|This page is part of the Military history WikiProject's online Academy, and contains instructions, recommendations, or suggestions for editors working on military history articles.|
While it is not one of the project's formal guidelines, editors are encouraged to consider the advice presented here in the course of their editing work.
One of the main stumbling blocks on an articles reaching featured status is its sourcing. While all articles are supposed to meet the reliable sources guideline, featured articles must meet criteria 1c, which requires that the article be:
well-researched: it is a thorough and representative survey of the relevant literature on the topic. Claims are verifiable against high-quality reliable sources and are supported with citations; this requires a "References" section that lists these sources, complemented by inline citations where appropriate
Recent changes to the criteria
The "high-quality" requirement and the expectation of a literature survey was added early in 2009, after a long debate at WT:WIAFA. Broadly speaking this means that in the writing of a FA, a person should gather a few books (high-quality scholarly texts), read them and then put them together.
While an expert who is familiar with the topic that they are writing about can get straight into the article by adding information from the top of their head and then look up sources, an amateur should do the opposite, otherwise they will almost certainly end up producing an article that is neither balanced nor comprehensive.
In the case of old FAs, many just used any old website, including amateur sites and blogs, or no sources at all. If an article is taken to FAR, the poorly sourced or unsourced sections can be switched to a proper book, but the article might still be missing some information if Google books etc were only used to search for statements that match up to the article text, but not for a survey of the literature.
Wikipedia articles should be based on reliable secondary sources. This means that while primary or tertiary sources can be used to support specific statements, the bulk of the article should rely on secondary sources.
Tertiary sources such as compendia, encyclopedias, textbooks, and other summarizing sources may be used to give overviews or summaries, but should not be used in place of secondary sources for detailed discussion. Wikipedia itself, although a tertiary source, should not be used as a source within articles.
Primary sources, on the other hand, are often difficult to use appropriately. While they can be reliable in many situations, they must be used with caution in order to avoid original research.
The "High-quality" requirement, judging by the discussion, expects people to consult a lot of classic books and weigh them up, rather than whatever websites are pulled up by Google. By many interpretations, this means that per RS, scholarly secondary sources should be preferred. Avoid relying on tertiary sources such as related encyclopedia articles, and other introductory guides/factfiles; instead, go for in-depth texts/journals and distill them. Primary sources are a trap for original research; we should report what the scholars think.
With regards to this, some users use academic citation indexes such as Google Scholar to determine which books have had the most impact, and ask for primacy of analytical space to be given to those works that were more influential in driving the scholarly debate. This was the driving theme behind the recent changes to WIAFA, which occurred after the FAC for John Wilkes Booth, the assassin of Abraham Lincoln, garnered criticism because websites were used while specialist textbooks were eschewed.
Examples of objections over the quality of sources and research at FAR/FAC.
Comment on 1c: This article is based on a single biography of Anne in addition to books covering a much wider scope (history of Britain sort of things). I checked around and there are several biographies of Anne. An FA version of the article would not present just one biographer's view of Anne, but the views of all of the major biographers. Awadewit (talk) 16:30, 12 July 2009 (UTC)
- Of this delisted article, the nominator said "There's far too large of a reliance on tertiary sources, and the article is lacking in its use of scholarly sources". At the start of the FAR the article was in such a state, with a heavy reliance on Banglapedia and a mini-biography/factfile.
The following is a quote from a journal:
Until the early 1990s, the most thorough descriptions of Diem's referendum to depose Bao Dai were penned by journalists or appeared in political memoirs, and thus cannot be considered scholarly accounts of the event.
This is in agreeance with the excerpt from the verifiability policy, which says
In general, the most reliable sources are peer-reviewed journals and books published by university presses; university-level textbooks; magazines, journals, and books published by respected publishing houses; and mainstream newspapers. Electronic media may also be used. As a rule of thumb, the greater the degree of scrutiny involved in checking facts, analyzing legal issues, and scrutinizing the evidence and arguments of a particular work, the more reliable the source is. Academic and peer-reviewed publications are highly valued and usually considered the most reliable sources in areas where they are available, such as history, medicine and science. Material from reliable non-academic sources may also be used in these areas, particularly if they are respected mainstream publications.
Thus, textbooks by scholars are generally regarded as being more reliable than newspaper articles, which vary depending on the seriousness of the news outlet. The freedom of the media also needs to be considered. Is the newspaper from an authoritarian country and is it under pressure to toe the government line? The bias it brings along.
Newspapers and news outlets
Among news articles, those by a long-term correspondent/specialist who focuses on a certain topic are more reliable than those written by an journalist who covers miscellaneous odds and ends. In addition, articles on miscellanea are more likely to be unattributed, and according to the reliable sources guideline, "Items that are signed are preferable to unsigned articles".
Most history textbooks written by academics are based on their PhDs or other labours that have often consumed more than four years of full time research, and many have been read, informally or formally reviewed and vetted by 10 or more other scholars, whereas newspaper articles are written on the run for tomorrow's edition. Here is a case of a newspaper that is considered an upmarket paper, where editorial oversight is supposed to exist, but of a sample of 60 written online pieces, no changes were ever made (if the editor actually ever read the piece); there were errors present. Newspaper articles are thus more prone to errors, or superficial and less sensible analysis due to immediacy, and are more prone to recentism, even the broadsheets. While in many cases, the raw data might still be correct, sensationalism, us v/s them mindsets and not-so-sensible analysis can be subconsciously imported into the article because of osmosis of the journalist's thinking.
The reliable sources guideline says that "Wikipedia articles should cover all significant views, doing so in proportion to their published prominence among the most reliable sources", which means that while the raw data presented by a newspaper and a scholarly textbook may be the same, the interpretation and analysis of an expert academic should be given more weight.
This is particularly due to the relatively sensationalist and less learned analyses by newspapers, especially with the urge to boost sales.
As an example, in one newspaper, after the First Test of the 2005 Ashes series, it ran the headline "Vaughan-again loses" after England was hammered. He also said that the rest of the world was hopeless and couldn't give Australia a decent game. England eventually defeated Australia 2–1 after five Tests, although one of the victories was by the second smallest margin of runs in 132 years of Test cricket, so it could have easily been an Australian victory. The same correspondent then said that Australia had been "exposed" and that they could only bully and beat incompetent teams. He predicted that it would take a decade for Australia to recover from the "ruins" of the campaign. A month later, Australia played a all-star ICC World XI, a composite team of players from the rest of the world, but as they weren't a proper team they went through the motions, played incoherently and were mauled in every match. The same writer then said that Australia were far better than the whole world put together, and so on.
Generally, such cheap headlines are employed to boost sales. A detailed book would not chop and change like this. While the raw facts were the same, news sources can be a pitfall in colouring the Wikipedia article with a silly tone. These dangers can be heightened for academic topics, such as science or archaeology.
Also, by the same token, books compiled by journalists based on years of their own work are more reliable than their day-to-day work, as factual bloopers are usually corrected, and time allows them to reflect on the longer impact of events.
During times of war, or at all times in the case of some countries, press censorship may be instituted in by the government. One needs to be careful when using contemporary references on wars, in case the news source was under censorship at the time, and if so, to consider what implications it might have on the reliability or neutrality of the article. For example, Defense of Sihang Warehouse, a former FA about an operation during the Second Sino-Japanese War, uses Chinese newspaper articles written while the battle was taking place. If censorship was in place then the contents may be unreliable or biased due to a government's desire to keep morale might by presenting a falsely optimistic picture of the operation.
The pieces in question were:
- "Our Determined Lone Army Makes Final Stand". Lihpao Daily 29 October 1937"
- "Su, Hua. "We Are Praying For You". Lihpao Daily, 2 November 1937"
The headline gives us an indication of what the agenda of the article may be.
High school textbooks and "popular" books
High school textbooks can be of questionable reliability. High school textbooks sometimes contain over simplifications that either introduce errors or a loss of technical rigour. In some cases, particular in science subjects, the authorities want to get more students, and one way is to introduce more "exciting" material into the syllabus. Unusual phenomena are often harder to explain, and to allow more difficult and "exciting" material to be accessible to students, the rigour is often diluted. High school textbooks are usually written by schoolteachers with bachelor’s degrees, whereas university texts are generally written by PhD/research scholars.
School textbooks can also be subject to interference from governments, in order to inculcate a certain propaganda slant into students, especially in history, which is why books in that field need to be treated with extreme care.
The same can apply to popular science textbooks. The editor from one astronomy magazine (with a proper publisher) apparently did not know the difference between linear velocity and angular velocity; he got confused when he used some rather loose language during a talk and was asked to clarify his statement. It turned out that he did not study physics at university. Later, he appeared on ABC NewsRadio to discuss an ongoing NASA mission and was described by the host as an "expert". Ouch..... This can apply to technical topics such as weaponry, equipment and fortifications.
Books that are targeted at the general public or children, tend to compromise rigour for accessibility. Such books tend to have a large proportion of pictures and other visual gimmicks. One way of checking such books is to type the name of the book into google, which will list libraries that stock the title, and its catalogues, which will categorize the book, and sometimes label it as "juvenile non-fiction" or "popular science".
Government sources can be reliable sources if they are peer-reviewed—a major criteria—such as a report by a planning committee, senate inquiry, royal commission, or a census. Some things on government websites that provide overview information can be of the ad hoc type and can fall into a grey zone.
On the recent FAR for Sikkim, one government website that was cited for a piece of general/basic information on forestry; there were more than five typos in one paragraph, and a lot of broken English. These sites might not be reliable at all as they seem to lack a proper peer review and inspection process. A lot of government websites might not be vetted properly and may merely be a means of providing a basic and informal customer service to the populace, whereas a formal government report is checked thoroughly—although of course it may have its own bias.
Needless to say, sites below newspaper quality, such as blogs, self-published material, hobby sites, websites of extremist organisations etc, aren't RS, except to record the opinion of the author or body in question. Given the requirement for "high-quality" reliable sources, there is effectively no chance that such material will be accepted. While self-published materials by an expert in the field, eg lecture notes on university websites by famous professors, are reliable, it will be available in a higher-quality form in a textbook, so the latter should be used.
Deciding on what is a reliable source
Although some sources have been agreed upon to be reliable by members within a certain WikiProject through a semi-formal discussion, or simply de facto acceptance due to a lack of objects, it is no guarantee of acceptance. In some cases, most notably in WikiProjects dealing with racial, nationalist or religious issues, if one ethnic or sectarian group has a majority, they will declare any sources that are favourable to them to be reliable.
Excluding POV issues, many WikiProjects feel that informal material by enthusiast societies are acceptable, but they are unlikely to be judged by an outside reviewer to be "high-quality"; again a RFC tends to attract interested parties much more than FAC/FARs and bloc pile-on votes are usually ignored anyway.
There have even been cases where the majority were supporters or sympathisers of a militant group proscribed as terrorists, and as a result, they declared by "consensus" that the said organisation's mouthpiece was a reliable source. Needless to say, it was not accepted by any neutral observers at the FAC. While it is relatively easily to break all manner of policies with abandon on a normal article, simply because only people of the opposite religion/race can be bothered at all, let along revert and argue against deliberate policy violations and misinterpretations all day. At FAC/FAR the reviewer generally doesn't have to patrol articles 24/7 to revert/fix policy violations, but only to object to a problematic, so neutrals can voice their opposition more efficiently, and therefore bad sources cannot be kept through attrition and conquest unless the nominators want the article to fail.
Aging of sources
Older works can become superseded by newer works that have access to more primary sources such declassified governments documents. These allow them more data from which they can reach their conclusion. As a result, all things being equal, a newer work can be more reliable and is often more representative of current scholarly thinking.
Different POV in foreign sources
Users need to be aware of the variations in historiography towards a certain topic across different countries and languages. This can extend to media outlets and can include explicit censorship or government interference, but may just be a reflection of a cultural or national bias. For example, writings about the History of Quebec may present strikingly different perspectives based on whether the writer is American, British, English-speaking Canadian, or French-speaking Canadian, as well as when they were writing. Under any of these circumstances (and especially if you are aware that your subject is in some way politically or culturally charged) it is important to have sources that represent a diversity of backgrounds.
Examples of red flags in websites that might be nominally reliable
- A lot of spelling errors. Even though the article might be affiliated with something reliable, like an academic institution or a government, spelling and grammar errors may be an indication that a particularly piece might not have been done thoroughly
- Kazi Nazrul Islam (delisted FA) used this conference discussion paper hosted by a university in this and many older versions. The article was full of English errors almost everywhere, and turned out to be an ad hoc presentation by a student that was not peer reviewed before being presented.
- Sikkim – per above mention (refer Section #4)