Sometime around next Sunday, Wikipedia will reach 2 million articles. It’s about eighteen months since the millionth article was added, and the number of new articles has stabilized at around 2000 per day. So the shift from exponential to linear growth (in article numbers at least) has taken place a bit sooner than I expected. Some disorganised thoughts follow.

The most obvious change in the past eighteen months is the way attention has shifted from the extensive margin (more articles) to the intensive margin (work on existing articles, metacontent such as categorization and classification schemes, and internal process such as the development and enforcement of policies on biographies of living persons, prompted by embarrassments like the Siegenthaler hoax and by the increasing propensity of politicans and others to edit their own entries).

There’s a natural economic logic here. With two million entries already, the typical new entry (ignoring the many short-lived attempts such as this one) is going to be something like List of state leaders in 1390s BC or KitaÅ?ji Station. The marginal benefit of adding an entry is declining, though certainly not zero. On the other hand, the demand for internal improvements builds on itself. A stroll through Wikipedia using the Random entry function shows that the great majority of entries are tagged as needing improvement of some kind.

This process of cumulative improvement is resource-intensive, but not nearly as much as the dialectical processes that operate for controversial entries (and on Wikipedia, anything and everything can be controversial). Edits are made, reverted, reverted, tagged as needing support or violating some Wikipolicy or other, until a single sentence can consume dozens of hours of work. Still, the result is often a drastic improvement in quality compared to a starting point in which one point of view or another is taken for granted. One obvious manifestation of this is the vast increase in referencing of claims, and the increasing pickiness of policy regarding sources for such claims. Blogs have been a particular victim, with only a handful of expert-written blogs being accepted as reliable sources on particular topics. Despite the merits of the process, it’s easy to get burned out defending an article like Global warming controversy against the sustained efforts of delusionists to include lengthy and uncritical presentations of their latest talking points.

One thing is clear though. Complaining about Wikipedia now is like complaining about the Internet. There isn’t going to be any alternative for quite some time to come. The much-debated comparison with Brittanica is ancient history in Wikipedia terms – the work was done in late 2005, when there were about a third as many entries as now, and many millions of edits ago. As regards rival projects, Citizendium has some appealing characteristics, but is not really going anywhere., having about as many articles in total as Wikipedia adds every day, and not many articles obviously superior to their Wikipedia counterparts. Meanwhile, Conservapedia is a bit more active, distracts some trolls from Wikipedia and is always good for a laugh