Openness, casts, and CT scans

Earlier this week, I wrote a little post referring to an article that pointed out why original fossils remain valuable long after their discovery. I got quite a lot of correspondence about it.

Some people were incensed that I had exaggerated the problems of using casts in research. Of course, since all of my research has involved casts in one way or another, this would really be pot-kettle territory.

In fact, I would argue that more good research is done with casts than with original fossils. Anthropology is a comparative science. For the most part, it is impossible to compare original fossils with each other, because they are housed in different places. It is much better to use a large collection of casts to test hypotheses about human evolution than a small collection of fossils. Sure, you have to make sure that the casts don't mislead your analysis -- but in practical terms most of us are very aware of the deficiencies of particular casts and reconstructions. You have to gain this familiarity to be a paleoanthropologist.

There is a lot of idol worship in the field of human evolution. Many of my colleagues won't take someone seriously until they have knelt in the presence of the sacred relics. This is symptomatic of a kind of gatekeeping behavior -- scientists who want to "keep out the riffraff," which generally consists of other peoples' students, but also professionals who haven't "put in the time in the field."

Another aspect of gatekeeping behavior is the availability of CT scans. One of my correspondents wrote that CT scanning will make casting irrelevant, because everybody will have CTs of all the fossils and will be able to make their own casts when they want to. Boy, it sure seems like this ought to happen. After all, CT scans are even better than casts in some ways -- they let you see internal details and allow computer reconstructions, for example. They're not perfect, particularly for close details beyond the resolution used in today's CTs. But they should be very cheap to distribute. A world that can disseminate Craig Venter's complete genome to anybody who wants it ought to be able to find some way to get a few hundred CT scans sent around.

A number of efforts are starting to make CT distribution possible -- notably, NESPOS, the Vienna Virtual Anthropology group, and a few others. I expect these efforts will improve, and we will see more and more students able to access the essential data of paleoanthropology.

But for those who've been reading the blog for any length of time, you'll remember I wrote about this problem two years ago.

During the two years since that post, there has been a great deal of progress in scanning fossils. Most papers about new fossils are supported by data from scanning. A small proportion of these scans have been made available to paying professionals, or soon will be. Most are locked away, with no long-term prospect of ever being distributed. Today, none are openly available. Not a single scan of a hominid fossil can be obtained in the open, free of charge.

Why do I argue so strongly for completely open access? I believe it is a matter of credibility. A fundamental principle of science is replicability. If someone else cannot replicate your results, they have no reason to believe you. You have no scientific credibility.

In practice, we don't replicate every piece of science. We accept the word of competent scientists. The greater their reputation, the more we tend to give them the benefit of the doubt. Their reputations makes it more likely that they will be published in journals like Science and Nature. These journals are invested heavily in making novel results. Open things by definition are not novel. In paleoanthropology, if your fossil has been published before, these newsmaking journals are much less interested. So secrecy yields dividends.

But today paleoanthropology faces a real credibility problem. A substantial majority in most of the world's countries believes that we are lying about human evolution. In the few nations that are exceptions, a substantial minority holds the same belief: human evolution is false. The human fossil record has been fabulated.

The beauty of science is that it is self-correcting. You can read this blog to find the obvious mistakes in papers published in those marquee journals. This self-criticism is essential to science's credibility. But it is hampered by secrecy. The reputations of many powerful scientists have been built on secrecy. In certain quarters, their secrecy enhances their power. But the science as a whole is like the crew of a wrecked ship, treading water. Some of the crew are climbing up on the shoulders of others, pushing them below the surface. Sure, these drowning people respect the power of those climbing up on them. But they're just drowning faster.

Is there a lifeboat on the horizon? Many of the influential scientists and curators dedicated to the fossil record understand this problem. Many of them have worked to make the human fossil record more and more open. They recognize that fieldworkers have a genuine interest in benefiting from their finds, that careful analysis of the fossils is more important than immediate access to them. But they balance these real needs with the recognition that the field's future depends on its credibility, that the next generation of students cannot be trained without seeing the important evidence, and that comparative science is not possible without comparative data. Along these lines, eventually there will be greater openness. I believe it is inevitable, because the field cannot survive without it.

Other fields have dealt with the same pressures in fundamentally different ways. Genetic data are mostly open and freely available. Physics papers are often developed entirely in the open, on the arXiv. These models have not been attempted in paleoanthropology, and they might never work. But it is clear that open access is directly contrary to the interests of the gatekeepers. I doubt we will see scans within the next 20 years. Someday that the blind will see and the lame will walk. But I think on that day, we still won't be able to get CT scans of hominid fossils.