User:J. Johnson/The Edit Function

The Edit Function – selecting for quality

Wikipedia seems to be the most successful, and most well known, example of the democratic ideal of "let everyone contribute". Yet as attractive as that ideal is, I find that it is too, well, idealistic. Wikipedia is also a prime example of the darkside of that idealism: vandalism, edit wars, sock-puppetry, plagiarism, and many other examples that not all contributors share the same ideals, nor even compatible purposes. This is further evidenced by the various levels of protection that have been found to be necessary.

I am minded of an essay from nearly four decades ago by publisher William Jovanovich. Titled "The Universal Xerox Life Compiler", it was written at the dawn of xerographic, or "dry", copying. (As distinguished from "wet" methods of copying that required processing with various chemicals.) This was the technological innovation that made quality limited-circulation publication (such as poetry magazines) cost-effective, and available to everyone (not just those who own the printing presses). The question Jovanovich asked was: what happens when everyone can publish?

This reminds me of the quip that the Web is the illustration of Eddington's idea of what you get after a sufficiently long spell of a million monkeys pounding on typewriters: possibly the complete works of Shakespeare, including the missing bits. But also every possible variant, and every possible permutation of all those letters. So how can you tell which is the authentic version? In brief, you can't; no one such production is inherently any better than any other.

This is where Jovanovich comes in. In preceeding eras, when the cost of publication was high enough that not everything written could be published, one of the roles of editors was to select what was published. For sure, the criterion for publication is often not literary, let alone scholarly, quality. But for whatever criteria are used, editors usually have enough to choices to be selective, and this tends to raise the quality of the material. (The lowering of quality attributed to the penny-newspaper and the dime-novel may have come about when the advances in printing technology suddenly created more demand for material than could be readily supplied. The Web seems to be similarly affected.)

Even though Wikipedia is not constrained by the cost of ink or paper, there still is an issue of selectivity. In particular, there are articles where editorial opinion is not just divided, but contradictory. E.g., why shouldn't the lunar landing deniers get "balanced" (i.e., equal) coverage? To say that they are a minority viewpoint, that the consensus viewpoint prevails, may be a "democratic" criterion, but is problematic. What if the majority is wrong? Should the article on 9-11 reflect the view of the 70% of the population that believe Saddam Hussein was behind it?

Allowing everyone to contribute anything is not workable. So by what criteria, and what process, do we select? "Truth" sounds good, but that only switches the question to "what is truth?" "Adherence to objective physical reality" sounds better, but the persistence of (among others) the lunar landing deniers and anti-evolutionists shows just how well that works. I think the requirement for citations is a good start, but it is only a start. And given that sources run from credible to out-right discredible, the requirement of citations only pushes the selection problem out a little ways.

The essential problem is one of epistemology – the study of "why we think we know what we think we know." It is not enough to collect and present supposed facts. We must work out how knowledge is to be derived from facts, and we must demonstrate that the process is legitimate. That, I think, is the essential issue that Wikipedia, and even society at large, needs to work on.