Wikipedia talk:Wikipedia is a tertiary source

Topic & primary source
The second full sentence in the article states: The topic of a Wikipedia article is never more than a primary source in the context of the Wikipedia article on that topic. Lets try substituting some topics for the phrase "topic of a Wikipedia article":
 * 1) George H. W. Bush is never more than a primary source in the context of the Wikipedia article on George H. W. Bush.
 * 2) The Brooklyn Bridge is never more than a primary source in the context of the Wikipedia article on the Brooklyn Bridge.
 * 3) Thermodynamics is never more than a primary source in the context of the Wikipedia article on thermodynamics.

Well, I suppose the phrase "is never more than" is an escape clause that makes all three substitutions technically correct, but substitute no. 2 looks a bit strange, and substitute no. 3 looks quite strange. --Gerry Ashton (talk) 19:16, 18 December 2007 (UTC)


 * the original sentence was obviously written by someone with poor writing skills. No matter how often I read it, it makes no sense whatsoever, and this is even without the 3 examples to demonstrate the point.
 * The sentence doesn't even work if I try
 * The topic of a Wikipedia article on thermodynamics is never more than a primary source in the context of that article.
 * "Thermodynamics" is never more than a primary source in the context of the Wikipedia article on thermodynamics.
 * So sucketh the prose of the whole shebang.
 * -- Fullstop (talk) 20:12, 18 December 2007 (UTC)


 * Never one to mince words, is Fullstop. Only time can tell whether or not such superbly punctuated opprobria will lead to PTSS. Nice points by Gerry by the way. I would say that the sentence is quite vague so I came to the talk page to ask Francis about it. Hi Francis -- maybe you can clarify its meaning/purpose here or improve it in the article? TIA. Avb 13:41, 19 December 2007 (UTC)

"[...] John Smith, the geographer, is a primary source for John Smith's work on geography [...]" (SlimVirgin 08:07, 29 November 2007, halfway through this talk page archive's section: Wikipedia talk:No original research/archive29)

Example: quote (by Tacitus) in the second paragraph of the Tacitus section.

Of course this also applies for declarations by a company quoted in a Wikipedia article on that company; for a politician's website quoted in a Wikipedia article on that politician; for the content of a film (even if a "secondary source" documentary film) in the Wikipedia article about that film etc.

Of course if the "topic" of a Wikipedia article is not a "source" (electricity is not a source; thermodynamics is not a source;...) there is nothing to worry about. If it's not a source, it's not going to be primary source in its own article, is it? Does that need to be made explicit in order to make it understandable for non-average readers? --Francis Schonken (talk) 20:03, 20 December 2007 (UTC)


 * When a reader has to ignore a statement from time to time because it makes no sense, that is poor writing. I don't know how many people there are like me, but due to my background in computer programming, whenever I read a rule, policy, etc., I always try to think of unusual, extreme, or special cases, and see how well the rule handles those cases. For people like me, rules that have to be frequently ignored because they make no sense are quite distracting.


 * (No doubt you've noticed that when computers encounter cases that make no sense to the program, the outcome is generally not good.) --Gerry Ashton (talk) 20:16, 20 December 2007 (UTC)


 * Humans are not computers. I've never assumed good faith on computers for instance, but I do on wikipedians (and computer programmers, who ususally apply something called "exception handling" in their programming, so that the program doesn't halt if you try to make it divide by zero, etc). Anyhow, I applied improvement - which might work better in the  "exception handling" department. --Francis Schonken (talk) 20:27, 20 December 2007 (UTC)


 * 1. Is it wise to make things understandable only for non-average readers?
 * 2. I don't understand the stuff about good faith of computers, which have nothing to do with the topic. Er, source I mean. No topic.
 * 3. I must be sub-average (or something). I don't understand the explanation of 20 Dec either.
 * How can a topic be a source? A topic is a topic even when it is a topic of a source, in which case it is then a topic of a source and not a source of a source, which is why it can't be a source of a topic on that source, which has nothing to do with the fact that a source of a topic of the source can't be a source of the topic of the source itself.
 * Now, everyone knows that. So I don't understand why Francis cannot see that when three people say a sentence doesn't make sense, then it is not a good idea to argue that it does (which, even postfix, sill causes my itsy brain to fault with an "unrecognized instruction")
 * -- Fullstop (talk) 21:25, 20 December 2007 (UTC)

Bertrand Russell's book A History of Western Philosophy is the main topic of Wikipedia's article History of Western Philosophy (Russell). Does that make sense to you? --Francis Schonken (talk) 21:52, 20 December 2007 (UTC)


 * I would certainly hope that that is the case. And not, for instance, "A History of Philosophy (Copleston)" -- Fullstop (talk) 22:05, 20 December 2007 (UTC)

Bertrand Russell's book A History of Western Philosophy is used as a source in, for example, Wikipedia's Anaximander article. The book is even listed as a secondary source in that article (see Anaximander). Do you agree with that? --Francis Schonken (talk) 22:10, 20 December 2007 (UTC)


 * What's your point? Why is it relevant to determine whether A History of Western Philosophy is/isn't what it is ostensibly about? And what does it have to do with the poor prose of a sentence on this page anyway?
 * To answer your question in anticipation of one that is on the subject of poor prose: if A History of Western Philosophy is an overview of the history of western philosophy, then it is a reference for the history of western philosophy, and legitimate in the article on a figure in the history of western philosophy.
 * Its completely irrelevant whether the book is primary, secondary, tertiary or a witches tit. Source-typing says nothing about the quality of a particular statement from a particular source.
 * So, if you don't mind, cut to the chase, or at least reveal the connection to the pitiful prose.
 * -- Fullstop (talk) 22:38, 20 December 2007 (UTC)

To me the current version parses as: "A subject can never be more than a primary source for Wikipedia content about itself" with a touch of WP:BLP, WP:SELFPUB and WP:SPS. Avb 00:38, 21 December 2007 (UTC)


 * hrm! BLP/SELFPUB were good hints. So, what Francis actually means is "An individual cannot him/herself be anything but a primary source for a WP article on that person" or "A book cannot be anything but a primary source for an article on that book or on its author." Right?
 * -- Fullstop (talk) 01:30, 21 December 2007 (UTC)
 * Yes. If you have any suggestions to offer on how this can be put in more legible prose I'd be all ears, I mean in a format that makes clear that this extends beyond persons and books too of course. --Francis Schonken (talk) 12:20, 21 December 2007 (UTC)
 * See --Francis Schonken (talk) 12:02, 22 December 2007 (UTC)


 * Can you give me an example of something that extends beyond persons/books? -- Fullstop (talk) 16:29, 22 December 2007 (UTC)
 * Sure,
 * Websites
 * Films (fiction and non-fiction), TV shows
 * Art, including lyrics and librettos
 * Anything written or printed on paper (including newspapers, magazines, flyers, postcards, leaflets and posters)
 * Public speeches (e.g. "I Have a Dream")
 * Things carved in stone (e.g. The Rosetta Stone)
 * Companies (I mean, the company as a whole, not the individual people involved in it)
 * Governments, political parties, and all sorts of organisations
 * Fictional characters (a ficitional character, for example representing a queen might in some circumstances be a secondary source on that queen, e.g. in a historical drama depicting how that queen allegedly reacted to certain historic events)
 * I'm sure I didn't cover everything yet. Anyhow, my point is that this is not some marginal effect for a limited range of articles: it occurs quite often. --Francis Schonken (talk) 17:27, 22 December 2007 (UTC)


 * What you're trying to say is:
 * "A subject (of an article) is not a secondary source for an article on that subject."
 * But we already know this because a secondary source is at least "one step removed."
 * -- Fullstop (talk) 17:55, 22 December 2007 (UTC)
 * How 'bout:
 * Any source that is the main topic or subject of a Wikipedia article, can only be used as a primary source for that article.
 * Yes, self-evident. That's what policies are for. Anyway, rather self-evident than "it makes no sense whatsoever", which was your original appreciation of the idea through my poor writing skills. --Francis Schonken (talk) 09:01, 23 December 2007 (UTC)
 * You can reiterate "primary source ... close to source", "secondary source ... one step removed" ten ways to Sunday, but it isn't necessary.
 * As such, the message is superfluous.
 * Anything to do with source-typing is not "self-evident." Even the title of the essay is not "self-evident." Even so, 'making no sense' refers to the sentence construction, not the message.
 * -- Fullstop (talk) 14:08, 23 December 2007 (UTC)

<<<Re. "the message is superfluous": "There seems to be a segment of the community who takes the view that primary sources are a certain type of source and the typology is per source [...]" This was written less than an hour ago. No, the Wikipedia community at large does not always make the simple deduction that "primary source ... close to source" for instance implies that the Encyclopædia Britannica is not a tertiary source nor a secondary source, but a primary source when used as a reference in the Encyclopædia Britannica article (and that, for instance, this reference is by all accounts a reference to a primary source).

No, not superfluous. --Francis Schonken (talk) 23:13, 23 December 2007 (UTC)


 * 1. if you were consistently going by what a segment of the community thought, there would be no need for WP:WITS.
 * 2. However, the sentence under discussion stands in WP:WITS, which notes:
 * "A secondary source is at least one step removed from the described topic."
 * There are no doubt dozens of ways to describe what is/isn't a primary/secondary/tertiary source. But only one is necessary.
 * Not that I care either way; If it makes you feel better, you could write the story of your childhood. Just like the sentence under discussion, it wouldn't further the point. -- Fullstop (talk) 00:30, 24 December 2007 (UTC)

Current discussion
See:
 * Village pump (policy)
 * Wikipedia talk:What Wikipedia is not
 * User talk:Jimbo Wales

Leaving no doubt:
 * Support promoting this essay to guideline. --Francis Schonken (talk) 08:20, 8 February 2016 (UTC)
 * Oppose: Think it needs more revision: For example I don't think anyone would seriously agree that a stub article on a notable painting would be better if the illustration of the painting was removed, but there's a section on image use that would require this. Leave images to the existing image use policies and guidelines, which handles them a lot better than this essay. Adam Cuerden (talk) 13:22, 9 February 2016 (UTC)
 * Re. stubs: straw man argument, stubs can be deficient in various ways (they're incomplete articles by definition), that doesn't mean they wouldn't be valid stubs. I don't see the problem adding a Primary sources template to a stub until the moment it complies to policies such as " base an entire article on primary sources, and be cautious about basing large passages on them." (WP:PRIMARY) and "Wikipedia is not a mirror or a repository of ..., images, or media files" (WP:NOTGALLERY). --Francis Schonken (talk) 10:07, 12 February 2016 (UTC)
 * Oppose: This would work better as an information page. It is not intended as a guideline, but an explanation of them. ViperSnake151   Talk  16:13, 9 February 2016 (UTC)
 * Strongly oppose (1) This proposal reads like yet another excuse to delete stuff. (2) If I am to be entirely frank, over reliance on secondary sources is a sign of incompetence. Moreover, the entire theory of primary, secondary and tertiary sources has been drawn from a particular discipline (history) and is wholly inapplicable to some other disciplines. (3) The concept of primary, secondary and tertiary sources is sufficiently obtuse that it is not likely to be applied properly, because most of our editors can't grasp it. James500 (talk) 18:04, 10 February 2016 (UTC)
 * Your objection No. 3 is the one that kept me from proposing an upgrade to guideline for so many years after it was written, it is certainly an argument I understand. However every once and awhile an issue arises which can't be properly dealt with without referring to WP:PRIMARY's " base an entire article on primary sources, and be cautious about basing large passages on them", which indeed is even more obtuse for most readers without a subsidiary guideline or essay that makes that policy passage somewhat more digestable. What I expect from a possible promotion to guideline is that in such case this explanation of the broader perspective would be easier to find (while now I suppose it is not found unless I'm involved in the discussion and point to it).
 * I can't subscribe to your objection No. 2 though: the guidance promotes more use of primary sources, respecting them for what they are (while other guidance sometimes leaves an editor with the impression primary sources are an "inferior" kind of sources). In this context: Sources – SWOT analysis maybe is better at explaining the specific "strenghts" of primary sources in Wikipedia context. That essay would also be my answer to your objection No. 1. Re. "drawn from the history discipline": that's an argument I completely reject since Wikipedia is largely in the business of history-writing (see WP:NOTCRYSTALBALL if you don't believe me that this is an actual project-wide policy). So if history-writing is a discipline that by and large applies to Wikipedia, it is almost unavoidable to encounter some basic principles of that discipline in the process, and as unavoidable to explain such concepts to editors that knowingly or unknowingly practice the "history" discipline, when it becomes unclear which edits are acceptable or unacceptable because of such basic principles of history-writing. It is about explaining that as widely understandable as possible. --Francis Schonken (talk) 10:51, 12 February 2016 (UTC)
 * Francis, there are fields of knowledge that bare absolutely no resemblance whatsoever to history. NOTCRYSTAL is irrelevant as I am not talking about predicting the future. Mathematics is, as Andrew rightly points out below, one of them, as the proof of a theorem is either right or wrong, and it will be self-evident whether it is true or not, because it depends only on the application of logic to a set of axioms. There is no scope for expressing any kind of personal opinion whatsoever. Whether a theorem is true is not a question of historical fact, and it does not depend on evaluating historical sources (what a historian does), because mathematics does not use any sources whatsoever for that purpose but relies instead on a chain of reasoning. That said, I can think of an even better example, a discipline that does use sources, but uses them in a way that bares no resemblance whatsoever to what a historian does, and which uses its own unique classification of sources (no primary, secondary or tertiary at all in that system). The idea that concepts invented by historians are necessarily applicable to other disciplines is WP:RANDYish. It is something that could only be dreamt up by a historian (or a young history student) who had never studied other subjects (or at least had not understood them). James500 (talk) 21:01, 14 February 2016 (UTC)


 * Oppose per WP:CREEP and WP:NOTLAW. This is a counsel of perfection which does not reflect actual common practise on Wikipedia.  For example:
 * Maths articles are commonly written without reference to any sources
 * Articles about places are often written from primary material such as census returns and geographical databases and then supported by original photography
 * Articles about current events such as politics, sport, weather and disasters are commonly compiled from news reports and specialist websites which are primary in nature
 * Articles about pop culture such as videogames, movies, music and TV are commonly scraped together from primary material such as interviews, sleeve notes, direct experience and the like.
 * Such work is not just exceptional or primitive drafts; you can find it regularly promoted on the main page as featured articles, in-the-news &c. For non-academic topics, such expedients are necessary and we should be content that some effort is being made to provide any kind of source rather than raising further barriers to entry. Andrew D. (talk) 19:13, 13 February 2016 (UTC)


 * Oppose 1/ The distinction between primary and secondary sources is not straightforward, and does depend very much on the subject. In addition to the examples already presented, consider the example of Samuel Johnson's 18th century essays on Shakespeare in hi edition of The Plays of William Shakespeare. They are secondary sources with respect to a discussion on the plays; they are primary sources with respect to s discussion of Johnson, or with respect to a discussion of the history of Shakespearian criticism. In a completely different field, take any scientific research article -- we normally consider them primary sources, and they are with respect to the experiments reported,  but the almost always contain a preliminary discussion of the state of the field, and that is a secondary source on he subject. Even in the subject of history,
 * There is no way to give a complete presentation of this that will answer the in advance all the possible questions that arise in the writing of a Wikipedia article.
 * 2/The extent of use of primary sources in WP is very large, and appropriately so. Some examples have been given by others above. But basically, the only thing we really need secondary sources for is interpretation or significance or judgment. And even some of the accepted uses are not as straightforward as they appear in the essay : we quote politicians for their political views, but that is exceedingly naive--what we can quote them for, is what they choose to say about their political views in particular circumstance.
 * 3/The use of secondary sources does not guarantee NPOV--in many cases, it merely conceals it. There is always the questions of what sources to use, how to summarize them, what to quote from them, what to attributes and in what manner. Objective writing is in many cases best done by giving the original primary sources as they are, and letting the reader do the interpretation. It is very easy to give neutral-appearing secondary sources there aren't actually neutral in the least. An excellent example of this is the use of the 1913 Encyclopedia Brittanica--it is in some areas   unreliable in the same fashion as the Great Soviet Encyclopedia. But we use it here for judgement extensively, and always wrongly, unless we say that According to early 20th century British views...
 * Despite all this, the essayis very useful--as an essay. It does present the basics, but the extent of qualification is mmuch greater than given there.  DGG ( talk ) 04:56, 15 February 2016 (UTC)