Friday, February 5, 2016

Reverse engineering the NLP Fast Phobia Cure

Phobias are pretty neat. They're a clear and powerful example of classical conditioning in action. The cool thing about true phobias is that they're so damn simple. There's not really any "secondary gains" and "ecology" stuff to worry about. No incentives preventing you from imagining whatever it takes to rewire your response. This is where mind hacking truly shines. People do all sorts of things, including hypnotic/imagined progressive desensitization. One of the more famous (and interesting) of these is the NLP "Fast Phobia Cure" (FPC) or "rewind technique" or "v/k dissociation"

So how do you do it and how do it work?



If you want more detail there's plenty on google, but the basic protocol is:

1. Prove that you can access phobic response.
2. From the projection room, watch yourself in the movie theater chair watching yourself respond to the stimulus on the screen in black and white.
3. Step into the movie, turn on color, and rewind quickly.
4. Repeat step 3 a few times.
5. Test!

As to why that should do anything interesting, it's surprisingly hard to find a good explanation, and people seem to disagree/emphasize different parts - which is funny, since it's a designed technique, not an evolved technique. Someone out there must know the answer that motivated it.



Let's back up a bit. Why do phobias even exist? If we know that a fear is irrational, how come we don't automatically stop fearing it? How would that work if it did?

Play a little imagination game with me. Let's pretend you have a phobia of baked beans. C'mon, really imagine it. You're sitting there doing your thing making dinner and when you open the drawer there's a can of baked beans there. You drop everything and jump back screaming. It's absolutely terrifying - with all the same fear as if it had been an axe murderer at the door. Close your eyes for a moment, go into it, and get that feeling. Baked beans are scary shit.

Now float out of your body and watch that scene again from the outside. What's that like? Personally, I laughed. It's a juxtaposition experience. After that, you can't really be afraid anymore. Since you're not exposed to the stimulus, your phobic response isn't triggered. Yet, you know what you're reacting to. You intuitively know that baked beans are harmless. Watching yourself flip your shit in response to them is comically out of place. Now that you see it as out of place, you discard that response and react however you feel appropriate next time. It's almost like you're just (system 1) thinking about your response for the first time and deciding what to do.

So how come that feels too easy to be true? How come that often doesn't work?

Well, people don't like that they have phobias. That's why we call them "phobias" instead of just "[rational] fears". The problem isn't just a phobic response, but a phobic response plus an ugh field type response to seeing yourself react that way - or see yourself subjected to unpleasant experience. If you've had the phobia for a while, perhaps people have said "they're not going to hurt you..." and looked at you like you're stupid. So when you see yourself reacting phobically, instead of laughing because you see the absurdity for the first time, you avert your eyes and think "ugh, I'm so stupid!". This flinch blocks the learning that you're being silly, because you're too focused on how bad it is to have this response to put it aside and let yourself realize that it is silly.

But this flinch is much like the original flinch about the baked beans. What if we float outside ourselves again and look at ourselves feeling embarrassed/ashamed/etc and flinching away from noticing how silly it is?

If this perspective is new - that is, you don't already have a "it's terrible that I'm embarrassed!" thing going on - then you get a chance to look at it and decide whether that's okay and whether you need to keep flinching away from seeing your fear. When that flinch feels unnecessary, it'll get reconditioned. Then you can go back down to the single dissociation, you can laugh at yourself again and be cured of your phobia. At least, in theory.

What about in practice? Well, Andrew Austin does a video of curing spider phobias without words. He does just about exactly this, and was successful.

Step 1: See the phobic stimulus first person - for real. Smile, you're being filmed.

Step 2: Watch the film of your reaction (and not the stimulus). Smile, you're still being filmed.

Step 3: Watch the film of your reaction to your reaction.

Step 4: Go play with the phobic stimulus. You're done.

(The order here is interesting though. You have to work yourself down the ladder too, not just up. Perhaps they did that on their own in their head?)

The first time I took a shot at changing a phobic response I took a similar approach. I actually made the very mistake he describes as common in the video where I put the stimulus on the screen (oops :P) and found that being able to see the black/white video of the stimulus from the projection booth (from where he could see himself watching the video and the screen) was enough to make him feel like he was going to puke. Hey, at least we know we can fire up the phobic response.

So I had him dissociate again - watching himself in the projection room watch himself in the seat watching the stimulus on the film. He noted that the single dissociated version of him was feeling bad for the guy he was watching (the guy watching the screen). But from this second dissociated viewpoint, he felt fine. If he were going through AA's real video thing, he'd puke, be sad for himself puking, then watch it neutrally. And he'd still have a phobia.

He was suitably dissociated and aware to see what's going on and not be freaking out - but the solution wasn't so obvious that it falls into place with a click and a laugh. So what I did was to have him replace the version of him watching the film with an ideal version of him - reacting however he would like to react. This piece held the magic. He saw the single dissociated him cheering the ideal him reacting fine to the stimulus. The ideal him even had his family spontaneously show up for support. He was so happy seeing this that he literally cried tears of joy.

He then jumped into the single dissociated viewpoint, and then the associated viewpoint and was the ideal version of him. Cool. So this one is pretty straight forward too - you dissociate until you can see the whole thing from a detached meta level, and then you have choice - even if not automatic choice. See yourself doing what you want to do and run back down the levels to make sure it worked.



It's strange that AA (and me, back in the day) likened it to the FPC, but his version is missing several pieces.

One bit that is missing is the submodality tweaking. In particular, the video is played in black and white the first time around. In this explanation of the FPC (which I have issues with), he talks about removing a baked bean phobia with a couple sentences, and seems to imply that the submodality tweaking performs the same function here as it does in the FPC - that is, it prevents (or lessens) the phobic response by taking away bits of the phobic stimulus that are necessary for the fear.

Anyway, his example is interesting, lets look at it

"I was once giving a public presentation and found a volunteer with a phobia of baked beans. It took about 20 minutes to find some beans (which duly produced screams of terror) and 30 seconds to cure the phobia: "So, how did you get here today?" Interrupt her fear state. "Now, if you saw yourself running away from blue baked beans, what would that be like?" She gave me a puzzled look and said "That would be silly." I anchored the response by touching her elbow. "Not any sillier than seeing yourself running from red baked beans." Touch elbow again. She looked surprised and said "I guess not." And while she stayed calm, I got the organiser to slowly bring the beans back. "I think I can touch them," she said. She picked one up and threw it at her friend who'd been teasing her with them."

Hmm... what function is the red->blue doing? What would happen if we removed that bit?

I'm imagining it might go something like this. "if you saw yourself running away from baked beans, what would that be like?" "bad!". Without the color tweaking, it's easy to focus on how much you like the situation rather than watching it for what it is and finding it funny. When you ask about blue baked beans, you get a new question that she doesn't have a "bad!" answer for yet - so she has to think from scratch. Once she's in the "thinking about what it would be like to run from something very similar to my phobic response" mode, he can add in the last bit without it changing the response - after all, red beans aren't any more dangerous. And her attention is already pulled to the "what would it be like???" question so the flinch has been preempted.

So does the B/W bit really do the same thing? Well, no. The B/W is in the representation, not in the territory. It might help prevent the phobic response by making it less real, but you can just omit the phobia from the screen to do that. Perhaps it's just a safe guard against accidental or required partial appearance of the phobic stimulus on the screen. But it doesn't seem to do that good a job at preventing phobic response. So that's one difference.

Another bit that is sometimes added to the FPC is funny music to the video to encourage seeing things in a funny way and hopefully anchoring it to that. However, it's not needed and may have been added by people who have no idea what they're doing just for sake of adding things (what a goober).



However, the bigger part of the FPC that AA skipped is the rewind bit. He skipped the rewind bit of what is sometimes called "the rewind technique"! And it worked! Perhaps it is also superfluous? Well, it didn't work for me by that point. Maybe it's a first stage that catches some fraction of phobias? What's going on here?

What would be perfect is to watch someone who knows his stuff go through a quick and stripped down version of the FPC. One where you can clearly see the effects on the subject's face. One where you can clearly see the difference between the pre-test and the post-test. A textbook example. While we're wishing for things, it'd be nice to have a follow up to make sure the change stuck for like... fuck it, make it 25 years.

Fortunately, we have exactly that. Here's the intervention... aaaand the 25 year follow up

Let's watch this intervention video carefully and see what we can see.
  • 0:35 she imagines a bee flying in and shows signs of a phobic response. It's not a full blown panic, but there's no actual bee and she's only starting to imagine it. The point has been made - she can access the phobic response in the lab.
  • 1:20 "we need to make a few preparations" - he see's the double dissociation thing as merely preparing for the real work.
  • 3:49 "it's hard to get to the end". Her video loops.
  • 4:49 "a little uncomfortable, but not bad" - still not super detached, but enough to not freak out.
  • 5:46 There's the magic. See the relief?
  • 6:11 Again, but much less intense
  • 6:21 and nothing the third time
  • 6:27 "that's all there is to it!"
  • 6:39 "I'm glad I didn't pay for this one!" - not the same as a witchdoctor or placebo. She was in disbelief
  • 7:42 She clearly anticipates it to be bad, but when he asks "you still have it", she says "no!?" all surprised like and cracks up again. Meta level magic here.
There is clearly no magic during the dissociated forward run. She's not even able to look at it emotion free like AA and I were aiming for. The stimulus is (probably) on the screen which she can see, so she's not even double dissociated. However, it doesn't appear that Steve Andreas doesn't know what he's doing. Of the people in NLP, he seems to be at the top of the list of people who get it. Even without that background, this is a very quick and stripped down version with almost no hangups (and the one slight hangup dealt with perfectly). More importantly, it worked. And he knew that the dissociated run through was "preparation", not the magic. It's not a first filter that catches some fraction.

During the backward associated run through, you can clearly see the magic at the end there. What's happening is "backwards conditioning". Backwards conditioning is where the Conditioned Stimulus comes as the Unconditioned Stimulus is ending. This teaches you that the CS is associated with the US ending, so it ends up being an inhibitory conditioning. No one ever seems to talk about this as a method to remove fear referencing the FPC or otherwise, but... After writing this I found where Steve Andreas pulls through again (towards the bottom).

When you play the video backwards, you see yourself calm, yourself starting to freak out, and then phobic stimulus comes around and calms you down real fast. To use the coherence therapy framework, this provides a juxtaposition experience that it wildly contradictory with the original experience - so it rewrites the original memory (again, notice her laughter when this happens). The dissociated run through is just priming.

This implies that the FPC is a Dark Art in the sense that you're updating on fictitious data (bees didn't magically remove her uncaused freakout and envenomated mess), and if the fear really is maladaptive that shouldn't be needed. It won't prevent you from relearning the fear or blocking the removal, but the technique alone does not discriminate. AA's method does - you have to see what you're doing as an overreaction - and so does the blue baked bean method. The method I used did too - you can only replace it with a response that feels right. But all of these, FPC included, should overwrite the actual memory and be more permanent and robust than systematic desensitization.

So when to use the FPC? When to use other phobia cures? Well, the FPC falls into the "dark" category, but only in a mildish sense - I expect your anticipations of danger will recondition your fear if need be. And while it's dark, it is not patchwork - it uses reconsolidation not inhibition. The other approaches seem light - you can only install what feels right. As to the class of problems you can apply it to, aim for anything classically conditioned - or for anything an overactive stimulus-response is a significant part. You'll know it is because the stimulus alone will give you a reaction. If the problem wouldn't be gone without that reaction, then FPC like things aren't the only thing you have to do - if useful at all.

No comments:

Post a Comment