[Martin Taylor 2012.04.15.09.24 CET]
[From Rick Marken (2012.04.14.2007 EDT)
In other words, you can have a model that works exactly as you intended, but is not valid.
The Ptolemaic epicyclic model of the Universe worked exactly as intended, and accurately predicted the movements of the planets to within the tolerance of measurement error. Before the age of telescopes, the Newtonian model didn't predict those movements sufficiently better to account for the shift of belief from one model to the other. Nor did the Einsteinian revolution depend on better description of planetary movements. In each case of a revolutionary shift in belief in the "validity" of the model, the shift was due to the ability of the new model to account for a wider range of phenomena than the older model. We now question the validity of the Einstein model, not because it fails to predict accurately -- it predicts extremely accurately under all tests to date -- but because there is another model which predicts equally accurately in its own realm (quantum chromodynamics) and which is incompatible with Einstein. That is why people are interested in the possibility that string theory might be more valid.
My own belief in PCT isn't based on the accuracy of tracking studies, but (apart from the philosophical and thermodynamic necessity of control of perception) on the way it is so easily applied in the everyday world outside the laboratory. The very names of the two current threads on CSGnet "School's in" and "Surfs up" attest to this. My own recognition that my theory of dialogue was a special case of the wider-ranging PCT led to my initial acceptance of PCT some 20 years ago.
All of which leads me to ask whether perhaps the problem in getting PCT generally accepted has nothing to do with the ability of PCT models to generate accurate simulations of lab experiments, and is more based on the fact that people studying the implications of the phase of n-th level epicycles have no interest in the bending of light by the sun. PCT doesn't (yet) have as much to say about the observed data of neuroanatomy, criminal rehabilitation, psychotherapy, peacekeeping among cultural and national groups, fads and fashions, predator-prey relations, competition and cooperation among plant species, etc... as do the people who study these important problems isolated from the rest of the world, but it does have a basic framework within which all of these begin to make sense, and within which it is probable that the specialists could make much better headway than they currently do.
PCT suggests why expertise in a specific domain leads a person to resist a theory that would require a fundamental shift in the foundations of their thinking about that of which they know more than almost anyone else. Would an expert in the astrological effects of the interplay of third level epicycles of Saturn with the retrograde movement of Venus have taken kindly to a Newton telling him that no such epicycles exist? Tim Carey has told some of us that when he used the "Method of Levels" in psychotherapy in Scotland he reduced the waiting list from 15 months to zero, but when he left Scotland, the method was banned and potential practitioners told they would be fired if they used it. It might be informative to run MOL with those who determined it should be banned, might it not?
It is now 40 years since I first argued mathematically that breadth as well as accuracy of description was an indicator of model validity, and that the complexity of the model as perceived by the individual, based on the individual's background, also determined its subjective validity (the argument from 1972 is at <http://www.mmtaylor.net/Academic/ockham.html>\). PCT has great breadth, and great accuracy and prediction power in the few areas where it has been stress-tested, but in most narrow well-studied areas the existing models provide at least as good predictive accuracy. Why should a person with years of specialized experience want to shift to a novel PCT base for thinking about the narrow topic? Why should a student take the advice of some one versed in PCT over the advice of many people believed to be authorities on the student's area of interest?
It's a conundrum for which I have no good answer.
Martin