I have checked The Emotional Brain out of the library, but haven't got a chance to look at it yet.

Let me make some comments from an NLP perspective on what you say above and see what you have to say about them.

Since NLP is not primarily theory-driven, different NLPers may have different ideas about the principles involved, but it seems to me that in NLP a distinction is usually made between feelings and emotions. A ``feeling,'' as I understand it, would be strictly somatosthetic. Most NLPers would think of a feeling as being a ``whole-body feeling,'' i.e. felt throughout the body (mainly in the muscular system?), as opposed to ``part-body'' somatosthetic experiences such as hunger or proprioceptive or tactile sensations.

``Emotion,'' on the other hand, as I understand it, more commonly refers to something with a much more complex structure. It might correspond to what NLPers would call an internal state. However an internal state might also refer to something like anticipation or curiosity, which I think are not commonly considered emotions.

I was taught that internal states have cognitive (visual and auditory) components as well as the somasthetic element. One of the exercises we did in the first NLP training I went throught was to elicit all the components of a particular internal state from another person (``Where in your body do you feel it? What do you see internally? What do you hear?'') and then to take on that state ourselves, so that I could find out what anticipation, for instance, was like for the subject and discover how different that experience was from the one which I myself label anticipation.

Now you say above that feelings are not stored in memory. Are you saying that memories commonly have only visual and auditory components, but not somasthetic ones? Or that the full-body sensations which as a NLPer I would call feelings are never stored in memory? This seems to me unlikely, because we can certainly identify various feelings fairly easily (I certainly can, at any rate), and it seems to me that identification requires some sort of recall.

In my NLP training, I was taught to distinguish between feelings which are ``part of'' a memory and feelings which are ``meta to'' that memory --- i.e. the feelings which the subject in the present has about that memory. This seems to be an important distinction. When I attempted to neutralize a traumatic memory and the subject was crying, it would be important for me to find out whether the subject was crying because she was re-experiencing the feelings she had had when the remembered experience happened, in which case I would know that my work had been unsuccessful, or whether she was crying because of feelings she now had about that past experience, in the same way that one might cry on hearing another party relate a very moving experience.

Now it seems to me that even if one grants the validity of this distinction, it is not necessarily incompatible with what you have said above. There is no doubt that when people remember intense experiences, they sometimes experience the same feelings as (or very similar ones to) the ones they had at the time of those experiences. But it seems to me that what you are saying above is that these feelings themselves are not actually components of the memory, but instead are a learned response -- a conditioned response -- to the cognitive components of the memory.

This would actually fit in rather well with some of the NLP approaches to neutralizing traumatic memories. The original NLP technique for doing this was a form of desensitization, given the name VK-Dissociation. The idea was to dissociate the visual (V) from the feeling, or ``kinesthetic'' (K) component, by having the subject imagine looking at a movie of the traumatic experience while simultaneously imagining that she has floated out of her body. This technique would actually make even more sense if what is really happening is that one is unlearning a conditioned response.

Another apparently essential element of VK Dissociation is that it involves a change in point of view. Rather than being ``inside'' the experience having it happen to her, the subject imagines watching the experience as an external observer.

A later refinement of this technique involves adding a second step where the subject imagines then ``associating back into'' the memory at the very end of it and then running the whole experience several times backwards extremely quickly. I don't know quite how this would be explained in terms of cognitive and behavioral psychology, but my experience with several subjects (friends and acquaintances) indicates that it works extremely well.

Another commonly used approach is to ``go back'' and change the memory in various ways. I.e. the subject imagines going back into the past and going through the experience again, but in a positive way. It is not necessary (or desirable, in my opinion) to actually create a false memory. The subject is still quite aware of what really happened. But the creation of an alternative imagined experience seems to be sometimes very effective in dealing with the trauma, even in some cases when the imaginary experience is clearly fantastical, such as having the subject become Superman and beat up his tormentors or whatever.

It seems to me that this might fit in quite well with what you are saying, namely that the memory itself contains only cognitive elements, not feelings. In that case, changing the cognitive content ought be be quite effective.

Another NLP idea, which I don't think the scientific psychologists have explored yet, is that of ``submodalities.'' NLP claims that people code their memories --- as well as beliefs and other subjective phenomona --- in terms of sensory qualities, which NLP refers to as submodalities. It is common, for instance, that a person might remember sad experiences with an overall bluish tinge. (Cinematographers sometimes indicate this sort of thing by the use of tinted filters). Or the size of the image, or its distance, or three-dimensionality, or location (whether the subject sees it on eye level straight ahead, or above eye level, or to one side) might be other ways a person codes the memory of an experience to associate it with a particular emotion. The Point-of-View I referred to above seems to be a very important submodality. Simply having a subject look at a memory ``from the outside,'' rather than re-experiencing the event from the first-person point of view, seems to often be very useful in teaching someone to cope with an emotionally troublesome memory.

Finally, there are the anchoring techniques, which involve having the subject maintain a particular positive internal state while remembering a difficult experience.

When I started my NLP training, I was extremely pleased to discover that there was a technique called Change History. At that time, I was obsessed with my past and all the negative emotions I had associated with past experiences. During my training, I went through all the approaches described above. Sometimes I felt really wonderful after a particular session, but I was always skeptical and assumed that any change was only temporary in making me feel better and that my problems still needed more work.

Eventually I reached a point where I started experiencing difficulties as a subject for the techniques, because someone would say to me, ``Think of an unpleasant memory,'' and at the moment I couldn't bring any to mind. This was extremely frustrating, because I knew that I had hundreds of the damned things, and I just couldn't understand why I couldn't recall any on command.

It was more than a year later when I finally set aside some quiet time for myself and started sifting through my memories and discovered that I simply didn't have any unpleasant memories left. I still factually remembered all the things that had happened, but remembering them no longer made me feel bad. They were now no longer important because they were now in the past. I remembered them as if they were things that had happened to another person. In fact, as far as I was concerned that person in the past was another person. I'm somebody different now.

--
Trying to understand learning by studying schooling is rather like trying to understand sexuality by studying bordellos.     -- Mary Catherine Bateson, Peripheral Visions