Hi Loren, Given this is the WITW website, I'm not going to explain or defend iWAM here - some of the other readers might dismiss that as "advertising". Still let me answer some parts of your mail which I expereicne as criticism.  a) Context When it comes to context that is a matter of chunking - of course, you can detect several subcontexts in work, and I even "believe" that a well trained NLPer should be able to adapt whatever metaprogram preferences needed to do a job (You can change state, etc). Purpose. However, that's not the point of iWAM.  The point is, even if I could change state and adapt my thinking to be in line with metaprogram distinctions, the question is whether I'm motivated to do so.  For instance, I could put me in a state allowing for some "peace of mind, appreciative of doing noting" (and thus building myself a resourceful state, for instance when I need some patience.  Fine, but in reality I couldn't care less.  I don't like to be patient and being able to change my state to appreciate "peace" is not going to change that.  Let me rather do something which fits my metaprogram preferences! What iWAM aims at it predicting to what degree, given ones metaprogram preferences, a person is going to be motivated by a job.  Its applications are in recruiting, career counseling, coaching and assessment. Limits: iWAM is a questionnaire, with all the advantages and disadvantages of questionnaires - and that is a discussion which has been done almost 50 years ago (see Meehl, "Clinical VS Statistical Prediction").  Is iWAM part of "NLP"? I couldn't care less.  I'm asking that for metaprograms in general, not for thinks which are clear applications, such as iWAM, LAB or some other variations on that theme. Also, i'm trying to figure out how "metaprograms" were discovered and developed from there.  John's mention of the task they give their students is the first useful clue I got. is iWAM modeling part of NLP? It's obvious that doing modeling using iWAM will be quite different from what Bandler/Grinder did.  In stead of needing to visit Erickson, Perls and Satir for dozens of days, all I would ask for is 30 minutes of their time (and I wouldn't even have to travel to be there with them - they just have to go on the net).  What would I learn?  What kind of metaprograms i would want when recruting people having the potential of becoming great therapists.  That's maybe very limioted to you, but then it will only take me a couple of hours as well.  And sure do learn lots of things from it (which may come as a surprise to you). My opinion on modeling: There are at least 3 ways of doing NLP related modeling: 2 are seen by the majority of people to be part of NLP, the third one is statistical analysis (no one has to start doing that, I'm okay to be the only one, that cuts down on competition).  The 2 generally accepted ways are what I call "analytical modeling" and "2nd position modeling".  Grinder's approach is a version of what I call "2nd position modeling". Grinder is dismissing "analytical modeling": looking at an examplar through the known distinctions already coded and seen by most as being part of NLP - to John that is "content" and "unethical".  Fine, that's one map of the world.  As Carmen & John say in WITW: "Let's have a multilogue and we shall see". Patrick