When I ask scientists what's the biggest misunderstanding people have about their work, they often talk about how they know what they know. People tend to think that a scientist's job is to gather every single datum about something in nature--a mountain, a species of jellyfish, a neutron star--and then, simply by looking at all that information, see the absolute truth about it in an instant. If science departments were filled with angels, that might be the case. But they're staffed by humans with finite brains, with tight research budgets, and with only so many years left before retirement or death. In order to tackle vast questions about the fate of the universe, the history of this planet, and the tangled bank of life on Earth, they have to live with uncertainty. To understand something, they can only gather a smattering of information about it, look for patterns within the data and use well-supported theories to come up with hypotheses about them. They can then gather more information in order to test the hypotheses again, and, if need be, alter their explanations to accord with the evidence. Their conclusions can only be tentative, but they can also be powerful. We were not around when the Earth formed, and we can only look at indirect clues in certain rocks and meteorites. And yet scientists have a good idea of when the Earth formed, how quickly the iron core settled to the center of the planet, when oceans began to appear, and so on.

Many bogus attacks on scientific research play on this common misunderstanding of science-as-revelation. If scientists don't know everything they can't conclude anything. Paleoanthropologists have found less than two dozen species of hominids from the past six million years--therefore they can't draw any conclusions about how humans appeared on Earth. Climatologists don't have a perfect temperature record for the planet--therefore they can't say anything about how man-made pollution is warming the atmosphere. In cases like climate change, these bogus attacks spread from science to policy based on science. To hear some people talk, we should only do something about climate change once we have tracked every molecule in the atmosphere since the dawn of civilizaiton and can predict its course for the next thousand years.

Extinction is a particularly good example of how confusion about the nature of science can cause serious trouble. In January I wrote a couple posts about some research that indicates global warming could cause a vast wave of extinctions in the next century, and how some critics deceptively emplyed the Imperfect Knowledge gambit. Today Science is publishing an important paper that may well attract the same specious criticisms, the same calls to ignore anything less than the wisdom of angels. (Here's the press release.)

Here's the lowdown:

British researchers have been working for the last decade to carry out the most ambitious analysis of changing biodiversity ever attempted. They took advantage of the fact that in the 1950s and 1960s, professional biologists and amateur volunteers began doing painstaking surveys of several groups of species around Britain. Mapmakers had carved up England into 10-kilometer-square parcels, and every year the surveyors would take a census of the species in each one. In the mid-1990s, researchers realized that this amazing database of biodiversity, unparalleled in the world, could let them track broad patterns of change. They picked out three very different groups of species to compare--butterflies, plants, and birds. They then chose the results from a few years in the early parts of these surveys to compare with results from the 1980s and 1990s. The scale of the comparison was staggering: every single species of butterfly, plant and bird known to live in England was tallied; the researchers analyzed 15 million records put together by 20,000 volunteers.

The results were bleak. Over a quarter of all native plant species had disappeared from at least one the survey squares in their range. Half the birds did. Butterflies fare worst, with 71% surrendering at least one square. But the average retreat was actually much bigger. The typical butterfly species vanished from 13% of its range, while the fastest-declining 10% of British butterfly species can't be found in over half of their former range.

There has been a lot of debate about how badly different groups of species are faring these days. Birds have been carefully studied around the world, partly because they're big enough to spot with binoculars, and they've been suffering significant losses. Plants, which can't move out of view, have also been pretty well studied (although not as well as birds). But could the same be said for insects, which number in the millions of species and are far harder to study? The answer, at least in this study, is yes.

What's particularly striking about this survey is that there are many good reasons to think that biodiversity has had it much easier in Britain over the past 50 years than man other parts of the world. Most of its forests had been cleared away many centuries earlier, so that the animals and plants living in 1950 had been living in fragmented habitats for a long time. Britain has suffered relatively few invasions of aggressive alien species that could have driven native ones extinct. Conservation is important to the British. And global warming, for the last few decades at least, has actually made Britain a better place for butterflies and plants to thrive. Nevertheless, the biodiversity of Great Britain is shrinking, and shrinking fast. That bodes ill for other parts of the world, particularly ones that are home to many species restricted to tiny ranges. (Many species in Britain can also be found elsewhere in Europe, which is why the population declines in Britain have not led to all-out extinctions yet.)

The scientists conclude: "If insects elsewhere are similarly sensitive, we tentatively agree with the suggestion that the known global exticntion rates of vertebrate and plant species may have an unrecorded parallel among insects, strengthening the hypothesis, derived from plant, vertebrate, and certain mollusk declines, that the biological world is approaching the sixth major extinction event in its history." (Italics mine.)

In a day or two I will update this entry. I am interested to see how this study gets digested by the media-punditry machine. I have a few suspicions of what we'll see:

Some environmentalists groups will trash the careful wording in the conclusion and simply say, we're doomed. If that does happen, it will be too bad, because it will undermine the care put into this research.

Some "skeptics" will say that you can't compare the surveys because different people made them, looking at different groups of species. That's actually untrue: the researchers did statistical tests to make sure that the comparisons could hold up even if there were some biases among the surveys.

I also predict that the skeptics will claim that British plants and animals have just retreated to the safety of refuges, where they can now live happily ever after. But the evidence points the other way. For example, the population declines took place "with remarkable evenness across the nation," the authors write.

The skeptics will also say you can't generalize from a small northern island nation to the world at large. But the results are actually in accord with other studies on extinctions worldwide.

I certainly don't mean to imply by all of this that this study is perfect. Perfection is, by definition, impossible to reach in science. But if critics say that we can't draw any conclusions--or any political decisions--until we are completely certain of how biodiversity is faring these days, and if they are sincere in their claim to be interested in protecting biodiversity, then I have a challenge. Go ahead and set up a project that would give us complete certainty. It took 20,000 volunteers to carry out these surveys in Britain alone. To survey the world, a few million more volunteers should do the trick.

Stay tuned.

Update here.