User:Caeciliusinhorto/Context considered harmful

In Wikipedia's formalised content review processes (WP:GA, WP:FA, and WP:PR), a reasonably frequent request is for the article to give more context explaining who a cited authority is, and why readers should trust them. In other words, that one should prefer Historian John Jones argues that... to John Jones argues that...

I do not believe that this advice is as universally applicable as some appear to think.

Giving context

Context about an authority is helpful when it helps readers assess how to interpret their views. For example:

• In an article about a political topic, it is often useful to know the politics of the authority being cited – comments on Hillary Clinton come across very differently depending on whether they are made by Republicans or Democrats.
• In a biographical article, it can help to distinguish recollections or opinions of those with personal connections to the subject from disinterested academic analyses.
• Where an opinion is particularly dated, it can be helpful to flag that – for instance, when citing Edward Gibbon on Roman history.
• Where views on a topic vary substantially between countries, this can be useful to note – mainstream British and French opinions of Napoleon are very different!
• In an interdisciplinary article, it can be helpful to know from which discipline a particular opinion is coming: e.g. in an article about the Black Death, in some cases it might be helpful to specify whether a particular opinion is that of an epidemiologist or a historian.
• To indicate when a particular view is included because it is notable but the person holding it is not an expert in the relevant field.

Not giving context

When none of these factors apply, this "context" doesn't help.

A common refrain is that we should give this kind of context so that readers know why the particular opinions we have cited are relevant. I am unconvinced.

The very fact that our article cites a particular authority's opinion itself conveys information. We should trust our readers to be able to infer that when we cite a particular source, it is because they have relevant expertise. Explaining that the viewpoint is that of a "historian" or "biologist" or whatever doesn't tell readers anything they couldn't already work out. It is better to be concise.

This is especially true as the context given is often entirely unhelpful. "Historian John Smith" is bad enough; this is a sufficiently wide category that it does not tell the reader much about whether Smith's opinion is relevant. If "historian John Smith" specialises in the causes of Second World War, his opinions on Women in Classical Athens are barely more useful than those of any other intelligent layman. Even worse (yet seen in at least one featured article) is "scholar Jane Doe". Scholar of what? Professors of biochemistry, economics, and French literature could all equally be described as scholars, but only one of their opinions is relevant to an article about Les Fleurs du mal.

On the other hand, even if we do give specific information about scholarly specialities, a lay reader still has no real way of knowing what that scholar's reputation in the field is. To take an extreme example, "Second World War historian David Irving argued that..." looks convincing to a lay reader who doesn't know that Irving was found to have "for his own ideological reasons, persistently and deliberately misrepresented and manipulated historical evidence".

Whether we give very specific information about scholarly credentials or none at all, the lay reader still has to fundamentally trust that we have selected relevant and representative opinions to cite. If the reader trusts our article, they will trust that the opinions cited are relevant regardless; if they do not trust the article, they will need to do extra-wikipedian research to ensure that the people we are citing are not fringe cranks whether or not we described them as scholars. Expert readers, by contrast, are presumably already familiar with the relevant names in the field and so our explaining that Martin Litchfield West is a classicist adds nothing to their understanding of the article.

How, then, does a reader know whether to trust the opinions that are cited in Wikipedia articles? The same way that they should be deciding whether to trust any claim we are making. If we are citing a particular scholar's opinion, then we should be giving a source for that claim. In many cases, that will be a direct source. We will be citing the claim "According to John Smith, Winston Churchill was a very good Prime Minister" to Smith, John. "Winston Churchill's premiership". Journal of Modern History 2012, and our readers can decide for themselves whether they think an opinion published in 2012 by the Journal of Modern History is likely to be relevant. In some cases, it will be an indirect source; something like: "Most historians think Winston Churchill was a very good Prime Minister, although Jane Doe's 2015 biography is critical of him" might be cited to Smith, Anne. "Winston Churchill: Historiography and Hagiography". Journal of Modern History 2018, and readers can decide based on the citation whether they trust Anne Smith's summary, or read the article to find the original source of Jane Doe's opinion.

Assuming that a topic is not particularly ideologically charged or a haven for fringe theorists, there is generally no need to add context tags to every opinion cited. They don't fundamentally provide any new information that isn't already obvious from the context; all they do is make the article less concise. If you are only citing one or two authorities this may not matter, but in an article like Corinna, which cites fifteen different modern scholars in under 2,000 words, the repetition of "classicist X" will get wearing very quickly.