Musings on “digital universe” and wikipedia: There is a weak underbelly to wikipedia-style demopedias, given that the know-nots will always, always, vastly out-number and out-shout the very few who actually have any idea of what they are talking about: That’s a gaussian inevitability. Peer-review was meant to counteract it, but demotic global graffiti-board free-for-alls risk drowning out the signal with the ambient gaussian noise.
Eppur, eppur, if that were altogether true, surely, wikipedia would be a lot noisier than it actually is. This could be just chance initial conditions, or it could be because most know-nothings simply don’t bother, so it’s only compulsive quacks that are the noise-makers, and they are not in the majority.
I am told that much wikipedia vigilantism seems to consist of deleting rather than inserting or over-writing. If that is representative (it might not be), it is analogous — but in a sinister way — both to evolution and to neural function: Most “selectivity”, in both evolution and neural function, is negative, eliminative. In adaptation, it is the failure of a mutation to out-perform the competition, and hence the mutation’s elimination. In neural function, the effect is in the form of both selective neuronal loss and active inhibition. A lot of neural growth, maturation, and even learning consists in the selective loss or “pruning” of connections (or even neurons), not in the positive creation of over-riding patterns. Similarly, a lot of our “skills” (e.g., motor skills) turn out to consist in the active inhibition of false moves. (This is unmasked in aging, as the inhibition begins to fail, and the gaffes again begin to prevail!)
I say the analogy is sinister, because whereas in evolution the “arbiter” and “enforcer” of the selective deletions is the “Blind Watchmaker” — i.e., adaptive success/failure itself — and in neural function too, it is functional success and its consequences that guide what is retained and what is selectively deleted or inhibited, in the case of wikipedia it is merely self-appointed vigilantes: Someone decides something is his “turf” and he deletes all interlopers (or interlopers with whom he does not happen to agree). It’s bad enough when supposedly qualified editors do this (when they do it badly), but when anyone can self-appoint — well, I suppose we are alive when this empirical question is being answered in real time, before our eyes. But since it is all happening anarchically (indeed, global demographic anarchy is one of the players that is under test in all this), it is not clear when or whether we have a definitive outcome. Wikipedia is limping along so far, globally, though I suspect there are local abuses that are much less rosy, and might bode ill for the project as a whole, further along the way.
But the medium itself might provide the fix! If the vigilantes are tracked (anonymously) and their changes suitably tagged, a user could, in principle, sample a few “diffs” for a given document, discern that the changes wrought by, say, “Rambo” make the document worse, not better, from his point of view; there could then easily be a means of selectively “extracting” the view that is Rambo- independent (i.e., selectively deleting everything in which “Rambo” had a hand). Since all drafts and diffs and vetters’ tags are stored, this would democratize Wikipedia even more: Not only could any vigilante come in and add, subtract, or over-write whatever he wishes: any user could elicit a “view” that was selectively purged of all contributions of that vigilante, if he so wishes.
But this has to be easy to do and transparent. The current versioning of Wikipedia is far too awkward and user-unfriendly.