The scientific method we're taught about in school is a cartoon: observation leads to hypothesis to prediction to experiment. It's easy to teach and lends itself to simple classroom exercises. But in real life, this kind of mechanical process is valid only for problems like solving a crossword puzzle or testing a circuit board. In real science, where much is unknown—certainly in geology—this method gets you nowhere.
When geologists go out in the field they confront a blooming, buzzing confusion of scattered outcrops, complicated by faulting, earth movements, vegetative cover, bodies of water and landowners who may or may not let scientists wander around their property. When they prospect for buried oil or minerals, they have to make sense of scattered well logs and seismic profiles, trying to fit them into a poorly known model of the regional geologic structure. When they research the deep mantle, they must juggle the fragmentary information from seismic data, rocks erupted from great depths, high-pressure mineral experiments, gravity measurements and much, much else.
A geologist in 1890, Thomas Chrowder Chamberlin, first described the special kind of intellectual work needed, calling it the method of multiple working hypotheses. He considered it the most advanced of three "scientific methods":
- The "method of the ruling theory" begins with a ready answer to which the thinker grows attached, looking only for facts that confirm the answer. It is suited to religious and legal reasoning, in large part, because the underlying principles are plain—the goodness of God in the one case and the love of justice in the other. Today's creationists rely on this method as well, starting in a lawyerly fashion from the bedrock of scripture and seeking confirming facts in nature. But this method is wrong for natural science. In working out the true nature of natural things, we must investigate natural facts before creating theories about them.
- The "method of the working hypothesis" begins with a tentative answer, the hypothesis, and seeks out facts to try against it. This is the textbook version of science. But Chamberlin observed "that a working hypothesis may with the utmost ease degenerate into a ruling theory." An example from geology is the hypothesis of mantle plumes, which is cited as an axiom by many geologists, although a spirited critique is beginning to put the "working" back into it. Plate tectonics is a healthy working hypothesis, being extended today in full awareness of its uncertainties.
- The method of multiple working hypotheses begins with many tentative answers and the expectation that no single answer may be the whole story. Indeed, in geology a story is what we seek, not just a conclusion. The example Chamberlin used was the origin of the Great Lakes: Certainly rivers were involved, to judge from the signs; but so was erosion by ice age glaciers, the bending of the crust under them, and possibly other things. Discovering the true story means weighing and combining different working hypotheses. Charles Darwin, 40 years earlier, had done just this in devising his theory of species evolution.
The scientific method of geologists is to collect information, stare at it, try a lot of different assumptions, read and discuss other people's papers and grope their way toward greater certainty, or at least figure the answers with the best odds. This is more like the real problems of real life where much is unknown and variable—planning an investment portfolio, devising regulations, teaching students.
The method of multiple working hypotheses deserves to be more widely known. In his 1890 paper Chamberlin said, "I am confident, therefore, that general application of this method to the affairs of social and civic life would go far to remove those misunderstandings, misjudgments, and misrepresentations which consitute so pervasive an evil in our social and our political atmospheres, the source of immeasurable suffering to the best and most sensitive souls."
Chamberlin's method is still a staple of geological research, at least in the mindset that we should always look for better answers and avoid falling in love with one beautiful idea. The cutting edge today in studying complex geologic problems, such as global warming, is the model-building method. But Chamberlin's old-fashioned, common sense approach would be welcome in more places.