Wikipedia talk:Reliable sources/Archive 49

Please move your question to

WP:RSN - this is not the right place for it, see framed recommendations on top of this page. --Francis Schonken (talk

) 02:44, 29 June 2015 (UTC)

If and when there is a discussion at RSN, ping me. I am not watching this page. I concur that this is not the place for this discussion.--

WP:WAWARD

) 00:04, 30 June 2015 (UTC)

If not, I'd like a section in this article describing why one can't use youtube. Thanks. 88.90.245.52 (talk) 02:35, 26 June 2015 (UTC)

Youtube itself is a

Wikipedia:Identifying_reliable_sources#Self-published_sources_.28online_and_paper.29). There's also potential copyright violation issues for material that would qualify as reliable, like a documentary. Ian.thomson (talk

) 02:40, 26 June 2015 (UTC)

And from there, it depends on the details, so you're better off posting a link to the exact YouTube video that you'd like to use, and the name of the article, at

WP:RSN, so that someone can help you figure out the specific situation. WhatamIdoing (talk

) 03:11, 26 June 2015 (UTC)

As always, context is king. For most situations, one simply should not use YouTube as a source for anything, for the reasons mentioned above, among others. That said, if you needed to support the statement that YouTube is a Google company then it might be perfectly acceptable to dig into YouTube's about pages and find that reference, although a book or even a news report might be a more authoritative source. Likewise, if you needed to support a statement that a particular person had announced his or her intention to run for a particular office during a particular election cycle by posting a video announcement to YouTube, it would certainly be appropriate to use a YouTube link to the announcement to help support that assertion. That said, the situations where YouTube or any other video sharing site are authoritative as sources are extremely limited, and generally there are better sources that can and should be used instead. Additional information can be found in the

WP:Twitter essay. Etamni (talk

) 09:15, 1 July 2015 (UTC)

Assuming your question is about sources hosted at YouTube, and not YouTube itself as the source (the way your question is couched, the opposite is implied, but I suspect you meant what I am assuming), then sure, depending on what it is. First, as noted, there are huge numbers of copyright violations there, which

can't be linked to. There's also truly vast amounts of YouTube content that is not reliable, user generated video. But if that's not the case (in disagreement with some of the above) there are also vast quantities of reliable sources that can be found at YouTube and linked to their hosting there. As an example, BBC News has an official channel with many videos, as does, well, almost any news source you can think of that has a TV broadcast arm. But remember that for this example, YouTube is not the source, BBC News is. The reasons I mentioned that it has an "official channel", is because you will also find that random user Kitten748 has a penchant for uploading BBC news videos. If it's not the organization's official channel, it's presumptively a copyright violation. So long as the context does not make us question whether the BBC News report is genuine, it would be evaluated as a reliable source just like any other, and it's irrelevant to that consideration where that video is hosted.--Fuhghettaboutit (talk

) 10:59, 1 July 2015 (UTC)

This RFC about the use of columns as sources may be of interest to editors here. Columns are opinion content, but they are typically written by professional journalists that work for the publication. I have a disclosed conflict of interest on that page. CorporateM (Talk) 21:17, 10 July 2015 (UTC)

It should be mentioned on

I don't see any value in adding this here. When and how to use a template is normally documented in the template's /doc page, not in sourcing guidelines. WhatamIdoing (talk) 16:48, 20 July 2015 (UTC)

Three related issues, and suggestion of what to do about them:

1. We do not have anything at all about citations to conference proceedings. Present practice is to cite these exactly like journals, and I see no evidence that they're being treated as less reliable, but they probably should be, on par with PhD dissertations, or perhaps even masters theses (depending on the prestige and exclusivity of the conference and the panel reviewing papers to be presented), since while they're subject to some degree of peer review (enough to be accepted, and enough, in reaction, to be criticized in later publications if they turn out controversial), it's usually less than would be required for publication in a peer-reviewed journal.
2. How should we treat non-academic, industry/trade journals? Most major (even most minor) industries have multiple trade-insider publications, and their content is usually written by subject-matter experts, but they're not subject to the peer-review rigors of academic journals. My feeling is that these should be treated with the same care as other primary-or-mostly-primary sources, but like academic journal papers are liable to be high-quality sources when they aren't presenting anything controversial. Where they contain material more in the form of journalistic reporting on their field, they should be treated as news sources as long as they're acting independent of the topic (i.e., not just regurgitating press releases, giving all-favorable product reviews, or otherwise acting like house organs).
3. For a
   Vlfberht
   , where an online Gedenkschrift source has been cited. The paper was written by a published expert, on a site devoted to papers by professional archaeologists in honor of one of their regional mentors who died recently. A Gedenkschrift like this is essentially a virtual conference. This is thus about as much peer review as a masters thesis at least, perhaps as much as a doctoral or journal paper; while we don't know the exact criteria, it is written and published by people with academic reputations to maintain, in a narrow field (Scandinavian archaeology), where errors or controversial claims would be noted by colleagues. The specific work in question is straightforward, mostly a matter of gathering raw data about the inscriptions on Viking swords, and drawing statistically-based inferences from it. I've flagged this material in the WP article as (presently written) being improperly cited to a primary source. It's thus probably subject to removal if not re-sourced some other way (I've only found one journal paper that can source some of it, so some material would still be lost). It seems preferable to me to attribute the work carefully, and state in WP's voice that it's a hypothesis by this specific researcher, not to state the claims in the article as if they're known facts. That should be sufficient. The hypothesis presented is noteworthy and relevant, and appears to have influenced a 2012 PBS Nova documentary on the topic, though they did not cite it by name. [I can't find any other source for the information, anywhere, so it seems highly probable.]
4. How do we feel about carefully attributed (or directly quoted) use of a masters thesis, e.g. "[Researcher_name], in a 2011 [University_name] masters thesis, drew parallels between the results found by [Peer_reviewed_paper_1] and [Peer_reviewed_paper_2]"? (I.e., secondary work, not the presentation of new data, which would be primary sourcing.) I'm especially thinking not of hard sciences (e.g. claims to have mastered cold fusion :-), but rather of non-controversial work in obscure topics in the social sciences, like archaeological, anthropological, linguistic, and mythographic details about which there is very little peer-reviewed material yet, and very little funding for further research (some subjects in these fields get examined once only in journals). Whether good encyclopedia writing or not, common WP practice right now seems to be to simply say "[Peer_reviewed_paper_1] concluded [X], and [similarly|dissimilarly], [Peer_reviewed_paper_2] concluded [Y]", letting the reader connect the dots. It seems better to cite someone else connecting those dots, even if indicating clearly that the synthesis is only from a completed thesis. This would frankly seem far better sourcing that much of what it done in pop-culture topics, where we cite random journalist connecting the dots between two other random journalists' work, without any peer review of any kind other than whether an editor thinks it's due-diligence enough to publish and will interest their newspaper/magazine readers. We regularly quote/attribute primary sources "with caution". Simple attribution in cases like what I have in mind would seem to be sufficient caution, for material that is not controversial, either in making extraordinary claims or contradicting prevailing scholarly views. As a concrete example, I've found a thesis, Prehal, Brenda (2011). "Freyja's Cats: Perspectives on Recent Viking Age Finds in !egjandadalur North Iceland" (PDF). New York University. {{cite journal}}: Cite journal requires |journal= (help) Among quite an array of material, there's a section suggesting continuity/verisimilitude of use, reported separately in previous peer-reviewed publications, of various Northern and Western European words for "cat" as vulgarities, simultaneously in reference to female genitals and to imply male cowardice. I haven't found a peer-reviewed paper that makes the same connection, which is only barely synthetic/analytic (namely that "puss" and its equivalents have a triple use with a long, multi-language history). I actually need the same paper for another pretty obvious, non-controversial synthesis of prior work, this time about cat demographics in historical Scandinavia (about which very, very little has been published in journals, even in the Scandinavian languages; one of the only two papers I can find on this is a doctoral dissertation, with only one journal paper on the subject).

Conclusion: It's important that we include something about academic (and tech, and other industry) conference proceedings and how to approach them as sources, because of the frequency with which they publish material we want to cite. My take: they are primary or mostly primary (unless just summarizing the state of current research, in which case they're tertiary), and should be used with caution, attributed as such presentation, and replaced with secondary, or at least peer-reviewed primary sources in journals, when possible. Tech and consumer conference presentations of new products, technologies, methods, and draft standards should be treated as strictly primary sources. But when inclusion criteria in academic conferences are very stringent, presentations can be treated the same as journals if publicly available in [e-]paper or recorded form. I think what I wrote above about industry journals can easily be directly adapted into guideline wording about them. Finally, the statement that "Masters dissertations and theses are considered reliable only if they can be shown to have had significant scholarly influence" is a bit overbroad. As long as they present nothing controversial, such a paper should be treated like any other primary or semi-primary source, if it"s completed, approved, and published/archived by an accredited university, in publicly available form, or is in the online equivalent of a Gedenkschrift/Festschrift or other scholarly compendium, and it isn't making controversial or extraordinary claims. This would of course not extend to undergraduate papers, unfinished theses/dissertations, and other pure-primary sources that are not from reputable publishers.  — SMcCandlish ☺ ☏ ¢ ≽^ʌⱷ҅_ᴥⱷ^ʌ≼  02:49, 3 July 2015 (UTC)

• @
  Fgnievinski (talk
  ) 05:59, 3 July 2015 (UTC)
  • Reliability is tied to reputability of publisher among other factors, including reputation of the authors, and of the reviewers (if any), the target audience, the breadth and depth of the work, the nature of the claims made and their relation to currently-accepted thinking on the topic, and much more. Some of this relates to notability, but does not equate to it. Disreputable publishers can be notable, and reputable ones can surely fall short of notability, since that requires significant coverage in multiple, reliable, independent secondary sources, and not many articles in off-WP publications are written about publishing companies themselves, meanwhile even coverage of universities does not often cover their publishing operations. The notability of an institution as a campus people go to does not automatically make it notable, or reputable, as a publisher; notability of any kind is never transferable. Whether something already has an article on WP doesn't equate to notability either, especially for something like conference proceedings, since we have hardly any articles on any of them at this point (less than 10!). Nothing in
    literature reviews, most material in academic journals is primary, or a mixture of primary and secondary, that we must "use with caution", because it is presenting novel claims for others to test and develop, and does not yet represent the mainstream, accepted view or even necessarily a noteworthy minority one. Much of it, especially in the hard sciences, is just experimental "noise". The second point of my original post (aside from the blind spot) is the "inverse" blind spot, of treating certain primary sources published as theses by universities as if verboten, when they're really just published primary sources like any others. They're not as useful as fully peer-reviewed ones, but in cases like I've outlined, the idea that they must have had a notable impact is probably too stringent, if we at least directly attribute them, and make it clear what kind of publication they are, in the article text, not just cite them as if authoritative. For some topic like the early history of domestic cats in Scandinavia (one of the above examples), it may well be that there will never be any further research. At bare minimum it should be enough that something in a peer reviewed journal has cited the thesis in question. I think the language we have about graduate theses was written with big, hard-science fields in mind, where new ideas are usually wrong and are very difficult to test. In many of the softer disciplines this isn't really the case, and much of the work in question is rote reporting of data with minimal synthesis. The "caution" needed is lower, and we can probably account for that in the guideline without much effort.  — SMcCandlish ☺ ☏ ¢
     ≽^ʌⱷ҅_ᴥⱷ^ʌ≼  08:23, 3 July 2015 (UTC)

• *Groan* This is a huge number of words that ought to just say 'case-by-case basis'. This stuff is so variable from one academic field to another than trying to write general guidelines is hopeless. And anyone who doesn't have a reasonable sense of the landscape in the field they're trying to write about shouldn't be writing about it anyway. How "we" "feel" about individual academic communities' publication norms is irrelevant. Opabinia regalis (talk) 06:32, 3 July 2015 (UTC)
  • I don't think the attitude is necessary, thanks. I never said anything about how we feel "about individual academic communities' publication norms"; you're just making up weird stuff and putting it my mouth. I'm talking about how we decide, internally, to approach particular types of sources in, e.g. anthropology and linguistics (I have one of those degree things in those fields, so you can keep your assumptions to yourself about what I know about their publication norms). I'm fairly confident I can write serviceable guideline text to cover all of this, despite the pessimism. I've written substantial portions of quite a few of our guidelines. I also decline to apologize for writing clearly, and with sufficient detail to cover what needed to be covered to forestall various (though clearly not all) distracting objections that miss the point, and for tying these matters to actual, specific examples (which is almost always required for proposals to go anywhere). Even if we conclude on "case-by-case basis", we don't presently indicate anything to suggest some of these categories of publication be treated on a case by case basis. For some we say nothing (leaving them in limbo, and with conf. proceedings, leading to over-citing), while with others we're being over-inclusive (not distinguishing industry "journals" from peer-reviewed academic ones, which has implications for our tech industry coverage in particular), and for another category we're a bit too exclusive. I'll just draft something I guess, in absence of any constructive suggestions.  — SMcCandlish ☺ ☏ ¢ ≽^ʌⱷ҅_ᴥⱷ^ʌ≼  08:23, 3 July 2015 (UTC)

• Regarding your point #2, could you post a link to an example of the type of journal to which you refer? As things stand, I'd agree that academic journals would be more reliable than trade journals, but even US Weekly is reliable enough. RS is a pass/fail criterion. We only need to establish a hierarchy when sources contradict each other.

As for masters' theses, if they've been reviewed, then why not? Though I concur that "case-by-case basis" is relevant here.

As for establishing rules/guidance/call-it-what-you-like for conferences and to a lesser extent rules for theses, I could support adding specifics to WP:IRS if the Wikieditors are having problems that adding such rules could solve. Are people citing conferences using a confusing mishmash of jury-rigged formats? Are people deleting conference-sourced material because they assume conferences are not RS? Any other problem? These questions are not rhetorical. Darkfrog24 (talk) 21:34, 3 July 2015 (UTC)

How they're formatted isn't relevant; that's a

WP:RS matter. Yes, the problem is that editors will delete citations of conference proceedings or masters theses, or tag them with dispute tags, or remove the entire passage, even when they are used appropriately and cautiously as primary sources, and will do so not because of any controversy about the material but because they believe they're not permissible sources (or that someone should have to argue with them for ten days on the talk page to make a case that there should be an "exception"). This guideline effectively encourages them to do so with regard to theses, and less directly does so (by omitting mention of them at all) with regard to proceedings.  — SMcCandlish ☺ ☏ ¢

 ≽^ʌⱷ҅_ᴥⱷ^ʌ≼  00:51, 22 July 2015 (UTC)

1. Conference proceedings are firmly discouraged, but not banned, at
   WP:DUE
   problems, since a lot of things that are presented never get mentioned again.
2. We should treat non-academic, industry/trade journals exactly like we treat other forms of non-academic media, e.g., political newsweeklies. InformationWeek == Newsweek for reliability purposes. Neither of them are stellar, but both of them are okay for supporting non-extraordinary claims.
3. We should treat a Festschrift exactly like we would treat any other compilation of academic writings, which means ignoring why it was written and focusing on the editorial process.
4. If a master's thesis is actually WP:Published, and it appears to meet or exceed the same standards that we would apply to any other source, then you can cite it. If it is not published, then do not cite it. If it is self-published, then follow those rules. If it is published by a publisher with a reputation for fact-checking and accuracy, then treat it like any other source that is published by a publisher with a reputation for fact-checking and accuracy. The fact that the publication was written in exchange for getting an academic degree rather than in exchange for money is really unimportant.

Conclusion: Classification as primary, secondary, and tertiary is largely irrelevant, since it usually changes what a source is reliable for, rather than whether it is reliable for anything at all. We don't need

Fgnievinski (talk

) 22:36, 20 July 2015 (UTC)

I agree that different fields have different standards. Also, "conference proceedings" encompasses a lot of sources. There is an obvious gap, for example, between "Here's the poster that I put up at the conference" and "Here is the peer-reviewed article about the conference, published in a respectable academic journal". WhatamIdoing (talk) 07:11, 21 July 2015 (UTC)

I'm with Fgnievinski on that. I mean "conference proceedings" in the formal sense. I'm not talking about posters, or about coverage of a conference in other publications, or a programme. I'm talking about the publication of the conference, the hardcopy of the material presented at it. Were the papers that were presented peer reviewed in detail? Reviewed to an extent by an acceptance committee on the basis of research quality? Just "reviewed" on the basis of whether they seemed like interesting presentations? There's a big difference between presentation a medical professional conference, TED, and a sci-fi or comic convention, in decreasing order of usefulness as a source about independent facts [con highlights can actually be a good source for things like actors' and writers' own statements about their work; that's not what I'm on about here].  — SMcCandlish ☺ ☏ ¢ ≽^ʌⱷ҅_ᴥⱷ^ʌ≼  00:51, 22 July 2015 (UTC)

FYI

 – Pointer to relevant discussion elsewhere.

This topic has had a lot of churn on various pages lately, so I thought people here might be interested in this discussion: Template talk:Citation needed#When to remove unsourced info vs. when to add this tag?.  — SMcCandlish ☺ ☏ ¢ ≽^ʌⱷ҅_ᴥⱷ^ʌ≼  23:40, 22 July 2015 (UTC)

Wikipedia articles are required to present a neutral point of view. However, reliable sources are not required to be neutral, unbiased, or objective.

I have a problem with the unqualified declaration made in the second sentence. I believe it should read, "However, reliable sources do not always have to be neutral, unbiased, or objective. (addition in italics)" My fear is that Randy in Boise will read that simplistic sentence and conclude, "Hey, I can just quote what somebody says as fact, because I got it from an RS!" (I have already encountered such a "Randy".)

I appreciate that "no RS is perfectly unbiased", but this is a special case where we must hold our sources accountable to some of the same standards to which we hold ourselves; at least we have to exercise discretion. The guideline must make clear that opinionated sources are only properly used in the context described: i.e., a topic has been recognized to be controversial, and the source is only to be used (ideally with in-line attribution) to verify the existence of the opinion (subject, of course, to

If you do not already, you need to realize that this Identifying reliable sources guideline is really a somewhat simplified interpretation or guide to policy, not the source which creates the policy. The actual policies involved here are (at least)

WP:CONSENSUS#No consensus). Regards, TransporterMan (TALK

) 19:08, 15 July 2015 (UTC)

• The key to understanding NPOV is this: it applies to Wikipeida, not to our sources. We (Wikipedia's editors) must be neutral in our writing... Our sources don't have to be neutral in their writing. Sources can state opinions on the topics they are writing about (in fact, we rely on them to do so). Sources can challenge, dismiss or ignore viewpoints that they disagree with. We (Wikipedia's editors) can not. We must report all significant views... even if we disagree with those views, or think they are wrong. Blueboar (talk) 22:20, 15 July 2015 (UTC)
• I think that the fundamental problem with this question is that it conflates "neutral" with "accurate". The key characteristic of a reliable source is its reputation for fact-checking and accuracy. They should also have other indications of reliability (see
  WP:RANDY can actually find a source that has a reputation for fact-checking and accuracy, professional editors, etc., then yes, even Randy is allowed to use it (usually) to present something as a fact. "Bias" isn't on the list: an academic journal may be strongly "pro" on the question of whether humans affect Earth's climate without becoming unreliable.
  On your example, a book about how astronauts' wives worried and had to play the devoted, patriotic wife in public probably does not have a reputation for accuracy about how NASA made politically significant decisions. A book can be reliable for one thing and unreliable for other things. WhatamIdoing (talk
  ) 17:00, 20 July 2015 (UTC)
• This exactly: 'The key to understanding NPOV is this: it applies to Wikipedia, not to our sources.'  — SMcCandlish ☺ ☏ ¢ ≽^ʌⱷ҅_ᴥⱷ^ʌ≼  00:14, 22 July 2015 (UTC)
  • Well, as with most bromides it only goes so far - it takes skill and wide reading of sources (generally in the tertiary voice) to recognize what neutral presentation is - you certainly won't learn it by following the partisan source. Alanscottwalker (talk) 00:25, 22 July 2015 (UTC)

Actually, you can learn neutrality by reading partisan sources... you just have to read lots of them, and (as a group) they have to express a wide range' - and ' of partisan viewpoints. Blueboar (talk) 14:08, 24 July 2015 (UTC)

Most unlikely, and not actually

well recommended for the anyone can do it encyclopedia. (see, Tertiary Sources section). Original misplaced emphasis, and lack of context is the bane of being neutral and of presenting the unoriginal - so the partisan source which is all about persuading one to place-emphasis-here and buy into a partisan context (or pretext) in ways that may in a solid tertiary presentation be unacceptable, are poor models - we actually have to exclude and not emphasize, that which should not be emphasized, and appropriately contextualize that which must be contextualized. -- Alanscottwalker (talk

) 19:21, 25 July 2015 (UTC)

WP doesn't actually care about "a reliable publication process"; that means a process that reliably produces a publication. We need to reword this to make some kind of sense that is the same kind of sense to anyone who reads it. I would strongly suggest we work the word "reputable" in, with regard to the publisher as an entity, and distinguish this from the editorial process.  — SMcCandlish ☺ ☏ ¢ ≽^ʌⱷ҅_ᴥⱷ^ʌ≼  23:44, 22 July 2015 (UTC)

Perhaps I am misunderstanding... but I think we do care about whether the publication process is reliable (certainly we don't want to cite a source that has an unreliable publication process)... but your point about "reptutable" is well taken. Perhaps we care about both? Blueboar (talk) 13:58, 24 July 2015 (UTC)

WP doesn't care as a policy matter about the reliability of publishing output processes in the production sense (e.g. if Elsevier or the Wall Street Journal or whoever, due to severe budget cuts, start putting out publications behind schedule, or on crappy paper, that's not of any concern to WP, as long as the quality of the material doesn't weaken. My point was that we're consistently using "reputable" everywhere else to refer to the publisher, and there's been noteworthy confusion in the past with people mixing up reliable source and "reliable" publisher. I think what this passage is trying, so vaguely, to say is something to the effect of "a [something] editorial process at a reputable publisher", or however we'd work that into the sentence, and whatever quality that [something] is ("trustworthy", "respected"?). I'm rather sure we shouldn't use "reliable" there. The problem with doing so is that it produces a circular, meaningless pseudo-definition, "a reliable source is one from a reliable editorial process", when the whole point of the exercise is to nail down what we mean by "reliable", by defining it in other terms relating to the reputation of the author(s)', the quality of the editorial process, the publisher's reputation, and the currency of the material. It just doesn't do anything useful to say a reliable source has a reliable author, a reliable publisher, with a reliable editorial process, putting out material that is reliably not obsolete. :-)  — SMcCandlish ☺ ☏ ¢ ≽^ʌⱷ҅_ᴥⱷ^ʌ≼  12:46, 1 August 2015 (UTC)

Wikipedia very much cares about a "reliable publication process", meaning "a process than produces a reliable publication". It is also not circular, because we do provide other definitions.

If you're finding it confusing (or perhaps only irritating), then the current text: Reliable sources may be published materials with a reliable publication process, authors who are regarded as authoritative in relation to the subject, or both. could be re-written to say something like this: Reliable sources may be published a publisher whose editorial process has led to a reputation for fact-checking and accuracy, by authors who are regarded as authoritative in relation to the subject, or both. WhatamIdoing (talk) 20:52, 1 August 2015 (UTC)

Moved to Help talk:Citation Style 1 § Merging Template:ArXiv

If what is published at

An argument can be made that they're self-published sources, since arXiv is acting as a self-publishing house for academics (in

Another argument can be made that it's

Exception made for material that is actually published in a journal by the time we need to cite it; the arXiv URL might be the only one we have, if the journal is behind a paywall.

(I asked this initially at

And I'll give you the same answer here as on

books

} 14:28, 24 July 2015 (UTC)

[plonk]  — SMcCandlish ☺ ☏ ¢ ≽^ʌⱷ҅_ᴥⱷ^ʌ≼  20:42, 24 July 2015 (UTC)

You could say that for any source, no matter how bad. Are you arguing that articles on arXiv meets the criteria "Articles should be based on reliable, third-party, published sources with a reputation for fact-checking and accuracy."? Because I don't think they do. If they are actually preprints that will be reliably published, then we just have to wait until that happens. The process of reliably publishing them might even produce significant changes. If they aren't going to be published, they're just a paper on a website. Doug Weller (talk) 18:23, 24 July 2015 (UTC)

Indeed. And like anything, they can be cited just as any other webpage on the internet can be cited, subject to our usual restrictions on

books

} 19:46, 24 July 2015 (UTC)

Arxiv is a simply a repository for pre-prints and does not by itself accord any reliability-points to a publication. As far as I can see, there are only two scenarios underwhich we could cite arxiv prints, :

1. When the paper has been peer reviewed and published by a reputable journal and the arxiv print is essentially a
   convenience link
2. As a self-published source, when the author, subject, secondary citations to the work, claim it is being cited for, etc justify such use under
   WP:SPS

Btw,

WP:RSN may be a better venue for such discussion if there is dispute over providing an arxiv link in particular instances. Abecedare (talk

) 18:40, 24 July 2015 (UTC)

The arxiv also has many postprints as well as preprints, btw. Altough as far as I understand SMcCandlish's questions, they just can't fathom why anyone would cite something from the arxiv at all.

books

} 19:46, 24 July 2015 (UTC)

@

TfDed. It's two legitimate functions are redundant with existing templates, and its potential for abuse is high.  — SMcCandlish ☺ ☏ ¢

 ≽^ʌⱷ҅_ᴥⱷ^ʌ≼  21:16, 24 July 2015 (UTC)

Here's a scenario under #2 (and there are many more of those scenario where one would cite the arxiv).

books

} 21:25, 24 July 2015 (UTC)

Already addressed under "Meanwhile, #2 can be done with ... if it's ever needed."  — SMcCandlish ☺ ☏ ¢ ≽^ʌⱷ҅_ᴥⱷ^ʌ≼  22:31, 24 July 2015 (UTC)

As for scenario 1, the solution would be to use {{

books

} 21:27, 24 July 2015 (UTC)

That's exactly what I said. If even you agree that the only purpose for {{cite arXiv}} is citing the unpublished version of something, then we don't need a separate template for that. {{Cite web}} exists for a reason.  — SMcCandlish ☺ ☏ ¢ ≽^ʌⱷ҅_ᴥⱷ^ʌ≼  22:31, 24 July 2015 (UTC)

I agree with Headbomb here: although appearing on arXiv is not sufficient by itself to turn something into a reliable source, there are nevertheless many reliable sources that appear on arXiv that we would want to cite (scenario #2), and the catchall {{

cite arxiv}} and {{cite web}} where the reference in question has been properly been published in a journal or a book as well as having an arxiv or web version (indeed I have frequently seen {{cite web}} with a url pointing to the official publication page for a journal paper or to a Google books entry), and these should generally be converted to {{cite journal}} or {{cite book}} with |arxiv= or |url= or |contribution-url= when they are discovered. —David Eppstein (talk

) 21:36, 24 July 2015 (UTC)

Please re-examine the argument. You're stating why someone could cite something at arXiv, when the question is why would we do it with this special template instead of the existing ones. Even Headbomb agrees (see above) that {{Cite journal}} is the one to use for citation of something at arXiv that has been reputably published. If the only use-case for {{cite arXiv}} is citing unpublished research, we have {{Cite web}} for that. The {{Cite arXiv}} template is redundant in either scenario, and provides as easily-abused avenue for source misuse. Due to rampant misinterpretation

WP:OWN every detail of citation formatting, it's extremely unlikely that checking uses of {{Cite arXiv}} on a case-by-case and replacing them with the equivalent, long-standing, general templates to indicate "this has been checked out and is a valid use of an arXiv citation" (using the appropriate replacement template for whichever of the two scenario types the use qualifies as), would not be met with revertwarring. There is thus no clear avenue for patrolling abuse of this template, which serves no purpose anyway. The obvious solution is to deprecate it, convert all of its in situ uses, and then TfD it.  — SMcCandlish ☺ ☏ ¢

 ≽^ʌⱷ҅_ᴥⱷ^ʌ≼  22:31, 24 July 2015 (UTC)

Please re-read my comment. As I thought I clearly stated, there are some citations that should go directly to arXiv rather than to journals (they are reliable, even though arXiv is their only publication) and {{

cite arxiv}} formats them better than the generic {{cite web}}. None of your hyperbolic attempts at justifying the removal of this template would be any different for {{cite web}}; both templates can be abused, but that's not a good reason to remove them. —David Eppstein (talk

) 00:14, 25 July 2015 (UTC)

As I've pointed out I think 3 times, we already have a template for precisely the situation you identify: {{

arXiv}} typo: {{cite arXiv}} template doesn't do anything special. It's just another identifier-specific template, which were all deprecated as a class, not because of exactly how they were coded, as you go on about above, but because they're redundant and their purpose was merged into the main citation templates. Same story.  — SMcCandlish ☺ ☏ ¢

 ≽^ʌⱷ҅_ᴥⱷ^ʌ≼  04:52, 27 July 2015 (UTC)

A further rationale, for at least deprecating this, is prior deprecation of all identifier-based (DOI, JSTOR, PMID, etc.) templates. See Template:Citation Style 1: "Identifier-based templates. ... All of these are deprecated." I'm sure I can dig up the actual deprecation discussion if someone thinks that's necessary. We're already handling arXiv identifiers in Template:Cite journal, making the arXiv template redundant, as well as use of a citation template approach that's already been consensus-rejected.  — SMcCandlish ☺ ☏ ¢ ≽^ʌⱷ҅_ᴥⱷ^ʌ≼  22:31, 24 July 2015 (UTC)

That's a completely unrelated issue. {{

cite arxiv}} is part of the cite series of templates, something to use within an article to provide a citation. Despite the name, {{cite doi}} is not. Instead, it is a way to tell a bot to set up the real citation elsewhere (usually using {{cite journal}}) and transclude it. That has multiple problems, among them making it difficult to detect vandalism (because the separate citation page is unlikely to be watchlisted) and making the citations hard to edit (especially when the |noedit= option has been used). Additionally, because {{cite doi}} is only used for citations that have dois, which in practice means journal articles, it's redundant with {{cite journal}}. The same goes for the other ones you list. —David Eppstein (talk

) 00:20, 25 July 2015 (UTC)

You're kinda making my point for me without realizing it: "Because {{Cite arXiv}} is only used for citations that have arXiv eprint IDs, which is already supported by both {{Cite journal}} and {{Cite web}}, it's redundant with them."  — SMcCandlish ☺ ☏ ¢ ≽^ʌⱷ҅_ᴥⱷ^ʌ≼  04:52, 27 July 2015 (UTC)

So basically, you object to {{

books

} 03:33, 25 July 2015 (UTC)

It's standard operating procedure to merge-and-delete redundant templates; that's around half of what

WP:TFD does (the other half being a mixture of subst-and-deleting extremely-low-use templates, deleting occasional totally inappropriate templates, and a few non-deletion/non-merge discussions). The ones you name, like {{Cite web

}}, etc., do particular special formatting and variances of that formatting for particular source types. This one does not (it does do one special linking thing that it shouldn't, which we'll get to). There is no difference at all between these two:

• {{cite arXiv |last=Sparling |first=George A. J. |eprint=gr-qc/0610068v1 |title=Spacetime is spinorial |date=2006 }} – Sparling, George A. J. (2006). "Spacetime is spinorial; new dimensions are timelike".
  arXiv:gr-qc/0610068v1
  .
• {{cite web |last=Sparling |first=George A. J. |arxiv=gr-qc/0610068v1 |title=Spacetime is spinorial |date=2006 }} – Sparling, George A. J. (2006). "Spacetime is spinorial; new dimensions are timelike".
  arXiv:gr-qc/0610068v1. {{cite web}}: Missing or empty |url= (help
  )

And you can't even do this, when the piece is finally published:

• {{cite arXiv |last=Sparling |first=George A. J. |eprint=gr-qc/0610068v1 |title=Spacetime is spinorial |date=July 2006 |journal=Nature Physics |volume=XIX |issue=2}} – Sparling, George A. J. (July 2006). "Spacetime is spinorial".
  arXiv:gr-qc/0610068v1. {{cite arXiv}}: Unknown parameter |issue= ignored (help); Unknown parameter |journal= ignored (help); Unknown parameter |volume= ignored (help
  )

because this template doesn't support |journal=, etc. But you can do the following, just leaving |journal=, etc., blank (or omitting them):

• {{cite journal |last=Sparling |first=George A. J. |eprint=gr-qc/0610068v1 |title=Spacetime is spinorial |date=2006 |journal= |volume= |issue=}} – Sparling, George A. J. (2006). "Spacetime is spinorial".
  arXiv:gr-qc/0610068v1. {{cite journal}}: Cite journal requires |journal= (help
  )

which insta-upgrades to the following when you add the journal info after publication:

• {{cite journal |last=Sparling |first=George A. J. |eprint=gr-qc/0610068v1 |title=Spacetime is spinorial |date=July 2015 |journal=Nature Physics |volume=XIX |issue=2}} – Sparling, George A. J. (July 2015). "Spacetime is spinorial". Nature Physics. XIX (2).
  arXiv:gr-qc/0610068v1
  .

(plus other parameters that could be added, like |doi=, that {{Cite arXiv}} can't handle).

So, this is even more redundant than I thought: There is never any need to even resort to {{

WP:AADD arguments. I'm looking for reasons to keep it. I thought "Surely, |class= does something important?" Nope.  — SMcCandlish ☺ ☏ ¢

 ≽^ʌⱷ҅_ᴥⱷ^ʌ≼  04:52, 27 July 2015 (UTC)