Here’s an essay on what we think we know and why we shouldn’t be so sure. After I wrote it I read and enjoyed Jonah Lehrer’s piece “The Truth Wears Off: The Decline Effect and the Scientific Method” in the New Yorker, which explores similar themes. As an aside, this essay might help explain why I was a lot less inclined than many of my lefty friends to mock Donald Rumsfeld for his famous “known unknowns” remark, though I did appreciate the rendering of that remark into free verse.
In 1813, the noted ornithologist John James Audubon witnessed a phenomenon that has become legendary, both for its size and for its somber lesson: the passing of a flock of passenger pigeons. It lasted for three days and nights, stretching from one horizon to the other, darkening the sky like an eclipse. Branches and small trees broke under their weight, and many were crushed by those that landed above them. Their poop fell like snow. These flocks were “more like storm systems than assemblages of birds,” writes William Souder in Under a Wild Sky: John James Audubon and the Making of The Birds of America. They inspired a gargantuan slaughter by the farmers and recreational hunters of antebellum America, one that continued for decades until the species was nearly extinct, no longer able to sustain itself. The last one died in a zoo in 1914.
Audubon later made a study of the birds’ digestion, from which he extrapolated their speed (60 miles per hour), based on the fact that they arrived in New York with rice in their bellies, which they could not have gotten farther north than the Carolinas. He extrapolated again to the physical size of the flock and finally—a fourth-degree extrapolation, like pushing a stick with another stick with another stick with another stick—their numbers, which, Souder says, work out to 25 billion or so.
In an aside, Souder comments that ornithologists believe the passenger pigeon was “adapted to survive and breed only in massive flocks of millions upon millions of individuals,” by way of explaining why it went extinct so rapidly, in contrast to the more usual pattern of gradual dwindling. Once flocks of the necessary size were gone, even very large numbers could not save them. He does not explain in detail how this adaptation was supposed to work, the evolutionary logic that might have driven it, nor what evidence supports this view.
A very different view of the passenger pigeons’ fabled population emerges from recent research on pre-Columbian American societies, described in Charles Mann’s 1491: New Revelations of the Americas Before Columbus. In a nutshell, Mann’s sources suggest that much of the mind-boggling biological extravagance the Europeans encountered—not just the passenger pigeons but bison, elk, mule deer and other species—was not “natural” in any sense, nor even characteristic of the situation before Europeans arrived.
A main line of this argument is that very little of the landscape or ecosystems that the Europeans encountered, from the great Eastern woodlands of North America to the Amazon jungle itself, was “natural,” if by “natural” we mean largely unaffected by human activity. Rather, they were the product of concerted human endeavor that included large-scale intensive agriculture and its attendant deforestation, burning that shaped both the forests and the prairies of North America, and building—mounds, terraces, cities.
But the case of the passenger pigeon is more complex. According to one line of reasoning, the preternatural abundance of these birds, the bison and other species was in fact an indirect consequence of the arrival of Europeans—more precisely, of the pathogens they brought with them. It is a fairly common view by now that disease was the principal cause of the decline in Native American populations, although the numbers themselves remain controversial, and deliberate slaughter was also an important factor. One consequence of this emphasis on disease as a, or the, primary factor in devastating Indian societies is the realization that plague ran ahead of the conquerors. In many cases, European settlers encountered an already depopulated landscape, one in which the most important species—Homo sapiens—had abruptly been reduced to a fraction of its former size.
Some provocative conclusions follow from these revisionist views of Indian population and the societies they built—bigger, more complex, more urban and more technologically sophisticated than previously believed. One is that many of the societies studied by anthropologists—like the Yanomamö, famous to generations of anthropology students—may actually have been remnants of much larger and more complex societies, devastated by smallpox and other invasive pathogens before, in some cases, the Europeans even encountered them. Mann likens conclusions about pre-Columbian societies drawn from the study of these groups to inferences about European culture one might make by studying the survivors of Dachau.
Another provocative deduction is that the astonishing abundance of certain species, epitomized by the passenger pigeon, was also an indirect result of the sudden collapse of Indian societies due to multiple plagues and slaughter. According to this argument, there is little evidence for huge flocks of passenger pigeons in pre-Columbian times. Researchers hypothesize that when Indian societies were decimated by European diseases and conquest, the dynamic equilibrium they had maintained with their environment was thrown out of whack. Species they had cultivated but also culled, such as bison (lured deep into the eastern forests by the practice of controlled burning, which cleared the undergrowth and created wide, park-like spaces beneath the mature trees), suddenly exploded, no longer checked by human activity. Other species, like the passenger pigeons, while not the object of Native Americans’ direct interest, nevertheless also benefited from their sudden disappearance.
This line of reasoning has a certain intuitive appeal. After all, it’s hard to see how a species as abundant as the passenger pigeon could possibly sustain itself over the long haul. Although hunting seems to have done them in, basic population dynamics suggests that, sooner or later, they would have experienced, if not a crash, at least a decline. They were eating farm crops, after all—in other words, using up not only the naturally available resources but whatever surplus human ingenuity could extract. What evolutionary advantage could such behavior confer? How could they have survived their own success, if their mere physical weight destroyed their habitat, breaking branches and even trees with trunks as thick as two feet in diameter? “The birds denuded and destroyed large sections of forest,” Souder writes. There is something not quite right about this picture of astounding fertility, something a little out of balance with the surroundings, that the hypothesis Mann reports helps bring into focus.
This argument does not require absolute skepticism regarding the legendary fertility of the land the Europeans first encountered. It’s certainly plausible that the Americas were more abundant than Europe after centuries of unsustainable farming, population growth, war, plague and famine, but a flock of birds that takes 72 hours to cross the sky is not just abundant, it’s freakish—not because it beggars the imagination (though it does), but because it doesn’t fit with the context. How could such a Biblical event survive?
If the research Mann describes is accurate, one explanation for the birds’ numbers would be the sudden release of immense wealth—that is, energy—that had previously been controlled or channeled by the human population. That population’s disappearance might well have triggered a sudden efflorescence as accumulated energy was released from the controls they had imposed.
Which brings me to the question of adaptation, and the ornithologists’ ideas about the size of the flocks. What if there were no adaptation involved? Way back in 1979 Richard Lewontin and Stephen Jay Gould pointed out that some things that look like evolutionary adaptations are really just accidents of structure, the forms that inevitably emerge in the spaces between real adaptations. Lewontin rehearses the argument in The Triple Helix, where he cites the example of the human chin. The chin, Lewontin notes, is a biological feature that our nearest relatives apparently lack, and so a great deal of effort has been expended in finding adaptive explanations for it. But, Lewontin writes, the real explanation may be that the chin does not exist. That is, it’s not a feature in its own right, scripted by DNA and subject to selective forces, but rather a by-product of the ratio of the lengths of two bones, which have changed over time at different rates.
If the passenger pigeon population exploded only as a result of the sudden collapse of the human population, then the search for an adaptive explanation for its vast numbers may be as misguided as trying to explain the chin in evolutionary terms. There may be no there there—nothing to explain. Or, to put it another way, what the population boom may evince is a simple tendency to breed as much as possible, for which no inherent check was ever selected, because external pressures had always been sufficient to keep it within bounds.
All this is fascinating in its own right. But it also provokes a certain epistemological anxiety which is, in a more abstract way, equally intriguing. Leaving aside the obvious holes in my research, and particularly the lack of detail in the putative adaptationist argument—granting, in other words, that some of these speculations may actually turn out to be accurate—it is, even so, disturbing how easy it is to adjust one’s notions of common sense and the obvious. It’s “obvious,” from one perspective, that natural selection need not be invoked to explain the bizarre fecundity of the passenger pigeon. It’s “common sense” that they would have multiplied in a frenzy once released from normal constraints. Certainly we have examples of other species that have done the same, often invasive species with no natural predators or parasites to curb their growth: kudzu, eucalyptus, the invasive snails in Hawaii that Gould wrote about, starlings… the list is long.
Some of this is run-of-the-mill, “everybody’s a critic” stuff: How easy it is to take potshots at experts you disagree with in a field you know next to nothing about. And some of it is simply trendiness, a matter of what everybody knows this month. Last month (or year or decade) we all knew that lemmings commit mass suicide; this month we know that butter’s not so bad for you after all. It’s annoying the way we parrot the latest factoid, however genuine the intellectual excitement or curiosity it may be grounded in, but it’s also absurd, and therefore funny—and it’s unsettling. Unsettling to think how much of the sense we make of the world is composed of these random fragments, jammed together in a kind of animated dadaist mosaic, subject to all the geologic forces at work in our mental world: drift, erosion, sedimentation, sudden uplift, subduction, metamorphosis, the works.
But Mann’s book, and Gould and Lewontin’s critique, provoke a deeper uncertainty, a more profound sense, not only that everything you know is wrong, but that whatever replaces it will turn out to be wrong as well.
It’s bracing to have so many received truths upended or seriously challenged as in 1491, to have a sweeping new vision offered of a subject whose basics you thought you knew, more or less—especially when the new vision accords so well with your values: pre-Columbian American societies were far more complex, sophisticated, unique and unpredictable than the usual dusty clichés about them. Nothing wrong with that. But the more received wisdom is challenged, the more you come to realize how much of your world view is nothing but received wisdom—and how much of the challenge is as well. After all, what you’re really doing is replacing one set of trusted experts with another. It’s not like I’m doing original research on passenger pigeon breeding habits or pre-Columbian animal husbandry. I’m not even getting my information from the horse’s mouth, but from a popularizer and interpreter. So another part of the epistemological anxiety is the realization of how thoroughly what we think we know is really a matter of trust, usually in people whose names we don’t even know, and in the abstract systems and institutions in which their work is embedded.
Still another dimension to this vertigo is the degree to which our so-called knowledge is built on inference—often astonishingly long chains of inference from a shockingly small original data set. Examples abound in every realm of science, from geology to archeology to astronomy to subatomic physics. Audubon’s fourth-level extrapolation about the number of passenger pigeons is small potatoes compared to some of these. For example, Natural History—in the same issue, as it happens, that carried an excerpt of Souder’s book on Audubon—had a fascinating article on the search for zircon crystals, which, it turns out, are some of the most durable objects on earth and may represent some of its oldest surviving materials. The only trouble is, this evidence is in the form of shards and fragments a few thousandths of an inch long, and the conclusions geologists reach about their age and the condition of the earth at the time they were formed are based on things like the relative quantity of two different isotopes of uranium, and the relative brightness of the original core of the crystal, as opposed to more recent overlays (each only a few ten-thousandths of an inch thick), when bombarded with high-energy electrons. I’m summarizing a summary, but even so it should be apparent that there’s an awful lot of if-this, then-that going on here, and you have to wonder how solid each of the original premises is.
I remember feeling the same way when I took an intro course in physical anthropology and became aware of the huge inferences drawn from mere fragments of bone, including assignment to, or even the definition of, a species—or an entire genus. The Lucy exhibit that toured the country a few years ago provided evidence of this process of revision in the phylogenetic trees it presented side-by-side, one from before Lucy’s discovery and the other a contemporary version. The contemporary version was far more diverse and ramified, thanks to the numerous fossil ancestors or cousins that have been discovered since, but what I remember most about it was the much larger number of dotted lines, signifying uncertain or contested lineages. The more you know, the more you don’t know.
Another example that came up when I was checking out the chin example is the difference between ontogeny and phylogeny—the difference between individual development and the development of a species. You find a fragment of bone with features that in modern individuals are associated with juveniles. Question: Is this fragment from a juvenile, or from an earlier form of the species (or from an earlier species) in which adults had this feature which is now only retained by the young? How do you know? When we accept the declarations of experts as truth, not only are we assuming that they have the facts to back up their claims—we are assuming that they’ve thought of all such questions and could, if necessary, provide satisfactory answers.
Analogous questions arise in astronomy, where they’re figuring out the mass and composition of a star that’s billions of miles away based on the color of its light, or fractional wobbles in its light, and in subatomic physics, where they infer whole dimensions, even entire universes, from the squiggles made by banging things into each other that they can’t even see. (One starts to feel a little less superior to augurs who saw the future in the cracks made by red-hot irons in the shell of a tortoise.) Huge inferences are drawn, and one can’t help feeling, at least at this distance from the subject, that they’re pretty shaky and subject to revision at a moment’s notice. The sensation is only amplified by reading the history of these sciences.
Of course there are ways of testing one’s inferences, through correlations with other data, and so on, but they remain inferences and as such they are always subject to a certain doubt. Are the premises valid? Are alternative conclusions being overlooked? Are other causal factors at work? The longer the chain of inference, the more such questions multiply, and multiply each others’ effects.
It’s not that I dispute the conclusions these experts arrive at, even if I think that, in some important sense, they don’t actually know anything. Just as some infinities are greater than others, the nothing the experts know is far greater than the nothing I know, which is why I can’t be comfortable with knee-jerk know-nothingism, the wholesale dismissal of experts and science in which some people take refuge. Even if I worry that in some important sense, their conclusions rest on some pretty shaky foundations, I still trust that, by and large, the chains of inference are solid. In this case, it’s not just trust in people but in method too. But that trust is accompanied, not always audibly, by a deep mental breath, not unlike the one I take every time I sit in a plane taking off and think, No way! This can’t work!
In fact, I think what I’ve been trying to get at with all this is the paradoxical combination of incredulity at the method and amazement that it actually does work.
On the one hand, science represents, arguably, one of the most powerful methods human beings have ever developed for making sense of the world, and its effectiveness is daily reconfirmed. On the other, it is grounded on a radical openness to uncertainty, skepticism, willingness to change one’s mind in light of new information or better explanations. Indeed, that is a major factor in its remarkable success. It succeeds through commitment to a higher level of abstraction—not to specific truths or substantive assertions, but to a particular attitude toward assertions as such, a certain habit of mind, such that even failures can be recuperated for its larger project. (And unlike the hermetically closed circle of certain self-reinforcing mythologies, where every inconsistency is accounted for through an elaborate but ad-hoc interpretive re-jiggering, such as when everything is found to be predicted in the Bible or explained by Marx or L. Ron Hubbard, this success has the virtue of falsifiability: again, a willingness to be proven wrong. It’s easier to feel confident that you’re right when you know what being wrong would look like.)
One common view is that, while science is indeed subject to constant revision, in a world where all knowledge is fundamentally uncertain, we can be more confident in an approach that makes transparent its way of arriving at conclusions, that is subject to testing and verification. It’s not perfect, but it’s pretty damn good, and it’s better than any other approach. Which, I suppose, I agree with. But I still can’t help feeling a kind of vertigo when I consider how what I think I know about the world seems ultimately built on sand. The alternative is faith in some specific idea—God, humanity, the Leader, whatever—which strikes me as nothing but a principled refusal to look at the sand. It works, but only so long as you don’t peek.
I suppose all this is banal; certainly it is not original. But it makes it a lot harder to be complacent when, mulling the ideas of anonymous ornithologists, I think I’ve detected an error.