Science and the Long Tail

In his new book The Long Tail, Chris Anderson takes a broad but unsystematic look at the myriad manifestations of “the long tail” in the modern world. One of the more striking anecdotes moves beyond economics into the realm of science: amateur astronomers, equipped by the thousands with high-tech telescopes, provide skywatching breadth to complement the depth of professionals. Amateurs often observe or confirm unexpected events that no one knew to watch for (such as the appearance of novae), and the astronomy community has accepted a two-tiered system in which amateurs often play an important part.

Anderson provides a useful scheme for analyzing how long tail systems emerge, which apply just as easily to science as to movies and music. The three critical elements are:

1. Democratizing the tools of production – making it so one doesn’t have to be in Hollywood (or a well-endowed university) to produce a successful movie (or do good science)
2. Democratizing distribution – bypassing the necessity of a marketing campaign (or publication in a high-impact journal) for one’s work to matter
3. Connecting supply and demand – providing a system so that a potential consumer (or fellow scientist) can quickly and easily find what will interest them within the sea of uninteresting/irrelevant cultural products (or scientific publications)

Science has made some strides toward the long tail in recent years, but in for the most part the often very undemocratic world of science is slow on the uptake when in comes to sociocultural change. The most obvious barrier in exploiting the potential long tail of scientific production and “consumption” is the continued dominance of big-name journals.

Journals, in their current form, are barriers to elements 2 (distribution/accessibility) and 3 (custumized search/filters) of the long tail. Especially the most-prestigious in individual fields, they serve an important purpose in tracking the overall important development in a field. But modern scientific disciplines are so highly specialized that every high-impact journal, almost by definition, publishes a smattering of (at best) tenuously related topics.

Publishing in the best possible journals is a necessity for scientific success, but the proprietary nature of nearly all scientific jouranls means that content is restricted to those who pay for access. While there are some efforts to change this, for the most part published scientific content is not nearly as freely accessible, cross-linkable, and modifiable as it should be.

Oddly, unlike in contexts like the entertainment industry, centralized control and restriction are actually quite at odds with the ethos of scientific culture. In their early manifestations, scientific publications were, de facto, freely modifiable and unrestricted; the main restrictions on what could be done with other people’s content were primarily social rather than legal. Journals gradually become economic entities as well as socio-scientific institutions, and are becoming more and more of an impediment to scientific efficiency. Most of the important channels for distributing scientific information created in recent years (such as the Protein Data Bank and newer online-only scientific journals and pre-print archives) break from the proprietary journal mold, but the journals are too much a part of the social systems of science to be easily supplanted. Efforts to add long tail services onto the existing system (the Chemical Abstract Service‘s SciFinder, for example) are useful, but suffer from the same problem of restricted access.

Element 1 (democratizing the tools of production) is a more complex problem in science. In part, earlier scientific publications are tools of production themselves, so again, journals are a bottleneck. Small universtities much choose carefully which journals and electronic services to subscribe to. But equipment and training (not to mention funding) are also crucial tools of scientific production. There are no obvious long tail fixes for these factors. But, as in video, the growth of a scientific long tail would probably involve shifting focus from traditional capital-intensive forms toward smaller units of scientific production that could be accomplished with less equipment and without training (and intellectual patronage) by best-of-the-best scientists.

Obviously, the potential for squeezing good science from the long tail would vary by discipline. High-energy physics is probably as long-tail as it can ever be, given the material requirements of the field (though regarding journals and division of labor, it is probably much more long tail than most other disciplines). A revival of natural history (assisted by modern digital video equipment and mass-produced all-purpose measuring devices) might be one powerful possibility. But overall, it’s hard to say how the scientific landscape might change if one valued aspect of science was how easy it was to do (i.e., how much could be done for how little).

If I were inclined to jump on board with Chris Anderson’s techno-utopianism, I might predict long tail science to usher in an era of social responsibility and a Renaissance in public interest in and understanding of science. As with the rest of Anderson’s rosy predictions for the long tail, that’s probably too much to hope for. But science certainly has plenty of room for improvement.