Um, yes, I think so. Actually, NLP is a grab bag of a whole lot of different things, some of which seem to me barely compatible with each other. But what you're talking about (I think) is the part of NLP that I believe has the greatest interest to cognitive science.

One of the main things a lot of NLP people have been interested in over the past ten years is what they call ``submodalities:'' those sensory qualities such as brightness, size, distance, loudness, pitch, weight, temperature, etc. Bandler's book USING YOUR BRAIN FOR A CHANGE or the book CHANGE YOUR MIND AND KEEP THE CHANGE by Steve and Connirae Andreas give a good introduction to this idea. (You have to get past the new age covers and titles. These books are written for a popular audience, but I think there are some ideas in them worth investigating in a more scientific fashion.)

Say that I remember something, and I have a particular feeling about that memory. Now what is this memory? In NLP terms, the memory has to consist of images, sounds, words, and feelings (in various senses of that word). Now how do I know to instantly have the emotion I do about that memory? One plausible hypothesis is that the memory is stored as a cluster of a lot of different things and the emotion I have is an integral part of that memory. The experience in NLP suggests that this is not the complete answer.

Another hypothesis -- and maybe this is just another point of view on the first hypothesis -- is that the emotion is a conditioned response to the memory. In NLP terms, though, one can't have a response to a thing in the world -- one can only have a response to a sensory impression. (Pavlov's dogs didn't salivate in response to the bell -- they salivated in response to the sound of the bell.) So when I person says ``Whenever I think about my mother, I feel sad,'' in NLP one tries to be more precise about the (internally generated) sensory data that evokes that emotion. ``Is it an image of your mother that comes to mind and makes you feel sad, or is it the remembered sound of her voice, or something else?''

Okay, so suppose it's an image. ``So how do you know how to feel sad when you see that picture of your mother in your mind, whereas you wouldn't have the same feeling if your mind brought up a picture of Madonna?'' (There is a presupposition in this question that there is a choice involved which is made almost consciously. In fact, this is false. The person doesn't ``know how to feel sad,'' it happens automatically. This question actually uses a certain language pattern that makes the client more receptive to the idea that the response can be changed. It's manipulative, as much of NLP is manipulative. But it's not manipulative in a dishonest way, in my opinion.)

Anyway, what I'm trying to get to is that the belief in NLP is that the learned response is not triggered by the totality of the image, but that the image is ``coded'' in sensory qualities -- submodalities -- that trigger the response. One starts asking questions like ``Where do you see this picture? How far in front of you do you see it? Is it bright or is it dark? Is it in color or black and white? Flat or three dimensional?'' And then by teaching the client to change these submodalities, one can change the client's emotional response.

For people like myself who don't visualize well, it can take a while to get the hang of this. But I think it's hard to go through an extended training in this material without becoming convinced that there's a lot of validity to this submodalities idea. I think it's certainly a hypothesis that's worthy of being investigated scientifically.

The other side of NLP, as it were (although there are really more than two sides) has to do with language. Dealing with language is almost as major a component of the human brain as dealing with sensory information. (Says me, who knows nothing about the structure of the brain.) And so in NLP, language is given a role which is somewhat equal to the role of the five senses. John Grinder was a linguist, and NLP started with ideas from linguistics -- specifically from General Semantics -- spelled out in THE Structure of Magic, vol. I.