Hi Clive Eating your cake and having it too, eh! On pages 328/329, you will find a brief characterization of the choices available for establishing a well-defined set - enumeration (extensive definition) and set membership rules (intensive definition). You are proposing deploying both in service of deep trance identification exercises (essential, for example, in modeling a la NLP) while simultaneously protecting yourself from negative consequences. Sure, seems like an intelligent approach to me. BTW, have you verified that your signals are still involuntary. If a person maintains the same set of signals for an extended period of time, there is a strong tendency for that person to come to a conscious mastery of those signals, thereby rendering them unsuitable as the method of communication between conscious and unconscious. Your second question revolves around a well-known phenomenon: if you attempt to become conscious of what you are doing while doing it and the skills involved are demanding enough, you will be unable to manage it as long as you are conscious. Simple example, years ago I purchased a full sized trampoline. My son Michael immediately began to do all sorts of interesting moves on it, including back flips. I managed to work out front flips but was having difficulty with the back flips (which, by the way, turn out to be easier than front flips). So, I made the interesting mistake of asking him (he was about 14 at the time) to teach me how to do it. This request, of course, caused him to attempt to make his tacit knowledge explicit with the result that for the next several days, he was unable to do the back flip. Of course, as soon as he stopped thinking about it (went unconscious again and simply did it), he was able once again to perform.  This phenomenon is simple to overcome. It just requires a judicious sense of timing and a clean third position. When performing, perform - with all your heart and soul. If you are intersted in understanding consciously (say, for purposes of instructing someone else to perform as you do) then perform, cleanly exit to third and examine that djembe player "over there" performing and digitialize his performance, chunking it down into various steps to be offered to your learner. Keep the two domains separate. As to modeling people who are not available, this is, strictly speaking by the criteria we present in Whispering not possible - except under extraordinary circumstances - for example, if there were multiple full video and audio records of the person you wished to model and, critically, full video and audio records of the audience, group or individual the model was interacting with. This last rarely if ever occurs and the result is a performance in a vacuum. Thus, when you would initiate duplicating the model's performance, you would not have a set of responses (from the audience, group or individual that the model was interacting with) to use as the basis for knowing whether your replication of the model's performance was effective. I trust you see the problem here. Words of wisdom - sort of out of my line but here you are: remember that the destination is simply an excuse to set yourself in motion and to discover what there is between where you presently are and where you believe your destination is located. Very much like climbing technical rock - there is absolutely nothing on top, so you had better enjoy getting there. John