Wikipedia talk:Frequently misinterpreted sourcing policy

Endorse

This should be required reading. John from Idegon (talk) 05:21, 30 June 2018 (UTC)

Let me know if you see anything missing or daft.  — SMcCandlish ☏ ¢ 😼  05:39, 30 June 2018 (UTC)

Lets get more eyes on this.....added to {{Wikipedia referencing}} .--Moxy (talk) 13:00, 30 June 2018 (UTC)

Looks good, and I agree that this should get higher visibility. Also, perhaps a bit of section structuring is in order. — AfroThundr ^{(u · t · c)} 14:54, 30 June 2018 (UTC)

Done. Copyedited a bit, too.  — SMcCandlish ☏ ¢ 😼  00:30, 1 July 2018 (UTC)

I'm also open to another title ("Frequent misinterpretations of sourcing policy"? "Common ..."? Something completely different?). I hope peeps realize this was (aside from recent tweaking) a one-sitting "brain dump". I wrote it in about an hour. It took way longer to copyedit and format it as an essay than it did to write it as a WP:VPPOL post.

As the HTML comment in the lead says, I put this in the Wikipedia namespace not userspace, because people should feel free to improve, without majorly changing the intent, deleting entire line-items they disagree with, etc. If someone hates this or something in it, talk it out here, or write a counter-essay.

If the MEDRS issue I highlighted gets resolved, I would probably shift the tone of the piece (or someone else could) more toward just being a list of sourcing-policy misinterpretations and their corrections, without the "what we should do" material, which I might move to a userspace essay as an opinion piece. There's a bit of a tension between this being an "information page" kind of essay and a "this editor's recipe" one. This is why I've not added it to "See also" at WP:V, WP:NOR, etc. Others are free to do so if they endorse it.
 — SMcCandlish ☏ ¢ 😼  01:11, 1 July 2018 (UTC)

I may endorse it--Ozzie10aaaa (talk) 11:41, 2 July 2018 (UTC)

"Phenomena"

Does anyone want to take on "term" versus "phenomena" in minor research programmes of sub-disciplines? Because I notice a lot of coatracking and "claimed term is phenomena" in humanities and social sciences where there is no weighty scholarly agreement that "phenomena" exists. Fifelfoo (talk) 03:23, 1 July 2018 (UTC)

Do you mean things like "The phenomenon of [jargony word here] has seen significant coverage in [field here] research in recent years. For example, [some primary journal paper]. Similarly, [some random, unrelated news story that mentions something superficially similar]."? I do see that sort of thing pretty often, and it's usually one to all four of novel analysis, evaluation, interpretation and/or (especially) synthesis. There are a lot of similar cases, including a "trend", "syndrome", "movement", "category/type", etc. They vary from assertion of a statistical pattern (and the direction in which it is moving), to assertion of "a thing", a discrete agency in the world or bounded classification, without any sources actually coming to such a sweeping conclusion. This can be less cut-and-dry. We often have to come to a conclusion along these lines in evaluating a huge source pile, e.g. to determine what the scientific/historian/whatever real-world consensus actually is. It's challenging to write out the difference.  — SMcCandlish ☏ ¢ 😼  10:38, 1 July 2018 (UTC)

Jargon Jargon is a [field] term used in [sub-field] by [seven scholars] to describe [phenomena]. Field review of sub-field's theory. Research article on theory. Research article on theory. Research article on theory. See also: [phenomena].

• Is fine. Jargon probably has notability as a scholarly concept.

Jargon Jargon is [phenomena]. Research article on case study of phenomena. News story using jargon, News story using homonym of jargon, New story using what an editor considers jargon.

• Is bloody awful.

Consider, for a moment, if the scholarly consensus position regarding the Soviet famine of 1932-1933 was that it was not specifically targeted at the population of the Ukraine SSR.[*The opposite is in fact true, but the distinct nature of the jargon term makes it a useful counterfactual.] We should then not have an article [Ukraine Famine of 1932-1933]. Our article on the Holodomor should solely describe the scholarly minority who view that the Ukraine Famine was specific, and the political jargon Holodomor in central Europe. The history of the famine, the case study, should not be sandwiched and coatracked into the article Holodomor.

The primary problem is POV-war editors. The secondary problem is better faith editors mistaking a "local consensus" in scholarly communities for a field wide scholarly consensus. Fifelfoo (talk) 10:52, 1 July 2018 (UTC)

Editorials, columns, and blogs are categorically primary sources

From Wikipedia:Frequently misinterpreted sourcing policy#Editorials, columns, and blogs are categorically primary sources: "Even when editorializing material has been written by such ostensibly reputable parties, it is [a] viewpoint (and often outright advocacy), not fact."

Statements by authorities such as the World Health Organization (WHO) or American Cancer Society (ACS), which are usually reporting health or scientific facts, are "often outright advocacy"? I disagree.

As for authorities' stance on something and that "it is actually highly politicized primary material," this is not always true. Maybe not even mostly true. Even if it is political, it does not make the matter merely an opinion. Authorities are clear about the detrimental effects of female genital mutilation (FGM), for example. Unlike with (male) circumcision, there are no known health benefits regarding FGM; that is not simply an opinion. And we don't use in-text attribution for it in the lead of the Female genital mutilation article. We don't state that the WHO states this; it would be misleading anyway since it would imply that only the WHO has stated this when other authorities have stated the same. And as for "when usable at all (with due weight), opinion material must always be directly attributed inline and often should be directly quoted. We don't repeat opinions as facts in Wikipedia's own voice.", are we to always do this in the face of scientific consensus? Do we always need to state "scientific consensus is [so and so]"? At Race (human categorization), we currently state, "Modern scholarship regards race as a social construct, that is, a symbolic identity created to establish some cultural meaning." At the Global warming article, we currently state, "The Intergovernmental Panel on Climate Change (IPCC) Fifth Assessment Report concluded, 'It is extremely likely that human influence has been the dominant cause of the observed warming since the mid-20th century.' " But we also state, "Multiple lines of scientific evidence show that the climate system is warming." And "The largest human influence has been the emission of greenhouse gases such as carbon dioxide, methane, and nitrous oxide." We don't qualify the global warming stuff each time with "scientists state" wording. Climate change deniers (usually discussed in the context of global warming deniers) are in the significant minority. Flyer22 Reborn (talk) 06:48, 1 July 2018 (UTC)

Scientific or other scholarly consenses are normally promulgated by peer reviewed scholarly publications which explicitly state the current dominant position held by field experts. I've seen this in field review books, chapters, articles and papers across HASS and STEM. I have also seen scholarly associations self publish reviewed reports as collective authors do likewise. Such publications are not op eds. on the other hand I can clearly distinguish these from press releases and forewords. Fifelfoo (talk) 07:04, 1 July 2018 (UTC)

Looks like good old misuse of "primary", here. Whether a source is primary or not is not a function of who publishes it or how, but of whether the information is based on analysis, evaluation, interpretation, or synthesis from different sources. That paragraph needs to be zapped completely. Jo-Jo Eumerus (talk, contributions) 08:33, 1 July 2018 (UTC)

Jo-Jo Eumerus, do you mind being clearer on what paragraph you want zapped? You are referring to this essay? Something from the WP:MEDRS page? Above, I cited text from three different bullet points in the essay. Flyer22 Reborn (talk) 16:50, 1 July 2018 (UTC)

No, I don't mind. Zap the section header "Editorials, columns, and blogs are categorically primary sources" at a minimum as it is completely wrong. Any mention of "primary" in that section is suspect, and I'd probably remove the first and second mentions. Jo-Jo Eumerus (talk, contributions) 19:16, 1 July 2018 (UTC)

Except it isn't. We're already making the point that such a piece may contain some secondary information, but the piece as a whole is primary by its very nature, if it's in that class. At a rough estimate, I'd say about 85 to 90% of regular, serious editors entirely understand this, but the small percentage who don't are dead set against this idea. Your tone suggests you're in that statistical cluster, and that you should go write a counter-essay, not try to derail this one.  — SMcCandlish ☏ ¢ 😼  08:01, 2 July 2018 (UTC)

I beg to disagree, from WP:OR author's own thinking based on primary sources, generally at least one step removed from an event. It contains an author's analysis, evaluation, interpretation, or synthesis. I don't think that the distinction between "opinion" and "analysis" is always clear cut. More generally I think that the primary vs. secondary source distinction has nothing to do with either reliability, or fact vs. opinion distinctions. They exist and are important but IMO they have nothing to do with primary vs. secondary distinctions. Jo-Jo Eumerus (talk, contributions) 11:13, 2 July 2018 (UTC)

A recommendation/position/stance is "the author's own thinking" (and "author" includes organizational authors, not just individuals). Anyway, rather than argue circularly about this, I'll try a revision later that gets to your and WhatamIdoing's criticisms of this segment. The current text really was a first-draft braindump, and is certainly not perfect. I think the fact that org. pos. statements contain and can be used for secondary material can be clearly separated from the fact that the recommendations/decisions they promulgate based on them are primary. This also holds for op-eds, which are often of the same bracketing nature: An intro summarizes a primary stance; secondary material is presented that's possibly reliable, though may be highly selective and interspersed with more primary editorializing; and a primary conclusion "lays down the law" according to the author. Many investigative journalism pieces also follow the same pattern, which I think the essay already makes clear. The central problem here is the "this is a secondary source" thinking. Particular content in it may be, but the framework and intent of such a piece is primary. This also means my wording about such sources being categorically primary is making the same error, a faux pas / brainfart I should have detected.  — SMcCandlish ☏ ¢ 😼  19:29, 2 July 2018 (UTC)

A source can also be 'primary' and 'secondary' at the same time depending on how it is used. For example, an NPR story on rising diabetes rates is a secondary source. The same source is a primary source in a study on media representations of public health issues. AugusteBlanqui (talk) 09:01, 1 July 2018 (UTC)

and despite being secondary and NPR having a fact checking system be utterly unusable for a claim in medical demography as the author / review system is not appropriately expert for an area of scholarly knowledge. Fifelfoo (talk) 09:07, 1 July 2018 (UTC)

Yep. Completely aide from the "that damned line in MEDRS" matter, this is a key concern that's almost never articulated clearly on Wikipedia. I'm glad Fifelfoo shoehorned it into the essay. I would've had way more trouble figuring out how to frame that.  — SMcCandlish ☏ ¢ 😼  11:34, 1 July 2018 (UTC)

On Flyer22's points, if there really is a scientific consensus on something, we'll have better sources (from a WP viewpoint, which is very different from med journal one) than organizational position statements, because secondary sources with no connection any agencies and association with funding to lose, etc., will have said so. If we have one of those, then it's also fine to cite high-quality primary sources (FDA, WHO, BMA, etc.) in addition, especially if they say something we want to quote. Part of what I made a point of including in here is that the mid-2000s idea that adding key primary sources after a secondary one is "over-citation" is completely dead in the water, especially in med/sci/tech articles. In such a case the secondary source is doing the synthesis, and it's the fact that all these orgs conclude the same way that establishes the consensus in the first place. But one org position by itself is just that org's position, and it's never going to be clear exactly what influenced their decisionmaking process. Moving on, the Race (human categorization) article badly needs work, but last I looked at it the statements about scientific consensus were not a determination people made on the talk page and then cited an organization statement or two on it; the idea that this general consensus that race is a social construct is coming from secondary sources. They certainly exist for it; I was reading them as long ago as the 1980s. I want to repeat: the fact that org. position statements are primary for the positions doesn't mean the sources are bad, only that they can't be used for any WP:AEIS. The same work can be used for AEIS when we're only pulling science claims from them – what is versus what should or must be. That little bit right there is it in a nutshell. The is–ought fallacy is at work here. PS: I don't think anyone's arguing for "scientists state ..." weasel wording. I'm not, and just added a criticism of this to the essay, actually. PPS: I haven't looked into the current state of the GW article; I steer clear of the whole topic area. But don't get me started on the ACS.  — SMcCandlish ☏ ¢ 😼  11:27, 1 July 2018 (UTC)

Without addressing all of your reply, I meant "scientists state" in a general way...as in using in-text attribution from an authoritative body, using the aforementioned "modern scholarship regards" wording, the aforementioned "multiple lines of scientific evidence show" wording, or simply "scientists state." Flyer22 Reborn (talk) 16:50, 1 July 2018 (UTC)

Sure. In the aggregate there is no (OR or otherwise) problem with noting that the 10 leading regulatory or professional academic organisations in a field all agree on X, one disagrees, and two are noncommittal or silent on the matter, and that for WP purposes this is a scientific or academic consensus; that's a big part of what we do.
The issue in a nutshell: That kind of Wikipedian aggregate RS assessment is completely different from declaring in Wikipedia's voice with a bare citation that "the easy availability of foo is a public health risk, and the costs outweigh the benefits, especially given ease of uptake, addictive potential, and marketing aimed at young users", just citing an AMA position statement to this effect, or declaring as fact that "foo, despite non-zero health risks and an appeal to youths, is overall a public health risk reduction because it supplants use of the more dangerous bazquux at a faster rate that deaths or illness caused directly by foo", just because the BMA's position statement said so, and just citing them for it. These are subjective (though reasoned and back-end sourced) claims that have to be attributed.

Yet many WPMED authors don't want to have to attribute them, just silently cite them, and this is emphatically not cool. It's not likely to happen in a dispute-prone topic like e-cigs; the concern is primarily about articles without a firestorm surrounding them, were it's mostly only WPMED editors working on the material, and convinced that subjective position statements are "ideal" and secondary.
 — SMcCandlish ☏ ¢ 😼  19:03, 1 July 2018 (UTC)

I'm questioning the need to state "modern scholarship regards," "multiple lines of scientific evidence show," "scientists state," or the like for consensus matters. There are some consensus cases where we should use in-text attribution or similar, but there are cases where we shouldn't either (meaning that stating the matter in Wikipedia's voice is fine). Flyer22 Reborn (talk) 05:37, 4 July 2018 (UTC)

To a first approximation, the claim about editorial content, columns, and blogs is true. Of course, it is equally true that, to a first approximation, the entire universe is made of Hydrogen, and that humans (therefore) don't exist. There is a lot more nuance than a simple yes/no here.

I think it's also interesting that this claim that all blogs are primary follows the section, ==Wikipedia over-focuses on publisher instead of author reputability==. So what the page says feels self-contradictory. On the one hand, we say not to pay so much attention to the fact that Alice Expert's meta-analysis was published in The Journal of Big Important Reviews, but we then say that if she publishes exactly the same words on her blog, it's not a secondary source. That's not how it works: a meta-analysis may be good or bad, it may be independent or not, it may be self-published or not, it may be current or outdated, but it is always a secondary source.

There also seems to be some over-confidence in the existence of high-quality peer-reviewed sources for some subjects. WP:FRINGE subjects and brand-new products tend not to be described in significant detail, or even at all, in the academic literature. The Science-Based Medicine blog or the Quackwatch website are perfectly good sources for describing some subjects, especially when nothing better exists. The fact that SBM uses blogging software doesn't mean that they aren't engaging in the analysis, evaluation, interpretation, and synthesis that are the hallmarks of true secondary sources. WhatamIdoing (talk) 00:24, 2 July 2018 (UTC)

As I said to JoJo, above, we're already making the point that a source that is primary in nature and intent may contain some secondary information. Perhaps this can be clarified in a way that eases these objections to an overliteral wording. This page is already getting longer and more detail-mired than ever intended. It's not meant to cover every imaginable scenario in pendantic TL;DR style, but hammer home some basics that too many people are getting wrong. They can internalize the exceptions to the general rule after they learn the general rule. This what humans do in every aspect of life anyway. Blogs: Yes, that will need disambiguation. Obviously, the intent is blogs in the "editorial self-publishing on the Web" sense, not in the "website, regardless of content or editorship, that happened to be managed and generated with a blogging software package" sense. Your "always" point may be true under the definitions of "secondary" in particular academic and professional fields, but it is not on wikipedia (and there's even a line-item in the page about this cross-disciplinary terminological confusion). In WP terms, all self-published sources are primary; it doesn't matter what they contain. This even includes ones that are not technically self-published by amount to it, including publication by a "you pay us to publish your book and get it onto Amazon" outfit, and user-generated content like most wikis; many of these contain meta-analysis. In this treatment of secondary (which Wikipedia isn't alone in using), it's not enough that it perform WP:AEIS on previously published material, it must also be subject to editorial review beyond they author. I think that somewhere we already have an essay on the difference between P/S/T source definitions at WP and in various different fields, but I don't remember where it is. Maybe our real articles actually are good enough for this, though the last time I looked them over in any detail (quite a while back) they only covered a few disciplines and had some PoV problems. Over-confidence in existence of peer-reviewed sources about fringe? Not sure what that's in reference to.  — SMcCandlish ☏ ¢ 😼  08:53, 2 July 2018 (UTC)

"In WP terms, all self-published sources are primary; it doesn't matter what they contain."

This is not true. As in, this belief is literally not found in any policy whatsoever. Go ahead and look for it, and feel free to provide direct quotations.

Confusing "sources we worry about" with "primary" has been a serious problem. You may have noticed that, for example, that WP:RSPRIMARY and WP:RSSELF are separate sections; that is because they're different things. The controlling policy for self-published sources is WP:V; the controlling policy for primary sources is NOR.

Trying to stop this persistent error is the reason that Wikipedia:Identifying and using independent sources, Wikipedia:Identifying and using primary sources, and Wikipedia:Identifying and using self-published works exist as deliberately separate pages. This isn't the "is–ought" problem; it's the "is–is like" problem. We normally treat self-published sources the same way that we treat primary sources, but that doesn't mean that self-published sources actually are primary sources. WhatamIdoing (talk) 15:44, 2 July 2018 (UTC)

agree(preferable not to use self-published source...IMO)--Ozzie10aaaa (talk) 17:27, 2 July 2018 (UTC)

Right: you shouldn't normally use a self-published meta-analysis, but you shouldn't use it because it's self-published, and we don't normally want self-published sources – not because you believe that historiographical classification depends upon how many people are involved in the publication process. WhatamIdoing (talk) 17:43, 2 July 2018 (UTC)

Background of the dispute (from SMcCandlish's perspective)

Just to address this as a general matter: We've had rounds and rounds and rounds of discussion of this at WT:MEDRS. It basically boils down to a belief (or pretend-belief) that medical professional associations and regulatory bodies when presenting a position statement on a medical issue are apolitical and robotic. Most of the regulars at that page engage in this fantasy (or put on a good show of it). Everyone else seems to clearly understand they are highly political (on multiple axes) packs of humans, full of biases and agendas just like everyone else. While they may be skilled at the research, and are not going to falsify it, they are still drawing subjective conclusions because of the research they select, and what weight they give to different facts and arguments, goals and outcomes. Even if 0% of the science is ever bad, plenty of the judgement is questionable or at least one-sided.

This is why, on e-cigarettes for example, you see very sharp splits between US and EU regulatory stances, and between different professional medical bodies in different subfields. If it were all simply objective and factual science, they would all come to, or near to, the same conclusions, but are often diametrically opposite each other.

Neither of the following is an issue: A) Whether the research they do and cite is reliable for the purely factual content; it surely is. If it's written properly, those aspects of it may even arguably be secondary, and some of this work is as thorough as literature reviews (not too sure about systematic reviews, though of course they cite those when available). B) Whether the position-taking aspects of it (which are the central point of these pieces – making recommendations, and sometimes more than recommendations, depending on their legal authority level) can be cited, with attribution; they certainly can, because the AMA, BMA and EMA, the FDA, MHRA and EFSA, etc. have a great deal of real-world weight, so it's WP:DUE to include them. It's not permissible to treat their stance/position – i.e. organizational opinion – as fact (other than tautological WP:ABOUTSELF fact that it is their stance). That material has to be attributed as subjective, organizational position, not advanced as truth in WP's own voice.

It just amazes me that so many people from the MEDRS/WPMED crowd either can't understand this or pretend that they can't. No one else on Wikipedia ever seems to have any such issue. If it's a pretense, I'd love to understand what the motive is. Professional pride and defensiveness? A more narrow version of same (field-specific, nationalistic)? Editorial territorialism about "their" guideline (WP:WPMED certainly do feel proprietary, in the aggregate, toward that site-wide guideline)? Just expediency (because not being able to use those sources in every single way they want to means doing more tedious source digging)? Just a difference in understanding of what "reliable sources" means in an encyclopedic versus medical-journal context? Hell if I know. I have come to the conclusion that it's going to take a WP:VPPOL RfC to resolve it, because I've been trying to periodically for over two years and it's the same stonewalling by the same editors. The fix is so very simple: Just move organizational position statements out of the "ideal sources" sentence and address them separately as secondary for factual claims, if their own sources are clear, and high-quality primary for the stance-taking. It drives me nuts, because this is otherwise our best and most important content-sourcing guideline, ever (not counting BLP, since it's a policy. >;-) PS: I bear no ill will toward any of WPMED/MEDRS shepherds; I'm just really frustrated with their blind spot (or stubbornness) on this.

 — SMcCandlish ☏ ¢ 😼  11:27, 1 July 2018 (UTC)

• "If it were all simply objective and factual science, they would all come to, or near to, the same conclusions, but are often diametrically opposite each other." Feyerabend. Fifelfoo (talk) 14:23, 1 July 2018 (UTC)

  Paul Karl Feyerabend?  — SMcCandlish ☏ ¢ 😼  18:50, 1 July 2018 (UTC)

I remember those conversations from about two years ago; you were claiming things like position statements were "almost invariably highly politicized primary source". I eventually gave up.

So here's the most recent position statement listed at PubMed whose full text is free online: Pelliccia, Antonio; Caselli, Stefano; Sharma, Sanjay; Basso, Cristina; Bax, Jeroen J.; Corrado, Domenico; D'Andrea, Antonello; D'Ascenzi, Flavio; Di Paolo, Fernando M. (2018-06-01). "European Association of Preventive Cardiology (EAPC) and European Association of Cardiovascular Imaging (EACVI) joint position statement: recommendations for the indication and interpretation of cardiovascular imaging in the evaluation of the athlete's heart" (PDF). European Heart Journal. 39 (21): 1949–1969. doi:10.1093/eurheartj/ehx532. ISSN 1522-9645. PMID 29029207.

I have spent less than ten seconds glancing at it, but it looks pretty typical to me. Can you tell me what about this source seems like it's the kind of politically motivated, opinion-stuffed primary source that you keep claiming is common? WhatamIdoing (talk) 00:32, 2 July 2018 (UTC)

Not right this instant; it's pushing 1:45 am my time. This one may be rather esoteric for me, and I'm not familiar with the bodies in question and what they're involved in as organizations, interfacing with other organizations. As an American with Wasinghton DC lobbing experience, it's easier for me to detect biases in FDA, AMA, etc. stuff.

But this kind of highlights the two key points, which you're swimming around instead of wrestling with: it's the advancing of a position, which is by its very nature subjective to the organization, not just a factual analysis, that makes it primary. I.e., it has an ought and well as an is, as part of its very fabric. That's primary. The parts of it that are blocks of opinion-free factual analysis without value judgements, recommendations, etc., are often secondary, but they can only be mined out, with care. It takes real work to separate those nuggets from the "we've made up our mind" rubble, and people may not trust it anyway because it came from a piece the purpose of which is, basically, to tell you want to do, not to describe an aspect of the world.

The second issue is that detection of politicization, fiduciary incentive, etc., is something that requires a considerable amount of either prior knowledge or in-depth research (much better spent on writing an article about the organization). These get teased out pretty well in an area that's high-profile (on- and off-site), like GMOs, vaccination safety, e-cigs, etc., at least with regard to organizations that are watchdogged closely by various press people, other organizations, etc. There's not just editorial scrutiny, but RS scrutiny that editors here can rely on. I don't know if that exists around cardiovascular medicine associations in the EU; I suspect there's little or any, and that they'll largely be cited in obscure articles where no one but the the "med association press releases are ideal secondary sources" crowd are the only ones paying much attention.
 — SMcCandlish ☏ ¢ 😼  08:55, 2 July 2018 (UTC)

But isn't there a danger here that you're requiring editors to do original research? If you start talking about things that "can only be mined out, with care", then that's too much work put on subjective Wikipedians. If you want to be a journalist exposing professional bodies, go for it. If you want to create a community-written encyclopaedia, then I think you have to accept compromises, including that medical professional bodies are, by and large, summarising the known science in a sensible way. Bondegezou (talk) 11:55, 2 July 2018 (UTC)

Evaluation of source reliability (including what is/isn't really secondary) and author/publisher reputability isn't OR, it's part of our "job". It's what we do every day on article talk pages. Sometimes its very easy, sometimes it's not. The obvious way to avoid wasting a lot of time on it (wasted if we're not also going to use that work to improve articles on the authors/publishers themselves) is to treat position statements as primary in nature by default except where particular material in them can be demonstrated to be secondary, same as we do with news reporting, op-eds, investigative journalism, and other stuff with this primary/secondary commingling problem. (Even research papers in journals, which are mostly primary: they frequently can be used as secondary sources for the background reviews of prior research that they typically include, bracketed between primary description of the research goals and methods, and primary conclusions being advanced on the basis of the new research.) It really is accepting compromises, as you put it, but not blindly.  — SMcCandlish ☏ ¢ 😼  19:57, 2 July 2018 (UTC)

I'm not in a rush. If they're advancing a position with the small number of "ought" statements present in that paper (I offer this as a time-saving example: "Elite athletes have normal aortic root diameter measured at the sinus of Valsalva and aortic valve annulus, with <1% of athletes showing increased dimensions. Clinicians evaluating athletes should therefore know that marked aortic root dilatation represents a pathological process and not a physiological adaptation to exercise"), then I'm sure that you'll be able to spot it in due course.

Of course, after you read through 20 pages of technical information about how the heart adapts to exercise and which medical imaging tests are most useful in which circumstances, you might also discover that an actual position statement (which differs from a policy statement) is not a lobbying document. In that case, I hope that you will reconsider your opposition to MEDRS's approach. WhatamIdoing (talk) 15:59, 2 July 2018 (UTC)

See reply to similar comments by Jo-Jo, in prior section (starting at 'A recommendation/position/stance is "the author's own thinking" ...'): I think this is resolveable, and will make a major revision in that direction. The gist is that I'm not really opposed to WPMED/MEDRS's approach, which is rarely directly abusive of sources on the part of med-specialized editors [in my personal opinion; several ARBCOM cases indicate not everyone agrees]. Rather, the problem is the wording suggesting that organizational position statements are "ideal" sources categorically. This signals to all editors – including thousands who are not primarily focused on medical research – that a) every single thing in an org. pos. statement is reliable and secondary (even that a bald-faced ought regulatory opinion of the BMA or the FDA trumps purely fact-focused and entirely secondary reliable sources), and b) that this same pseudo-principle applies to all topics (if it's good for medicine, it must also be good for politics, religion, technology, etc.). It's a massive argument to authority, without that being the original intent. The essay just isn't getting this across very well yet; it's not be absorbed because of the inclarity of the material leading up to it, and I'll try to fix that later today [got busy with other stuff]. The med editors have been super-resistant about this, however, no matter how it has been phrased or in what venue. It seems to be because they're only considering their personal and wikiproject intent and rejecting any consideration of interpretation, inference, and effect just because they're pissy about being criticized at all. "How dare you question our assessment of medical sources?!" [sigh] It's really a critique of over-generalized, un-nuanced wording about the assessment, not the assessment itself.  — SMcCandlish ☏ ¢ 😼  19:57, 2 July 2018 (UTC); updated: 07:00, 5 July 2018 (UTC)

The entirety of Wikipedia is based on argument to authority. That is, on summarising what reliable sources say. If you want investigative journalism or research, then that's great, but that's not what Wikipedia is. Or there's Citizendium! But encyclopaedias, including Wikipedia, maybe especially Wikipedia, are not written by experts always able to critique every source in detail. I don't believe that's achievable.

Yes, we should try to have nuance and recognise when any source requires to be considered in context. No source is 100% reliable. I agree with that. Bondegezou (talk) 22:57, 2 July 2018 (UTC)

Appropriately citing sources is not argument to authority, in the fallacy sense, which is clearly what I meant: accepting as prima facie true an opinion simply because of who expressed it.  — SMcCandlish ☏ ¢ 😼  07:00, 5 July 2018 (UTC)

I don't think that the problem is people being upset about being criticized. I think that the fundamental problem is that you have failed to convince me (or apparently anyone else) that you actually know what MEDRS means when it talks about a position statement. It does not mean friendly little press releases, of the sort that a charity might send out in the hope of influencing legislation or drumming up donations. It means detailed, peer-reviewed papers that tell medical professionals how to translate scientific facts into daily medical practice. I see no hope for resolving this situation until you actually sit down and read a few proper position statements. I've already given you an example of one, and I can provide more upon request. WhatamIdoing (talk) 04:32, 3 July 2018 (UTC)

@WhatamIdoing: position statements can be highly politicized. For example, in 2010 the American Academy of Pediatrics published one distinguishing between what it saw as harmful and harmless forms of FGM: "the ritual nick suggested by some pediatricians is not physically harmful". [1] There was an outcry and it was retracted. Its position now (and since 1998) is that it opposes all forms of FGM. Nothing about the science had changed to prompt the 2010 statement. SarahSV ^(talk) 05:36, 3 July 2018 (UTC)

SarahSV, the key words are at the top of that document, under the date and above the title: "Policy Statement". A policy statement is not a position statement. WhatamIdoing (talk) 15:13, 3 July 2018 (UTC)

No source is perfect. There are position statements that have problems. But then there are also published systematic reviews that have problems. I don't see any reason to pick on position statements as being hugely worse than the other sources recommended by WP:MEDRS. And I don't think we should throw the baby out with the bath water. Bondegezou (talk) 10:58, 4 July 2018 (UTC)

A distinction between position and policy statements isn't well-maintained in the real world. Nor – and this is important here – is the supposed virtual synonymy of "position statement", "guideline", and "medical consensus", at least as to their reliability level. It takes no time at all to find in-field material not only dispelling their conflation but laying out specific definitions that strongly support what I've been saying. To quote the most relevant definition of three from the first example I ran across: "A position paper is a document that presents an opinion about an issue, typically that of the author or another specified entity. ... Position papers can be used by organizations to communicate their specific beliefs and recommendations. Position papers promote discussion on emerging topics without the experimentation and indicate original research needed to develop a guideline or present in an academic paper." [2] I.e., they damned well are editorials. Just one example. Then you get "public policy position statements", too [3]. Not just politicized but explicitly so ("public policy" = "politics"). The whole problem here, as I've been saying all along, is treating "position statements" as a mystically special class, when – like all other sources – they need to be evaluated as individual sources. A position statement from an organization that takes a hardline internal stance on the difference, and works to keep organizational opinion out of its position statements and stick to a systematic-review-style approach is going to be more reliable and more secondary that one from an organization that draws no such distinction, and injects organizational viewpoint into them, while reserving its just-the-facts approach for consensus statements. And so on. These organizations and the people behind them are not automatons, and they use language differently. The desire of some 'pedians to stick invariable definitions on the terms doesn't match reality.  — SMcCandlish ☏ ¢ 😼  07:17, 5 July 2018 (UTC)

Sure, there are groups (especially minor advocacy groups) that use the name for anything they want. But that's not a position statement in the sense that MEDRS cares about. There are groups that call their POVs "guidelines", too, even though that term has a legally regulated meaning (and only sources that are guidelines in that legally binding you-could-lose-your-medical-license-if-you-don't-do-this sense are actually supported by MEDRS).

I've given you a link to a recent position statement. That is the kind of position statement that MEDRS favors. Please go open the PDF, read through it, and tell us if you find content that sounds like it's just someone's opinion. WhatamIdoing (talk) 18:41, 5 July 2018 (UTC)

I wasn't pointing to "minor advocacy groups", but to medical ones, and there are many more. Those were just a couple of examples. I appreciate that the MEDRS crowd think the distinction is very clear, but it isn't in the real world, and it wouldn't be to our editors even if it were in the real world. I want to re-stress that I don't think MEDRS is wrong-headed in what it wants – I repeatedly say it's our most important sourcing guideline – but rather that its wording is muddled and there's resistance to doing anything about it out of some combination of the experts half of Dunning–Kruger effect and simple territorial stubbornness. When the assumption that everyone thinks about it the way you do commingles with refusal to absorb the concerns others are raising, the result is a counter-productive stonewall. [That's the rhetorical you, I don't mean "WhatamIdoing in particular".] I would be resolvable to moving "position statements" out of the "ideal sources" sentence and having a separate sentence with a footnote, like "Factual, research-based positions statements by major medical associations and medical regulatory bodies are also usually high-quality sources.{{efn|Position statements in this sense cite their own sources and are similar to literature reviews. These are distinct from policy statements, and from opinion releases that happen to be labeled as position statements. Not all organizations use exactly the same terminology; see [some article or something here].}} S'not difficult. But getting anyone from that crowed to even consider a change to the "ideal" sentence is like trying to argue with a boulder.  — SMcCandlish ☏ ¢ 😼  22:42, 5 July 2018 (UTC)

Making this into policy

For the love of all that is holy, if you ever try to make this into policy focus on existing policy pages, shorten them and accentuate the correct and expected behavior. Do not add more "supplements" that will remain unseen and unread. Bright☀ 09:45, 6 July 2018 (UTC)

Lay summary

What is the purpose of |lay-url= as used at pages like Macroketone? Is it related to the part of this essay which says

Our policies do not yet codify this, but current community best practice on technical topics (scientific, medical, IT, legal, etc.) is to cite, when available, both a high-quality secondary source and the original primary-research source, for our claim. Non-experts will understand the secondary source, and Wikipedia itself wants it as the main source. But technical readers in that field or a related one aren't interested in that; they want to read the gory details in the original journal paper. Similarly, in a computer science topic, we definitely want secondary sources for, say, the history of an Internet protocol; but IT people want to read the actual specs, too.

Emir of Wikipedia (talk) 20:23, 27 July 2018 (UTC) (please  mention me on reply; thanks!)

@Emir of Wikipedia: The lay parameters are documented here. (I think I'd like to see that functionality removed, simply because anyone interested in citing a lay-source should really be interested in citing both documents, for which two templates can be used.) --Izno (talk) 20:55, 27 July 2018 (UTC)

Some technical or scholarly sources provide their own lay summaries – think "impenetrable report, plus separate executive summary" or "jargon-filled medical guideline, plus patient information page". These are perfectly reasonable uses of the lay source parameters. I've more often seen it used to cite a major report plus an independent news article that explained it.

As for the advice of just citing them separately... well, I don't think that's a practical recommendation for medical content these days. One of the self-appointed MEDRS "enforcers" will simply remove the second citation without a second thought. WhatamIdoing (talk) 20:12, 28 July 2018 (UTC)

Another option might be converting these to be used for linking to an abstract-only page. One thing that chaps my hide is how few academic citations actually go to usable material for the average editor; the cited details are often in a page that cannot be accessed without jumping though hoops. While wave have ways to specify those hoops they're a bit of a hassle, so no one does it. It would be easier if |url= were reserved for cases that are either full-text, or sufficient text to source the claims, with abstracts that are neither being confined to a different parameter.  — SMcCandlish ☏ ¢ 😼  01:44, 1 August 2018 (UTC)