I retired at the end of 2020 after thirty years as a university professor of geology, but I am in many ways extraneous to the earth sciences, or at least to the orthodoxy within the profession. I remember thinking during one of my first introductory geology labs, more than half a century ago, that I did not belong there. Putting names on minerals felt a bit like collecting stamps, to use Ernest Rutherford’s description of everything that is not physics. My calling was mathematics, or perhaps theoretical physics, with emphasis on theoretical, for I am equipped (cursed?) with a deeply abstract, logical, and indeed impractical mind. There is nothing abstract nor logical about learning to identify microcline, nor is it really interesting. And is doing so an intellectual accomplishment that one could call “science”? It is no more science than, say, telling a VW Bug from a Mini Cooper. So why was I in that lab, and why did I stay? Because my other love was – is – the outdoors, and as a clueless and abstract-minded teenager I had decided that geology would give me the best of both worlds – doing scientific work in the outdoors. I discovered how wrong I had been when, after graduating unenthusiastically as a geologist, I spent about a decade in the minerals exploration industry, with even less enthusiasm. Working outdoors was destroying my love for the outdoors, because what I was doing bored me to death. What I was doing had nothing to do with science. It gave me no intellectual satisfaction whatsoever. And the anguish that arose from my unfulfilled expectations very nearly made me hate nature and the outdoors.
From that nadir followed a decades-long tortuous path, that included a Ph.D. in Geosciences from the University of Oregon and three decades of swimming against the tide, both as a researcher and as a teacher, at the University of Georgia. This path eventually led to what is my only contribution to science: a book published by Cambridge University Press in 2011 with the title Thermodynamics of the Earth and Planets. My goal in writing this book was to present a rigorous and somewhat comprehensive mathematical analysis of planetary processes, beginning from general concepts and equations of physics and building up to the complexities of a messy planet. This is a reflection of my conviction that only mathematical formalism can reveal how the universe works. The leitmotif of the book could be summarized like this: there is only one objective and rigorous way of understanding our planet, and it requires that we stop thinking of geology as a discipline that stands in splendid isolation of physics. For physics is, like it or not, the queen of sciences, the one and only path to understanding nature. And physics, and by implication the universe of which Earth is a minuscule part, cannot be understood without mathematics, which is the only universal, immutable and unambiguous language. Most geologists, in contrast, rely on words to make their arguments, but words can be twisted to mean anything, a fact that is clearly displayed in the temporal evolution of languages. And in the ability of a very few true geniuses to build entire worlds out of words.
Geology is the physics of a very complex assemblage of interacting systems. As all physics, geology should begin with objective, reproducible and quantifiable observations, which should then be placed within a formal mathematical framework. And if geology is to be considered a science, then it requires testability. Geology shares in this respect two handicaps with astrophysics. First, the subject of study, or at least most of it in the case of Earth, is not accessible. Second, the timescales of interest are many orders of magnitude greater than a human lifetime – in fact, than civilization itself. But geology has an additional burden. Whereas astrophysicists can draw from hundreds of millions of samples, for instance in the form of stars of different masses, compositions and ages, geologists have only one Earth, plus a handful of other terrestrial planets. Given this state of affairs, we should recognize that there are substantial limitations to what we can objectively learn about our planet. For every planetary system or process there is a level of detail beyond which claimed findings lack physical significance, because they are not testable. For instance, we have a solid understanding of the fact that the Earth’s mantle is a heat engine. This heat engine transfers heat between a deep reservoir of thermal energy and the surface of the planet. As any heat engine, the Earth’s mantle converts a small fraction of that thermal energy into mechanical energy: plate motion and ultimately gravitational potential energy stored as topography. We know enough about the material properties and phase transitions of magnesium silicates to understand why present-day convection of the Earth’s mantle gives rise to a regime in which the top thermal boundary layer is fragmented into plates that move with some independence from one another, and why, in contrast, the top thermal boundary layer is largely unbroken in the other terrestrial planets. Maybe it is possible to increase the level of detail somewhat beyond this. But I wonder what future philosophers of science will make of the conjuration of ancient continents on the basis of no more than a few radiometric ages, some subjective observations, and the aesthetical inclinations of the author. This is what practitioners of “Tectonics” do, and it is no better than Aristotelian physics or medieval alchemy. My decades of swimming against the tide were spent advancing logical arguments for why a not insignificant portion of what we think that we know about our planet is scientifically suspect. Consider some examples.
Arguments that rely on a few chemical analyses and meaningless adjectives such as high and low (or elevated and depleted, if one wishes to sound more scientific) are repeated ad nauseam, in a seemingly infinite number of papers that are impossible to tell apart from one another, in order to assert that a particular magma was formed in one way or another. The authors of these papers never consider something as fundamental as whether what is claimed is possible from a simple energy balance argument, and very seldom take the trouble to discuss whether the phase relations of the system are known well enough to make their conclusions testable. In the absence of testability, such conclusions are not all that different from astrology. Even more infuriatingly, rather than striving for a general understanding of igneous processes, the authors of these studies go out of their way to emphasize the (supposedly) unique and mind-numbingly trivial details of their pet volcanic field or pluton. This obsessive preoccupation with minute details that lack scientific significance is a defining trait of the geological profession, and the chief reason why I stopped talking to, and collaborating with colleagues, a very long time ago.
There are also claims that it is possible to determine crystallization conditions of metamorphic mineral assemblages with accuracies that a simple mathematical analysis of the underlying thermodynamic equilibria shows to be impossible. In fact, the uncertainties in equilibrium locations that arise from uncertainties in standard state thermodynamic properties and in excess mixing properties in complex mineral structures are often of the same order as the purported values of intensive variables during crystallization. And this is even before one accounts for analytical error, crystalline structure defects, and the ever present unknowable: did the mineral assemblage attain equilibrium, and were those equilibrium compositions then effectively frozen in the minerals that one is analyzing? The fact that one can spend innumerable hours in front of a microprobe and then plug mineral analyses into some sort of black-box software package, and receive at the other end numbers that are supposed to represent the values of intensive variables during crystallization, does not mean that those numbers are worth any more than the bits that represent them on a computer screen.
These are examples taken from the areas of geological research that I am most familiar with, but one can find similarly illusory results everywhere across the earth and planetary sciences. Statistical analyses are used to “prove” results while lacking a solid theoretical foundation for the processes, that can establish a causal relationship between variables. As if simply using some sophisticated statistical technique, deciding on an arbitrary value for a Bayesian “prior”, and setting arbitrary confidence intervals, constituted scientific understanding. Or using fallacious logical arguments. If one finds an outcrop of a certain age in, say, a mountain belt, then rocks of that type formed during that time in the spatial location that is now that mountain belt. But the failure to find rocks of that type and age in another mountain belt does not mean that they did not form there. This is elementary logic: “if p then q” does not mean that “if no p then no q“.
Why are the earth sciences stuck in this nineteenth century frame of mind, in this pursuit of cataloguing? Why do most geologists fail to understand that geology is no more than the physics of an astonishingly complex and fascinating system? Geology attracts a large share of wide-eyed yet deeply mathematically-challenged students, who believe that recording observations (naming rocks or fossils, drawing outcrops on maps, measuring rock unit thicknesses, generating tables of chemical analyses) and drawing cartoonish and untestable interpretations of these observations, is science. Most geology programs perpetuate this trend by establishing risible course requirements in mathematics and physics. This is not only because a large proportion of geology faculty are themselves mathematically and logically challenged, but also because otherwise enrollments would all but disappear and departments would be forced to close. Given the currently popular view of Universities as businesses, this is unlikely to change. I am contented to have done my minuscule part to show how we can do better, and resigned to the certainty that it will not make any difference.
