Patrick I take modeling ala NLP (using the description of modeling offered in Whispering as a model) to be a content free strategy for identifying, assimulating to criterion, coding and testing a set of patterns known as the resultant model. If this is acceptable, then the answer to your proposal is simple enough. Your proposed three levels: application modeling of excellence modeling of modeling (meta modeling in your terms)   fail in my opinion - since modeling ala NLP is content free, it doesn't matter whether the modeling involved is modeling of Erickson's hypnotic patterns or Bandler and Grinder's modeling of Erickson's hypnotic patterns. Additional comments: 1. you are arguing that NLP modeling and the philosophy of science are if not identical at least at the same logical level and of the same logical type.  Interesting proposal and one which Carmen and I discuss in RedTail Math but such a leap requires an explicit argument - please present it. 2. You state that meta programs are NOT patterns but distinctions. OK, Whispering offers an explicit representation for pattern, what is your explicit representation for distinction? 3. I am still in the dark about what brand of meta programs you are requesting that I offer you historical information about. I have requested from you several times a relatively sensory grounded description (including the suggestion that you present the "distinctions" called meta programs in the presentation of pattern format detailed in Whispering). You have indicated that in your opinion, this format is inappropriate. Fine, since my requests to you for you to present a description concrete enough to identify what you are referring to have gone unanswered, I take it that you are either unwilling or unable to do this. Others have attempted to assist in developing these threads by posting what they take to be meta programs - these amount to lists of "distinctions". These list themselves consist of a string of  nominalizations (as well as other logical types of language) - none of them defined, without concrete examples. While I appreciate the positive intentions behind such efforts, they simultaneously reveal one of the maladies that Whispering was written to challenge - the lack of precision in description with the resultant enormous waste of heat and time in exchanges that never touch ground and for which it is literally impossible to determine what is being proposed and what, if any, differences actually exist between the parties to the exchange. As long as NLP accepts such wind as meaningful, it is unlikely we will advance our work. This thread (now present in several different threads on this website) originated with your request about the origins and development of meta programs in NLP. So, for whatever it is worth, here's the story. Bandler and I had finished a session of some 36 hours (essentially writing the complete - 95% at any rate - of volume I of The Hypnotic Patterns of Milton H. Erickson, M.D.). Frank Pucelik dropped by early in that morning to check on how we were doing. In fact, we were totally blitzed and enormously pleased with what we had succeeded in doing. When Frank arrived, I decided to play a little game (as posted by someone in these threads, as much to entertain myself, Richard and Frank as for research purposes). The game consisted in exploring the interface between language and the primary representational systems (VAK). I proceeded by asking Frank to put himself in a quiet state with excellent connections to his internal representations and simply report his experience when presented with various verbal stimuli. I then went through a series of prompts and received from Frank his descriptions of the internal processes by which he processed (or "understood") these prompts. The results were fascinating. The three of us enjoyed the game immensely. This process went on for some hours. Afterwards, we excitedly decided to pass the task of continuing (and deepening) the patterning we had been exploring by assigning the task to some of our students. In the period that ensued, we were amazed by the inability of our students to maintain the distinction between the processes by which these various prompts were processed and the near compulsion of the students to seize upon the prompts themselves as conscious points of reference - that is, content that offered the illusion of stability. We joked among ourselves at many points in the activities that at least we had discovered that our students were not capable of consistently distinguishing between form (or process) and content. The same appears to be an accurate observation for the present situation as well. I will simply invite the readers to re-read the sections in Whispering concerning the distinctions between content and form (process)and especially its importance for the ethics of the practice of NLP. To decide as Dilts does that there is some hierarchy of (what - values???) nominalizations that represents the ideal or standard organization of a human being - equally in the application of meta programs - and act as if these content models are part of NLP is: 1. unethical 2. a violation of what Carmen and I propose in Whispering as the defining characteristic of NLP. Arguments such as the number of executives or companies that have (presumably) happily bought services based on such content distinctions (Bailey, Merlevede...) or that they are successful in inducing change are entirely beside the point (see especially pages 305 - 308 in Whispering. There are enormous amounts of money wasted by companies around the world on nonsense. Finally, I want to be clear that the original "game" played by the three of us some decades ago is an exquisitely important "game". What is clearly missing - not only in NLP but in Cognitive Psychology and Cognitive Linguistics as well is precisely the connection between language and experience (remember - FA). Perhaps in part because meta programs can be easily applied and create (in my opinion) the illusion of patterning, the original intention of the "game" has not been pursued. I can think of few pieces of research more important than the mapping of the processes at the interface between FA and initial partitions over FA created by language structures - alas, this critical topic lies idle while we prattle on about content distinctions. All the best, John