Response to Shawn Carson’s blog post on the swish pattern

Steve Andreas


I appreciate all the time and effort that Shawn put into his response to my blog post on the swish pattern. This has provided a unique and refreshing opportunity for me to think through many issues, and develop more clarity. Accordingly, it has taken me some time to compose my response. This kind of exchange of views and understandings is woefully infrequent in the field, and I think it’s absolutely necessary if NLP is going to become a coherent and systematic set of understandings and methodology—rather than a “herd of cats.” Shawn’s response appears below, with my comments in brown type interspersed, so that you can see exactly what I’m responding to.


How to Save the Swish: A Thoughtful Response to Steve Andreas

NLP in a State of Change

By Shawn Carson Aug 07, 2016


In this article we will comment on a blog post by Steve Andreas, entitled How to Ruin the Swish Pattern: “Let me count the ways”’.

I believe the NLP and hypnosis community shares a common goal of ‘raising the bar’ in respect of change work and coaching, and healthy debate about NLP techniques, principles and the merits of different philosophical viewpoints supports this goal. No one can doubt the many contributions Steve Andreas has made to the field of NLP and it is with a due sense of the debt the NLP community owes to Steve that I respectfully disagree with some of the premises in Steve’s blog post.

In his blog post Steve critiques (although it might be more accurate to say ‘criticizes’) various YouTube videos of the swish, including videos featuring Michael Carroll, Tony Robbins as well as yours truly.

Steve has two problems with the various swish techniques or variations he critiques:

  •    Firstly they reflect a “lack of ability to learn and follow the steps of the pattern”
  •    Secondly they reflect a “lack of understanding of the principles underlying each step”


“Lack of ability to follow the steps of the pattern”

Let’s deal with the first point first. NLP has never been about following the steps of a pattern. Richard Bandler (co-founder and arguably the creative force behind the birth of NLP) defines NLP as “an attitude of wanton experimentation that leaves behind a trail of techniques,” meaning the steps of the pattern are something that are left behind after the application of NLP. If you are simply following the steps of an NLP pattern, it does not mean you are necessarily ‘doing NLP’.


Firstly, Bandler’s quote is at least partly a marketing statement designed to distinguish what he does from the rest of us. Grinder has made similar statements describing his work as “real NLP,” implying that the work of others is not.

Secondly, if we apply Shawn’s statement that, “NLP has never been about following the steps of a pattern,” to the context of cooking, then anyone who uses a recipe would not be a cook; only someone who creates a new combination of ingredients would deserve that title. While it may be useful to distinguish between creating a new recipe and following an existing one, I think most people would consider both to be “cooking.” Does Shawn really mean to say that someone who uses the swish, or 6-step reframing, or “mapping across” with submodalities is not doing NLP?


Neither of the co-founders of NLP (Richard Bandler and John Grinder) teach NLP using the deductive teaching approach of ‘follow these steps’.


I trained directly with Bandler for a dozen years, and Grinder for about half that, and I can assure you that they both offered many “follow these steps” patterns, from predicate matching to 6-step reframing. Anyone who doubts that can easily verify it by reading Frogs into Princes, Trance-formations, Reframing, or Using Your Brain—for a Change. In several recent Bandler videos he uses the same stepwise process to cure a phobia, “Take that picture, shrink it down to the size of a quarter, and blink it black and white 25 times.”


Rather they teach by encouraging their students to step into a state of personal excellence, and then to work with their ‘client’ from that state. As a result, their students all have unique experiences.


The principles underlying the Swish

I totally agree with Steve Andreas that it is vital to understand the principles underlying the NLP patterns. Once you understand these principles, you can step out of the confines of the ‘steps of the pattern’ mind-set, and enter the dance of change with your client. This is why we write our NLP Mastery books to fully explore the principles underlying each of the core NLP patterns (including ‘NLP Mastery: The Swish’ by Jess Marion and Shawn Carson).

Therefore I will re-analyze each of the videos Steve Andreas critiques, but with the intention of pointing out how these videos reveal some of these principles, especially where they vary from the “standard” swish Steve Andreas describes, i.e. from the perspective of ‘what is right’ rather than ‘what is wrong’ with the videos.

Before we get to that, I would like to point out a few areas of disagreement that I have with Steve’s discussion in his “Background” section.


What is a ‘swish’???

Steve defines the swish as “a rapid way to change any troublesome habit or other unwanted response”, and it’s certainly the case that the swish was originally described in the contexts of problems such as smoking and nail-biting (we will talk more about this ‘classical’ swish a little later).

*However, I would define the swish as a technique that chains or links two representations (typically two pictures) using a sliding submodality shift, in order to redirectionalize the mind.


I think both descriptions are accurate, each points out different aspects—and each is also incomplete.


As an example any technique that links a trigger picture with an outcome picture by making the trigger picture shrink, as the outcome picture gets bigger, is in my mind a swish pattern.


That definition omits the very important criterion that the outcome picture is a desired self-image, not a specific behavior (more on this later).


Obviously techniques that are swish patterns under my definition, would not necessarily be swish patterns under Steve’s definition, but this is semantics.


I think it is much more than semantics. If someone describes a suspension bridge as one that is suspended from cables, and later someone else describes a bridge built entirely of bricks as a suspension bridge, that is simply not true, even if it’s a useful bridge. And if you are building a suspension bridge, but you leave out the cables, it won’t work very well. The swish was described in great detail about 30 years ago, so that word has a very specific meaning. Using the word “swish” for an avocado and cheese sandwich isn’t useful, even if the sandwich is nourishing.


Does this matter? Well, certainly if you are teaching an NLP Trainers Training, then it is important. Trainers should be able to ‘speak the language’ of NLP in a way that allows them to communicate with each other. But if you are working with a client, or even teaching an NLP Practitioner course, it’s less clear that overly ‘standardized’ NLP terminology provides any substantive benefit to your client or students.


If I’m working with a client, I wouldn’t use any NLP terminology at all.


‘Classical’ or ‘standard’ swish

In the classical swish (Steve refers to this as the ‘standard’ swish), the issue is an unwanted behavior (say smoking). The trigger picture is whatever the client sees out of her own eyes immediately before she loses conscious control (e.g. the cigarette packet, or the newsagents where she buys her cigarettes, or the cup of coffee that precedes her first cigarette of the day, etc.). The outcome picture is how she sees herself being as a person (identity level dissociated picture) when the habit is no longer an issue for her. So far, so good, we all agree on this basic foundation.


I agree with this description. However Shawn’s earlier description (marked with an asterisk* above) omitted the “identity level dissociated picture.” My main criticism of the swish videos I reviewed was that this crucial piece was left out of all the demonstrations (except for Tony Robbins,’) replacing it with a specific behavior. This change is what I described as “replacing the engine in a Lamborghini with a hamster wheel.”


So why is the desired self-image dissociated? Richard Bandler told me that the classical swish was intended to maintain the client’s state while changing the context. Meaning the client has a desire for the cigarette, and the classical swish is intended to maintain the state of desire but to then apply that desire to the self-image. The client maintains the state of “I want” but changes from “I want to smoke” to “I want to be her”; meaning her ideal future self.


This is an interesting alternative understanding of principle. Perhaps someday when NLP has become scientific, someone will design an experiment to decide between Shawn’s view and mine. Until then I can only offer my rationale. Yes, the cue image triggers desire, but it also triggers a conflicting urge to not yield to the desire, a state of unpleasant incongruence, which is why the client wants to change it. If Shawn is correct in saying that the feeling state triggered by the cue picture is transferred to the self-image, that would mean that this incongruence would be transferred. I don’t think that would be useful, and I don’t think that is what happens. I think it is much simpler and more accurate to think of it as moving away from the unpleasant incongruence of the cue to the congruent desired self-image.


This is why the image is dissociated in the classical swish, because (as Steve Andreas rightly says), if it is associated there is no longer the ‘wanting’, rather a ‘being’.

I discussed this idea (that the classical swish is intended to maintain a state of desire) with my teacher and mentor, John Overdurf. John shrugged and said “well, maybe”, again because he sees the swish as a much wider pattern than mere changing of unwanted habits. In any case, if we limit our discussion to the classical swish as described by Dr. Bandler then there is no state change (the state is ‘desire’ for both the trigger and self-image pictures) and the state changes that Steve Andreas describes do not take place, i.e. they are inconsistent with Dr. Bandler’s original swish.


Shawn states that “the classical swish is intended to maintain a state of desire.” This does not appear in Bandler’s original description in Using Your Brain—for a Change. Since the state in response to the cue image is incongruent (both desiring and not desiring) and the state in response to the self-image is congruent desire, it doesn’t make sense to say that the first state is “maintained,” or to say that “then there is no state change.” If that were true, the incongruence would be maintained. It makes more sense to me to think it as a shift from incongruent desire to congruent desire, propelling behavior away from an incongruent state, and toward a congruently attractive self-image.


Self-delusion of change

Now Steve makes a strange and in my view a mistaken claim:

“if the self-image were associated, that would assume that the client had already become it, so there would be no motivation to change, only a self-delusion that change had already happened.”


Please see my next comment below, and also my more extensive comments clarifying what I mean by the word “delusion” in response to Shawn’s first video below, and in response to David Shepard’s video later.


The problem with this statement is that if the swish (however it is done) works then change has already happened; it’s no delusion, it’s reality. The reason to keep the image dissociated (in the classical swish) is to utilize the client’s (existing) state of desire, not to avoid self-delusion.


Perhaps we are talking about different time frames as if they are the same. The desired self-image is dissociated to create motivation, because at this point in time the client is separated from the image that they want to become. “If the swish works” is a bit later when the cue triggers the desire to become like the desired self-image, and this change is very fast, so they become it. But if the desired self-image were associated to begin with, there would be no motivation, and no change, because the client would think they had already changed, and this is what I referred to as a delusion. Again see my more extensive comments on delusion in response to Shawn’s first video below, and in response to David Shepard’s video later.


State based coaching

Substitute behavior

Steve suggests that seeing the self-image in a context, “doing a specific behavior” is a mistake. If so, then it’s a mistake that Richard Bandler made with the original swish as he describes getting smoke-free clients to see themselves happily co-existing with smokers (because he claims he did not want to create anti-smoking crusaders). Indeed it’s impossible to see a self-image absent any behavior; standing is a behavior, sitting is a behavior, so how can a self-image have no behavior?


Firstly, the word “behavior” usually means a movement is involved. A still image is not a behavior, even if the person is standing or seated, but a movement like standing up or sitting down is. So it is not at all “impossible to see a self-image absent any behavior.”

Secondly, I agree that Bandler often made statements like, “getting smoke-free clients to see themselves happily co-existing with smokers (because he claims he did not want to create anti-smoking crusaders),” and that could be understood to indicate a specific behavior. However, “Happily co-existing” is a general statement that could include a huge variety of specific behaviors. My understanding is that he wanted the self-image to have that attitude or capability, not a specific behavior.


Now in the HNLP (Humanistic Neuro Linguistic Psychology: Overdurf and Silverthorn) coaching model we are always looking for ‘context-trigger-state-behavior’, both for the problem and for the change. If the behavior has not changed then typically the problem hasn’t either. Therefore, there is no way to test change unless the coach sees (and the client experiences) a new state and a new behavior.


I’m OK with that; that is why all good NLP work is tested to be sure there is a change in behavior. The question boils down to whether the change is chosen in advance by a conscious mind (client or practitioner, or by the client’s unconscious mind as a result of wanting to become the desired self-image.


State-based coaching

State based coaching is the basis of HNLP. State based coaching assumes that behaviors and other responses are based on:

  • Your state,
  • In a context,
  • Presented with a trigger.

For example, if you are in your cube at work (the context), and you notice your boss looming over you (the trigger) then your response will depend upon the ‘state’ you go into when you notice your boss. For example if you go into a ‘flapping panic’, then you are likely to respond inappropriately, while if you go into a state of ‘unflappable calm confidence’, then you will likely respond appropriately.

Change work is therefore about collapsing a problem trigger so that the trigger becomes an anchor for a resourceful state, rather than an unresourceful state.


Although I don’t think the word “collapsing” is a good metaphor (I think “linking” would be better) I otherwise agree with this.


The major issue I notice, when watching unsuccessful swishes (or any other NLP pattern for that matter) is that the coach fails to elicit a resourceful state or fails to attach that resource to the trigger.


In the case of the swish, that would be to make sure that the desired self-image elicits a resourceful state.


But you will all-ways create change if you follow the four steps of John Overdurf’s Meta-Pattern (associate into the problem, dissociate, associate into the resource, and collapse i.e. attach the resource to the trigger so the trigger becomes an anchor for the resource).


Earlier Shawn disparaged “following the steps” as “not necessarily doing NLP”; now he is advocating following “the four steps of John Overdurf’s Meta-Pattern.” This particular meta-pattern that Shawn attributes to Overdurf is one that we learned from Bandler many years ago, way before the swish.

The steps Shawn proposes are one useful sequence, but not the only one. Many other patterns don’t follow the Meta-pattern sequence, for instance context and content reframing, simply directing attention to a different scope of time or space, or a different categorization. “Mapping across” with submodalities, the compulsion blowout, and other patterns of change also don’t follow the Meta-pattern outline.


See The Meta Pattern by Sarah Carson and Shawn Carson for more.


The description of this book states that, “The Meta Pattern is at the heart of all successful influence whether in therapy or business.” “All” is an overgeneralization that is not true. It is one useful meta-pattern, but not the only one. Thinking that it’s the only meta-pattern puts a severe limitation on the “wanton experimentation” and “dancing with the client” that Shawn advocates.


So, if the new self-image triggers a sufficiently powerful resource state, and the swish attaches this resource to the trigger, the technique will be successful, if not, it won’t.


Agreed. Which is why it is so important to be sure the self-image is very desirable.


If you get hung up too much on the image, rather than the resource state, or ‘end state energy’, you are ‘focusing on the finger and missing all the heavenly glory’.


I would need to see a denominalized translation of that last sentence in order to respond to it.


‘State’ versus ‘end state energy’

Understanding the concept of a resource ‘state’ versus ‘end state energy’ (using HNLP terminology) is important. Typically, if you ask a client for a resource state, how they would prefer to be feeling in their problem context, they will offer a word that is more-or-less the ‘opposite’ of their problem state. For example, if they feel afraid of speaking in public, say, they might say they want to feel ‘awesomely confident’.


Yes, I agree. Clients often ask for the other half of a polarity, rather than a state of integrated resolution.


These big resource states are great for breaking down problems, but not so good for generative change.


“Big resource states” are not necessarily the best; they are much less stable, more likely to flip back to the other polarity, unless they two are integrated.


Why is this? Because these states are difficult to maintain over time; they are too high energy. An accomplished speaker will probably not say they feel ‘awesomely confident’ when they speak in public, they’ll say something like “I feel free, relaxed, open…”.


I couldn’t agree more with the paragraph above. When you approach a door, twist the knob and open it, you don’t say to yourself (or someone else) “Wow, I can open the door!!” In fact, you usually don’t even notice it at all. That is why it’s useful to say that the desired self-image is of “the you for whom smoking is no longer an issue,” in contrast to, “the you who is triumphant about having conquered the habit,” or some other “big resource state.”


These lower energy ‘end states’ are typically associated with ‘values’ or ‘identity’ level states, such as “freedom”, “love”, “being myself”. Think about it, it’s pretty easy for you to feel “free” all the time, but to feel “awesomely confident” all the time would be exhausting!


I wish Shawn, or somebody else, would convince Tony Robbins of this! He is very good at getting people to flip to the other side of a polarity (for example see him turn a stutterer into a speaker like Tony). This can be a useful first step, but Tony seldom takes the next step to integrate the polarity, so that, for instance the stutterer can talk like an ordinary person. There are several videos in which Tony asks for someone who is suicidal, and they end up saying something like, “I can conquer anything.” A statement like this is a universal quantifier that reality will soon refute.


Now it’s important to know that if you as a coach, generate a big powerful resource state in your client (“awesomely confident”, say) and use that to collapse the trigger, change will happen. And, over time, that “awesomely confident” will transform into a lower energy (but more sustainable) ‘end state’.


I would be interested in Shawn’s understanding of how this transition from a big energy state to a lower energy state occurs. I would say that sometimes that happens. Other times the change will fall apart, and the person will revert back to the other side of the polarity.


The swish pattern typically uses ‘end states’ (remember end states arise at ‘identity’ level, or ‘self-image’). BUT a good swish can (and should) also layer in more energetic resource states such as laughter, excitement, confidence, and so on. When you watch the videos, please look primarily for the state of the client, and how their state changes, as they go through the swish.


Video Review

Before we get to the videos themselves, I think it’s important to bear in mind the purpose of a YouTube video. Most of the videos discussed are short videos, not extracts from lengthy trainings, and even the ones from trainings do not contain all the teaching that preceded and followed the demo. Such videos are never intended to teach chapter and verse on a specific pattern; wouldn’t life be easy if you could master an NLP pattern by watching a three minute video! I would suggest that you watch each video as a summary of the swish, more intended to raise the profile of the trainer than to lay out all the details of the swish.


I can agree with all this; in fact I wrote much the same in my blog post. However, what is presented in a short video should be accurate. Making sure that the cue is associated and that the outcome is a desired self-image are essential pieces, not “chapter and verse” minor details, and most of the videos did not get these right.


Michael Carroll

Michael describes the swish using the ‘slingshot’ method. This is not a client demo, and is not really even a ‘covert demo’ (compare this to the Tony Robbins video discussed below).

It does utilize the ‘slingshot’ swish, where the trigger image is sent out to the distance and returns as the new self-image. In HNLP we consider the slingshot to be often superior to the ‘standard’ swish because of two important underlying principles:

  1. Different clients react to different submodality shifts. The swish Andreas describes uses two of the principle submodalities (size and brightness), while the slingshot uses three (size, brightness and distance), so if your client’s driving submodality is distance, she would get a better result using the ‘slingshot’ than using the ‘standard’ swish.


There are two issues here; one is the use of distance. I totally agree that distance will be a better driving submodality for some people, and using three submodalities rather than two is fine. If one of the submodalities truly is a driver, it will carry one or more other submodality shifts along with it spontaneously, without having to mention them. This is why when we teach the “designer swish” pattern we teach people how to identify which submodalities are drivers for the client, and use those.

The other issue is the way distance is used. The slingshot first sends one image into the distance, and then brings the other image back in, what I have called a “butt joint.” If one image goes off into the distance at the same time that the other comes in from the distance, this creates a continuous change, which I have described as a “lap joint,” which I think makes a more dependable connection.


Michael could have done a better job of stressing this in the video, but again it’s a short 6 minute YouTube video.

  1. NLP is ultimately theater; it relies on communication between coach and the client’s unconscious (this by-the-way is the basis of New Code NLP, which Michael practices – to maximize unconscious involvement in each key step of transformation). The slingshot swish is often better geared toward a theatrical presentation by the coach, which is more likely to engage the client’s unconscious.


I agree that good NLP work depends heavily on nonverbal cues (“theater”). However, you can be just as theatrical (or even more so) using both hands going in different directions simultaneously, gesturing with one to indicate the cue image moving away, and the other hand to indicate the self-image moving closer. Given Shawn’s emphasis on the importance of nonverbal “theater,” it’s curious that his two videos below use still images of cartoon characters, leaving out theater altogether.


It’s true that Michael does emphasize behavior over state in the video (the antithesis of New Code), but again I suspect this is more to do with the ease of describing the swish in a short 6 minute video, than the way Michael would actually work with a client.


Possibly; we have no evidence on this. Shawn writes that “Michael does emphasize behavior over state.” However neither behavior or state is a desired self-image.


Tony Robbins

This is a great video.

The first thing to note about Tony Robbins’ video is that his definition of the swish is much wider than the classical swish, designed to change a habit, that Steve Andreas seems to be focused on.


I think Shawn means “application” rather than “definition.” I specifically wrote “troublesome habit or other unwanted response” (not just “habit”) which I think is pretty inclusive.


The video is excellent (of course, being Tony Robbins). I will point out a few of the principles Tony is using:

  1. Tony is doing a ‘covert demo’, meaning he is overtly talking about his own experience, but covertly is leading his ‘client’ (in this case his audience), to actually experience the swish. This means that when they actually come to practice it their unconscious minds have already been through the swish at least nine times!


I agree. Good point.


  1. Tony not only uses the visual association of chaining the two pictures, he also uses the physical behavior (raising the hand to the nail-biter’s mouth) to trigger the swish. So when the client actually raises their hand to their mouth, this physical motion (as well as the visual cue of the hand) will trigger the swish.


Excellent observation; I missed that nice utilization of the hand movement.


  1. Tony triggers the swish by reaching down with his right hand to a resource (in this case Tony’s identity level self-image as a communicator). For a normally organized individual, this reaching down-right accesses their kinesthetic resources, and moving the hand up toward the chest associates into the chosen resource by ‘pulling it’ into the body. If you are familiar with Tony Robbins’ work you will know this is a common feature of his ‘swish’ patterns (and one described in detail in Jess and my book ‘NLP Mastery: The Swish’).
  2. This reaching down also sets a gestural anchor that can be used by the client whenever they need. Tony refers to this at 7.40 of the video.
  3. Tony also plants a post-hypnotic suggestion that the change he has installed will transform the audience’s life in unexpected ways (generative change).


Agreed. But it is the desired self-image that drives the generative change.


The really great thing about this demo is watching Tony use the theater of the pattern to create a big positive state in the audience. You will see this in the faces of the audience around 7.00 of the video. Tony creates this energy using speed and tempo, and voice tonality, as ‘sliding anchors’ as he repeats the swish (getting faster, more up-tempo and more energetic each time).


Yes, Tony does a nice job (excluding the role-play clips of the father and daughter) in using his nonverbal expressiveness to amplify his verbal instructions—the best of all the videos, in my opinion—and impossible to do in a cartoon animation.



Alkistis (Steve Andreas refers to ‘Akistis’), discusses a pattern she calls a swish, but which is closer to a map-across. As such we will not discuss this video further.

Note that Ms. Alkistis does not claim to be an NLP trainer, so it seems petty to fault her on calling a ‘map-across’ a ‘swish’.


Firstly, what she did is not even a “map across.”

Secondly, I didn’t research the trainer status or background of any of the people in the videos. However, this doesn’t seem to be a relevant variable. For instance, Mark Hayley’s web site states that he apprenticed for 5 years with Bandler, is licensed by Bandler as a Master Trainer of NLP, has been “attendee, assistant, apprentice, or the principal trainer of 150+ NLP and hypnosis seminars, has practiced for over 15 years, and is personally endorsed by Bandler as follows: “Mark Hayley is thoroughly trained, highly skilled and very elegant with my most up-to-date teachings. I highly recommend him.” Despite all that training experience, his demonstration of the swish (next video) is almost as bizarre as ‘Alkistis.’


Mark Hayley

This is a file download. I was not comfortable downloading this to my computer so did not open and review the video.


Shawn Carson (oh, wait, that’s me!)

This is part of a series we did using an animation program called ‘Two Minute NLP’ (although this video runs an impressive 3.17).

This sets out the classical swish, but using the ‘slingshot’, in a (hopefully) fun way. I lead the client Sophie into choosing a self-image not a behavior.


This is flat out false. The caption states: “Sophie, what if there were a way that every time you saw a donut, you felt motivated to work out?” “Felt motivated” is a feeling and “working out” is a behavior; neither is a desired self-image. The sketch is of Sophie with a barbell and exercise ball — doing a specific behavior. Sophie responds, “That would be fantastic!!” which is a clear indication that she hasn’t thought this “solution” through at all.

Firstly, it means that she can never enjoy choosing a donut, because if she has to exercise instead, she will be just as choiceless as when she’s not able to refuse it. That is replacing one robotic response for another — not necessarily progress, and certainly not generative.

Secondly, whenever a specific behavior is selected as an outcome, it’s important to expand the frame and ask, “How well will this work in the real world?” In this case, the answer is “Not very well!” If Sophie is offered a donut at a meeting, and feels motivated to work out, she won’t be able to, so now she’ll be frustrated by not being able to (on top of being tempted by the donut!). There will be many other contexts in which she won’t be able to work out in response to seeing a donut. This is a great example of the limitation of a specific behavior chosen by someone’s limited (and often downright stupid!) conscious mind. Another example is in David Shepard’s video, where the desired behavior is a dissociated image of himself going around a tight turn on a motorbike (see below).

The next frame says, “Make a picture of yourself as you want to be . . . your ideal self. . . Make it big and bright. . .” The image is of her with barbell and exercise ball again, looking slim, and the balloon over her head says, “OK I look fantastic!!” This image is either her behavior, or the result of her behavior (being slender and fit) but it’s definitely not an image of someone for whom responding to a donut is an easy and effortless choice.


To represent this in an animation there has to be a picture of something, in this case Sophie in work-out clothes (Steve Andreas describes this as a ‘behavior’; so be it).


The barbell and exercise ball clearly indicate a context in which the behavior is exercising.


Steve Andreas makes the point that the ‘standard’ swish uses a submodality change that allows the self-image to (for example) get bigger as the trigger picture gets smaller. The changes take place simultaneously. In contrast the slingshot swish, which is used in several of the videos including this one, uses sequential submodality changes. The trigger image gets smaller as it moves further away, followed by the self-image picture getting bigger as the picture returns.

We talked about a couple of the advantages of the ‘slingshot swish’, versus the ‘standard’ swish when we analyzed Michael Carroll’s video above. But what about Steve Andreas point that the simultaneous change in submodalities of trigger picture and self-image is like a “lap joint” and therefore “stronger and more lasting”?


This is another difference that perhaps can only be decided by experimentation when NLP becomes scientific.


As you watch all the videos, you will see that the swish pattern begins slowly, to allow the client to acclimatize, but ultimately each swish is being run in a fraction of a second. This does not allow time for one state to decrease and the other to increase.


That last statement is offered without proof or rationale, and in any case it is the images that increase/decrease, and those changes elicit the feeling changes.


States take up to a minute to ebb and flow, not fractions of a second.


If that were true, the swish couldn’t work, because the feeling elicited by the cue wouldn’t have time to transform into the feeling of desire for the self-image. If you vividly imagine that you are furiously angry at someone, and then that person points a loaded gun at you, with a facial expression that indicates they are quite willing to pull the trigger, and you will find that your anger response changes to fear in a very short period of time — certainly less than a minute.


In fact, the swish is run so fast that the client realistically does not have time to even change the pictures in a meaningful way;


That is a conscious-mind statement, without evidence or rationale. One of the reasons for doing it fast is to force the client’s unconscious mind to make the connection.


I would argue that the swish neurologically wires the end-state to the real-world trigger, via Hebb’s Law. As a result, the trigger becomes an anchor for this new state.


I agree, but that will happen with either a lap joint or a butt joint; it’s only a question of which is stronger.


Finally, Steve Andreas talks about images being ‘realistic’ (see discussion above or Steve’s swish video at It’s pretty clear that in reality the trigger is not going to actually reduce in size. Imagine walking into Dunkin’ Donuts and seeing the donuts actually get smaller before your eyes (in reality). Ain’t gonna happen; not realistic.

Given the length and very basic nature of the video, there are no ‘important principles’ revealed here.


Thanks, I had forgotten about that little clip. When I used the word “realistic,” I meant “believable” to the client (which I think is clear from what I say in the video) not “realistic” in the size of the image. Luckily, readers can click on the video and decide for themselves — one of the great things about having a video to observe.


Shawn Carson (me again!)

This excellent video (did I just say that?) uses the swish to deal with difficult people. I learned this pattern (which Jess and I refer to as the ‘New Behavior Generator or NBG Swish’ in our book ‘NLP Mastery: The Swish’) from John Overdurf. When understood it does reveal some important and useful principles, including:

  1. Creating patterns by Integration. Some great hybrid patterns can be created by incorporating elements of one NLP pattern with another. For example we use Deep Trance Identification or New Behavior Generator as Steve Andreas rightly points out, within the swish. This can be a particularly powerful approach with clients who can’t imagine their ideal future self, but can find a model (Superman, Oprah Winfrey, James Bond, Emelia Earhart or whoever) with the qualities they wish they had. Seeing, then stepping into, this image can create powerful change. Just ask Steve Gilligan who used DTI to model Milton Erickson. Steve Andreas refers to this image as “unreal”, and “a delusion”; it’s actually called a “positive hallucination” in hypnosis!


It is fine to use the DTI or NBG (both are patterns with steps that Shawn earlier disparaged as “not necessarily NLP”) to access qualities for a desired self-image. However, the NBG has an important step that Shawn has left out, namely to adjust the image of the model before you step into it, so that the face is mine, the body shape is mine etc. This both adapts the quality to make it appropriate to the individual person, and also provides a kind of reality check.

Regarding delusions, I think there are two related issues. One is the selection of an imaginary model like superman, in comparison to a real person. Superman relies on “superpowers” which don’t actually exist in the real world. If you identify with Superman, and think that bullets will bounce off your chest, I hope you don’t have an opportunity to test that delusion! Some kids who decide that they can fly like superman, so they put on a cape and launch themselves off a top bunk. There are a number of people in mental hospitals who believe that they ARE Christ or the Virgin Mary, and that is not particularly useful. Quite a few NLP trainers are deluded in much the same way.

When using DTI or NBG with a real person, it is relatively easy to access useful general qualities or attitudes like persistence, empathy, courage, kindness, etc. It is somewhat more difficult to access specific abilities and behavioral skills, particularly those that require systematic practice over time. Identifying with Usain Bolt may be useful, but it probably won’t get you to the Olympics. Gilligan spent a long time in his very thorough identification with Erickson, and learned a great deal from it, but even now, over 35 years later, he hasn’t reached Erickson’s level of perceptivity, subtlety, and skill — and I think he would be the first to agree with that.


  1.     This pattern is structurally different to the classical swish. In the NBG Swish, the client steps into (associates with) the new self-image. This is particularly effective when the pattern is being used to generate a specific new state and behavior (rather than to eliminate a compulsion as in the classical swish). So in this case, the client steps into a state of (say) confidence so he can deal with his difficult boss.


I agree that stepping into an image can be useful to link to a specific behavior, but then it is no longer a swish (just as a bridge without cables is no longer a suspension bridge), and it will no longer be as generative.


  1. The coach presupposes the change has taken place (as Steve Andreas notes), so that the client can search for and notice what is different.
  2. It is vital that there be a break state in any pattern

(see NLP Mastery: The Meta Pattern by Sarah Carson and Shawn Carson).


I strongly disagree with that statement. I don’t think there is a break state in the compulsion blowout pattern, or in a content or context reframe, mapping across, or many other patterns. The break state is essential in the swish pattern because we using repetition to teach the client a certain sequence of experience, a direction and not a loop.


However here it is not the ‘blank-the-screen’ break state that appears the classical swish. This is presumably why Steve Andreas, focused on the classical swish, misses the break state. The break state at 2.12 of the video when the client sticks the post-it (or postage stamp) onto Burt’s forehead; if there is someone who makes the client feel bad, and you ask them to imagine sticking a post-it on that person’s forehead, you can get them to laugh (laughter is a break state). That’s why I crack a joke at this point in the video.


I didn’t laugh, so I guess it wasn’t a joke or a break state for me. However, let’s assume it’s a break state. The purpose of a break state is to interrupt one state to keep it from being connected to a following state. However since the next step is to have the “picture leap out of the forehead,” a break state at this point will only interfere with the sequence that Shawn wants to establish. A bit later (2:40) the instruction says, “Repeat several times to condition the change.” “Repeat” isn’t very specific about what exactly to repeat, and there is no mention of a break state in between the repetitions. This is where a break state is needed, because if there is no break state between repetitions, that can result in a yo-yo effect. Instead of creating a single direction from the problem image to the desired image, it may oscillate back and forth.


  1. Many patterns use physiological tricks to enhance their effect. This one is no exception: the act of looking at someone’s forehead (i.e. above their eye level) is associated with social dominance. Having the client stick the post-it (and hence look at) the forehead of their nemesis tends to put them into a more socially dominant physiology, and hence will tend to shift the dynamics of the encounter in their favor.


That’s an interesting point that I can agree with. However, my end goal would be equality, rather than either submission or dominance, which are opposite polarities.


Anthony Beardsell

For me this technique was more like a ‘double map across’ (such as is used in the NLP Belief Change) than a swish. But what the heck, who’s splitting hairs now!

I enjoyed this video because it shows a really important principle, namely that change arises primarily from the rapport between coach and client. Rapport is much more important than ‘following steps’. Anthony creates very nice rapport with Michelle and she gets her change quickly and easily as a result.


I can easily agree with all that, but what he did was not a swish, and it was not generative.


David Shepard – The Performance Partnership

The great thing about this video is that you can see how the coach enters into the dance with the client. As Steve Andreas notes, the client responds with several problems throughout the pattern. The coach responds to each with an elegant reframe, for example using submodalities, meaning reframes and hypnotic suggestions. The end result of the intervention appears to be extremely positive for the subject.

Steve Andreas does raise some excellent issues regarding the video which, I believe, come from the fact that driving a motorbike round a tight bend on a race-track is primarily a kinesthetic experience, while the swish is primarily a visual pattern. Perhaps this demo would have been better using a different NLP pattern?


Perhaps. But the swish can also work fine if done appropriately; the desired self-image would be someone with exquisite kinesthetic sensitivity and balance — far better than a dissociated image, or the delusion that the motorbike is actually on rails, so it can’t possibly slide. This is another example of expanding the frame to find out how well a specific behavioral outcome will work in the real world. In this case, the answer is that it would likely be catastrophic. The client would happily make an image of himself going around the turn fast “on rails” and have a nasty crash — and no one would realize that it resulted from an incompetent NLP intervention!


The video does not show the ‘demo selection process’ that preceded the demo, but in an ideal world I might have saved this problem to demo a more kinesthetic pattern, such as Bandler’s ‘backward spin’. At the end of the day, the change appeared to come more from the client’s feeling of ‘being on rails’ than the change in the (visual) picture.

In any case, this demonstrates the adage than any NLP Pattern can be used to address any problem,


Another universal statement that I strongly disagree with. If you use the phobia cure on grief (or the grief resolution process on a phobia) it will not work, because a phobia has a structure that is the opposite of grief.


assuming rapport between the coach and client, and the coach’s ability to ‘dance’. Again David Shepard’s rapport with the client is excellent, as is his dancing!


Jevon Dangeli

This is a very rich demonstration with lots of great learning points. I’ll point out a few of these:

  1. Utilization: at an early point in the demo there is the sound of construction, an electric drill or saw, from outside the seminar room. Jevon utilizes this by commenting, ”we are cutting through already!”
  2. Jevon elicits a nice ‘clean language’ metaphor for the problem behavior, a ‘machine gun’ (although he doesn’t utilize it further in the demo).
  3. Jevon associates the client (Rene) into the problem state and calibrates Rene’s physiology nicely (so he can test his change later on).
  4. After associating Rene into the problem state, Jevon breaks state nicely by first ‘wiping’ away the internal image, then re-associating Rene back into ‘now’ by getting her to focus on the light coming into the room through the leaves of the trees outside. He uses this break state several times through the demo.
  5. Jevon elicits a ‘last time and place’ trigger very nicely (the look in the other person’s eyes; 11.40 through 13.00). Those familiar with John Overdurf’s work will recognize this as the first step in the Meta pattern. This is not surprising as Jevon is an HNLP coach.
  6. Jevon first finds the client’s desired state, and uses this state to elicit an image. This presupposes (correctly) that if the feeling elicits the picture, then seeing the picture will elicit the feeling. At the end of the day, in my opinion the swish relies on the ‘end state energy’ feeling associated with the new self-image.


That is what results from keeping the desired self-image dissociated. If the client associates into the self-image deliberately, the motivation to become it is lost.


  1. Jevon gently challenges Rene’s change using Overdurf’s ‘testing loop’ (“Are you sure…”). When the client responds with “90% sure”, Jevon utilizes the Ericksonian technique (‘90%? Not 91% or 89%?’).
  2. In the Q&A Jevon gives a good explanation of the positioning of chairs (or relative position of client and coach) for the swish.

One issue with the demo is (as in the last video), the swish might not be the best NLP pattern to use in Rene’s context. I say this because Rene had a workable strategy (the ‘machine gun’ strategy) for dealing with the problem. Simply changing the state does not necessarily offer the client a new strategy that will necessarily work to give her the desired outcome. To complete the change, Jevon might have used e.g. a strategy installation, to make sure the client had a new strategy.


I can agree with this. However Shawn’s “Meta-pattern” described earlier “state-based coaching” states that a resource state is all that is needed, and doesn’t say anything about strategies.


Of course, he might have done some additional work later in the course that we didn’t see. In any case Rene’s post course feedback indicates that the change was effective.

For those still uncertain as to the difference between the swish and the map-across, there is also an interesting discussion between a couple of the students in the audience.


Mel Cutler

Again Mel does not claim to be an NLP trainer, so it’s perhaps a little unfair to judge the video according to some strict standard of whether the pattern is a swish or not, or the fact that Mel does indeed appear to be consulting his notes as he speaks. Mel’s focus is clearly to get change for the client (which he appears to do).

This video is a pretty good straightforward demo of the swish with a nicely responsive client. Watch it for a nice, simple, clean demo of the swish.


No, whether good or bad, effective or not, it is not a demo of the swish; what he did was quite different from what Bandler described as a swish in detail some 30 years ago.


The one point I found interesting was that Mel used the image the client provided and chunked up on that, “what will that do for you, and what will that do…” He went as far as values (“make more money, grow your business” – Mel is after all an ‘entrepreneur’s coach’), rather than identity, but it’s a nice approach to get a new self-identity from a client who otherwise has difficulty finding that new self-identity.


“Make more money, grow your business” are both goals, not identity, and not even behaviors that would result in achieving the goals.


Pip Thomas Edge NLP

To me, this is more like a cross between a swish and the Coaching Pattern, because the new image is associated. Pip uses submodalities to boost Louise’s resource state. I would have liked to have seen more of a ‘pop’ in Louise’s state, although this may have been because the demo was more ‘staged’ than the typical demo in a seminar where the client brings a bigger issue.

It’s a nice demo of using submodalities within an NLP pattern (although not a typical swish pattern).


George Hutton – Mind Persuasion

This is not a demo, but rather a ‘hypnotic product’. George sets anchors for positive and negative states using blue and red circles, and then ‘swishes’ by changing the size of the circles. It’s a cute video (and has lots of fun state breakers). It will not teach you the swish but if you run through the video with a problem in mind you may find yourself getting some change! Enjoy!


Alex of

This video contains one real gem of information. Unfortunately it is skipped over so fast it’s easy to miss. It’s this: for procrastination, one of the most powerful resource states you can bring to your client is the feeling of completion. Nobody likes doing their taxes (or their homework as in this case), but we all enjoy the feeling of having mailed our tax return off to the IRS and knowing we are done till next year.


I can easily agree with all that — changing the scope of time to elicit a useful state. Another key piece is to be able to chunk down the task into small enough pieces, and to have that good feeling of completion after each chunk is done, so that there are motivating rewards all along the way. If they only get the good feeling at the end of the entire task, it won’t work nearly as well.


This feeling of completion creates the new self-image of being a person who ‘gets things done’ so that you can enjoy the feeling of being free.

This is all ways a great starting point when using the swish for overcoming procrastination strategies.


The nice feeling of completion is great, but it doesn’t necessarily result in a change in identity. It can remain simply a useful learned behavior unless the client generalizes to “I’m someone who can get things done.” All of us have behaviors (both good and bad) that are not part of our identity. A client may use a change in behavior to conclude that they now have a different identity, but you can’t depend on it. If you assume that will happen, you (and your client) will often be disappointed. For how to work with many different aspects of self-concept, see my book, Transforming Your Self.


Ved Prakash –

I have to say the Caribbean or maybe Pacific background music does not add value to the quality of this video! I was unable to watch the whole thing, sorry Ved!


Terry Elston –

This is a demo from a seminar. Again, for me it’s a very clean demo of the Coaching Pattern rather than a swish. Terry doesn’t use a trigger picture or an outcome (self-image) picture, but rather attaches the client’s (Rachael) resource state to the ‘trigger’ by ‘stealing anchors’.

Terry discusses the importance of attaching the resource precisely at the trigger point, rather than when the client has already dropped into their negative state. He gives a good demo of rewinding the client’s story, although for me he did not actually identify the trigger point (meaning an external event that lets Rachael know it’s time for her to book her train and hotel). The trigger becomes ‘six o’clock’, rather than “I look at my watch and realize it’s six o’clock”.

Terry does give a good demonstration of ‘stealing anchors’ by mirroring Rachael’s physiology in her resource state. It’s worth watching this demo for that alone.


Keith Livingston –

This is a very short (2.40) description of the swish. As such it is a brief overview and will not add much to your understanding of the swish, if you have watched all the other videos so far.



It is my belief that NLP is a living, evolving discipline. The co-founders of NLP continue to develop their own techniques as do the those that I admire most in the field, such as Tony Robbins and John Overdurf. The moment we say “NLP is this and only this, so that is not NLP”, we remove the creative spark, the ‘attitude of wanton experimentation’ that created NLP in the first place.


That is chunking way up from “The swish is this and only this” which was my focus. Knowing how to do a swish correctly doesn’t prevent anyone from “wanton experimentation.” Neither does it prevent someone from proposing a new pattern, or a change in how the swish is done. But none of those is a swish, any more than a rabbit is a robin.


The moment we say “the swish is this and only this, so that is not a swish” we limit our ability to ‘dance’ with the client, and the dance is where we find the magic of change.


That is not a logical conclusion. You can dance with a client all you want to, or experiment wantonly all day long. But if you are communicating with someone else, and you say you did a swish, it would be nice to know that you both agree about what that means. If you order a chocolate cake from a bakery, you probably would expect that it had chocolate in it (not carob!).


Please do study other practitioners who are courageous enough to post their material publicly. Notice what they do well and absorb that. Notice the mistakes they make and avoid them. Focus on the true aim of this wonderful art, which is becoming more of the person you were born to be.


That is all well and good, but how does the average person “Notice what they do well and absorb that. Notice the mistakes they make and avoid them”? I may be a much slower learner than most, but I needed a lot of examples of both good and bad pointed out to me, along with some rationale for why something was useful or not useful. Sometimes learning what is a mistake is far more important than what to do correctly.

People often learn most easily from contrast — red looks much redder when it is next to green, for instance. Contrasting what to do with what to avoid clarifies both. That is why I wrote my original post, and why I have taken the time to respond to Shawn’s post. I hope this exchange of understandings has been useful, or at least identified interesting questions to be explored further.

Again, thanks to Shawn for his thoughtful post.