Hey JG, I think the distinctions you've made are excellent and very powerful...I agree that design is not modelling (based upon your distinctions)...and that both have their uses (as in Lewis' paraphrase)...but could you please expand upon the following points? JG wrote: "In design, no such explicit criterion is obvious." mmm...I agree that not ALL design has such an explicit criterion...yet some MAY.  In my experience the design process/designer often has an abstract exemplar they are aiming for...(as you described it may even be synthesized from a number of models over time).  While I agree that this may not have the strong foundation of being based upon a specific and "available" model, these people are not always available.  To limit NLPers to only people they can access for modelling may be a little geographically or socially punitive 8) (although step 1 can often be a nice exercise in building flexibility - I particularly liked your example with Erickson) JG also wrote: "Modeling  further, distinguishes itself in that the person (or team) being modeled has already demonstrated that he/she/they are capable of consistent behavior of the highest quality in the field in which the patterning occurs - phase 1 of the 5 phases is expressly the identification and securing of access to such performances of genius." I think this is an particularly strong distinction...yet based on my point above I'd also like to ask are you suggesting we should only limit ourselves to behaviour and patterns that already exist? And one last question.  Are you binding design explicitly to f2?  Isn't the new-code process of 6 step reframing that your unconscious mind so eloquently wrote on the blackboard an example of unconscious design (specifically step 4)?  This defnly isn't modelling...yet I think this step (and the way the process was created) is an excellent example of design that's largely free of f2 filters...until it is expressed of course 8) When I perform design (in a number of different fields) I often just absorb a range of related input...then forget about it for a while until my unconscious mind let's me know I'm ready to dump some ideas.  I defnly don't think this is modelling...but it's also not necessarily limited to f2 either. roBman