Some implications of "digital" for scholarly writing and publishing

“Wikipedia is based on evidence, not anecdotes”

Reading Time: 3 minutes

This claim is common in Wikipedia discussions; most recently, I read it in an article  by Andy Tattersall on the LSE Impact of Social Science blog. I don’t disagree with the sentiment, but is it true of Wikipedia? You don’t have to look far to find anecdotes rather than evidence.

For example, the Wikipedia entry for Albrect Durer, Renaissance painter, and engraver, discusses his relationship with the Holy Roman Emperor Maxmilian. It includes the following anecdotes:

Dürer manifested a strong pride in his ability, as a prince of his profession.[36] One day, the emperor, trying to show Dürer an idea, tried to sketch with the charcoal himself, but always broke it. Dürer took the charcoal from Maximilian’s hand, finished the drawing and told him: “This is my scepter.”[37][38][39]

In [sic] another occasion, Maximilian noticed that the ladder Dürer used was too short and unstable, thus told a noble to hold it for him. The noble refused, saying that it was beneath him to serve a non-noble. Maximilian then came to hold the ladder himself, and told the noble that he could make a noble out of a peasant any day, but he could not make an artist like Dürer out of a noble. [40][41][42]

[extracts from Wikipedia, accessed 22 February 2022]

Wikipedia, as I am sure you know, has a policy of no original research, but every statement to include a reference to a source. However, the sources used are frequently questionable as to their authority. It seems to me that quoting a very old and unreliable source is complying with the Wikipedia rules but not with their spirit.

For the first claim above, the sources (numbered 37 to 39 in Wikipedia) are:

Ignore the “retrieved” date, which is irrelevant. Glancing through the Wikipedia article, it looks to me as though a sizeable proportion of the references are to books published over 100 years ago – I could see at least 20.

Andy Tatersall’s point in his blog is that references to articles from Wikipedia should be to open-access content. For these references above, they are to books that are not only not open access, but likely to be difficult to find in any library. My guess is that these references have been taken wholesale from one of the many sources Wikipedia has included to create its core content, notably the 1911 edition, now in the public domain; Wikipedia even has an article on how to incorporate EB content. They state:

Starting in 2006, much of the still-useful text in the 1911 Encyclopaedia was adapted and absorbed into Wikipedia. Special focus was given to topics that had no equivalent in Wikipedia at the time. The process we used is outlined at the end of this page. Since then, a lot of the text has been incrementally improved or replaced by more modern information as editors slowly made Wikipedia better. A small number of articles are still missing details that could be usefully imported, but they are few. If you want to do this, refer to the tips at the end

While it makes for easy content, my suggestion in this case is that it would be better to remove all the 1911 content as unreliable and difficult, if not impossible to verify. Just making the attribution more accurate does not improve the accuracy or reliability of the content.

What is the moral for digital publishing? Something like this: just because some content is in the public domain, it is not necessarily a public service to include it in the world’s most widely used (and, in many cases, most up-to-date) general reference source. Does it matter that Wikipedia is full of dubious anecdotes? Perhaps few readers notice, but thousands of students seem to treat Wikipedia as immutable truth, and it is worrying.


How to talk about books you haven’t read


Researcher to Reader: the first face-to-face event for two years


  1. Michael J. Amphlett

    “…but thousands of students seem to treat Wikipedia as immutable truth, and it is worrying.”

    Thousands Michael…? I’d suggest tens of millions, and that is much, much, more worrying!

    The glib statement that people have used wikipedia for their “research” is also very worrying, they seem not to be capable of differentiating between ‘research’, and just ‘search’.

    • Michael Upshall

      Yes, you are I’m sure correct – the number is much higher than I stated. Thanks for your comment.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Powered by WordPress & Theme by Anders Norén