Wikipedia is where we go first to look something up: you use it when you know nothing about a subject. That’s why Google, until recently, always promoted a Wikipedia article to the top of search results when you searched for a term on Google. But one recent trend in Wikipedia has been to include links to primary research material from articles. That, in my opinion, is a mistake; it goes against one of the basic principles of reference editing:
Do not link to primary research.
Here is an example. I was in Florence recently, visiting Fiesole, a mile or two away, today a small village, but actually predating Florence – it was an Etruscan settlement, with a fascinating archaeological site that displays the remains of a temple building, Not knowing anything about the Etruscans, I looked it up in Wikipedia.
Wikipedia’s coverage begins with an overview, entitled “Etruscan civilization”, containing about 3,000 words. There are separate, longer articles for topics such as Etruscan history, and Etruscan origins; this is a very good use of the reference hierarchy (explained below). But within a few sentences of starting to read the introductory article I found myself in a polemical debate, about the origin of the Etruscans.
It appears the origins are disputed. The author of the Wikipedia article, or its most recent editor, was convinced that we need convincing about the indigenous origin of the Etruscans, and included references to prove his or her point – no fewer than five citations, and 2,800 words are devoted to proving that the Etruscans were indigenous. The links are to such primary research articles as “The female ancestor’s tale: Long-term matrilineal continuity in a nonisolated region of Tuscany”, from the American Journal of Physical Anthropology. If the reader isn’t satisfied, there is a longer article on “Etruscan origins” with 6,900 words, much of it repeating the introductory article, with 102 references, mainly to primary research.
What’s wrong with primary research?
There’s nothing wrong with primary research per se. Researchers contest the scholarly record, report on new findings, and so on – but they are not writing for the general reader. Perhaps they should, but that is another discussion. When we reach the level of primary research, we are so immersed in the subject detail that we cannot see the wood for the trees. At that point, we have lost the whole idea of reference. This is similar to the Borges story describing a map so detailed it was as big as the world it represented.
Now, for Wikipedia editors, this principle may seem counter-intuitive, and I can understand why it might raise a few eyebrows. Perhaps the reason it happens is that for some years many commercial publishers, including Oxford University Press, have provided access to scholarly resources free of charge to Wikipedia editors. The reasoning is ostensibly to make the scholarly literature more available, but of course readers of Wikipedia do not have the privilege accorded to the editors. On the Wikipedia side, providing links is in accordance with the stated principle of Wikipedia editing, that there should be a citation for every assertion. Nonetheless, I believe it is a mistake to link to primary research in the majority of Wikipedia articles. It confuses authority with reference.
A possible explanation
Wikipedia editors are free to make any changes they like, but of course, that leaves open the question of authority. There is little curation after the event, except by other Wikipedia editors, and this is intermittent. Usually, in higher education, a statement would be reviewed by an external reader (for a book) or by a peer reviewer (for an article). This provides a level of credibility that does not exist in Wikipedia circles. Wikipedia recognises the challenge that free access to create information leads to the danger of low-quality content creeping in. In response, the Wikipedia authorities try to limit the damage by asking for a citations for any assertion. Some enthusiastic Wikipedia editors get involved in a scholarly debate and justify their work by providing lots and lots of citations to primary research. The result, however, can be that a reference article, which is intended as an introduction, has lost its way and become part of the scholarly literature: it is no longer an introduction, but a discussion forum with peers – not for the general reader.
The reference hierarchy
As a general principle, by removing primary references, I believe the reference work remains more closely aligned to its primary purpose, which is providing signposts for the general reader. Better would be to point to a number of well-regarded secondary works on the subject, and then leave it to the reader to go to the sources of these works if they choose.
Reference can best be viewed as a series of levels, assisting the understanding of a concept at several levels of difficulty, and providing links up and down at each stage – see the diagram above. Typically, the reader will move up or down one level, but rarely, if ever, beyond. The purpose of an encyclopedia is typically before the textbook or introductory book. The encyclopedic entry might miss one or two details of current debate, but to include these would overburden the reader.
Within scholarly publishing, there is an ongoing debate about making scholarly content more accessible. I am completely in agreement that the scholarly literature should be open to all – but that doesn’t mean it is appropriate to refer to it in all cases. I’m sure there is a highly detailed technical manual that describes and illustrates every nut and bolt on the trains and planes I travel on, but I don’t need to read it for myself.
Leave a Reply