RobMan Look - I enjoy responding to what you are posting and I am sorting for difference. If my style seems argumentative to you, cool - let's both blame the Jesuits. 1. You wrote, "I still believe design CAN have specific criterion as Lewis stated" I am still waiting for a compelling example. 2. You wrote, "I'm getting the feeling that I've ended up in an argument about modelling vs design...which is not what I want or believe in at all.  I also think that the AI split you talk about is a useful, yet totally artificial construct...because once you get to the detailed software development level you are only left with creating functionally equivalent elements...after all what you are building is not the human you are imitating...(to follow your metaphor)." Yes, I am reasonably sympathetic to your point. I would not however give up the possibility that for some modeling (especially given the relative process precision in reading strategies (NLP strategies - ordered sequences of representational systems) projects (especially ones with high cognitive components) either of the approaches might succeed - it would be stunning to do this with two very competent modelers independantly and then compare the resultant models. 3. You wrote, "After all when we attempt to imitate another person (with f2 suspened) then surely this is a good example of our unconscious mind "designing" new behaviours/reps...because as you've said we don't have any access to what/how the other person is really representing...or even their real behaviour...just our reps of it.  So I feel that in phase 2 the difference between my unconscious mind imitating an observed behaviour and it imitating a powerfully imagined behaviour can at times be quite small...whereas in phase 3 there is a world of difference." Interesting - imitation and design are non-overlapping in my lexicon, so for me, this doesn't work. However your point that the difference between your unconscious imitating an observed behavior (one from the source, let's say) and an imagined one can be quite small is well taken. The difference is that the observed behavior (assuming that the source really is a genius and this is one of the behaviors that exemplifies this genius) is guarenteed to move in the direction of excellence whereas the imagined one may or may not. 4. You wrote, "But I sense that you may feel more comfortable if I simply replace the word "designing" in the paragraph above with "creating" or "generating"... One of the major objectives of Whispering about which Carmen and I were committed was an effort to stabilize some portions of the terminology that grounded NLP - thus, my insistance on the distinctions in our exchange. All the best, John