The Meaning of Consensus in Science

David Morrison

Four Revolutions in the Earth Sciences: From Heresy to Truth. By James Lawrence Powell. New York: Columbia University Press, 2015. ISBN 978-0-231-16448-1. 384 pp. $35.

The scientific skeptic recognizes that even generally accepted scientific ideas might be wrong, but as skeptics we also need the intellectual tools to understand how consensus is achieved in science and what it means. James Lawrence Powell1, who has written several excellent books on contemporary issues in the geosciences (including Night Comes to the Cretaceous, Grand Canyon, and Dead Pool) and global warming (The Inquisition of Climate Science), was challenged to write this book by a friend who asked him how he could be so sure of the reality of global warming. After all, “science has been wrong before.”

Four Revolutions in the Earth Sciences book cover

The simple answer is that scientists accept theories when the data demand that they do so. However, the process is not simple, and there have been times when entire scientific communities pursued dead ends and persisted in error even in the face of transformative new data.

To explore this question, Powell has analyzed four fundamental discoveries in the geosciences that are central to understanding our planet: the age of the Earth, the formation and evolution of its major landforms, the role of cosmic impacts, and the stability of our atmosphere and climate. Powell has written a compelling narrative of the ideas and personalities that shaped these big questions over the past two centuries. This is first-rate story telling, with heroes, villains, and the often-unexpected discoveries that created revolutions in our concept of our planet.

Some major themes recur in these histories. One is the conflict between traditional geoscientists and their outside critics, especially physicists. The descriptive and historical scientists often closed ranks against anyone who was not steeped, as they were, in the details. The outsiders, in contrast, insisted that their critics simply did not understand the requirements of “natural philosophy.” Another recurring issue is whether descriptive data can be trusted in the absence of a full theoretical underpinning. The history of continental drift provides an example, with the compelling empirical arguments of Wegener discounted because he proposed no viable way for continents to move. Finally, there was the influence of some senior scientists who, having achieved positions of influence or authority in universities or government agencies, effectively closed the door on innovations that challenged their magisterial position.

Let’s look very briefly at each of these revolutions before returning to fundamental questions about whether we can be confident in any of these areas that we have finally got it right. This is especially important when Powell discusses climate science, where clearly many members of the public don’t think we have it right.

Age of the Earth

In eighteenth – and early nineteenth-century England, James Hutton and Charles Lyell defined the principles of uniformitarianism, asserting that the same processes that act today have always done so, in an endless cycle, with “no vestige of a beginning, no prospect of an end.” To them, the question of the age of the Earth was irrelevant. The uniformitarian philosophy was challenged by Darwin’s concept of biological evolution, but the most fundamental critiques came from physicists led by Lord Kelvin. He asserted that uniformitarian principles violated the laws of thermodynamics and that a sort of geological perpetual motion machine was impossible. Kelvin went further and used the principles of thermodynamics to estimate both the age of the Earth (the time needed for it to cool) and the age of the Sun (its lifetime generating energy by contraction). Both numbers were in the tens of millions of years, and the physicists asserted that older ages were “impossible.”

At the turn of the twentieth century, the situation changed dramatically with the discovery of a new energy source. Radioactivity falsified Kelvin’s calculations for the age of the Sun and provided ways to calculate the age of the Earth. These calculations were not easy; it required decades of improving technology to assemble the tools to do the job, and ultimately the age of the Earth was tied to the age of the oldest meteorites. The current best value is 4.56 billion years.

Plate Tectonics

One long-lived offshoot of uniformitarianism was the assumption that continents and ocean basins were permanent and immobile features of the planet. In the first decade of the twentieth century, Alfred Wegener challenged this assumption by his investigation of the Atlantic Basin, where not only do the coastlines of Africa and South America line up like pieces of a jigsaw puzzle, but the detailed geology and fossil records on both sides demonstrate continuity across what is now a 4,000-km ocean gap. The conventional explanation for these similarities (if they were acknowledged at all) was that a long land bridge had connected the two continents. Although such fanciful land bridges violated the principles of physics by alternately rising and sinking, this problem was ignored. Wegener, as an astronomer and meteorologist and even—God forbid—an Arctic explorer, became a non-person. By midcentury, the words Wegener and continental drift had been purged from geology texts. Geological history was all a matter of rising and sinking motions, never sideways.

A way out of this intellectual dead end came from new disciplines of geosciences and new people entering the field, using the remarkable advances in data collection that followed World War II. Sea-floor maps clearly showed the mid-ocean ridge where the two sides of the Atlantic had parted, and paleomagnetic studies mapped the alternating bands of magnetism that traced the movements of continents. Today we can even measure their motion directly using GPS. In 1965, the sea-floor spreading idea was extended to plate tectonics, now the most fundamental concept of terrestrial geosciences. However, as late as the 1970s, several of the most respected senior scientists refused to accept any of this “nonsense.” Powell writes, “those who continue to insist they were right the first time, in spite of accumulating evidence to the contrary, are stuck forever with their original belief.”

Cosmic Impacts

Although pioneering work at the end of the nineteenth century had suggested that lunar craters were impact scars rather than volcanic features, these ideas had almost entirely disappeared by the mid-twentieth century. It was geological dogma that geologists should look at what was happening at their feet and not indulge in speculation that some cosmic intervention could solve their problems. But after mid-century a new generation of interdisciplinary “planetary scientists” was stepping up. Ralph Baldwin measured thousands of lunar craters and concluded they could not be volcanic; Gene Shoemaker’s detailed field work at Meteor Crater demonstrated beyond a doubt that this particular feature, at least, was an impact scar; Nobel Laureate Harold Urey changed his major research direction to study theories of lunar origin; and a young Bill Hartmann used astronomical data to correctly estimate an age of the major lunar features of three billion years. However, most geologists ignored these upstarts. In 1964, on the eve of the first NASA lunar missions, attendees at a major conference on Geological Problems in Lunar Research were nearly unanimous that the lunar craters were volcanic. One leading speaker declared that meteorite impact is a “trivial process in affecting both the geneses and development of almost all major lunar features.” In a matter of months the Ranger, Surveyor, and Apollo missions began to arrive at the Moon, and the rest is history. New data again provided the key.

Recognition of the role of cosmic impacts in terrestrial and planetary history is now nearly universal. Powell devotes the second part of this story to ideas that are more controversial: the proposed giant impact origin of the Moon, and the role of cosmic impacts in the biological history of our planet. His earlier book, Night Comes to the Cretaceous, is one of the best accounts of the controversies surrounding the Chicxulub impact and the KT mass extinction. There are few scientists today who do not credit an asteroid impact with the extinction of the dinosaurs, but questions remain about the details, particularly if the KT is the only mass extinction associated with an impact. It is dangerous to generalize from a sample of one.

Global Warming

For many readers the section on global warming will be the most interesting topic in Powell’s book. Here he steps back from today’s headlines for a unique guide to the early history of the science behind global warming—ideas that stretch back to the nineteenth century.

The greenhouse effect was first suggested by the discovery that light rays and heat rays have different ability to penetrate glass, and were thus capable of heating the inside of a glass enclosure exposed to the Sun. In 1836, this idea was extended to the role of the atmosphere in heating the Earth, and the gases carbon dioxide and water vapor were identified as playing a similar role to the greenhouse glass. At the end of the nineteenth century, Svante Arrhenius first calculated the magnitude of the greenhouse effect and estimated that doubling the amount of CO2 would cause a temperature rise of about 5°C. There was a problem, however: water vapor and carbon dioxide had overlapping infrared spectra, and water vapor seemed to be the dominant absorber. Therefore most meteorologists early in the twentieth century concluded that CO2 made only minor contributions to the atmospheric greenhouse.

Meanwhile, a key insight emerged from joint consideration of radiative and convective processes in the atmosphere. When moving heat, convection in a fluid dominates over conduction or radiation, and the lower atmosphere of the Earth is convective. Water vapor is largely confined to the lower regions, but CO2 is the dominant absorber in the stratosphere, above the convection layers. This leads to the following simple picture of the effect of CO2 on temperature.

1. On average the Earth must be in equilibrium, radiating the same energy it receives from the Sun, which is equivalent to that of a “bare rock” planet about 30°C cooler than the current surface temperature.

2. The effective radiating layer is primarily dependent on the carbon dioxide concentration in the upper atmosphere, where there is little water vapor. In the troposphere, heat is transferred primarily by convection and is largely independent of the infrared absorbing gases.

3. Therefore, the equilibrium surface temperature can be found by using the temperature gradient (lapse rate) in the convective troposphere to maintain the correct temperature in the stratosphere at the effective radiating layers.

4. One of the key signatures of greenhouse heating is that while the surface temperature rises, the stratospheric temperature actually declines.

Unfortunately, by the time this concept was developed the meteorology community had largely abandoned the idea that CO2 was a major contributor to surface temperature. They took note only when evidence accumulated that the temperature was in fact rising in parallel with the increase in atmospheric CO2, which in turn was tied to the release of carbon from burning fossil fuels.

The actual task of modeling a complex system like the Earth’s atmosphere is formidable, taking into account both global atmospheric circulation and radiation, modulated by ice, clouds, and weather. Today’s climate scientists are working at levels of detail unimaginable fifty years ago. But it is valuable for Powell to point out that the fundamental concepts of global warming were established even before World War II.

Powell’s original question was how we know anthropogenic global warming (AGW) is real. The simple answer is that the data demonstrate it. It has become almost impossible to deny that the planet is warming, based on multiple independent evaluations of the measured surface temperatures, global remote sensing from orbit, and large-scale environmental changes, especially in the Arctic. But how do we know that the primary cause of global warming is CO2 from burning fossil fuels? Because both theory and data show that temperatures rise with atmospheric CO2 concentrations, and the isotopic analyses of the added CO2 demonstrate that it came from the burning of old carbon (fossil fuel).

To go beyond these simple truths, particularly to predict the course of future warming, scientists must delve into complex calculations and computer models that are difficult for laypersons and even most scientists to judge. Here we have an advantage over the historical examples Powell cites, since there are thousands of climate scientists around the world working on these problems, checking each other’s work. For example, the Intergovernmental Panel on Climate Change reports represent researchers from 195 countries, and UN rules require unanimity in approval of the report language on climate change. Today we are flooded with climate data, and ultimately it is data, not models, that matter in science.

In his final paragraph, Powell summarizes:

Of course scientists have been wrong in the past, but they did not stay wrong. As new data arrive, scientists changed their position, some enthusiastically, some begrudgingly. A few were unable to make the transition, going to their graves clinging to their long-held positions. A new generation, with no allegiance to the old ways, came along to replace them. Thus data transform heresy into truth.


Powell, a geologist and geochemist, has been president of both Reed College and Franklin and Marshall College, acting president of Oberlin College, and director of the Los Angeles County Museum of Natural History. He also served twelve years on the National Science Board, and he is now president of the National Physical Sciences Consortium.

David Morrison

David Morrison is a NASA space scientist and Skeptical Inquirer contributing editor.