following is from an email I sent on 22/03/2007. I thought I might as well
keep it here as anywhere.
people have pointed out that wikipedia articles often say true things.
That may very well be so. In fact it might even be the case that, in
general, it so happens that wikipedia articles are as accurate as
Encyclopedia Britannica articles; but that is really quite irrelevant to
the question of whether wikipedia is a suitable reference for academic
work. The real question is whether wikipedia can satisfy the accepted
epistemological criteria which define the collective enterprise of modern
If I decided to estimate the number of people in a room by rolling dice
and adding the spots, it might happen that I got a true answer; and if I
decided whether to take an umbrella to work by reading the entrails of a
chicken, it might again happen that I got a true answer. But we do not
accept that these are acceptable ways to arrive at the truth. What we
prefer in these cases is to count the students or to look at the sky.
Error is still possible, but experience has shown that the latter methods
are vastly more reliable than the former and they are therefore to be
preferred on epistemological grounds.
Through bitter experience and disciplined thinking modern Western
scholarship has determined a set of standards and methods that are
believed to maximize the reliability of the outcomes for this collective
enterprise. These are the standards and methods with which we are
familiar. One of these standards is that we prefer to trust someone who
has been accepted as an expert in some area by others who are interested
in that area. Another is that we test new scholarship by peer review by
accepted experts. Another again is that when we refer to some supposed
fact we support it by appealing to the agreement of some trusted expert.
None of this is guaranteed to find Truth: experts will disagree, frauds
and fads will occur, and accepted facts will change. It is only our best
Wikipedia, by contrast, deliberately flouts these standards. Does that
make it unacceptable? Well, that depends upon whether you think that the
standards and methods embodied in wikipedia are as likely to yield
reliable results as the standards and methods that they replace. Is that
plausible? Is it plausible that the rejection of any preference for
expertise in determining trustworthiness will lead to more reliability in
scholarly results? Is it likely that the preferences of the most
persistent editors will provide results as reliable as the preferences of
the generally accepted experts? I would suggest that that is vanishingly
unlikely. Wikipedia would replace the present accepted standards with
worse, less productive standards.
The wikipedia model should therefore be rejected as a possible amendment
to the epistemological criteria of modern Western scholarship; and
consequently, we should insist that students avoid references to wikipedia,
because we have a duty to train them in the best traditions of modern
Western scholarship, and wikipedia is opposed to those traditions.