Consensus might be an effective means of choosing leaders, but as a method of discerning the truth, philosophers have long found it wanting. The standard critique is simple enough to state and grasp: The fact that everyone decides to act as though something is true does not make it true. Nevertheless, consensus is an important facet of modern science; the notion looms especially large in climate science. It’s become a sort of best-guess assessment of how the world works when a controlled test of a master hypothesis — for example, that humans are the leading contributors to climate change — is not possible. It’s not perfect, and to many philosophers not even desirable, but it has proved a useful concept, especially as human society can’t afford to test the models by waiting for events to vindicate them.
Consensus is also the defining characteristic of epistemology on the internet, thanks to the awesome growth of the wiki. From the humble beginnings of the WikiWikiWeb, the open-source wiki movement has achieved near-total saturation of public consciousness on the strength of its most famous progeny, Wikipedia. Much of the world thinks of it as an indispensable first step for researching anything, but it still has its critics: Some have called Wikipedia “a public toilet” and its editorial style “digital Maoism,” and more temperate concerns hold that the bottom-up model of content generation at the heart of the wiki, and its mass-consensus model of knowledge, seem to threaten our ability to verify facts and know the truth. As scientists move toward embracing the wiki — which they are doing, through sites such as OpenWetWare and Proteopedia — it would be reasonable to worry that a consensus of the uninformed could overwhelm the knowledge derived from hard-won expertise. The possibility of verification would seem undermined, the reality of truth and facts buried under an epistemological sludge. But, whatever the flaws of Wikipedia, those worries seem unfounded. It is not scientific rigor that is accommodating the wiki, but the wiki that is accommodating science.
One of Wikipedia’s most persistent critics is the man who first proposed the creation of a wiki-based reference guide. Larry Sanger, cofounder of Wikipedia and self-described “wiki-apostate,” thinks the encyclopedia exists in a state in which the “know-nothings can drive off the know-somethings” with a bankrupt founding principle of “radical egalitarianism.” That a meaningful consensus can derive from “a project that has millions of participants is ridiculous,” he says. One of the system’s biggest supporters, Clay Shirky, author of Here Comes Everybody: the Power of Organizing Without Organizations, holds the opposite opinion of its value, but a similar opinion of the wiki’s strength. Shirky argues that authority and expertise are only “social facts.” “Works are only authoritative because everybody agrees they are,” he says. “It’s circular. That’s what’s going on with Wikipedia now. To the degree that people begin to take it seriously, it’s going to acquire the patina of authority.”
Many scholars dispute the notion that wikis will eliminate more traditional models of scientific authority, but the wiki model is nonetheless exerting a gravitational influence on some of the core principles of epistemology. “I think it’s changing how we think about authority,” says Tyler Cowen, a professor of economics at George Mason University. But that may not be the case for science wikis. The opinion of an authority, however well informed, is no more a hallmark of the truth than consensus is. Two primary meth- odological pillars of science — peer review and the pursuit of replication of results — and its philosophy of falsifiability, rather than being threatened by the wiki, might actually be well served by it.
For Seth Lloyd, MIT professor of mechanical engineering, the wiki model is “reasonably good” at discerning, if not what is true, then what is false. The notion of falsifiability, thanks to the massively influential work of Karl Popper, sits at the core of many conceptions of how science operates; scientists may not be able to prove what is true, but by repetition of prediction, trial, and analysis, scientists can at least discern what is false. Indeed, Shirky’s description of the wiki as a rapidly evolving “process, not a product,” speaks to the power of the wiki to act as a chronicler and creator of repositories for science, to create a home for what we haven’t proved to be false and for what we think to be true. Constant checking, a ceaseless selection of good information and argumentation over bad, could accelerate falsification.
The key is who gets to participate. Sheila Jasanoff, a professor of science and technology studies at Harvard’s Kennedy School of Government, argues that those tools are undergoing more of an evolution than a revolution. “It’s not that the peer-review system is breaking down so much as getting remade into some different kind of thing,” she says. According to Jasanoff, one problem wikis face is the absence of clear methods for quality control. That uncertainty, Jasanoff argues, has meant that wikis have inadvertently enhanced the status of peer-reviewed journals. “People want to know who’s good and who’s bad and who’s better and who’s worse,” she says, and thanks to the prestige of journals such as Science and Nature, they remain the place to find out who’s what. And if scientific wikis were in fact like Wikipedia, then those journals would probably forever remain the best places to find those answers, and the wikis, insofar as anyone used them, would present real problems for notions of truth and reasonable consensus, for figuring out what work is better and what work is worse. Fortunately, the people building scientific wikis have taken those concerns into account. Proteopedia, for example, was introduced to the world this past August in the journal Genome Biology, and its founding principles were laid out for inspection. Would-be contributors have to petition the site’s editors for the ability to create and edit pages, and even then they work only on specific topics in which the editors have accredited them as expert. Furthermore, as is also the case with OpenWetWare, a wiki established by Drew Endy and Tom Knight at MIT in 2005, now host to more than a hundred research laboratories, no one writes from behind a veil of anonymity. However little Wikipedia cares for authority, and however much that encyclopedia has dispensed with peer review in favor of a dictatorship of the persistent, authority and peer review are concepts built into the core of science wikis.
Sanger himself is using those same principles in his latest venture, the Citizendium. Shirky calls the failure of Sanger’s new project a “foregone conclusion,” because it wants mass participation while eliminating the openness that has encouraged Wikipedia’s mind-boggling growth. Indeed, for something to get as big as Wikipedia, there can be few barriers to entry. For wikified science, Sanger may just have the ticket. — With reporting by Matthew Power
Originally published February 9, 2009