Extraordinary Science; Incommensurability of Scientific Paradigms Don L. Anderson Seismological Laboratory, California Institute of Technology, Pasadena, CA 91125 dla@gps.caltech.edu

“"...conventional wisdom is difficult to overturn. After more than 20 years some implications of plate tectonics have yet to be fully appreciated by isotope geochemists… geologists and geophysicists."

“"A myth is an invented tale, often to explain some natural phenomenon...which sometimes acquires the status of dogma...without a sound logical foundation. It is a dogma that has distorted thinking about the Earth for decades. In science this is an old story, likely to be repeated again, as the defenders of conventional wisdom are seldom treated with the same scepticism as the challengers of the status quo...the dogma has been defended with false assertions, defective data, misconceptions and misunderstandings, and with straw-man arguments... The justification...boils down to a statement of belief, an opinion, rather than a deduction from observations."”

“"[scientists]  are reluctant to abandon cherished concepts they grew up with and have vigorously defended during their education and research careers."”

Richard L. Armstrong, 1991

The distinction between myth and science is subtle². Mythic thinking places special emphasis on a selective conjecture, based typically on the initial observation or explanation of a phenomenon, which is thereafter given privileged status over alternate interpretations. This also describes a paradigm. In myths, phenomena are attributed to things, substances or beings, usually unobservable, rather than processes. Examples include gods, demons and fairies; in science we have had aether, calorics, phlogiston, sea people, strings and plumes. Concepts in geoscience are often mythic²; fixity of continents, fixity of oceanic islands, uniformitarianism, geosynclinal theory (vertical tectonics), mantle plumes, mantle reservoirs, undegassed lower mantle, and so on. Geomyths  and paradigms can focus thinking but can retard further progress by diverting attention away from the real cause of a phenomenon and dismissing contrary views. An abandoned paradigm is easily recognized in hindsight as a myth, or a collection of ad hoc stories; failures of current conventional wisdom are not so obvious.

Myths and vague theories are impossible to falsify. In fact, logicians have shown that no theory can be falsified by data; observational evidence cannot decide between conflicting myths or paradigms. Data, logic, rational discourse, simplicity, and good intentions–the essence of progress in ordinary science–have little to do with the overthrow of a popular faith-based paradigm, or myth³.

Some of the implications of plate theory are still not fully appreciated: plates are not really rigid, the mantle is not homogeneous or entirely subsolidus, magma fracture and lithospheric architecture control locations of volcanic chains,  plates organize mantle convection–not vice versa, delamination of continental crust and subduction of aseismic ridges create fertile streaks in the mantle, recycling creates low melting point blobs and so on. Some of these are simple reversals of cause and effect from conventional wisdom. By dropping such adjectives as homogeneous, uniform, rigid, undeformable and well-stirred the plate paradigm explains many of the features that were thought to be outside its purview.  But the words plate tectonics, plate, convection, mantle, upper mantle and lithosphere mean different things to different workers, and therein lies the rub.

We are currently on the cusp of a major paradigm shift in Earth sciences. Ideas about continental growth and breakup, mantle structure and dynamics, reference systems, geochemical reservoirs and magma chemistry are changing; weaknesses and paradoxes of the Standard Models of mantle geochemistry and geodynamics are multiplying. Most of this is invisible in the flagship publications, which document perceived progress and consensus views rather than problems in reigning paradigms, or, heaven forbid, questioning of conventional wisdom. Peer review and editorial and funding policies–the infrastructure of a paradigm–are effective in trimming both tails off the Gaussian that represents the allowable range of publishable wisdom. How do we recognize and deal with an impending  paradigm shift–a scientific revolution⁴?

Philosophers of science^4-6 analyze extraordinary science; the great intuitive leaps that spice up the history of science. This is not deductive science or even discovery science.  Extraordinary science consists of bold formulation of hypotheses, which are then subjected to attempts to falsify them. Thomas Kuhn⁴ saw the history of science as a succession of paradigms that are pitted against one another rather than against objective scientific truth. The scientific method does not apply when conventional wisdom starts to fall apart. Defenders of an existing paradigm use their own ground rules and assumptions to attack the new models³, and rational discourse plays little role^4-6. New conventional wisdom requires new terms, new assumptions, and, quite often, different people.

One clue is that the words paradox, enigma, surprising result, problem, anomaly, crisis and unanticipated start to show up more frequently in published papers⁷. When paradoxes and complexities multiply it may be time for a new paradigm. A certain amount of humility and befuddlement is reflected in these words. However, the arguments that are used to defend the status quo in spite of overwhelming evidence to the contrary are also telling; “most people believe…”, “truth, like beauty we suppose, is in the eye of the beholder…”,  “the approximations–or the data–are not good enough”, “…the effects are too small to detect”, “…evidence to the contrary does not disprove a theory”, and “do not throw out the baby with the bath water”. Many of the arguments are based on ignorance of the field or on tradition; "I admit there are a lot of problems with the hypothesis, but I'm going to stick with it until something better comes along", “I spent a lot of time trying to find alternatives…”, and “we must point out, however, that many other geophysicists have interpreted the available data as supporting plume emplacement..”. Also, after considering selected pieces of several strawman models, but not any of the mature alternative models “based on an unusually comprehensive interdisciplinary database, we maintain that the plume model…explains better the available evidence than any of the suggested alternative models….”. This covers the range of rationalisations to which the defenders of a faltering paradigm appeal.

Most productive scientific theories are simple, or as simple as they need to be, and they explain more than they assume. If every new observation results in a new modification to the theory, and the theory becomes more and more complex, more vague, less general, and less testable, then it is time to reconsider the options.

Scientists can debate different theories, but those who have different worldviews talk past one another; they cannot a priori arrive at agreement given their different theoretical languages and assumptions. One group argues from the perspective that the phenomena or object they are studying actually exists, be it aether, phlogiston, caloric or plume, and their job is to refine the details. The other group knows that it is silly to think about processes as things. They know about waves, oxidation, thermal vibrations, intrusion, recycling, stoping–and about scaling relations such as homologous temperature; they worry about self-consistency. They are not ready to accept that approximations such as homogeneity, uniformity, steady-state, fixity, linearity, elastic and Boussinesq capture the essence of phenomenon such as mantle dynamics. The defense discounts problems by saying all approximations have limitations to reality or that falsified predictions do not falsify a theory; they are assuming they have a good approximation to reality, and that the thing they have constructed in their minds actually exists. But aether is not a good approximation to EM theory no matter how many bizarre properties are assigned to it. Ignoring pressure or entropy is not an approximation to mantle convection, or any thermodynamic cycle. We need to keep in mind that lithospheric stress and fertility variations in the asthenosphere might not be well approximated by rigid–or flexible–tubes to the core. Even if our current ideas about crustal growth, geochemical reservoirs and mantle dynamics are correct, it is useful to remind ourselves of what the philosophers say, and, perhaps, bring back the paradigm of Multiple Working Hypotheses.

References & Notes

1. Armstrong, Richard L. 1991. The Persistent Myth of Continental Growth, Austr. J. Earth Sci., 38, 613-630
2. Dickinson, William R. 2003. The Place And Power Of Myth In Geoscience: An Associate Editor’s Perspective, American Journal of Science, 303, 856 – 864
3. Debate about a paradigm is different than rational scientific discourse; it can get ugly, viz.
   http://www.mantleplumes.org/TopPages/GreatPlumeDebateTop.html
   http://www.mantleplumes.org/TopPages/MagazineTop.html
   http://www.mantleplumes.org/FUA.html
4. Kuhn, Thomas, 1962. The structure of scientific revolutions, University of Chicago Press, Chicago.
5. Lakatos, I. 1970. ‘Falsification and the Methodology of Scientific Research Programmes’, in Lakatos, I & Musgrove, A. (eds). Criticism and the Growth of Knowledge. Cambridge University Press, Cambridge.
6. Feyerabend, Paul,  1975. Against Method: Outline of an Anarchistic Theory of Knowledge ,New Left Books, London, 1975.
7. Try combining these words with plume or geochemistry in a Google search.