Beth Cooper says: “I think that MFgeo’s insightful observations should be reposted in yr thread at
Climate Etc, Scottish Sceptic.”
Happy to oblige
Copy of MFgeo’s original posts:

I believe that the fundamental reason that sceptics and non-sceptics can come to radically different conclusions from the same information is that they are operating under fundamentally different, and incompatible paradigms. Despite all of the criticism, much of it apparently valid, that has been directed at some aspects of Thomas Kuhn’s “Structure of Scientific Revolutions” over the past 50 years, his description of “normal science” as exploration of a specific set of identified problems, pursuant to a paradigm has held up extremely well. A Kuhnian paradigm not only provides the theoretical framework for the discipline during the intervals between “revolutions”, but also defines the limits of what constitutes acceptable subjects for investigation within the discipline. Except during “revolutions” the principles embodied in the paradigm are not subject to change or criticism by practitioners of the discipline. A point Kuhn illustrates with numerous examples is that, during the period of crisis that precedes each “revolution”, the two sides may engage in debate that is totally inconclusive because they are talking past each other — either by using the same words to mean different things or by adopting premises that the other side is unable to comprehend. The current situation in climate science is just such a crisis.
The group you categorize as sceptics are either from disciplines outside of climate science, and are therefore interpreting the climate data, model results, etc. under paradigms radically different from those of climate science, or are climate scientists who have recognized the impending crisis that may (hopefully will) lead to a paradigm shift within climate science.
My own situation may shed some additional light on this distinction. My academic training was in the geosciences, but I have spent my career as a design engineer in the computer industry. As one of a small number of completely computer-literate geoscience graduate students in the early 1970s, I was involved in the early days of this sort of model development. I KNOW the excitement of coding a model and getting it to the point where it produces results good enough to present with a straight face at professional conferences; and I know the optimism we all felt about how much better our model results would be with more computing power, smaller grid cells, and better source data. However, I also knew that we were building arbitrary and un-confirmed assumptions into our model code (which is far more pernicious than the explicit modeling assumptions, because the implicit assumptions are not documented and rarely get reviewed or questioned so long as the model is “working”); that we were using parameterizations outside the domain of their known validity (which is not to say the usage was invalid, but rather than the validity was undetermined); and the fact that low-level implementation issues, such as floating-point roundoff, were likely to degrade the accuracy of model results to be far worse than the quoted error bars.
The net result what that I understood — but was unable to convince my colleagues — that the models we were building were superb qualitative tools; however, this whole approach to modeling (of a poorly-characterized system that was known to exhibit chaotic behavior) was INCAPABLE of producing quantitatively meaningful results, regardless of improvements in computing power, grid size, and/or source data. Furthermore, I became increasingly concerned that our results, presented as flashy computer animations (at a time when color computer graphics itself was new and exciting) were extremely persuasive, which created a major risk that people would misinterpret, or, worse, misrepresent, those results as being quantitatively meaningful. While not the proximate cause of my career change, my growing unease about the future of model-based “science” was a significant contributing factor.
It is also worth noting that back in the early and mid 1970s, at least where I was, there was no AGW (or even GW) dogma — temperatures had been declining since the 1940s, and speculative discussion about climate trends ended to be about whether the current interglacial was coming to an end. For the first 20 years after my career change, I thought very little about GW, but in the late 1990s I noticed that lots of people were starting to take AGW as proven. I resumed reading the professional literature, looked as some of the actual data, as was aghast at the degree to which the data failed to support the conclusions. About ten years ago I seriously considered reversing my career change, in hopes of helping to correct the increasingly absurd direction the field was taking. Such thoughts ended as soon as I talked to some former classmates who had stayed in the field — they (and/or I) were no longer the same people. We could scarcely communicate. I regarded the climate system as a subject for study, they regarded it as a solved problem that needed only the last few decimal places to be measured. They would make assertions that to me were physically implausible, statistically unjustified, and/or empirically unsupported; and if I questioned the validity they would direct me to papers that varied between implausible and unconvincing (and, as subsequently shown by climategate, the work of Steve McIntyre, etc., in some cases demonstrably incorrect).
[Thanks – it would be useful if you could elaborate a bit more about the problems you encountered. Mike]
To the extent I remember the details, I believe the problems I encountered were mostly due to differences regarding what sorts of data we considered conclusive, as well as the list of subjects where we felt that conclusive data were necessary.
I questioned whether the concept of “global average temperature” was physically meaningful; they saw no problem even while admitting that temperature is an intensive variable, hence non-additive.
I questioned how they knew that warming was actually occurring, due to issues with the methodology used to calculate surface temperature, as well as the fact that the “unprecedented warming” began in synchronism with the number of surface stations declining by two-thirds; they assured me that the methodology was able to achieve 0.05C accuracy, but were unable to cite any evidence that I found remotely acceptable (physically or statistically) to support this assertion.
I questioned the attribution of late 20th century warming to anthropogenic CO2 when there had been previous of warming, greater than at present, during the Holocene; they disputed whether the previous warm periods (other than early Holocene climate optimum) were actually warmer and/or global, and when I cited archaeological evidence, rejected that as being “anecdotal” rather than a valid climate proxy. Finally they fell back on the fact that with the increased CO2 forcing in the late 20th century it could not be otherwise — that there was a impending, anthropogenic catastrophe even if previous warming and cooling were totally due to natural variation.
I questioned why, during the Pleistocene, temperature shifts occurred 500-1000 years before the corresponding changes in CO2 levels. They admitted this was curious, but were certain that the discrepancy would go away with better measurements.
Since every line of discussion eventually ended back at CO2 forcing, eventually I asked what had elevated this “working hypothesis” (one of several possibilities that got discussed when I was a graduate student) into a “ruling theory” (to use T. C. Chamberlin’s terminology). The answers were all over the place, but in one form or another always came back to physics precluding it from being otherwise. I responded that, in engineering, when nature did not behave according to theory, we might blame our instruments before questioning the theory, but that we never blamed nature!
Another point that seems worthwhile to mention is that another characteristic which Thomas Kuhn identifies in “Structure of Scientific Revolutions” is that practitioners NEVER abandon their paradigm due to empirical results that “falsify” it; they only abandon their paradigm in order to adopt a replacement paradigm that they consider better at explaining the particular empirical results that provoked the crisis. From this viewpoint, the behavior of non-sceptical climate science is understandable: Nobody from has yet proposed an alternative paradigm that is sufficiently superior in explanatory power; or the crisis has yet to get sufficiently severe, from the non-sceptical viewpoint, that a replacement paradigm is worth considering.
A paradigm shift is not necessarily to a new paradigm that is objectively “better” than the old paradigm. What the new paradigm must do is to explain the anomalies that caused the present crisis better in the view of those holding the old paradigm. An example of this distinction is the paradigm shift to plate tectonics (PT) that occurred in geology during the second half of the 1960s. PT explained the very surprising (to then-mainstream geology) features of the ocean basins discovered in the two decades following World War II. For many aspects of geology on the continents, PT provided worse explanations than its predecessor, or no explanations at all; but the focus of attention at that time were ocean floor ages, mid-ocean ridges, volcanic island chain tracks, fracture zones, etc. In just over one decade the situation in geology switched from it being practically impossible to publish papers favoring “continental drift” to being just as difficult to publish papers challenging PT. The later situation became so severe that geologists concerned with the shortcomings of PT had to form their own journal (The New Concepts in Global Tectonics newsletter, http://www.ncgt.org) in order to publish their results. Maybe there should be a “New Concepts in Climate Theory” journal that provides a non-blog outlet for alternative climate theories to be disseminated.