Faultlines in geology

The theory of plate tectonics is the geosciences’ grand unifying theory, but it was discovered in bits and pieces. So does it count as a scientific revolution?

Illustration by Ben Coy

Illustration by Ben Coy


Thomas Kuhn's The Structure of Scientific Revolutions covers science prior to 1962, but what about later transformations in science? In this column, we look at what might be considered modern scientific revolutions (here's our 101 on what makes a revolution). This month, Diana Crow investigates how tectonic plates shook geologists' world.

After nearly six weeks of drafting and redrafting maps of the mid-Atlantic seafloor, Marie Tharp was finally ready to show her boss what she had found.  It was 1952, and the Lamont Earth Observatory at Columbia University in New York had sponsored six years’ worth of sounding expeditions by the research ship Atlantis. Oceanographers of the previous century had discovered a Mid Atlantic Ridge the old-fashioned way — by tying heavy objects to long ropes, throwing said heavy objects over the side of the boat, and seeing how much rope it took to hit the ocean bottom.

But after WWII, ambitious American oceanographers wanted to use sonar to map the underwater ocean range in detail. Tharp’s task was to translate the data into a chart showing the elevation of the seafloor. But what she found seemed impossible.

Between the peaks of the ridge, lay a vast and profoundly deep rift valley in the ocean floor.

“When I showed what I found to Bruce, he groaned and said, ‘It cannot be. It looks too much like continental drift,’” Tharp later wrote. “At the time, believing in the theory of continental drift was almost a form of scientific heresy... Bruce initially dismissed my interpretation of the profiles as ‘girl talk.’”

Marie Tharp, Bruce Heezen, and Doc Ewing documented the features of the Atlantic seafloor in detailed maps like this one.   NOAA  (public domain)

Marie Tharp, Bruce Heezen, and Doc Ewing documented the features of the Atlantic seafloor in detailed maps like this one. NOAA (public domain)

The Earth was supposed to be eternal. Up until the twentieth century, most people thought of the continents as permanent features of the Earth’s surface. And yet, the land is pockmarked with mountains, volcanoes, and other signs of stone in motion.

A few explanations had been suggested. In the late 1800s, Austrian geologist Eduard Suess proposed that the Earth — once a ball of incandescent molten rock — was slowly cooling and contracting inward. Mountains were essentially wrinkles caused by the Earth’s shrivelling. Others had proposed that the continents had a tendency to bob up and down, like apples floating in a barrel, without moving laterally.

But in 1912, a German meteorologist and Arctic adventurer named Alfred Wegener published a paper advocating a different idea. Wegener was no geologist, but it didn’t take a geologist to notice that the eastern coast of South America looks very much like it should fit into Western Africa. What if, Wegener asked, all the landmasses of the world were once a single 'urkontinent' that drifted apart over millions of years.

Most geologists were hesitant to endorse an outsider’s notion. Wegener’s musings about mechanisms didn’t make much sense, and it was difficult to imagine such enormous landmasses migrating without someone noticing. Critics jokingly nicknamed the hypothesis continental drift.

The idea sparked fierce debates in the 1920s. Historian Naomi Oreskes attributes this standstill to a lack supporters in a key faction of 20th century science — the Americans.

In an anthology of essays called Plate Tectonics: An Insider’s History, Oreskes argues that American scholars in the late 19th and early 20th century had a very particular vision of what science ought to be. “Americans believed good scientific method was empirical, inductive, and modest, holding close to the objects of study and resisting the impulse to go further,” she wrote. In this view, theories built on inference were symptoms of European arrogance and therefore ‘bad science’. Continental drift made big claims about the nature of the earth beneath our feet without matching mountains of data. And perhaps most damningly of all, the idea had come from a man best known for gallivanting around the Arctic in hot-air balloons.

In 1930, Wegener died on a Greenland ice shelf while trying to resupply a research expedition, but continental drift lingered on as a fringe hypothesis. Students of the 1930s and 1940s heard about 'continental drift', but usually as the butt of a joke.

But no matter how much Americans wanted to dismiss it, continental drift did address several unexplained observations of the sort science philosopher Thomas Kuhn would later call ‘anomalies.’ (For more detail on Kuhn and his ideas, see our Paradigm Shift 101 explainer.)

Palaeontologists found fossils of the same species in regions separated by thousands of miles of modern ocean. These discoveries supported the continental drift hypothesis.   Osvaldocangaspadilla/Wikimedia Commons  (public domain)

Palaeontologists found fossils of the same species in regions separated by thousands of miles of modern ocean. These discoveries supported the continental drift hypothesis. Osvaldocangaspadilla/Wikimedia Commons (public domain)


Anomalies, in Kuhn’s view, are the first sign that a reigning scientific worldview is about to fall — and geology had anomalies aplenty. Continents moving vertically could account for marine fossils found on mountain tops, but it didn’t explain why coastlines thousands of miles apart had matching fossils. And no one was quite sure where earthquakes fit in.

But, anomalies can also persist in plain sight for centuries before the scientific consensus changes.

In the mid-1950s, new evidence emerged that prompted a few scientists to take a second look at continental drift. However, most geologists didn’t change gears right away. When the Lamont Observatory announced the discovery of the rift in the Mid Atlantic Ridge in 1953, they presented it as evidence in favour of the expanding earth hypothesis, and Marie Tharp didn’t receive credit for the find.

Meanwhile, researchers from the Scripps Institute for Oceanography were measuring the magnetic polarity of rocks in Pacific seafloor. To their surprise, they found rocks’ magnetism alternated in a zebra stripe pattern.

Understanding the zebra stripes required combining two separate hypotheses. In the early 1900s, French physicist Bernard Brunhes measured the magnetism of ancient basaltic lava flows and found that the direction of the magnetism seemed to vary with the rocks’ age. He surmised that the Earth’s magnetic field must flip every several thousand years. Decades later, in 1960, a geologist named Harry Hess proposed that mid-ocean ridges might form in an area where magma from the Earth’s mantle oozed up through cracks in the crust. However, he had no direct evidence for this seafloor spreading. The Scripps team published their “zebra stripe” paper in 1961, and, at first, no one put the pieces together.

Two years later, a team from Cambridge University — Frederick Vine and Drummond Matthews — proposed that the zebra stripes formed when magma rose up through a rift in the seafloor, immortalising the magnetic field’s direction when it cooled into rock. (Unbeknownst to Vine and Matthews, a Canadian geophysicist put forward the same idea a few months earlier, only to get this response: “His idea is an interesting one — I suppose — but it seems most appropriate over martinis, say [rather] than in the Journal of Geophysical Research.”) Even after Vine and Matthews’ paper, few were convinced. Matthews, who was a first year grad student, started looking into other potential thesis topics.

Then, in 1965, the stars aligned. Both Hess, the originator of seafloor spreading, and Canadian geophysicist J. Tuzo Wilson took sabbaticals at Cambridge during the same semester. Wilson was one of the most eminent voices in favour of continental drift. Two years prior, he had proposed that the Hawaiian archipelago could have been formed by an oceanic plate slowly sliding over an especially turbulent hotspot in the Earth’s mantle. Again, there was no direct evidence for hotspots, but it did account for volcanic islands that sprung up far from any alleged plate boundary. By 1965, Wilson was working out another key idea: plates sliding past each other might be responsible for earthquake-prone zones.

The San Andreas Fault in California is one example of a transform fault, where two plates are sliding past each other.   John Wiley/Wikimedia Commons  (BY-CC-A-3.0)

The San Andreas Fault in California is one example of a transform fault, where two plates are sliding past each other. John Wiley/Wikimedia Commons (BY-CC-A-3.0)

Wilson, Hess, and Vine discussed the idea. If Wilson was correct, there should be an underwater ridge marking a transform plate boundary near the border between the US and Canada. Hess pointed out that someone (the Scripps team) had already made maps of the seafloor there, so Vine ran upstairs to the library and brought the maps down where all three could see. And there was Wilson’s ridge. “All three of us stared at it in amazement,” Vine later wrote. “Despite the fact that this diagram had been in the literature for four years, no one seems to have noticed this symmetry.” The zebra stripes had struck again.

After 1965, the idea built up momentum. Earthquake data from the Mid Atlantic trench backed up Wilson’s transform fault idea, and other researchers were able to reconstruct continents’ past trajectories. Even then, the theory of plate tectonics wasn’t complete. Two more key papers — one by Dan McKenzie and Bob Parker of Cambridge and one by Jason Morgan of Princeton — debuted in 1967, independently confirming the calculations behind proposed plate movements.

Rather than being driven by direct observation (like a theory in physics or biology), the young geologists in 1960s plate tectonics research seemed to latch on to the idea because it just made sense. “I certainly would not describe Jason’s and my activities in 1967 as hypothesis testing; as soon as I realized that earthquakes and their mechanisms were expressions of plate tectonics, I knew I was right!” McKenzie wrote in Plate Tectonics: An Insider’s History. “The great and immediate success of the theory was the result of everyone else reacting in the same way.”

Of course, not quite everyone was reacting with enthusiasm. It took time for the plate tectonics camp to work out the mechanics, and they didn’t have direct data confirming the inching movements of continents until geodetic surveys conducted via satellite in the 1980s. Many geologists were justifiably skeptical.

But by 1968, the plate tectonics paradigm had arrived at its mature form: The Earth’s surface is covered with stone plates that float on top of a layer of molten magma. However, those plates aren’t permanent: they drift, split open, crash into each other, and can even melt in small hotspots. These movements explain earthquakes, mountain ranges, volcanoes, the shapes of the oceans, and many aspects of the fossil record.

Mid-ocean ridges mark the boundaries of many tectonic plates.   NOAA/Wikimedia Commons  (public domain)

Mid-ocean ridges mark the boundaries of many tectonic plates. NOAA/Wikimedia Commons (public domain)


Today, it’s hard to imagine the geosciences without plate tectonics. Before the 1960s, geology had been all about traversing the planet, gathering rocks, and studying them. Geophysicists were few and far between. By the end of the decade, geophysicists were leading the way with unifying models of plates’ motions and reconstructions of ancient supercontinents.

Plate tectonics differs from many older paradigms in that it can’t be traced back to a single leading thinker or a single exemplar paper. Many accounts tend to focus on Wegener, but his ideas on continental drift were insufficient. Instead, new ideas had to be hammered out through a series of papers, conference presentations, and conversations between scientists.

Thomas Kuhn’s book The Structure of Scientific Revolutions first appeared in print in 1962, right around the time that seafloor spreading was picking up steam. Many adopted the dramatic Kuhnian paradigm shift language. Naturally, observers labeled the rise of plate tectonics as a scientific-revolution.

Some historians and philosophers of science have pointed out that plate tectonics isn’t a 100% match for Kuhn’s description of a scientific revolution, largely because plate tectonics grew out of a series of incremental papers.

Still, there is no doubting that the theory of plate tectonics is monumental. And, in broad strokes, its rise does bear most of the hallmarks of a Kuhnian revolution. As geologist Peter Molnar cheerfully observed in a 2001 essay: “When Kuhn wrote a book about scientific revolutions, and others deemed plate tectonics an example, it felt good to be part of one, especially a nonviolent revolution.”

Edited by Tessa Evans