to create motivation. This is an important counterexample to the Meta Pattern.

         Now let’s examine the NLP fast phobia cure. Here the second step is to dissociate from the problem state. This dissociation is also the resource state, rather than a separate “step” as in the Meta Pattern. And in the phobia cure there is no collapse step, another significant counterexample to the Meta Pattern.

         One can find even clearer examples of change that don’t fit this meta-pattern in conversational change work. For instance, often a client will say that they feel bad because they should be able to do something about a problem, presupposing that they can. At another time they also say, “I can’t do anything about it,” often because it is simply not under their control. If you say to them, “How is it possible that you should be able to do something you can’t do?” you’ll often see a “smoke coming out of the ears” response, as they realize they have been torturing themselves by thinking that they should do the impossible. When the “can” and “can’t” presuppositions are brought together in the same sentence in a moment in time, the two opposite presuppositions “collide,” and one of them has to go—usually the “should” that has been making the person’s life “shoulddy.” The two states don’t integrate; neither do they chain; one of them vanishes. If you think of the “should” as the problem state, and the “can’t” as the resource state, putting them in the same sentence simultaneously combines the two, creating a contradiction. This might be an instance in which the word “collapse” is appropriate, though only one of the “states” actually collapses.

         Another example. In the visual squash pattern, the two parts of a conflict are both seen dissociated, but in the next step the resource state is created by the integration of the two parts—each of which is part of the problem state conflict. This is a collapse (step 4 in the Meta Pattern). So in this case the collapse (step 4) comes before the resource state (step 3).

         Furthermore, before the collapse there is a step in which the meta-outcome of each part is elicited. The Meta Pattern is a fairly good fit for simple reanchoring when there is no positive outcome being served by the problem state—what used to be called “first order change.” But when the problem state has an outcome (which is most of the time) resolution requires that it be satisfied—what used to be called “second order change.” The Meta Pattern doesn’t provide a step to include either an outcome, or a meta-outcome in the choice of a resource state to bring resolution.


Shawn continues:

However, the Meta Pattern is not a pattern in the sense of the Swish or the Compulsion Blowout. Rather it represents the structure of all change work, so Steve’s comment that I am introducing “steps” is not really correct. What is important are the transitions contained within the Meta Pattern, i.e. moving from one step to another.


Steve’s comment:

         Shawn writes “one step to another” just after writing that, “Steve’s comment that I am introducing ‘steps’ is not really correct,” a puzzling contradiction. Shawn continues to use the word “steps” in what follows below.


Shawn continues:

Let’s take a look at a typical client session:




Perhaps first I build rapport with the client, and perform an intake that asks about their hobbies, interest and favorite places (among other information). This is accessing resources, so I am starting on ‘step 3’ of the Meta Pattern.

Now at some stage, I am going to have to ask the client what they want to work through. You can think of this as a mini ‘collapse’ if I have succeeded in making them feel secure (say) with me, Step 4 of the Meta Pattern. I am then going to begin associating my client into the problem, perhaps by saying “Tell me about the last time this happened, where are you…”. We have finally reached Step 1 of the Meta Pattern.


Steve Comments:

         Asking the client “what they want to work through,” is not a collapse, it elicits the problem state, which could be either associated or dissociated.

Alternatively, you can skip eliciting the problem state entirely, by asking for the desired outcome (resource state) in a particular context. If the resource state is elicited behaviorally, step 4 (collapse) becomes unnecessary, because the resource state is already associated with the cues in the problem context.


Shawn continues:

We then begin some change work, which itself will follow the Meta Pattern through any number of iterations.


Steve comments:

Any process with iterations requires an exit in order to avoid infinite looping, but there is no exit in the Meta Pattern, as there is in the TOTE (Test, Operate, Test, Exit) model of Miller, Galanter and Pribram’s book, Plans and the Structure of behavior. All NLP processes need to have a test of the intervention, in order to know when it has succeeded, and stop the iterations. In Shawn’s diagram, there are arrows between all four circles, with no indication of when to stop cycling.


Shawn continues:

So the Meta Pattern is not a fixed sequence of steps like in Steve’s interpretation of the Swish, it is the ‘dance’.


Steve comments:

         Firstly, I didn’t present an “interpretation” of the swish pattern, I presented it exactly as Bandler originally taught it, without interpretation.

         More important, when Shawn says, “the Meta Pattern is not a fixed sequence of steps,” that is somewhat ambiguous. Shawn’s diagram has arrows from one circle to the next, clearly indicating a sequence from one step to another, and not the reverse. “Not a fixed sequence of steps,” could mean that you can start anywhere in the diagram and follow the arrows.(Shawn’s example below is consistent with this meaning.) Or it could mean that you can skip steps in the sequence. But if it means that you can do the steps in any order, then the arrows in the diagram are misleading.


Shawn continues:

It simply indicates to the coach or hypnotist what the next step in that dance ought to be, but does not spell out how to take that step. As an example, if you have your client feeling good, perhaps laughing (Step 3: R/S) and you as the coach don’t use that opportunity to go to Step 4: Collapse by saying “Hey, what was that problem you used to have” with a twinkle in your eye, then you as the coach have wasted an opportunity to shift the problem. If you want to see this approach in action watch Nick Kemp do provocative therapy!


Why Does the Meta Pattern Work?

To understand why the Meta Pattern underlies all change work, you need to first understand the nature of a typical client problem. In HNLP we say there are four types of problem: something a person feels, which they don’t want to feel; something they don’t feel, which they would like to feel; something they do, which they don’t want to do; or something they don’t do, which they would like to do.


Steve comments:

Since it is common to divide experience into three aspects, thinking, feeling and doing (or internal computation, internal state, and external behavior) I would add thinking to this list, as in obsessive thinking, or lack of ability to think in certain ways—though of course thinking can be described as a kind of doing.

         More important, whether the problem appears to be one of thinking, doing, or feeling, it is the resulting unpleasant feeling that motivates the client to seek help.


Shawn continues:

Now, a ‘state’ only lasts (let’s say) about a minute if left to its own devices; that is the amount of time it takes for the chemicals that wash through the body. So if I can only feel bad about my problem for longer than a minute if I ‘throw logs on the fire’, meaning by thinking about the problem, making pictures of the problem, saying nasty things to myself and so on.


Steve comments:

The foregoing appears to equate “state” with emotional feeling, omitting other (VAK) aspects. I agree that longer lasting states are maintained by some process that repeatedly elicits them. However, that is true of many very problematic states, such as depression, moods, and anxiety, so Shawn’s Meta Pattern solution below would not apply to those.


Shawn continues:

Otherwise my bad feeling is pretty much over in one minute; not much of a problem, right? This is how most of us deal with most of our problems:


Steve comments:

This may be adequate for trivial problems, but even when that is true there is no permanent resolution of the problem, only an avoidance of the problem. And quite often the “decide what to do” step chooses an activity that is harmful to the person, for instance drugs or other risky distracting behavior, hardly a good solution.




Shawn continues:

But I could use the negative feeling to mess my life up, by assuming the feeling is just going to get worse and worse and last forever.


Steve comments:

Forecasting a troubling future (usually with an internal voice) can certainly be a problem, but avoiding a feeling by finding something else to do can be just as problematic. An unpleasant feeling, like a “trouble light” on the dashboard of your car, indicates that something needs attention; ignoring it is generally not a good idea.


Shawn continues:

I might therefore avoid the context of the feeling (“I can’t speak in public because I get scared”). There is no change because the Meta Pattern is not complete.




Steve comments:

I think it is more accurate to say that there is no change because there has been no intervention of any kind, Meta Pattern or otherwise. I agree that avoidance prevents change; previously Shawn advocated avoidance in order to “decide what to do” and “get on with it,” but now he presents it (accurately, in my view) as “problem avoidance.”


Shawn continues:

Alternatively, I give in to the feeling by doing something I know I shouldn’t (as in a compulsion). Again either the Meta Pattern is not complete, or even worse I install a meta-problem by feeling bad about my addiction…

Dr. Jeffrey Schwartz and Compulsions

There is plenty of research showing change follows the Meta Pattern. For example Dr. Jeffrey Schwartz successfully treated OCD simply by getting his clients to say, “That’s not me, that’s my OCD” whenever they experienced their compulsion, and then go and do something else, perhaps gardening say.


Steve comments:

         Firstly, Schwartz’s instruction to the client to dissociate from the OCD is very crude (“take a step back”) and secondly he also tells clients to think of the OCD thoughts as “not real,” which may be even more useful. Finally, Schwartz’s method takes a lot of conscious effort, and a very long time. I have written a blog post in which I go into much more detail about his work.


Shawn continues:

They would then be asked to notice how much better it is for them to be gardening than compulsively checking that the front door is locked.


Steve comments:

Even if/when “doing something else” works, the problematic feeling of compulsion is left unchanged. That leaves the symptom intact, in contrast to changing the cause of the symptom so that the symptomatic feeling is no longer elicited. This fundamental error is also true of many other approaches, such as “stress reduction” or “anger management.” Rather than changing the process that causes the stress or anger, interventions are directed at modifying the resulting symptoms. In medicine, prescribing an aspirin for a fever caused by an infection is considered malpractice—unless there is no known effective treatment for the underlying cause.


Shawn continues:

By the way, this is an example of the Meta Pattern. Take a look and you’ll begin to see how the Meta Pattern does indeed underlie all human change.


         This is the fifth time Shawn has made this assertion. Repetition does not establish truth; this is an informal logical fallacy, sometimes called “argumentum ad nauseum,” for good reason.




Shawn continues:

Ultimately both feelings and behaviors are driven by ‘states’, these problems boil down to the client going into an inappropriate state in a particular context, and this state lasts one minute left to its own devices. I have been talking about ‘states’, so what is a state? A state is a combination of emotional biochemistry, thoughts, behaviors (including physiological behaviors), and sense-use (what I focus on when I’m in that state, and how I focus).


Steve comments:

Now Shawn is using a much more inclusive definition of the word “state.” As I described earlier, the word “state” is a kind of “place holder” for a package of ignorance, unless all the elements that Shawn mentions (and more) are specified, which is very difficult. Fortunately, many very effective process interventions are directed at only one or two elements of a state. For instance, “Expand your view to include what surrounds what you see,” can elicit “the big picture,” which often results in a more balanced emotional response. In this example there is no dissociation step, the instruction elicits the resource directly, and there is no collapse. When the client sees the larger context, they often spontaneously have a much more useful response.


Shawn continues:

How to Install a Compulsion

Dr. Schwartz explains how a compulsion is installed in the first place. I’m writing this at Hypnothoughts Live in Vegas so let’s use gambling. I think “Vegas Baby!” and I go into a state, including a desire to play some Blackjack. Then I ‘add logs to the fire’, I say, “I’ll have so much fun”, and I make a picture of flashing lights and pretty waitresses so I go to the tables. I win a hand, I get a shot of dopamine, “This is awesome”. Then I lose a hand, that’s no fun and I get a dopamine crash. So I picture myself winning again and get a shot of anticipatory dopamine.


Steve comments:

I believe the current understanding is that all dopamine release is anticipatory, rather than a reward following a behavior. It is that anticipation that keeps people at the gaming tables; the occasional win only keeps the anticipatory images refreshed—and it is those images that create and maintain the state of anticipation.


Shawn continues:

I bet again and win, more dopamine… Ultimately I become addicted to my compulsion, because my brain gets addicted to the dopamine rush.


I don’t know exactly what happens neurologically. But experientially what creates the compulsion is a change in one or more submodalities used to represent the image of the object of desire, not the content represented. Often this will involve an increase in size, closeness, or zooming in, height in the visual field, brightness, etc.


Shawn continues:

Note that installing a compulsion cause me to change (obviously); I didn’t have the compulsion before and now I do. As I said above, all human change involves the Meta Pattern. The dopamine crash following my loss becomes the ‘P/S’, putting that loss behind me to see myself win the next hand becomes the ‘Dissociation from the P/S’, the dopamine rush I get when I win is the ‘R/S’, which is ‘collapsed’ onto the idea of Blackjack.


Steve comments:

Since the dopamine rush is anticipatory, the description above has to be incorrect. Dopamine doesn’t surge in response to a win, or crash in response to a loss. It surges in anticipation, and sinks when behavior commences—before there is a win or a loss as a result of the behavior.




Shawn continues:

Compulsion Blowout

Now we are ready to explore the Compulsion Blowout. For those unfamiliar with the pattern it involves taking a compulsion and actually making it worse. Make it bad enough, and it ‘blows-out’, it essentially disappears.




Steve comments:

The feeling of compulsion disappears, but the representation doesn’t; there is a spontaneous change from submodalities of reality to submodalities of unreality, as a result of a submodality passing a threshold.


Shawn continues:

How do you make a compulsion worse? Well, the easiest way is to get your client to see a therapist for 10 years to talk about it (just joking), alternatively you can amp up the submodalities of the compulsion; that’s a little quicker!

Obviously this compulsion blow-out does not contain a ‘break-state’, a dissociation, right?

Well actually, it does. Remember a state will last one minute unless I throw logs on the fire. My brain may make a picture big, for example, as part of its ‘log throwing’, but not too big. By asking a client to crank up the submodalities of a problem beyond this threshold, the client’s thoughts are taken out of the pattern normally associated with the problem, and are distracted from running other parts of the problem pattern. This is a mental dissociation; the client is dissociated from their usual thought patterns. Starved of further fuel, the state runs for another minute then disappears. The disappearance of the feelings associated with the compulsion (the ‘blow-out’), is the second part of the dissociation; dissociation from the feelings.

Now, you may argue that we are doing more than just breaking the client’s normal thoughts patterns when we super-amp the submodalities, and that may be true. But either way we’re now at step two of the Meta Pattern.


Steve comments:

I agree that the compulsion blowout creates an experience of crossing a threshold. After the threshold is crossed, the image will be in submodalities of “unreality.” (Schwartz asks his compulsive clients to think of their compulsion as unreal, but he doesn’t tell them how to do it.)

         But I disagree that this process involves dissociation. If a client’s compulsion, for instance, involves a martini that is 12” tall, he is asked to continue to see it and then to rapidly increase its size to be “larger than the known universe,” in order to cross a threshold. This is not dissociation; if anything it is an example of “throwing logs on the fire,” increasing the feeling response until the size exceeds a threshold for “reality.”

         With a driver submodality that doesn’t have an infinite range, you have to use a different process, called the ratchet method, which uses a repeated increase of a submodality, for instance, zooming in repeatedly (described in our book, Change Your Mind, chapter 5). Both methods are fully associated, contrary to what Shawn asserts. Furthermore, there is no step 3 or 4 (associate resource state, collapse). The compulsion blowout is one of the few NLP patterns in which a problem response is eliminated without eliciting a substitute resource state. That is why it is necessary to follow it with a swish, in order to provide that. Shawn’s follow-on comments below seem to agree with this need.




Shan continues:

If that was all the coach did, I very much doubt the change would take.


The change will “take,” but you can’t predict what new response will take its place.


Shawn continues:

Any good coach will associate their client into how they will be without the problem e.g., “How will you be as a person when this is no longer an issue?” (remember this can be done at the start of the session – the Meta Pattern is not restrictive). And will collapse the outcome state (at least in the testing phase) into the trigger, “Now as you’re feeling free, see those losers around that blackjack table and realize what that freedom means to you and your family”.


Steve comments:

I don’t think it’s useful to install superiority and disdain for “those losers”; most people are already burdened by many judgments of others—as well as judgments about themselves. Judgment is a problem, not a solution.




Shawn continues:


The problem with thinking too strictly in terms of NLP ‘patterns’ without understanding how the patterns are working (e.g. within the Meta pattern of change) leads to inflexible thinking (in my opinion). Keeping the Meta Pattern in mind (or some other frame that is larger than just the NLP pattern being used), facilitates the coach-client ‘dance’, leading to greater flexibility (i.e. the Law of Requisite Variety).


Steve comments:

         I totally agree that understanding how interventions work, and having flexibility in applying them is important—something that I emphasized in my original post on the swish. In cooking, if you know that the function of vinegar in a recipe is to provide sour flavor (or to react with baking soda) you can substitute lemon juice, tomatoes, etc. The alternative is to apply a pattern by rote without understanding it, which leaves you helpless when something unexpected happens. However, it still makes sense to follow a recipe that is known to dependably get the results you want, and I don’t see how trying to force all interventions into a single Meta Pattern increases flexibility in the “dance” with a client. (Procrustes had a bed like that!)


Shawn continues:

Thank you Steve for continuing this dialogue.  It is a pleasure to engage in such thought provoking discussions.  Exchanges like this benefit all of us, including the greater NLP community.


Steve comments:

Thanks to you Shawn, for raising issues that I would never have thought of on my own, and for taking the time and effort to respond to my posts. Responding to your views has been a unique opportunity to think about them and clarify my understanding. I’ve hoped that others would also chime in and offer additional views, but perhaps you and I are the only ones reading these.   ;-)



         Every way of generalizing about a data set will focus attention on certain aspects and ignore others. People talk about “thinking outside the box,” but I don’t think that is possible. However, you can expand the scope of the box, and you can think inside a different box, and that is what the different patterns of reframing do.

         I view the Meta Pattern as a schema that is useful for beginning students to understand the basic structure of simple anchoring formats for problems that don’t involve outcomes. But there are too many counterexamples for me to agree that, “all human change is based on the Meta Pattern, all NLP patterns follow the Meta Pattern,” as Shawn repeatedly claims. Rather than present yet more counterexamples, I’ll offer an alternative outline of how all change can be understood as the consequence of three experiential variables. Few—if any—of the interventions below follow the four steps of the Meta Pattern.


Reframing Patterns: A Reorganization

Steve Andreas

Every reframing pattern changes one or more of the following: a scope of experience in time or space, the categorization of a scope, or the logical level of categorization. This organization helps you understand how all the different reframing patterns are related, what kind of change of experience will result from each, and points out ambiguities in earlier presentations of reframing patterns. Whenever a pattern has previously been named (for instance in Dilts’ “sleight of mouth” descriptions) that name is used. Dilts lists 14 different patterns; the list below contains 26 patterns, but some are different names for the same kind of scope/category distinction, and some differ only in content. The number of patterns is not written in stone; that depends on how specific you make distinctions in creating categories. A simple sentence stem is used to exemplify each intervention to make it easy to distinguish differences (sometimes this restriction results in somewhat awkward sounding sentences).

  1. Change of Scope:


Expand frame (larger scope) “And the larger context around that is. . . ?”

Shrink frame (smaller scope) “And part of that is. . . .”

Change frame (different scope) “And something entirely different than that is . . . .”

Perceptual Position (self, other, observer) “And how s/he would see this is. . . ?”


Prior cause (earlier scope) “And that’s because. . . ?”

Consequence (later scope) “And the result of that is. . . ?”

Expand frame (larger scope) “And if that still picture were expanded into a movie. . . .” Shrink frame (smaller scope) “And the most significant part of that event is. . . .”

Change frame (different scope) “And a very different time is. . . .”

  1. Change of Categorization: (at the same logical level)

Redefinition or Redescription “And how else could you describe that. . . ?”

  1. Change of Logical Level of Categorization:

       Eliciting a more general category (higher logical level)

Meta-frame (The prefix “meta” alone has been used ambiguously in the past to indicate either scope or category, but “meta-frame” has usually indicated a shift to a more general category, rather than a larger scope.) “And that is an example of. . . ?”

There are many such meta-frames. Some of the more useful and well-known ones that have been described previously are listed below:

Positive Intent “And his/her positive intent is. . . ?”

Model of the world “And so the way you see it is. . . ?”

Learning “And what you learned from that is. . . ?”

Curiosity “And what was most interesting to you about that is. . . ?

Hierarchy of criteria “And what is more important to you than that is. . . ?” Analogy/Metaphor “And that is like what. . . ?” (Metaphor creates a category, and often also creates a prototype example that represents the category.)

       Eliciting a more specific category or an example (scope)

Category to more specific category “And that is what kind of. . . ?”

Category to scope And an example of that is. . . ?”

Counterexample category (Category to specific category with negation) “And examples of when that isn’t true are called. . . ?”

Counterexample scope (Category to scope with negation) “And an example when that wasn’t true is. . . ?”

       Looping between category and example, or between category and more specific category (the category includes itself as an example) These patterns are seldom applicable, but very useful when they are, because they are logically “airtight.” Both of these loop between logical levels;

Apply to self (applying a category to itself) “And is that true of what you just said?” “You said that you hate complaining; is what you said a complaint?” (See Six Blind Elephants, volume 2, chapter 5)

Paradox (apply to self with negation) “You said, ‘I won’t communicate with you,’ but what you said is also a communication.” (See Six Blind Elephants, volume 2, chapter 7)

       Ambiguous Reframing Patterns (in addition to, meta -frame, or meta above) Each of the categories below is an example of one of the previous categories described.

Outcome “And the outcome of that is. . . ?” An outcome can be either a scope of experience (a specific new car) or a category of experience (status). Asking about an outcome could shift from one scope or category to another, or from scope to category, or vice versa (four possibilities).

Another Outcome Just as an outcome is ambiguous, another outcome could also yield the four possibilities listed above.

Meta-outcome (outcome of the outcome) Again, asking about a meta-outcome could also yield the four possibilities listed above. When the prefix “meta” a used in other ways, it is also ambiguous in regard to scope and category.

“Chunk down” can mean either going to a smaller scope or to a more specific category. “Chunk up” can mean either going to a larger scope or to a more general category.

Reality Strategy “And the way you know that is. . . ?” asks for the evidence (the epistemological basis) for their experience. The responder may tell you a category (“That is one of the things my parents told me.”) or a scope of experience (“I saw it happen,” or “It’s in the Bible.”).



There are many ways to use the information in this very brief outline; here are just a few:

You can write down examples of what you have said to clients, and discover which patterns you typically use. The ones that you don’t use indicate ways to expand your range of skills and flexibility.

Better yet, make a transcript of a short segment of a session with a client, and notice both what patterns you use, and which patterns your client uses in response. Do they respond appropriately to the pattern in what you said, or not? If not, did you repeat what you said, or accept their inappropriate response?

You can do the same with transcripts of different therapy sesions, to notice how they are biased, and how this limits what they can accomplish. For instance, psychoanalysis mostly asks for an earlier scope of time, and the tired old, “How do you feel about that?” asks for a higher category, what is often called “meta” (All emotional feelings are “meta” to the experiences they evaluate.)

You can practice using different patterns together for greater effect, for instance: “When you see the larger context around that event, how else could you describe that situation?” asks first for an expanded scope of experience, and then for a recategorization of that scope.

As an exercise, pick any two patterns at random, and create a sentence that uses both; then notice your response to that sentence, how it changes your experience.

You can say to a client, “Tell me more about that,” which is ambiguous, and notice which patterns they use in response to it, which will tell you something about how they are naturally organized—or how a previous therapist has trained them to respond in the therapy context. They might tell you more detail, they might tell you about its history, or its consequences, they might recategorize it, or tell you their intent or outcome. Generally speaking, the patterns that they don’t use will be more impactful in changing their experience, so they are more likely to be useful.