Jump to content

Wikipedia:Reliable sources/Noticeboard/Archive 479

Add links
From Wikipedia, the free encyclopedia
Archive 475←Archive 477Archive 478Archive 479Archive 480Archive 481→Archive 485

HonestReporting

Greetings everyone. Is HonestReporting.com a reliable source to use on Wikipedia? I've seen an user citing this source to prove that Hasan Piker promotes terrorism... I have also used this source myself to support a specific sentence in 2025 killing of Israeli Embassy workers in Washington, D.C..

According to all information i was able to find, this site belongs to a pro-Israel advocatory group named HonestReporting that monitors media to combat "anti-Israeli prejudice". It was involved in a lot of controversies, two times for accusing people of bad things without evidence. What do you guys think about it? WhoIsCentreLeft (talk) 16:49, 28 May 2025 (UTC)

It's pretty blatantly an advocacy group, so no usage without attribution/in Wikivoice. The Kip (contribs) 16:58, 28 May 2025 (UTC)
Given the controversies (which amount in some cases to clear evidence of outright falsehood, putting journalists lives at risk), there are absolutely no circumstances where we should consider HonestReporting a legitimate source to support assertions that anyone 'promotes terrorism'. And I'd be wary of citing them for anything that couldn't be independently verified - which would make citing them redundant anyway. Partisan pressure groups can sometimes be cited as sources, but not when they are on record for making shit up. AndyTheGrump (talk) 17:01, 28 May 2025 (UTC)
Regardless of the reliability or otherwise of this source, we should never be describing a living person with a phrase like “supports terrorism” in our own voice, unless the overwhelming preponderance of reliable sources say so. We should always attribute it, especially if it comes from an advocacy organisation. The same is of course true for phrases likes “is an Islamophobic extremist” (as with the SPLC above) or “is an antisemite” (as with the ADL). There seems to be a trend towards declaring sources unreliable because they make controversial judgements about such subjective labels, rather than simply applying our policies of not repeating such claims in our voice. So hopefully editors can chip in here as to whether HonestReporting has a reputation for reliability about facts and not whether its labels are accurate. BobFromBrockley (talk) 08:18, 29 May 2025 (UTC)
That site sounds like it was founded by one of those users with "truth" in their id. Who are these people who are so gullible as to believe a name? Phil Bridger (talk) 08:33, 29 May 2025 (UTC)
All of these pro-Israel advocacy "watchdog" groups like NGO Monitor, CAMERA, etc are unreliable. They are not interested in providing sober analysis, but deflecting criticism of Israel and attacking those that criticise it. There are enough sources covering this topic area that we can go without using them. Hemiauchenia (talk) 11:33, 29 May 2025 (UTC)
It may well be that these three advocacy groups are unreliable (I believe we have strong consensus that NGO Monitor is, can't recall if CAMERA has been discussed), but they're not unreliable because they're pro-Israel, any more than pro-Palestinian advocacy groups are unreliable because they're pro-Palestinian. It's far better for us to make a robust judgement based on our policies than a partisan decision based on our positions in relation to Israel/Palestine.
However, one indicator of unreliability is being amplified by and amplifying material by other unreliable sources, so if HonestReporting shares material by other advocacy groups we've already found to be unreliable, that would be a relevant data point. Is that the case here? BobFromBrockley (talk) 15:51, 29 May 2025 (UTC)
Absolutely, I wasn't saying that all "pro-Israel" outlets (which I suppose would include newspapers like the Jerusalem Post, etc) are unreliable necessarily, but there are certainly quite a number of pro-Israel "watchdog" groups that I would consider unreliable. There are obviously pro-Palestinian sources like Electronic Intifada that are also unreliable, and avocacy groups aren't necessarily unreliable, see for example the SPLC. Hemiauchenia (talk) 16:06, 29 May 2025 (UTC)
thanks Hemiauchenia, I agree completely. BobFromBrockley (talk) 09:19, 30 May 2025 (UTC)
LOL, any source with that name rings alarm bells as to how honest they are. Slatersteven (talk) 11:37, 29 May 2025 (UTC)
Having looked, I can't see any WP:USEBYOTHERS evidence of HonestReporting being taken seriously by RSs. On the other hand, the only reliable source coverage I've seen are these three damning articles from Reuters: https://www.reuters.com/company/honest-reporting-com-inc/ BobFromBrockley (talk) 15:55, 29 May 2025 (UTC)
blatent advocacy group. bobFromBrockley link seems to indicate they are mostly used to spread misinfo against perceived ideological enemies. even dismissing lack of reliability, they probably aren't due. better to look for other sourcing. Bluethricecreamman (talk) 16:30, 29 May 2025 (UTC)
Additional clarification, but should not a proper RfC be prepared or could consensus be reached here in this discussion? Lf8u2 (talk) 01:55, 30 May 2025 (UTC)

Deutsche Presse-Agentur

I noticed that there have been quite extensive debates about reliability of high-profile state news agencies such as Associated Press (US), Agence France-Presse, Anadolu Agency (Turkey), TASS (Russia) etc. but basically none about the Deutsche Presse-Agentur (German News Agency), or DPA. Even though DPA is cited in a lot in English Wikipedia articles.

Here is a short list of examples where Deutsche Presse-Agentur is currently cited:

Again, this is just a sample, and as you can see, it is really large and there are all sorts of topics. I tried to point out rather high-profile topics, but as you can see DPA is used in articles about politics, sports, entertainment, sensitive and non-sensitive topics, within Germany and international, you name it.

Now, at first glance it seems to be a no-brainer, and there is no reason to question its reliability. And in general, news agencies from Western Europe seemingly have a solid reputation. However, this section of the article about DPA makes me question this, as apparently there have been accusations of agenda-pushing, potential non-transparency and not-so-good fact checking. That makes me question how freely can we cite articles written by DPA, especially when it comes to sensitive topics.

Nonetheless, I'd like to know what is the opinion from the more experienced Wikipedia editors, and whether we could consider DPA reliable for:

  • politics within Germany, Europe and worldwide
  • wars and terrorist attacks happening within Germany, Europe, Africa and Middle East
  • other high-profile incidents within Germany and internationally
  • sports and entertainment, whether it's Germany-exclusive or of international merits

Kacza195 (talk) 19:34, 24 May 2025 (UTC)

The DPA reliability is similar to Reuters, Associated Press or Agence France-Presse and all of them are roughly as reliable as reputable newspaper or journalistic publications. So at first glance I see no issue here and I'm not aware of any specific "agenda-pushing, potential non-transparency and not-so-good fact checking"-issues. Rather than making a vague insinuation here, you'd need to be more concrete for this to be assessed. Wo is accusing DPA over what?--Kmhkmh (talk) 12:39, 25 May 2025 (UTC)
I left a link to the Criticism section of the acticle about DPA. I'll leave the link again here. Kacza195 (talk) 13:31, 25 May 2025 (UTC)
Now, at first glance it seems to be a no-brainer, and there is no reason to question its reliability. And in general, news agencies from Western Europe seemingly have a solid reputation. Well there you have it, it is indeed a no-brainer. These agencies do not "seemingly" have a solid reputation, they just do. This one is as reputable as the others, meaning that they sometimes mess up, but possibly less often than almost any other media outlet. Do you have a specific concern to address? It is not because discussions have taken place about other agencies that we have to discuss all the other ones. Choucas0 🐦‍⬛⋅💬⋅📋 13:56, 25 May 2025 (UTC)
I already provided the link to the Criticism section of the article, but I can re-print some snippents here:
  • Regarding agenda-pushing: "There have been ongoing accusations of the dpa abusing its power within the marketplace, using its position for agenda-setting and to manipulate the general public. These sentiments have led to isolated recommendations for restrictions to be placed on the agency's power. Such concerns have been reported as early as 1970 with articles published in ZEIT and SPIEGEL generating discussion on the agency's close affiliation to the government and of the dpa colouring its reporting"
  • Lack of transparency: "Journalist Stefan Zickler included the company as part of his criticism of the structure of the German Press in a publication in which he challenged the belief that total privatisation of the agency by its around 170 shareholders prevent manipulation of its content. As a company owned by around two hundred shareholders who are responsible for ensuring its independence, the total privatization can be seen as a drawback, as it prevents any state and majoritarian involvement. Furthermore, the ownership also places great power in the hands of the Editor-in-Chief, who can shape the media landscape by controlling how information is disseminated." Basically there is no way to check if the ownership/editor-in-chief can be held accountable in case DPA consistently prints misleading or outright biased information.
  • Not-so-good fact-checking: "In subsequent years, the agency was forced to apologise for inaccurate reports regarding the protests against the G8 summit in Heiligendamm (2007) as well as the news of a scandal involving the then Federal Economics Minister Karl-Theodor zu Guttenberg (2009). The role of the dpa in the Bluewater affair in 2009 led to new internal regulations regarding the sources of the disseminated news. The Otto Brenner Foundation conducted a large-scale study in March 2010, led by Hans-Jürgen Arlt and Wolfgang Storz. This study, named "Business Journalism during Crisis - The Mass Media's Handling of Financial Market Policy", evaluated the working procedures of the dpa from spring 1999 to autumn 2009. The ultimate conclusion of the study was that German business journalism failed to provide proficient and informative coverage of the financial market and its related policies prior to the onset of the global financial market crisis. The evaluation of the dpa's contribution to financial market policy journalism was described as being "highly deficient" and that it gave a sense of confusion rather than offering orientation."
Kacza195 (talk) 15:32, 25 May 2025 (UTC)
I've read the article and some of the linked sources. I don't get the overall impression that dpa is any less reliable than most of the other major privately-held news agencies. It has made errors but it has corrected them. The Bluewater Affair (no English article) was a deliberate and fairly elaborate hoax designed (in part) to illustrate the power of disinformation and the relative ease of creating a false narrative that would be picked up and disseminated by the news wires... numerous other high-profile organisations picked up the story and published it after dpa did, and dpa promptly published a retraction. Great as an example of media fallibility, but not that compelling as a reason to doubt the overall reliability of dpa over the long term. Pretty much all news organisations get criticised, the existence of criticism doesn't automatically make them unreliable. YFB ¿ 16:01, 25 May 2025 (UTC)
Thank you, I did read the page, no need to copy paste an entire section here. The mere existence of a criticism section on the article of a media organization does not really factor in reliability discussions of it as a source. If two different editors tell you that they think your question/complaint is unfocused, you should maybe try to engage with what they said instead of assuming that they somehow did not read what you said in the first place. Choucas0 🐦‍⬛⋅💬⋅📋 16:08, 25 May 2025 (UTC)
My point wasn't that the mighty Criticism section exists and it's big, it's what the criticism is about. I'll try to elaborate. My concern based on the allegations in the section is that the agency seems to be closely affiliated with the German government, and that the Editor-in-Chief has a lot of power within the agency to push whatever agenda he wants/whatever agenda the government wants him to - and it already showed several times. So, for example, should I refrain from citing DPA in articles about Gaza war, considering the fact that the current prime-minister is considered a staunch supporter of Israel? Should I think of a better source when I want to add something about Alternative for Germany and right-wing extremism, considering the current government openly opposes AfD? Or am I being overly critical and I can cite DPA without too much concern? Should I refrain from using DPA if I want to expand articles about sports considering that DPA might publish misinformation to either a) distort the impression of the viewer about Germany's stance and influence in the world of sports (whether it's football, ice hockey etc.) b) simply stir some drama which means higher customer engagement? And so on. Kacza195 (talk) 17:05, 25 May 2025 (UTC)
Do you have any evidence at all that dpa has published inaccurate information about Gaza or the AfD, let alone being involved in publishing "misinformation" about sports or stirring "drama"? Otherwise this discussion seems a bit of a waste of time. YFB ¿ 18:37, 25 May 2025 (UTC)
The silence speaks for itself. There is no proof, therefore there is no reason to doubt DPA's reliability. PierroPawleczko (talk) 15:14, 27 May 2025 (UTC)
Please WP:AGF and remember that everyone is a WP:VOLUNTEER here so response time may vary. Silence does not speak for anything, and what you say is somewhat of a non sequitur as well. No need to pile on after several editors already pointed the same thing out. Choucas0 🐦‍⬛⋅💬⋅📋 17:44, 27 May 2025 (UTC)
While I was unable to find much regarding Gaza war, this link provides some other examples of DPA being either misleading or biased:
  • Greta Thunberg: DPA has been reporting that Thunberg "is using the reach of her social media "to draw attention to the devastating situation of the Palestinian population in the Gaza Strip" but repeatedly ignored her accusations of genocide against Palestinians and her boycott campaign against Israeli singer Eden Golan
  • The riots caused by left-wing extremists after the 2024 French legislative election were reported as "clashes between anti-fascists and the police"
  • Calling Masoud Pezeshkian a moderate politician: while he can be considered moderate compared to the rest of the Irani government, he is still an ultra-conservative and strongly anti-Israel
  • Their portrayal of climate activists' protests in Lützerath in 2023: DPA published a photo of peaceful protesters facing police officers in full gear and titled it "the contrast could hardly be greater", but apparently protesters resorted to violence from time to time. Again, not necessarily a lie - those protesters in the snapshot probably were indeed peaceful, but the report distorted the perception of what was happening in Lützerath.
  • In June 2023, DPA reported that "UN was concerned because German authorities were investigating climate stickers". Apparently that wasn't true.
Kacza195 (talk) 20:44, 28 May 2025 (UTC)
To be honest at this point it is starting to look like you have an axe to grind against the DPA for some reason, because your approach seems to be to try to validate a posteriori your initial assumption of lack of reliability. This latest message appears to be another attempt to throw random stuff at the wall in the hope that something sticks. Even the cherry-picked examples you cite are very unconvincing, reflecting only what based on my knowledge of these issues are completely banal editorial choices, not errors, let alone disinformation. Your repeated use of the word "apparently" like you just found out about some of these things does not help with this impression either. I am not sure what you are trying to accomplish here, because it is quite clear you are not presenting things likely to move the current consensus about the DPA's reliability. I think accepting that and moving on would be best as a result. Choucas0 🐦‍⬛⋅💬⋅📋 22:27, 28 May 2025 (UTC)
Where did you get the impression that I have some personal beef with DPA or something? Like I said at the beginning, there has been some criticism regarding DPA's work. Considering the criticism, I started to doubt whether DPA is trustworthy enough to use as a reliable source. At the same time, DPA is used a lot across Wikipedia, even English Wikipedia, and that could be a problem if DPA's track record is worse than we initially considered. You have repeatedly questioned where I'm coming from, so I provided everything I could. I'm not out there to prove at any cost that DPA is shady, but to simply discuss the matter and conclude whether a) the available info is enough to put DPA's reliability into question and maybe we should think twice before using DPA as a source and if possible, use another source instead b) these controversies are too minor or simply not enough to warrant these doubts. And that's exactly what I did. And I see that the consensus that DPA is a reliable source still stands. I'm fine with that and will move on. Kacza195 (talk) 15:15, 29 May 2025 (UTC)
Just to note that if this discussion relates to the current situation in Gaza only extended confirmed editors can take part, per WP:ARBECR. That means only editor who have 500+ edits and who's account is over 30 days old. -- LCU ActivelyDisinterested «@» °∆t° 20:58, 28 May 2025 (UTC)
Your opening premise is flawed. The AP, AFP, and DPA aren't government-affiliated. Anadolu and TASS are state-owned. State control is a major factor in evaluating reliability. Mackensen (talk) 13:04, 27 May 2025 (UTC)
Mackensen, you are, of course, correct that DPA is not state-owned. But (and whatever comes before but can be ignored) in my view it is not as independent as AP. Do I have a source for that? No, not at all. Let me clarify that I am not German and have no stake in what happens there, but I know a thing or three about Europe, and speak some languages used there. DPA is effectively controlled by the "big money muscle" industrialists who would like to keep the status quo. And that can be done in Germany, unlike a place like Italy where the word "control" has no meaning. But the big money muscle has enough PR people to make sure we have no conclusive evidence to establish that. I think Kacza195 is not going to go very far in getting his ideas implemented in Wikipedia. C’est la vie. Yesterday, all my dreams... (talk) 00:07, 29 May 2025 (UTC)
Assessment of source should be done by policies, guidelines and what secondary sources have reported, not editors personal opinions about a sources. Unless you have sources to show "big money muscle" control, it's a none starter. -- LCU ActivelyDisinterested «@» °∆t° 17:31, 29 May 2025 (UTC)
Well, what I said was "Do I have a source for that? No, not at all." Did you read that? My goal was to clarify that there are no solid sources to support Kacza195's views, even if those views may have some small basis in reality. And I hoped the C’est la vie comment would suggest to him to stop. I may have just succeeded in that. As far as Wiki is concerned DPA is reliable, true or not is another issue. Yesterday, all my dreams... (talk) 23:53, 29 May 2025 (UTC)
You miss my point, if you don't have a source for that then it doesn't matter. There are many sources that are considered reliable, and I've argued for the reliability of, that I personally believes are compete trash. My opinion doesn't matter, whether I believe it to be true or not doesn't matter, editors should try and judge sources impartially not by their own personal biases. -- LCU ActivelyDisinterested «@» °∆t° 00:42, 30 May 2025 (UTC)
What is funny is that we are in full agreement, but you do not seem to be aware of it. I also think that the personal opinions of all users have the same value, namely zero, and Wiki content is based on sources. Anyway, we must end this "heated agreement" before it gets out of hand. Yesterday, all my dreams... (talk) 09:46, 30 May 2025 (UTC)

Sources pertaining to Balochistan's alleged self-proclaimed independence

Picture to make this section harder to miss while scrolling.

Three sources are currently used at Operation Herof 2.0:

  • Arab News: "Pakistan blames 'state patronage' from Afghanistan as separatist insurgency intensifies in southwest". Arab News. Retrieved 2025-05-17.
    Used in the infobox to support Pakistan accusing Afghanistan of supporting separatists in Balochistan, which Afghanistan denies.

The rest have only been cited in Republic of Balochistan when it was still an article:

The only reason I can mention these four is because I had made a source assessment table for that article before it was deleted. Unless older versions of this redirect are somehow restored, I can't recall exactly what these had been cited for, but the titles alone do seem to suggest a match with the facts I remember being present in the article. – MrPersonHumanGuy (talk) 21:47, 20 May 2025 (UTC)

It's not totally clear what your question is here. Could you restate it, please? Furius (talk) 07:58, 25 May 2025 (UTC)
It kinda boils down to this: Which of the sources listed above, if included on a source assessment table for a Balochistan-related topic, should receive which color (  Yes   No   Partial   Unknown) under the Reliable? column? I'm aware that Indian newspapers (especially those that are part of Godi media) tend to be regarded as "trash", but I want to make sure whether or not the sources listed here are trash. – MrPersonHumanGuy (talk) 10:42, 25 May 2025 (UTC)

Morphological Characteristics of Endemic Species of Araceae to Al-Jabal Al-Akhdar, Libya (Field Simulation)

At Arum cyrenaicum, I have used this 2024 source to provide this rare species with a morphological description. It is the only such description made to modern standards, and for parts of the plant the only description ever published. There has been a discussion on the talk page of the article between myself and @Headbomb. The allegation is that, because Beall identified the publisher of the journal as potentially predatory about a decade ago, the source cannot be cited. The editor in question initially removed the source less than a minute after a similar edit, and at that pace may have understandably missed an evaluation of the authors' work. Salih and Abdulrraziq are both plant physiologists at Omar Al-Mukhtar University, and have been cited in BMC Plant Biology,[1] Izvestia KSTU, and many other indisputably non-predatory journals. More importantly, it is not being cited for any evaluation made in the paper. Only for the morphological description of A. cyrenaicum. Ⰻⱁⰲⰰⱀⱏ (ⰳⰾ) 05:21, 25 May 2025 (UTC)

To be clear, the allegation is not that Beall considers these journals predatory. These journals ARE predatory. They're published by IDOSI and ISROSET, both publishers that advertise fake impact factors and that take pride in having ISSNs and being indexed in CrossRef (one of the agency that issues DOIs), Google Scholar (which scrapes everything it possible can from the web) or Wikidata, which is utterly inconsequential.
These are not reliable publications, these have not been cited by BMC Plant Biology (how could it, the BMC paper is from 2023, whereas the predatory papers are from 2024), and should not be cited, per WP:PREDWHEN.
Headbomb {t · c · p · b} 07:34, 25 May 2025 (UTC)
Only the ISROSET citation remains disputed, as Headbomb appears to have accepted my argument that, since the IDOSI paper was cited by a reputable source (WP:UBO) and was never used on its own in the article anyway, the fact it was published in such a journal no longer matters. The main rationale I have for including Morphological Characteristics of Endemic Species ... is in the WP:RSCONTEXT of the research in the paper being a straightforward description of a species lacking such detailed treatment from any other source. No biographical or medical statements were included, and no analysis either. The description could have come from the personal blog of an amateur botanist, and it would still be important until alternative sources could be found.
As for BMC Plant Biology, I did not say either paper was, only that their authors had published in such journals (in the case of BMC Plant Biology a paper by Abdulrraziq), and were professionals in the relevant field. WP:PREDWHEN is an essay, not a policy, but even it states, "Wikipedia cannot conduct such an expert analysis of sources, and must instead rely on the analysis of other experts in the field", which justifies the inclusion of the paper not under discussion. My contextual argument for the paper under discussion is that, as far as WP:VERIFIABILITY is concerned, it is neither "material whose verifiability has been challenged", "material whose verifiability is likely to be challenged", nor in any way a matter of the authors' interpretation, and so applying the "Predatory open access journals are considered questionable due to the absence of quality control in the peer-review process" guideline of the policy WP:NOTRS is excessive for now. It is a morphology paper and includes only morphological descriptions and some distribution statistics. Ⰻⱁⰲⰰⱀⱏ (ⰳⰾ) 11:23, 25 May 2025 (UTC)
"as Headbomb appears to have accepted my argument that"
I have done no such thing. Headbomb {t · c · p · b} 14:37, 25 May 2025 (UTC)
Has Headbomb not challenged the verifiabilty of the material? Are they not likely to? Phil Bridger (talk) 18:25, 25 May 2025 (UTC)
@Headbomb Do you challenge the verifiabilty of the material? Ⰻⱁⰲⰰⱀⱏ (ⰳⰾ) 18:54, 25 May 2025 (UTC)

De-archived. Headbomb {t · c · p · b} 17:37, 30 May 2025 (UTC)

References

  1. ^ Zhou, Hang; et al. (2023-11-16). "Transcriptomics and physiology reveal the mechanism of potassium indole-3-butyrate (IBAK) mediating rice resistance to salt stress". BMC Plant Biology. 23 (1): 569. Bibcode:2023BMCPB..23..569Z. doi:10.1186/s12870-023-04531-1. ISSN 1471-2229. PMID 37968598.

Puck.news, also known simply as "Puck"

Wikipedia Article: Puck (media company)

URL: puck.news

Puck bills itself as the place to find the "inside story happening at the nexus of Wall Street, Washington, Silicon Valley & Hollywood."

Authors include: Matt Belloni, Julia Ioffe, Julia Alexander, Tara Palmeri, Peter Hamby, Dylan Byers, Tina Nguyen, John Heilemann, Leigh Ann Caldwell, Baratunde Thurston and Lauren Sherman, as well as number of other journalists who do not have articles on Wikipedia.

Roster list: https://puck.news/authors/

Editors: Jon Kelly[1] and Danny Karel[2]

Questions for the Noticeboard:

  1. Is Puck reliable? Such as for a BLP article?
  2. Is Puck considered a "Self published Source"? Or is it an online magazine/newsletter?
  3. If Puck articles are reliable, are too podcasts produced, hosted and staffed by Puck?

The New Yorker wrote extensively about how Puck operates in this article published in 2022, if that's of help to anyone. This article by the New York Times describes Puck as a newsletter.

Any Noticeboard regulars commentating would be highly appreciated CeltBrowne (talk) 21:46, 26 May 2025 (UTC)

It's Danny Karel's substack and associated podcast. It is absolutely inappropriate for controversial BLP claims like what you seek at Taylor Lorenz. Simonm223 (talk) 21:50, 26 May 2025 (UTC)
I think Puck is reliable. They've got a team of professional journalists, many from well-known outlets like Politico, and they have an editorial structure in place. They’re widely-cited (WP:USEBYOTHERS), and they’ve built a reputation for solid fact-checking. Obviously, we should pay attention to whether a piece is reported journalism or more of a newsletter-style opinion, especially when it comes to BLPs. But I haven’t seen any concerns from WP:RS about their fact-checking or overall reliability, which is what really matters for a WP:GREL source. Podcasts, however, are more journalists' opinions and should be used mindfully and carefully for facts. Longhornsg (talk) 07:18, 27 May 2025 (UTC)
I generally agree Puck is reliable to the extent it's fine to cite basic Who What Why Where to it. In looking at its WP:USEBYOTHERS, this is consistent with how its reporting is referenced. It may be problematic to cite analytical statements to Puck as these may be of inappropriate tone or content for our encyclopedia and are not consistent with USEBYOTHERS For instance, in this story [3]:
  • Citing Puck to establish that Chanel reported $3.4 B earnings is probably fine.
  • Citing Puck to establish that Chanel only reports its earnings for public relations purposes ("only publishes its earnings to appease a nosy press") is probably not fine.
Under this standard, and in the hypothetical case of a BLP, I think it would be fine to cite Puck to establish that "John Smith served as communications director for Senator Jane Doe" but not to establish that "John Smith was a driving force behind all of Senator Doe’s legislative successes". Chetsford (talk) 07:52, 27 May 2025 (UTC)
The question here is that Puck claims that Taylor Lorentz was dismissed from the Washington Post for violations in editorial standards coming out of her posting a photo on social media calling Joe Biden a war criminal over his repsponse to the Israel / Palestine conflict. I am saying that Puck is not reliable for this claim as it is directly contradicted by more reliable sources that say the Washington Post released no findings from their investigation of Lorentz and said that she quit rather than being dismissed. Simonm223 (talk) 12:39, 27 May 2025 (UTC)
I don't want to sprawl on-going discussion from Talk:Taylor Lorenz onto this page: This reply is for the benefit of Noticeboard readers rather than as a direct reply to the above. I will try to focus on staying on the topic of the reliability of Puck.news, which is what we are here to establish.
My view is that other sources do not "directly contradict" Puck.
There is a (stark) difference between a source noting the Washington Post's statement on the matter and those sources endorsing that statement as "what happened" in their narrative voices. For example, the New Yorker ([4]) cites Lorenz' statement on her exit, the WaPo's statement on the exit, but also cites Puck/Byers' contradicting reporting that WaPo "determined Lorenz violated the paper’s standards and had no desire to continue employing her". In the case, The New Yorker is presenting all versions of the story and not endorsing any single one as the factual "what happened". They are leaving that to the reader. Other reliable sources do not state "Lorenz left of her own accord". They state "Lorenz says she left of her own accord".
Sources noting the statement of one side of the story do not "contradict" the other. They are not endorsing the statements, merely quoting/noting them.
Puck should not be presented as "unreliable" on this matter simply because other reliable sources quote press releases from the other side. Presenting a different version of events than official statements does not make a source "unreliable".
If Puck is reliable, I believe this situation (per WP:BLPPUBLIC) is resolved by noting Lorenz' official statement on the matter, but also the contradicting reporting, just as the New Yorker did. CeltBrowne (talk) 15:22, 27 May 2025 (UTC)
Regardless of what you may want, reliability is always contextual and the context of how you want to use Puck and for what matters, as does the fact that this edit crosses multiple contentious topics (BLP and Israel / Palestine) which indicates we should be seeking best sources, not settling on dubious ones because they present a convenient picture. Simonm223 (talk) 15:36, 27 May 2025 (UTC)
I agree that everything on Wikipedia should be properly sourced, which is why I am engaging with the Reliable sources Noticeboard to determine the reliability of Puck. I am happy to abide by whatever the noticeboard concludes.
So far, the responses to Puck outside your own have been favourable, and do not seem to agree with your assessment that Puck is a SPS. CeltBrowne (talk) 16:11, 27 May 2025 (UTC)
I don't think Puck media in general should be considered self-published, however the specific source used in the Lorenz article is from two of the owners of Puck media. As with the RFC on WP:SBM this means it probably shouldn't be used for BLP purposes. -- LCU ActivelyDisinterested «@» °∆t° 08:49, 28 May 2025 (UTC)
That’s understandable. I used the podcast as a source as I was having difficulty finding an exact print source for Puck’s reporting (their website is behind a paywall). I will endeavour to find a printed version on Puck’s website. CeltBrowne (talk) 12:46, 28 May 2025 (UTC)
This discussion has now become important for another reason: Today several news sources are citing Puck.news reporting that an allegation of rape has been made against NBA player Zion Williamson and a lawsuit is beginning. If one users would like to comment on the reliability of Puck.news, now would be a good time to do so. CeltBrowne (talk) 16:09, 30 May 2025 (UTC)
Update: Williamson has confirmed allegations have been made against him and confirmed the lawsuit. Puck.news is being credited for getting the scoop on the story. CeltBrowne (talk) 20:10, 30 May 2025 (UTC)

reporteconfidencial.info

Near as I can tell (from page inspection) this is a Wordpress hosted publication without a visible publisher. I'm not fluent in Spanish but it appears to focus on Venezuelan pop culture for which it's used in some Wikipedia pages like bios for actor Lourdes Valera and beauty pageant contestant Valeria Cannavò. I'd like to know if it's suitable either for BLP details or for establishing notability. ☆ Bri (talk) 13:33, 31 May 2025 (UTC)

Red Feather Journal

Hello! I have been collecting sources to potentially work on the Melaka Fray article in the future. I found the following article from the Red Feather Journal: International Journal of Children in Popular Culture, which provides scholarly analysis on the character.

For some background, I have found out that the Red Feather Journal is now known as the Journal of Children in Popular Culture, and it is seemingly connected to the University of Winnipeg. I am uncertain if this journal would be reliable for Wikipedia in general, or would be considered appropriate for a potential FA as I would be working on this article with the FAC process in mind.

Apologies if this is obvious. I have never heard of this journal before, under either name, so I wanted to err on the side of caution and get feedback on it. Thank you, and I hope you all have a wonderful rest of your day/night! Aoba47 (talk) 19:37, 30 May 2025 (UTC)

They have an editorial board that does a peer review of the submissions. It seems to be a reliable source that can be used. TurboSuperA+(connect) 14:41, 31 May 2025 (UTC)
Thank you for the help! Aoba47 (talk) 15:56, 31 May 2025 (UTC)
Seems like a good source for the article you mentioned. Ramos1990 (talk) 01:52, 1 June 2025 (UTC)

Hugues de Payens

On the Hugues de Payens article (or more specifically the sources section of the talk page), we are having a disagreement over whether several sources dating from the 15th to 19th centuries (but largely from the 17th century) can be used to support the claim that Hugues was Italian rather than French. They include Filiberto Campanile (1610), Antonino Amico (1636), Constantino Cajetan (1638), Blaise Francois Pagan (1669), Pierre Dupuy (1654), and Bernardo Giustiniani (1692), among others.

There is also a disagreement over the reliability of the modern Italian author Mario Moiraghi. He is also a proponent of the Italian origin for Hugues, but in particular, Moiraghi is apparently the source of the statement that a medieval text was "altered" by a 19th century editor (this statement is also under dispute).

All of the arguments have already been laid out on the talk page over the past several weeks. Now we are looking for comments from the rest of the Wikipedia community. Are these sources reliable or not? Adam Bishop (talk) 01:25, 31 May 2025 (UTC)

Adam, here is my take on the situation:

  • I am certain that Mario Moiraghi is not WP:RS for any historical items. He is an engineer, with no academic fame in history. He is in fact in the "fringe book business" for all I can see. I had never heard of him before, and I am a book lover. Umberto Eco he is not. And let us recall that with a population of 60 million in Italy, authors have to write several books to make any money because the market is much smaller than English language countries. Moiraghi seems to be in the far off fringe multi-book business. I wonder if his naext book will question if Totò was a descendent of George Washington.
  • I have only glanced at the article and the sources, but the situation is pretty confused. I do not know much aboutt the Templars, but know that they are a controversial group. From what I can see there is no certainty as to where Hugues was born.
  • Finally, why on earth is there a brouhaha about where Hugues was born?

I think the best way is to say we are not certain where Hugues was born. Yesterday, all my dreams... (talk) 08:07, 31 May 2025 (UTC)

I'm pretty sure with sources that old, WP:RSAGE would recommend against using them at all. Sources from centuries ago shouldn't be considered reliable secondary sources. It takes real historical expertise to understand how and when to rely on documents that old. Ideally we should base articles on modern scholarship where professionals interpret, analyze, and summarize those older sources. Red Rock Canyon (talk) 03:24, 1 June 2025 (UTC)

Pinkvilla

This source Pinkvilla, which mostly posts box office numbers and entertainment gossip, is considered reliable at Wikipedia:WikiProject Film/Indian cinema task force. I'm posting this RfC here because if I post it at WP:ICTF, the result is only going to be in favour of this source. The editors there defend this source with passion. Here is a brief history of the source:

First thing first, here is the disclaimer they give out at the end of each article, as pointed out by the editor Black Kite in a previous discussion here:

The figures can be approximate, and Pinkvilla does not make any claims about the authenticity of the data

That should say enough but here's more:

Pinkvilla was previously considered unreliable at ICTF and was moved to the list of reliable sources in February 2022 without proper discussion, as seen here.

The Entertainment Editor at Pinkvilla, Himesh Mankad, welcomed the current admin of their box office section and his associate to Pinkvilla in December 2021, as seen here.

In this tweet, the admin of Pinkvilla Box Office confirm that they started the blog Cinetrak and moved to Pinkvilla from Cinetrak. Now Cinetrak is considered unreliable. Although the admin say they have not been involved with it since moving to Pinkvilla, they continue to endorse Cinetrak and call it the "best box office tracking site in South India", even in this recent tweet. The box office figures given by Cinetrak and Pinkvilla are also exactly the same for every film.

Now, it's not as if there aren't other sources reporting on the box office of Indian films. Many sources including newspapers like The Hindu, The Indian Express etc. publish box office figures. And they do this without citing any of these other dubious sources. If a movie has indeed earned a specific amount of money, then other sources will publish about it, as Indian films are often widely covered in the media.

(2405:6E00:223E:AF6E:A85D:7FF:FEB8:8AC6 (talk) 03:41, 31 May 2025 (UTC))

Note: Moved from WT:RSP per that page's uppermost talk banner. Aaron Liu (talk) 05:16, 1 June 2025 (UTC)
This discussion was 3 years ago. Has there been any discussion since that woul satisfy WP:RFCBEFORE? TurboSuperA+(connect) 05:50, 1 June 2025 (UTC)
There have been many discussions since then, both at ICTF and at RSN. The latest happened around 3 months ago. You can see the link to the latest discussion here. 2405:6E00:2800:5F7E:DC54:41FF:FE2C:7CD5 (talk) 07:21, 1 June 2025 (UTC)
Thanks. I was mainly asking because I couldn't find this quote "Pinkvilla does not make any claims about the authenticity of the data", it is pretty damning. I would support labeling the source generally unreliable and adding it to WP:RSP. However, your RFC does not contain a (neutrally-worded) question, per WP:RFCBRIEF. You can look at this example (or any other on this noticeboard) to see how others have done it. Here is how you could do it:
Name the section RfC: Pinkvilla and put this in the body (copy-paste or modify):
{{rfc|media}}
What is the reliability of [https://www.pinkvilla.com Pinkvilla]?
* Option 1: [[WP:GREL|Generally reliable]]
* Option 2: [[WP:MREL|Additional considerations]]
* Option 3: [[WP:GUNREL|Generally unreliable]]
* Option 4: Generally unreliable, with [[WP:DEPREC|deprecation]]
Then you post a notice on the WP:ICTF page and on the talk page of any involved editors with the following template:
{{subst:rfc notice|Pinkvilla|Wikipedia:Reliable sources/Noticeboard#RfC: Pinkvilla}} ~~~~
The current RFC should be withdrawn, and you can do it since you started it, per WP:RFCCLOSE. The text you wrote to open this discussion is more of an argument for the source's deprecation, so that should be posted as a WP:!VOTE or comment in the RFC.
Hope that helps. TurboSuperA+(connect) 08:33, 1 June 2025 (UTC)
  • The problems mentioned at that last RSN continue (I think it is fine if box office returns using PinkVilla are clearly marked as "estimated" but it appears that too many are not). However, there is a far bigger problem with Pinkvilla and that is the lifespan of their news stories - go to this page, and try clicking the links. Most of them simply lead to Pinkvilla's custom Error 404 page. Pinkvilla is used 5,088 times and I suspect at least half of those links are dead. Black Kite (talk) 07:54, 1 June 2025 (UTC)
    The first couple I tested were backed up at the Internet Archive. There roughly 2,500 articles with Pinkvilla links,[5] many of which already appear to have archived links. Pinkvilla articles do seem to have a very short lifespan, with even articles from late 2024 no longer appearing on their website. I'm not sure that a RS issue though, even if it is odd. -- LCU ActivelyDisinterested «@» °∆t° 08:29, 1 June 2025 (UTC)
    I think it is indicative of them trying to save money on web hosting, but as you said, not an RS issue. Regardless of what the source ends up being labeled as, a note should be written that any time an article is used it should be archived and the archive link provided in the citation. TurboSuperA+(connect) 08:39, 1 June 2025 (UTC)
    GreenC might be able to help ensuring that any current uses get archive links added. -- LCU ActivelyDisinterested «@» °∆t° 10:47, 1 June 2025 (UTC)

Classical sources (Herodotus, Plutarch etc)

On a lot of pages related to classical history, it is not uncommon to see direct citations to ancient writers such as Herodotus and Plutarch, often taking the claims of ancient writers as face value. (See Ptolemy IV Philopator for an example of direct heavy use of classical sources). Citations to classical writers are also found on a lot of other pages. In my opinion, uncritically citing writers who lived well over a thousand (or two) years ago is pretty dubious, and ideally instead we should really be citing the opinions of modern (or at least within the last century or so) scholars who comment on the remarks and claims of these writers, and citing directly should only be done very sparingly and extremely cautiously (i.e. always attributing to the writer in question), if at all, as it is easy to engage in WP:OR using them. I think it would be good to have something at WP:RSP about it like we have for religious texts (WP:RSPSCRIPTURE). Hemiauchenia (talk) 22:47, 11 May 2025 (UTC)

Full agreement from yours truly. And I would extennd the time frame and include sources such as Galen. And given that some of the "supposed" writings only survive in translations and later quotations, the whole issue is often less than clear. I mentioned Galen, given that there is even a question as to how many Galens there were. Regarding Plutarch, given that at times he differs from Suetonius, just quoting him alone is obviously a mistake. We need modern comments on these ancient works, for sure. Yesterday, all my dreams... (talk) 00:59, 12 May 2025 (UTC)
While adding commentary or analysis from modern scholars is of course always ideal (as is simply using modern, scholarly sources in place of, or in addition to, ancient ones, wherever possible), I would point out that the "dubiousness" of using ancient sources on their own doesn't necessarily apply across the board: mythology is an exception, as it isn't something that ancient authors can be "wrong" about. The usage of such sources in a mythological context is of course still subject to WP:PRIMARY, but I say this just because I wouldn't want to end up with a guideline that inadvertently prohibits directly citing Apollodorus for the claim "Apollodorus calls Zeus the son of Cronus", for example.
But yes, I would agree with the general point here, especially in the context of historical events. I would also note the existence of the essay Wikipedia:Primary sources in classics. – Michael Aurel (talk) 01:52, 12 May 2025 (UTC)
I wouldn't want to end up with a guideline that inadvertently prohibits directly citing Apollodorus for the claim "Apollodorus calls Zeus the son of Cronus", for example.
Even in this sort of seemingly simple case there are several caveats that editors would do well to be aware of, however:
  • Apollodoros might support the claim that "Apollodorus calls Zeus the son of Cronos", but we need to be careful to attribute those claims to Apollodorus; the fact that a surviving source gives one particular version of a myth does not mean that was The Canonical Version. Apollodorus says that Asopus was the son of Oceanus; other sources say that it was the son of Zeus or Poseidon.
  • Especially for a subject like Zeus, for which there is a lot to say, the fact that a story is told by one author does not necessarily mean that it's important enough to include on its own; we need to base which stories about Zeus are important to include primarily on secondary sources (though this is more a question of WP:WEIGHT than WP:RS)
  • If you are going to cite an ancient source, don't rely on the 19th-century translations which are freely available online. In the case of poetry, translational fidelity was frequently sacrificed for metrical reasons; in the case of the fragmentary lyric poets editors and translators just made stuff up to fill in the gaps; more recent discoveries or scholarship have sometimes changed the agreed-upon text. Otherwise you end up with: "In the Ode to Aphrodite, Sappho describes Aphrodite as riding a chariot pulled by swans (Sappho 1 Edmonds)" – the Greek actually says the chariot is pulled by sparrows but in Edmonds' edition he just substitutes swans. Even when the editor is more generally reliable than the famously bad Edmonds, scholarship moves on: "Sappho dedicates a hymn to Kypris and the Nereids (Sappho fr.5 Voigt)" is entirely verifiable by checking the cited edition – and there are 50 years of scholarship for which this is the agreed text – but a 2014 discovery proved that the beginning of the poem doesn't mention Kypris at all.
Caeciliusinhorto-public (talk) 08:55, 12 May 2025 (UTC)
Certainly. I agree with every part of that. – Michael Aurel (talk) 09:11, 12 May 2025 (UTC)
Yes, and I don't mean to suggest that you aren't aware of these caveats. I thought it was worth explicitly putting in the discussion, though, because I see all of these issues in our articles. It is both in our older content (which was often either copied wholesale from out-of-copyright encyclopedias or gradually accumulated by random users adding whatever details about a subject they could easily find online) and in newer content added by enthusiastic new users who clearly haven't thought too deeply about these issues (in many cases I suspect young people without any academic training in a humanities discipline and without access to sources other than popular history books and easily-available translations of the major primary sources aimed at a general rather than academic audience). Caeciliusinhorto-public (talk) 09:22, 12 May 2025 (UTC)
Yes, all good. In some sense, all of those points derive from other guidelines or from general good practice (WP:INTEXT, and the general principle that statements which aren't factual or widely agreed upon should typically be attributed to their source; WP:WEIGHT; and WP:RS and WP:RSAGE), but they're all good points, and it's worth articulating them in a discussion which might result in a guideline being written up. I think I was treating them all as implicit truths (and I wouldn't say that any of them necessarily go against the permissibility of citing ancient sources for simple mythological statements, as the same points – or slightly more generalised versions of the same points – would probably apply equally to a number of other kinds of content), though maybe they aren't quite as "implicit" as I think they are, seeing as it would be entirely accurate to say that the majority of our Greek mythological articles currently don't adhere to those principles.
One could probably add a handful of other points (or caveats) to the above:
  • Primary sources can be misinterpreted: texts can have corruptions and lacunae, and in some cases the original work may simply have been poorly written (such as the Fabulae), resulting in the translated text being stilted or difficult to interpret. Some readers might even be thrown off by flowery translations. This is aside from basic content misunderstandings such as not realising a name is an epithet for a deity, or not knowing whether the words "Ocean" and "Night" merely represent figurative language or refer to specific deities.
  • You should of course also make sure you actually read a primary source before citing it, per WP:SWYRT, and because some secondary sources – I'm looking here at the DGRBM, in particular – can get their citations wrong at times. There can also be discrepancies between the source itself and how it's presented in a scholarly discussion (the swapping of Greek and Roman names of gods would be a common example). A website such as Theoi.com also has the pernicious tendency of adding certain information (especially names of mythological figures) in square or round brackets to passages of quoted text, when this information isn't present in the original.
  • Knowing the age of certain sources – as well having some general contextual awareness about them – is also important. Repeating a genealogy from Tzetzes (and citing only him for the claim) as though it is representative of ancient Greek belief is misleading. This is not to mention the issues with repeating, without context, the glowing endorsements of the Greek gods found in Pseudo-Clement, or the "creative" mythological material in Natalis Comes.
  • What's discussed at WP:CLPRIM#Transmission (which you've indirectly touched upon) is also worth noting, as I suspect it's something that most editors aren't aware of. "Hesiod, Theogony 600", unaccompanied by any sort of bibliographic information, doesn't refer to an exact, definite text, so making clear which edition or translation you're working from is essential.
– Michael Aurel (talk) 13:20, 12 May 2025 (UTC)
Careful citation clearly distinguishes between what sources, ancient or modern, have to say about a subject, and what we know to be the case. There should be no problem in quoting or citing Greek and Roman writers for what they say, provided that their opinions aren't treated uncritically. Indeed, in history generally, and classics in particular, most modern references rely primarily on what historical sources report, and so those sources need to be cited, provided that the citations aren't needlessly cumulative or misleading. There's not much more frustrating than seeing a claim about historical events made by a modern writer with a strong opinion about a subject, but having no way to determine what the sources being analyzed are.
Naturally, anything that goes beyond mere citation should come directly from secondary sources to whatever extent is possible. Sometimes secondary sources are cited generally in articles that rely heavily on extensive treatments from just one or two of them, even though they may cite several different ancient authors internally. This can give the false impression that an article depends chiefly on primary sources, but the fault here is with the manner in which the sources are cited, not the fact that they're cited.
My rule of thumb is to cite the most authoritative sources for each point, both ancient and modern, and prune the ones that have the least detailed or helpful information when they become cumulative. Generally ancient sources are good for bare facts and what the ancients knew or thought they knew about things; modern sources are good for compiling and analyzing those facts and making broad statements about them.
These sources should be complementary, not exclusive; part of the value of detailed citations is allowing readers to find the ultimate sources and consider them in light of what later writers have to say on a topic—and given that classics has been a major subject of scholarship since the late 1700's, opinions vary widely and have changed over time—sometimes repeatedly, along with the availability and accessibility of sources, both ancient and modern. When carefully written, an article will make clear both what the original sources for any statement are, and what contemporary scholarship has to say about it. When an article fails to do this, the remedy is to supply what is missing in the ordinary course of editing. P Aculeius (talk) 02:25, 12 May 2025 (UTC)
It's an ongoing problem, yeah. I agree with what others have said so far. And I find that mostly the people at WP:CGR and other history-related wikiprojects usually have a pretty good sense of this, but there are a lot of contributions still out there from (mostly) newer editors who didn't necessarily come from an academic history background who don't understand this and don't necessarily have it explained to them before they make a good number of contributions.
A couple caveats:
1. Though we should cite modern scholarship wherever possible, primary sources are important as well and often we'll want to cite both, so it's not always the best idea to remove them. In the absence of a handy secondary source to cite, I often find that it's possible to reword a claim such that it's *not* wrong per se even if it's factually inaccurate. e.g. If I write Pherecydes of Syros drank from a magic well and predicted an earthquake in WP:Wikivoice like it's totally normal and cite Diogenes Laertius, that's a potential problem. If on the other hand I say Diogenes Laertius states that Pherecydes of Syros drank from a magic well and predicted an earthquake instead and tag it with Template: Primary source inline, that's not quite as bad, the source does in fact verify that Diogenes said that. Ideally it would be prefaced by a modern source saying something like Several miraculous and apocryphal deeds were attributed to Pherecydes by ancient biographers.... It can be more difficult when we have primary sources making less fantastical claims that also happen to be wrong, like when someone was born or who their teacher was, but in those cases if we *know* it's wrong it shouldn't be too hard to dig up a modern source that says so.
2. Many of our articles were copied from public domain sources such as the 1911 Encyclopedia Britannica and Dictionary of Greek and Roman Biography and Mythology, which do frequently cite primary sources without any accompanying secondary source (other than themselves). These sources, while written by some of the best scholars of their time, are generally rather outdated and often far more credulous than modern scholars, and most of those articles do need to be updated. I tend to think it's better than nothing in most cases, though if you encounter anything that's copied from there, make sure the article has the relevant attribution template, such as Template:DGRBM; many of them have been removed or were never placed on the articles in the first place and in addition to being important to not commit plagiarism, it helps to be able to track them to have the template on there.
I'd also second the essay Wikipedia:Primary sources in classics as a great resource to point people to if you get pushback; it's not policy but it articulates the problems pretty clearly. Perhaps there's something that could be done to put a condensed version of that on WP:RSP? Psychastes (talk) 02:46, 12 May 2025 (UTC)
Thanks for linking Wikipedia:Primary sources in classics that's a really fantastic in-depth essay on the topic that I wholeheartedly agree with. I would support AD's suggestion below to link to this essay at WP:RSP. Hemiauchenia (talk) 13:07, 12 May 2025 (UTC)
Separately, I think that in a lot of cases with these stories they really should be in the narrative present to indicate that they are stories. Ifly6 (talk) 17:31, 12 May 2025 (UTC)
I don't know if a specific guidance is needed, but I do agree that we should rarely cite ancient texts on their own. WP:PRIMARY is already a relevant guideline here. Traumnovelle (talk) 02:52, 12 May 2025 (UTC)
It depends on what the source is being used on. Wikipedia:Primary sources in classics gives some helpful information. But in general there are divisions in articles about what a classical author states and what scholars today think about the matter. Context usually matters. WP:PRIMARY is helpful with handling Classical sources. I also know that there are modern commentaries from scholars on Classical sources for example, on Julius Caesar's "Civil War". Also I find that many of the books that are used as primary sources such as from Loeb Classical Library, they tend to have introductions by translators or experts which explain and interpret the Classical author's text, giving important context and secondary source material essentially before the primary source material. For example, Plato's Complete Works. Sometimes there are footnotes by the translator or expert on the primary source text too giving modern commentary as you read the ancient authors works. Ramos1990 (talk) 07:33, 12 May 2025 (UTC)
The idea of having a summary of (and link to) WP:Primary sources in classics sounds like a good idea. Such sources shouldn't be used without attribution. -- LCU ActivelyDisinterested «@» °∆t° 09:02, 12 May 2025 (UTC)
Concur. WP:PRIMARY is the guiding document and the essay about primary sources in classics is a good one. There may be times when it's apropos to include a quote. For instance if you are describing something Plato said directly you might, you know, quote Plato. However beyond that limited use case it is best to cite living classics scholars. It's a vibrant academic discipline and there's plenty of contemporary work. Simonm223 (talk) 13:14, 12 May 2025 (UTC)
(Also I wish I knew about this essay when I was engaged in that dispute over Vinland a while back - it would have saved me a lot of typing.) Simonm223 (talk) 13:15, 12 May 2025 (UTC)
I think it is OK, as long as modern scholarship does not challenge the claim, if so. Go with the modern source. Slatersteven (talk) 13:22, 12 May 2025 (UTC)
Agree with those who mentioned WP:ATTRIBUTION too. Ramos1990 (talk) 13:57, 12 May 2025 (UTC)
The (imo pretty obvious) source, pun unintended, of a lot of the bad articles that the OP talks of are inexperienced or negligent editors deciding to add material to the encyclopaedia by turning to Livy or Dion Hal first and then just paraphrasing whatever translation was on hand. This was really evidently with the pre-rewrite Founding of Rome and Overthrow of the Roman monarchy articles (or the current Pyrrhic War). Setting a P&G that simply precludes doing that would drive such editors to better sources that would be preferable not only for the encyclopaedia but also for their own self-development. Ifly6 (talk) 17:23, 12 May 2025 (UTC)
No issue with the idea of a notice or some such. Slatersteven (talk) 17:26, 12 May 2025 (UTC)
Many of the commenters here contributed to WP:CLPRIM, as did I. I would support an addition like WP:RSPSCRIPTURE which links to CLPRIM, along with essentially the same guidance that directs primary sources be used only with citation to reliable secondary sources and prohibits editors' analysis as WP:OR. I also completely agree with Caeciliusinhorto's comments on transmission, which in another form were incorporated into CLPRIM: it establishes also the folly of just quoting or citing primary sources for what an internet edition appears prima facie to say.
At the same time, I see absolutely no need to banish parallel primary source citations (eg MRR 1.123, citing Livy 1.23, Suetonius Julius 12.3, Dio 1.2.3, and Plutarch Caesar 1.2.3). Such parallel citations are of great use to specialist readers – especially so when they are complete surveys of all the relevant sources – and I think we should encourage their inclusion; editors, however, should not seek to create their own surveys: only secondary sources can tell us, usually by speaking for themselves, whether a citation is due or not. Ifly6 (talk) 17:17, 12 May 2025 (UTC)
Yes its inappropriate to provide a truly outdated source as a stand alone citation, at best its simply not due if it hasn't been mentioned by a modern writer. I think the line is also a lot closer to a hundred years than a thousand or two... Horse Eye's Back (talk) 17:36, 12 May 2025 (UTC)
Looking at Ptolemy IV Philopator in a bit more detail now, I'm not sure if that page actually has an issue; most of the paragraphs with primary source citations in them seem to cite Hölbl, Günther (2001). A History of the Ptolemaic Empire. London & New York: Routledge. ISBN 0415201454 at the end. Routledge is certainly a reputable publisher and Hölbl seems to be a reputable Egyptologist. If I were to nitpick, it would be better to have more than one modern scholarly source supporting the majority of the claims, but admittedly I also don't have the background to assess if Hölbl is enough of an authority that it doesn't matter. Taken under the assumption that all of the secondary source citations at the ends of the paragraphs support the primary sources cited, I don't think there's any glaring issues with that article, it looks pretty close to the ideal we're describing, though it may help if anyone else wants to take a closer look and weigh in. Psychastes (talk) 18:06, 12 May 2025 (UTC)
If the citations there should be taken to mean this content is supported by Holbl 2001 p ### which cites Polybius #.## and Suchandsuchius ##.##, they should be joined up into the same citation, and the text should convey that dependency, rather than be put in separate citations like Polybius #.## and Suchandsuchius ##.## // Holbl 2001 p ###. I haven't read Holbl 2001, so I don't know whether they should be joined up or not. Perhaps we could ping or tag the original author(s) for clarification? Ifly6 (talk) 18:14, 12 May 2025 (UTC)
Looks like the main contributor to the article is @Furius, who has been active recently and might remember how he cited the claims in Ptolemy IV Philopator, though admittedly most of the article seems to primarily date from 2018. In general my default assumption with this type of citation style would be that it's probably parallel given that all of these secondary citations are at the ends of paragraphs (which is technically all that's required), though I certainly wouldn't change it to be explicitly so without checking the source. Psychastes (talk) 18:32, 12 May 2025 (UTC)
When I was doing these Ptolemaic articles, I read Hölbl and other secondary sources. When they cited a primary source, I double checked the citation and then included it as well. Furius (talk) 10:24, 13 May 2025 (UTC)
I'd be slightly resistant to "as cited by" for passages from classical authors that are regularly cited for a given fact, because that ties the primary source inescapably to one secondary source, which may or may not prove to be the best one. Furius (talk) 10:26, 13 May 2025 (UTC)
I'm unconvinced by this rationale. We won't know if something is the thing cited without checking multiple sources, in which case you could just add them on. Also consider form like Whateverius #.##, cited by: A 2000 p ##; B 2000 p ##. Ifly6 (talk) 13:34, 13 May 2025 (UTC)
Yeah, perhaps that page is not the best example if I've just misread the citation style. I think Demetrius of Phalerum is considerably worse, given that it almost entirely relies on classical sources (note the 1925 source is actually a translation of Diogenes Laertius who lived around 200 AD) Hemiauchenia (talk) 18:34, 12 May 2025 (UTC)
Ah, yeah, I agree, that's a good example, looks like it's mostly still got the text from the DGRBM entry, with some apparent WP:OR from people digging back into those same primary sources that the DGRBM cites to do their own research. Seems to have been mostly a slow bleed by a number of uninformed editors with minimal pushback, certainly ought to be cleaned up and updated though.
And yeah the dreaded Template:cite LotEP probably ought to be deprecated or at least modified immensely, the way it is now it looks like a source from 1925 which, while still not *ideal*, is a far cry from Diogenes Laertius, who might as well be a cautionary tale in the unreliability of ancient sources, given how he mostly just pastes together extracts from contradictory testimonies haphazardly. Psychastes (talk) 19:22, 12 May 2025 (UTC)
I'd be glad to see a statement in WP:RSP. Article creators and editors may be unfamiliar with these problems and even think that an ancient historian is, being a historian, a secondary source. Thus we have for example many articles on battles based solely or largely on Livy (e.g. in Category:4th century BC in the Roman Republic) though as Battle of Saticula#Modern views rightly says, "Livy's battle-scenes for this time period are mostly free reconstructions by him and his sources, and there is no compelling reason why this battle should be an exception." It's understandable that editors with an interest in, say, military history might be unaware of this and meticulously summarise freely available Livy rather than seek out modern WP:SCHOLARSHIP but it's not good for our readers. NebY (talk) 12:06, 13 May 2025 (UTC)
Entirely agreed. Ifly6 (talk) 14:28, 13 May 2025 (UTC)
  • (1) Some classical sources are secondary because they are history books describing events that happened hundreds of years before they were written. (2) Some classical sources are reliable for some purposes (but not all). (3) Primary sources, such as eyewitness accounts, may sometimes (but not always) be more reliable than secondary sources, such as history books written long after the events they describe. Some classical sources are very unreliable precisely because they are secondary sources trying to describe events from hundreds of years earlier that they know nothing about. See, for example, the comments of Betty Radice in the introduction to The War with Hannibal. I cannot support the veneration of "secondary" sources that is based on a profound misunderstanding of what the word "secondary" means, and what actual primary and secondary sources are and are not reliable for. The community needs to stop trying to weaponize the words "primary" and "secondary" by treating them as synonyms for "unreliable" and "reliable" respectively, which they are not. I certainly oppose the inclusion of this discussion of classical sources in RSP. James500 (talk) 16:52, 13 May 2025 (UTC)
    No, that's not correct. All classical sources are primary sources for the purposes of wikipedia. Analyzing ancient sources for yourself to determine which ones are reliable or not is considered WP:OR. This should be done by professional historians, not wikipedia editors. Please do not do add information to wikipedia based on your own interpretations of classical sources. Psychastes (talk) 17:17, 13 May 2025 (UTC)
    If it was, in of itself, original research for a Wikipedian to !vote on whether or not a source (whether ancient or modern) is reliable, then everything that has ever been said and done at WP:RSN would be original research, including everything you have said in this thread. Please do not make WP:NPA personal attacks by falsely accusing me of adding "information to wikipedia based on your own interpretations of classical sources", as I have not done that. James500 (talk) 20:41, 13 May 2025 (UTC)
    WP:TROUT Psychastes (talk) 22:34, 13 May 2025 (UTC)
No, all classical sources are primary sources. Further examples of primary sources include: ... religious scripture; medieval and ancient works, even if they cite earlier known or lost writings. WP:OR. Ifly6 (talk) 20:12, 13 May 2025 (UTC)
This issue here is not whether the Wikipedia policy purports to define classical sources as "primary". I can see that it does. The issue is that the Wikipedia policy's definition does not match the definition used by professional historians, and accordingly the wording of the policy is therefore causing confusion and obfuscation that disrupts the project. The solution is to simply rewrite the policy, which is what I am suggesting we should do. James500 (talk) 20:50, 13 May 2025 (UTC)
Re the definition used by professional historians. I think you'll find that all classical historians call Plutarch's Vitae Parallelae or Livy's Ab Urbe Condita primary sources even though they wrote of people who lived centuries before their time. Ifly6 (talk) 21:11, 13 May 2025 (UTC)
On a search for livy+"is a secondary source", the first thing that comes up is The Princeton Guide to Historical Research (2021), which says he is secondary for "events that took place centuries before [his] birth": [6]. The author is a history professor. It would be helpful if I could see your sources, if you are telling me that classical historians use a different terminology from other historians? James500 (talk) 21:27, 13 May 2025 (UTC)
Taking a look about Google Scholar, I concede that it is not universal to call the ancient sources "primary sources". (An explicit counterexample is Worthington, in Alexander the Great: a reader, whose focus on "real" primary sources elicits explicit clarification in the BMCR review.) However, this terminology is very common: see the first heading under WP:CLPRIM; a Google search will yield yet more library pages putting Plutarch and Livy in their primary sources tabs. The next immediate thing that comes to mind are source books like Gary Forsythe's Primary sources for ancient history, with a BMCR review and excerpts. If you flip through the chapters and look at what the excerpts are, you'll see that they are mostly ancient literary sources. From the review:

His collection of sources includes all of the greatest hits of Roman historiography. We start out with Plutarch’s Life of Romulus, followed immediately by Livy on the end of the monarchy. From then on, the narrative history of the Republic is told almost entirely in excerpts from Livy and Polybius, supplemented by biographical sketches from Plutarch. Polybius is also utilized, as he should be, to explain the Roman constitution. Appian, Dio, Sallust, Caesar, Cicero, and Suetonius round out the sources for the late Republic.

Similarly, Mellor and Podany's The world in ancient times which contains Plutarch's Pericles (written centuries after Pericles' death) and Mathisen's recent Ancient Roman civilization: history and sources (described as "primary sources" in this review) which contains Plutarch's Romulus and substantial portions of Livy. This usage is remarked upon in another review:

When reading those introductions, students probably will observe that the phrase “primary sources” on the title page is appropriate only if it refers to ancient texts written in Greek or Latin... This is how a classicist will understand the phrase, but such traditional stories as an anthropologist of our time would classify as primary source material—retold orally and unaltered by literary ambition and intellectual reflection—are mostly out of reach to the student of classical mythology. The sources at hand are typically several steps remote from their supposed originals.

Reviews regularly discuss poor referencing of the primary sources (eg Plutarch and Arrian), the presence of many citations to primary sources (in CAH2 13), or authors' credulousness (Zosimus lived long after the Palmyrene empire) towards primary sources. Context indicates they are referring to the ancient literary historical narratives.
Moreover, most books will also include a section discussing problems with sources and their citations, which will often refer to ancient literary sources collectively as primary sources (eg Tempest Brutus our primary sources are the secondary historians of their own day and Wilson Dictator Primary sources often stated... Dig. Pomp. 1.2.2.18... Livy 2.18.8... Livy 4.13.11... Dion. Hal. 5.70.1... Plut. Fab. 3.5.) and also direct you under that name to some scheme that contains abbreviations for authors (OCD: Plut. = Plutarch, Suet. = Suetonius, Polyb. = Polybius) who could not have been around to describe the events they wrote about. Ifly6 (talk) 05:24, 14 May 2025 (UTC)
Support adding this to RSP, per my comments above, and this is evidently a problem many editors struggle with. Psychastes (talk) 17:32, 13 May 2025 (UTC)
Agreed. Ifly6 (talk) 20:12, 13 May 2025 (UTC)

PLEASE NOTE: the above "Wall of text" is well worth reading. Paul August ☎ 12:28, 22 May 2025 (UTC)

RE There can and should be no objection to citing a passage of Livy or Polybius or Cassius Dio for the facts that they record, provided that it is clear that they are the sources, and the articles citing them do not present their opinions on various matters as objective fact; and that a general reference from modern historical writing is cited for anything that analyzes or provides context for the facts that they report., I don't disagree, but I believe this is the exact thing that is often unclear to newer editors or those less familiar with classics, and, while I also believe it follows from already existing Wikipedia policies, I do not believe it follows non-trivially, and we would benefit from an explicit statement of (basically the quoted text here) at RSP. Psychastes (talk) 16:54, 14 May 2025 (UTC)
RE a simple account of them is as reliable no matter which of a hundred respected scholars of the last two centuries wrote it, I respectfully disagree; this may be true for (some parts of) Ancient Athens in the classical period and the late Republic/early Empire, but, for example, the vast majority of critical scholarship on Hellenistic philosophy has been done since the early 1970s and for the late antique Commentators on Aristotle since the late 1980s, scholars from prior often didn't have critical editions of the texts, and even someone like Hermann Diels or Eduard Zeller is often just flat wrong about the basic contents of the vast corpus of texts they discuss, let alone their interpretations! and Hegel or Friedrich Schleiermacher are so rarely correct it makes sense to expunge them entirely because they didn't even *have* the texts. Even a large part of historical Plato scholarship has been overturned by the discovery of early papyri with what 19th century historians alleged were *late* pseudepigrapha! But for the more thoroughly-studied parts of the classics, I find there's almost no shortage of reputable scholars from the past fifty years willing to re-make basically the same (valid) interpretations of Cicero or Virgil. So while we shouldn't go through and purge every cited source who published before 1925, if someone is insisting on an older source because it's the *only* place they can find verification for a claim, it somewhat begs the question why no more recent scholars have tried to make the same claim. Psychastes (talk) 16:54, 14 May 2025 (UTC)
Agreed on the earlier and later points. The only way to determine whether inclusion of some primary source citation is WP:DUE (or not) is by reference to modern reliable sources. Ifly6 (talk) 17:27, 14 May 2025 (UTC)
You missed the first part of that sentence: "them" refers back to "the basic facts", not analysis of things, much less philosophy. It is unlikely that anyone is going to discover that the Pelopponesian War was actually fought over cake recipes or that Socrates didn't really die but went into hiding, that Spain was invaded not by the Vandals but by the Sacks, that Caesar crossed the Rubicon a week earlier than all the authorities report, that the patricians were made of cheese, etc. These kinds of facts do not change because of new scholarship revealing long-hidden secrets; unlike an analysis of Aristotle, these sorts of things are rarely affected by archaeology, and it does not matter what in the last two or three centuries they are cited to.
Obviously in the rare instance that a new discovery does shed new light on something, then that should be cited—but if the only change is that two scholars differ in their interpretation of something, and the only basis for their disagreement is the passage of time, not the discovery of new and previously unknown material, then it would be irresponsible to claim that whatever the latest opinion is must necessarily be correct, and all other opinions must be wrong. For that matter, changes in the interpretation of anything over time are also noteworthy, and should be mentioned when they fundamentally alter our understanding of those things.
To be clear, what I am arguing against is not the inclusion of new or recent sources on any topic, but rather the desire, or perceived desire, of some editors to purge cited sources: ancient ones because they may be subject to interpretation—even where the cited passage makes no attempt to do so, and is supported by modern scholarship; and modern ones not because they can be shown to be wrong, but based solely on their age—whether or not a more recent source on the same point is available, or disagrees materially with the cited statement. P Aculeius (talk) 20:41, 14 May 2025 (UTC)
These kinds of facts do not change because of new scholarship revealing long-hidden secrets Accepting that this is true for some facts, I don't know that we can reasonably define the types of facts for which it would be true in a way which would be in any way useful as a guide to editors. I can think of plenty of examples of seemingly basic facts which modern scholarship is doubtful of. Plutarch says unequivocally that Aspasia was put on trial; the majority opinion today is that this never actually happened. Ancient sources absolutely get dates wrong – there's an ongoing scholarly debate about whether the ancient sources are even in the right century when dating Corinna's floruit, for instance. Roman and Greek sources absolutely do misidentify or conflate various groups of people, or indeed the origins of specific people, and they absolutely claim causes for events which modern scholarship disagrees with. Caeciliusinhorto-public (talk) 12:18, 16 May 2025 (UTC)
I think you've laid out the argument for restraint in WP:Creeping out a new sourcing caveat--at least to the extent of using a light touch--quite effectively above, Aculeius, but I feel you are falling into a bit of epistemological trap here: the extent to which we judge a given source to be an accurate rendering of past events is not itself subject to the function of whether it is likely to be disproven. So, to illustrate that point: say we are considering some claim Suetonius makes about the death of Geramnicus, which for some point of historiocity or general skepticism, we (either as modern interpretors generally or a subset of the Wikipedia editors working in this area in particular) feel a cause to take with a grain of salt. Now you're suggesting we will not typically be faced with a situation where new evidence is likely to disrupt the standign weight we give to the statement of that fact, beyond the extent that it was already in doubt by logn established conflict with another source or the influence of long centuries of critical analysis. Which I think we can grant is true (at least for a large number of cases), but which is also kind of to the side of the arguments you were responding to, if I interpret them correctly. Because the point that we cannot count on their being further disrupted does not directly influence how reliable they were as its own matter.
So I'm not altogether indifferent to the notion that it might make sense to qualify the use of classical sources a little more than we currently do. But I share your skepticism of some of the arguments that have been advanced as to particulars. For example, I am definetely not in favour of the proposals below, which seem very reductive and problematic to me in a number of respect. To begin with, attempting to copy-paste our approach to biblical sources as a starting point seems to me a very hamfisted and in-apt starting point. classical historians may not have had access to the fraction of the knowledge and methodologies of their modern counterparts, but they largely had some sense of histioricity and what we would regard as a sense of editorial discretion as to their sources. At a bare minimum, they are certainly qualifiably different texts than those produced for liturgical purposes, so this transplant of polict language feels very ill-advised. As is any proposed new policy language that attempts to paint all "classical" texts with the same brush in terms of basic reliability. I have to think whatever the proper solution here, it's got to be more restrained than that which has been proposed so far, and likely also more nuanced as to the particulars. Now, I grant that this is a tall order; I honestly hesitate at where to even begin. It's certainly one of those situations where I appreciate that there is WP:NORUSH. SnowRise let's rap 13:40, 18 May 2025 (UTC)
On a separate procedural note, if the below is being considered as a serious WP:PROPOSAL, then the the discussion should be moved to the talk page of the relevant polcy namespace, or else to WP:VPP. Or at a minimum, notices should be placed in both those spaces and relevant WikiProjects. Actually, I think this would be a substantially relevant enough change in the sourcing policy to warrant a listing as WP:CENT as well. SnowRise let's rap 13:46, 18 May 2025 (UTC)
I would note that I don't think the following "proposals" (or "suggested starting points", depending on how you want to label them) yet represent anything particularly set in stone. As to the discussion should be moved, the suggestion here is (and I think has always been) to add something to WP:RSP, so I'm under the impression that this is the relevant place for this discussion. – Michael Aurel (talk) 15:07, 18 May 2025 (UTC)
Oh, I see: that makes sense. Notices probably still would not hurt though, once discussion has moved on to well-refined proposals. SnowRise let's rap 16:46, 18 May 2025 (UTC)
Sure, that seems very sensible to me. – Michael Aurel (talk) 01:47, 19 May 2025 (UTC)

Just copy RSPSCRIPTURE?

The following is the text of WP:RSPSCRIPTURE (after correction of a typo).

Scriptural texts, like the Bible and the Quran, are primary sources only suitable for attributed, relevant quotes and in compliance with other Wikipedia content policies and guidelines. Content that interprets or summarizes scriptural passages or narratives should generally be cited to appropriate scholarly sources (for example, in the academic field of religious studies) and attributed when appropriate. Analysis of scriptural content by Wikipedia editors is prohibited by the Wikipedia policy regarding original research, and a 2020 discussion found no consensus on whether unsourced summaries of scriptural texts should be allowed under MOS:PLOTSOURCE.

Something here could be as simple as the "copy homework" meme. Changing a few words and omitting the (seemingly irrelevant for us) tail:

Ancient literary sources like Cicero or Plutarch are primary sources only suitable for attributed and relevant quotes in compliance with other Wikipedia content policies and guidelines. Content that interprets or summarizes those sources should generally be cited to appropriate scholarly sources (such as those in classical studies) and attributed when appropriate. Analysis of those primary sources by Wikipedia editors is prohibited by the Wikipedia policy regarding original research. See also WP:CLPRIM.

I also added a see also signal to CLPRIM. I used Livy and Plutarch because they are commonly cited in middle and late republican scholarly works. They need not be pilloried so; a non-history source example might be worth adding. Ifly6 (talk) 19:23, 15 May 2025 (UTC)

I think this makes sense. Maybe Cicero for another non-history example? And possibly Aristotle, I often see him cited for a lot of claims about prior philosophers that he can't really be trusted on. There's also Diogenes Laertius mentioned above, but I think we *shouldn't* mention him because I worry about WP:BEANS as he's not that well-known amongst people who don't already know better. Psychastes (talk) 19:36, 15 May 2025 (UTC)
I added them to the draft. I'm not 100pc about the long name list but if we want to reduce to two exactly I think I would prefer Cicero and Plutarch now (original was Livy and Plutarch). I also corrected what I think is a sentence structure error in the original text: only suitable for attributed, relevant quotes and in compliance with other Wikipedia content policies and guidelines → only suitable for attributed and relevant quotes in compliance with other Wikipedia content policies and guidelines. Ifly6 (talk) 19:48, 15 May 2025 (UTC)
Put the list in brackets to more clearly set it aside. Ifly6 (talk) 19:52, 15 May 2025 (UTC)
As proposed this suggests that classical authors can be quoted but not cited without quotation. In scriptural matters, precise wording is usually crucial, even in translation; in history it is not. If the reason we know of something is because it's reported by Cicero or Plutarch, it should be fine to cite them alongside secondary sources that verify or comment on the relevant material—though not necessarily each detail, provided the attribution is clear. For example,

According to Suetonius, Vespasian then moved his troops to a high point overlooking the city.<passage in Suetonius> Cassius Dio reports that he had two thousand cavalry patrolling the area.<Cassius Dio> He then received a messenger from Bolonius warning of a possible ambush in the pass, and sent scouts to investigate the situation.<Suetonius><Pauly-Wissowa><Blarghson>

No quotes; just a straightforward account of the facts and where they come from, without being too repetitive. The secondary sources could possibly be cited for each fact, but I'm not certain that's necessary. P Aculeius (talk) 22:54, 15 May 2025 (UTC)
The secondary sources could possibly be cited for each fact, but I'm not certain that's necessary. the problem here is the number of wikipedians who actually know when they could "get away" with not citing the secondary source or not is probably a few dozen or so at most, we can't make policy on that, and it's meaningless in a content dispute. There are numerous places all over the encyclopedia that cite ancient primary sources by themselves, yes, but there are also numerous places that have verifiable claims that aren't cited at all. In both cases I'm not likely to tag it if it's not a Good Article unless I actually doubt it that it's true or plausibly think someone else might challenge it. But for the purposes of "material likely to be challenged" the primary source is worthless. "Suetonius said X" okay , i can verify that in a primary source, but is it true? When I say "West says Hesiod lived and wrote before Homer" with a citation that doesn't just imply that you could go look in West's book and find the claim, it also says something about West, as a classical scholar, making plausibly reliable claims about reality. If you want to challenge that claim, you need to furnish another modern scholar who disputes West (of which there are many). But Suetonius cannot do this, no one should ever weigh Suetonius against Cicero or Suetonius against any modern scholar, and if any claims about reality made by Suetonius that are likely to be challenged (or, known to be wrong) ought to have a secondary citation. Psychastes (talk) 23:16, 15 May 2025 (UTC)
And yes, many biographers will say things like "Philosopher was born in the 85th Olympiad and their teacher was FamousGuy" and if you consult a modern source you learn that FamousGuy lived 100 years earlier and also never had any students. and Philosopher was born in the 78th olympiad. There is no claim trivial or mundane enough that an ancient source can't fumble, and no guarantee even that the greek-illiterate scribe who copied the sole surviving manuscript of their work won't mangle the truth beyond all recognition anyways. Psychastes (talk) 23:27, 15 May 2025 (UTC)
Aculeius' hypothetical passage could well really be this:

According to Suetonius, Vespasian [but actually it was Titus because this was in AD 69; the source just uses "Flavius" but the editor read it wrong] then moved his troops to a high point overlooking the city.<passage in Suetonius> [which geographers all agree doesn't exist because the only high point in the area is the city] Cassius Dio reports that he had two [but the now-accepted emendation is seven for MS reasons and because Joseph BJ and Tac Hist both say seven] thousand cavalry patrolling the area.<Cassius Dio> He then received a messenger from Bolonius [but now people think it is Epolonius rather than hapax Bolonius] warning of a possible ambush in the pass, and sent scouts [now rendered as "spies", almost like translations are themselves interpretations, because the consensus is now that Titus learnt of the ambush because Smartypants 1917 connected Suet to passages in Frontin Str and Joseph BJ] to investigate the situation.<Suetonius><Pauly-Wissowa 1909><Blarghson>

Sure, the possibility that this original research is so egregiously wrong is probably low. But beyond the OR, nowhere in this example did Aculeius engage with the possibility that using Suetonius and Dio as the main sources for this narrative might not be the appropriate weight for their testimony what some website's translation says is their testimony.
WP:PRIMARY is not just about whether something is verifiable (and, notably, verifiable ≠ truth). It also incorporates the non-negotiable WP:DUE, which requires presenting historical material in proportion to its acceptance. This cannot be judged by internet-translation-of-primary-source alone or even by flipping through an 181-year-old book. Ifly6 (talk) 00:54, 16 May 2025 (UTC)
Something along those lines would probably work. To critique some of the wording, I do have to say that the first sentence sounds a bit as though it's only ever acceptable to cite primary sources for quotes, whereas I think that most editors here agree that citing primary sources in addition to (ie., in the same ref tags as) modern, scholarly sources is acceptable, assuming that those primary sources are the ones cited by the modern source (parallel primary source citations, as it was called above). On the second sentence, I would probably protest grouping "interprets" and "summarizes" together. While there's perhaps some disagreement here over the degree to which summarising a primary source is acceptable (when only that primary source is cited), I think everyone is in agreement that any interpretation of a primary source must be cited to a secondary source. I would also note that I think appropriate scholarly sources (such as those in [...] ancient history) could sound as though it's referring to "scholarly sources" from ancient history (ie., ancient works written by scholars). I do also have some thoughts on the content of the proposed text (and, in particular, what might be worth adding), but I'll comment on that separately. – Michael Aurel (talk) 14:15, 16 May 2025 (UTC)
I omitted ancient history (such a distinction problem also recurs with Roman historian meaning both the modern one and the ancient ones who were themselves Roman). I also omitted authors not Cicero and Plutarch. I'm not exactly sure how to word exactly what I think the consensus is. Ifly6 (talk) 14:41, 16 May 2025 (UTC)
As an immediate set of thoughts, if I were writing just what I wanted it might be something like this:

Ancient literary sources like Cicero or Plutarch are primary sources. Such primary sources may be cited (1) only alongside a reliable secondary source also citing that source or (2) for attributed and relevant quotes. Content that interprets such primary sources must be cited to appropriate scholarly sources (such as those in classical studies). Analysis or synthesis of such primary sources by Wikipedia editors is prohibited by the Wikipedia policy regarding original research. See also WP:CLPRIM.

I would strengthen should generally be to must for interpretations: this I think reflects the consensus here. I omitted mention of summaries, but if we are to have two exceptions: (1) dependent parallel citations and (2) attributed quotes you can't write such a summary anyway. If we really want to keep the low quality borderline-OR content that is According to Livy, ...; according to Plutarch, ...; according to Dionysius of Halicarnassus, ...; etc we can add "non-synthesising summaries" (or even a whole sentence as to what kind of summaries are acceptable). I'm also not a fan of the density of this language but I think we do want to keep things short. Ifly6 (talk) 14:57, 16 May 2025 (UTC)
In general, that version looks better to me, and I think it solves the issues I had with the previous wording. And on strengthen "should generally be" to "must" for interpretations: this I think reflects the consensus here, yes, I'd agree. As to summaries (or "non-synthesising summaries") of ancient sources, I suppose it depends on how explicit and specific we want to be. I have to say I quite like CLPRIM's mention of "relevance and veracity" for why we should generally avoid citing primary sources on their own. It hints at relevant policies such as WP:DUE and WP:PROPORTION, and I think it reasonably allows for primary sources to be treated slightly differently in cases where their veracity isn't relevant.
For example, while I'd agree that ancient sources shouldn't usually be summarised without secondary sourcing in articles on history proper (or ever? I don't really edit such articles, so I wouldn't say I have strong opinions on the matter), for mythological articles I can vouch that it can be necessary to summarise ancient sources without secondary sourcing in some places. A simple example would be a scholar analysing a story or passage about a particular figure without summarising that story or passage first. Assuming there isn't another secondary source which provides such a summary, I think it's fine to, for example, summarise that story/passage in the first part of a paragraph (citing just the ancient source) and then include the scholar's discussion of it afterwards. That a scholar has chosen to analyse/interpret/etc. that story or passage in relation to the mythological figure should be sufficient to indicate its relevance, and we don't need to be concerned about the source's veracity. (All of this is of course still subject to the same points about transmission, attribution, DUE, etc., brought up above.) – Michael Aurel (talk) 16:32, 16 May 2025 (UTC)
One of the acceptable cases I wrote in to WP:CLPRIM#cite_note-38 was actual non-existence of reliable secondary sources. I think the only time this actually came up in my editing, which is mostly late republican political history, was with the Alexandrian war#Battle of the Nile. (Whether that is due or not wasn't something I considered at the time: it may be sufficient to say "Caesar won; Ptolemy died" as all the sources I checked do. The section aside, at lot of these articles on ancient battles should be deleted or moved to a higher-level article because there is so little coverage.) Do you think that is an acceptably restrictive standard? I worry that mere nonexistence would invite irrelevant cruft which nobody other than the transcribing editor cares about.
Hmm... perhaps a sentence such as Summaries of primary sources should be short, not synthesise any material, and be avoided: consider adding one only if familiarity with the summarised material is assumed in secondary sources.? Ifly6 (talk) 19:42, 16 May 2025 (UTC)
Hmm, yes, I'm inclined to agree that the non-existence of secondary sourcing on its own probably wouldn't be sufficient, and might end up encouraging the sort of thing we want to avoid (inclusion of irrelevant or UNDUE ancient sources). The additional requirement if familiarity with the summarised material is assumed in secondary sources might be quite good. The first example that came to mind where I'd summarised an ancient source without secondary sourcing was the first paragraph of Nyx#Early sources, though I seem to remember that was less about actual "non-existence", and more because summaries in secondary sources didn't contain enough detail for the story to make sense to the average Wikipedia reader (so there was probably still some "assumption" happening there).
I also wouldn't have an issue with essentially borrowing the aforementioned sentence from CLPRIM: In general, do not add statements based on ancient primary sources without a reliable secondary source establishing their relevance and veracity. (I've removed "citations", because I think(?) that was referring to parallel primary source citations, covered by (1) in your "immediate set of thoughts" proposed text.) This would also apply to summaries, though we could make this explicit if we wanted – doing this would give us something like:

Ancient literary sources like Cicero or Plutarch are primary sources. Such primary sources may be cited alongside a reliable secondary source which cites the ancient source, or used for attributed and relevant quotes. In general, do not add statements based on such primary sources (including summaries of those sources) without a reliable secondary source establishing their relevance and veracity. Content that interprets such primary sources must be cited to appropriate scholarly sources (such as those in classical studies). Analysis or synthesis of such primary sources by Wikipedia editors is prohibited by WP:NOR. See also WP:CLPRIM.

(Note that I've made a few other wording/concision changes.) We could perhaps be more explicit about when exceptions might apply (eg. "Because such primary sources can often be wrong, ...", or even "though in some contexts an ancient source's truthfulness is not relevant"), but probably this isn't necessary. – Michael Aurel (talk) 03:06, 17 May 2025 (UTC)
I think this is probably a better start than my adaptation of the existing text relating to scriptural sources. Could you start a new section with that text proposed so we could discuss it more cleanly? It might be that staying closer to the existing text on scripture would be more defensible in terms of a mos maiorum. On the other hand, the existing text on scripture may be more restrictive than what we want. But, as to that kind of framing strategy, I defer. Ifly6 (talk) 04:09, 17 May 2025 (UTC)
Great. I've done so. – Michael Aurel (talk) 06:03, 17 May 2025 (UTC)
  • Oppose. (1) This proposal implies that all ancient literary sources are equal. They are not. Some are more reliable than others. I am under the impression, for example, Polybius is relatively well regarded and, at the other extreme, Geoffrey of Monmoth (not ancient, but I need an exceptionally bad example) makes up total fantasy. (2) [It is questionable to characterise ancient sources written hundreds of years after the events they describe as "primary", as I pointed out before] (3) This proposal seems to imply that recent history books are more trustworthy than primary sources. This is not obviously always true. See, for example, the comments made by G O Sayles in The King's Parliament of England, in addition to those by Betty Radice I mentioned above. You seem to be telling me that, for example, if a recent history book tells me that King Edward's motto was "keep troth", I can cite that history book and the inscription on his tomb, even though Sayles warns me that the inscription was written two hundred years later and has nothing to do with Edward, that contemporary literary sources (which you would have me ignore) say Edward was in reality a cheat, and that the history book is lazily and ineptly copying from lots and lots of earlier history books that all contain the same mistake. (4) This proposal includes restrictions on the use of ancient sources that are not contained in WP:NOR, and which are in fact new and additional restrictions, not applied to contemporary primary or secondary sources. (5) WP:NOR should not be changed in this way without an RfC. (6) These restrictions are not used by historians or history books. If we are going to change policy, we should move towards the historical method used by historians, not away from it. James500 (talk) 19:54, 16 May 2025 (UTC)
    The historical method is, tautologically, WP:OR. Psychastes (talk) 19:58, 16 May 2025 (UTC)
    Deciding whether a particular source is reliable for a particular claim is not, in of itself, OR. James500 (talk) 20:16, 16 May 2025 (UTC)
    as the article you linked notes, the historian's skill lies in identifying these sources, evaluating their relative authority, and combining their testimony appropriately in order to construct an accurate and reliable picture of past events and environments. And we should be relying on reputable modern historians (i.e. peer-reviewed WP:SCHOLARSHIP) to make these determinations. not wikipedia editors. our task is to paraphrase and summarize the "accurate and reliable picture" not to construct it ourselves. Psychastes (talk) 20:39, 16 May 2025 (UTC)
    and really there is no fuzzy boundary here, it's quite clear-cut. Reading Polybius for yourself and evaluating whether or not you can "trust" him is, fundamentally, a different sort of activity from consulting book reviews from sites like Bryn Mawr to decide whether or not a new book on the 3rd century Roman Republic accurately represents the consensus of modern historians or not. Editors who do the former should either be coached on how to do the latter, or encouraged to find other ways to contribute to the encyclopedia. Psychastes (talk) 21:05, 16 May 2025 (UTC)
    Regarding "Editors who do the former should either be coached on how to do the latter, or encouraged to find other ways to contribute to the encyclopedia", I have already asked you once to stop making comments that sound like personal accusations that I have actually done something. If you do not stop this, I will refuse to interact with you. James500 (talk) 21:18, 16 May 2025 (UTC)
    I've rarely interacted with Psychastes and never interacted with you before. I'm unsure as to why this comment is necessary since I don't see any personal attack or insinuation thereof. Ifly6 (talk) 21:24, 16 May 2025 (UTC)
    Re ancient literary sources not being equal. Even in the example of Polybius there are substantial questions: the tide at New Carthage and the many criticisms of Polybius' description of the Roman army (anachronistic) and constitution (eg, as overly schematic, Mourtisen 2017). Intraset reliability is irrelevant anyway; WP:PRIMARY paints with a broad brush.
    Re bad modern history books: WP:FRUIT and ... the normal balancing of sources in content dispute resolution. Picking between Holland Rubicon and Gruen LGRR is an easy choice. Re historical method, operating as if we were historians writing history books free to compare Sall Cat and Plut Cat min is OR and not how Wikipedia works. Last, can you explain what these new and additional restrictions are rather than vaguely asserting their existence? Ifly6 (talk) 20:35, 16 May 2025 (UTC)
    As far as I can see WP:OR does not require that "primary sources may be cited (1) only alongside a reliable secondary source also citing that source or (2) for attributed and relevant quotes." That does appear to be a new restriction. I cannot see in WP:OR (a) any requirement for attribution or quotes (beyond what is required by the attribution guideline, which RSN and RSP are not concerned with); or (b) any requirement that a use of a primary source be backed by a secondary source, in cases where the use is not "analytic, evaluative, interpretive, or synthetic" etc. The proposal does seem to be changing the policy. James500 (talk) 20:59, 16 May 2025 (UTC)
    Would you object to WP:RSPSCRIPTURE on essentially the same grounds? If not, how are these different? Ifly6 (talk) 21:23, 16 May 2025 (UTC)
    I would say that RSPSCRIPTURE contains restrictions ((a) attributed quotes and (b) backing summaries with secondary sources) that are not found in WP:OR. In that case, however, there was an RfC in 2020, and RSP only claims that "analysis of scriptural content by Wikipedia editors" is actually forbidden by the policy. I would like to withdraw from this discussion now. James500 (talk) 21:36, 16 May 2025 (UTC)
    I want to respect your indication of withdrawal. At the same time, thank you for making clear your concerns about going beyond WP:PRIMARY. It does appear to me now that RSPSCRIPTURE's prescriptions are probably substantially more restrictive than PRIMARY, inasmuch as they are: (1) you can use, say, Matthew only for quotes and (2) essentially everything, including summaries, should be – though not must be – cited to secondary sources first. I think we would definitely want to be able to say that the 1915 Tuebner of Plut Vit says what it says. Ifly6 (talk) 04:18, 17 May 2025 (UTC)
    Historians and scholars of ancient sources never agree with each other nor do they agree with methodology. It often boils down to a scholar, their sources they select and their argument. I am thinking of Socratic problem as an example of how even scholars cannot determine historical facts between each other even when having decent sources available. Much less between wiki editors like us who tend to be not experts let alone published experts. WP:ATTRIBUTION seems to be the only middle ground solution. Ramos1990 (talk) 21:03, 16 May 2025 (UTC)
    ah, so historians sometimes disagree, I guess we should just make our own determinations for ourselves on the solutions to the socratic problem? might as well take the opinion of folks on the street, how does anyone know anything about history anyways, it's not like they were there themselves. or maybe if we're going to have an article on an actual academic dispute, we can cite modern scholarship on what the key arguments are? Psychastes (talk) 21:10, 16 May 2025 (UTC)
    I was in the middle adding to my response. Attribution to scholars and attribution to primary sources, if used, would be good. Ramos1990 (talk) 21:28, 16 May 2025 (UTC)
    okay yes, I agree. but please don't say Historians and scholars of ancient sources never agree with each other, it's rather disrespectful and trivializing. There's certainly disagreement about how much Plato/Xenophon/Aristophanes/Aristotle accurately represent Socrates' actual opinions, but it's generally agreed that he existed, lived in the 5th century BC, challenged the values of his day, was tried for impiety, etc. Psychastes (talk) 21:38, 16 May 2025 (UTC)
    What I meant was that they never agree on all or, in some cases, most details. If they did, then no more work for future historians. I was trying to bouce off of your comment in that historical studies are tautological. And that because of that it should be leaned to experts by appropriate weight, not non-experts like wikipedians. That is why I cited the scoratic problem in that even though you have pretty decent sources, there are divisions on who Soctares really was in scholarship. I never mentioned existence or other basic facts on Socrates. Basic stuff like you mentioned, is not really disputed as far as I can see. I think the parallel wording proposed intially by Ifly6 based on WP:RSPSCRIPTURE was ok as those require careful interpretatrions. But also I am thinking that most editors will never look to RSP for guidance on this. I never even knew of WP:RSPSCRIPTURE. Most look at WP:RS. Shouldn't this go there in some way under WP:Primary? I think it does address this in borad brush strokes there already, though.... Ramos1990 (talk) 23:36, 16 May 2025 (UTC)
    I quite agree that we need to base our work on "experts by appropriate weight" - applying WP:DUE and WP:SCHOLARSHIP, to put it another way. But let's not be snide about present-day historians always disagreeing for the sake of it or for personal gain. So much great work's been done in the last fifty years by asking new questions, applying new methods, and examining new evidence, and historians do move on from endlessly rehashing unknowables like the Great Rhetra. Disagreement's not new either: consider Dionysius on Thucydides, Thucydides on his predecessors, Plutarch on Herodotus, Polybius on Timaeus, Callisthenes, Demochares, Agathocles, Zeno, Antisthenes and more; the discrepancies between Livy's chronology and Varro's, between Suetonius and Tacitus; the detail lost in Livy's paraphrasing of Polybius. Indeed, part of our problem is that too often, only one account or one side has survived; Xenophon's reputation suffered so badly in the 20th century partly due to more critical reading but partly also because the discovery of the "Oxyrhyncus historian" indicated how selective, biased and even simply forgetful a memoir the Hellenica is. The application of WP:DUE and WP:NPOV requires the use of modern scholarship even more because of the sparseness of surviving ancient sources.
    So though we do already have footnote (d) to WP:PRIMARY Further examples of primary sources include ... medieval and ancient works, even if they cite earlier known or lost writings, I agree we should provide editors with clarity at WP:RSP too. NebY (talk) 14:49, 17 May 2025 (UTC)
    This isn't a change to WP:NOR; the inclusion of ancient works in footnote (d) goes back to 2011 at the latest. (We can also note that Herodotus et al were self-published, and so on.) It's entirely appropriate to provide clarity at WP:RSP too. Personally, I'd be open to an argument that we should describe them as "primary for Wikipedia's purposes" or some such, but that discussion should be at WT:NOR and may indeed have already been discussed there many times - there are not only 64 main archives but also 5 archives of a single long discussion Wikipedia talk:No original research/Primary v. secondary sources discussion. NebY (talk) 15:28, 17 May 2025 (UTC)
  • Comment: I carefully refrained from replying to what I consider an absurd take on my example above, which assumed that everything in the sources cited was wrong, but that no reputable secondary sources were cited; and that this was a reason why Greek and Roman writers should not be cited at all, unless every word is backed up (not uncontradicted, but positively affirmed) by secondary sources written in the last two or three decades. This would be untenable in the field of classics, and greatly restrict the ability of editors to write articles saying anything of value about all but a small fraction of topics. But I'm not going to waste time arguing it point by point, because I believe there's a fundamental flaw in the argument being made for at least some of the proposed language.
The language proposed prohibits the citation of any historical source for any purpose—not just analysis or opinion, but even just for a statement of what they contain—unless cited alongside a secondary source citing the specific text in question for whatever is being cited. This is a standard not generally applied anywhere else in Wikipedia, except perhaps in biographies of living persons. It sets the bar higher than for articles in general, and creates a new policy that differs from and is far stricter than the general rules about primary sources. It makes no exception for context or the reasons for which a source is being cited; and with this language it would encourage—even be seen to mandate—the deletion of relevant sources, simply because no secondary source addressing the particular point was included or found—something far more likely to be the case when only very recent secondary sources are considered, since the bulk of classical scholarship is older than that preferred by some of the editors crafting this policy.
This reminds me of the legal doctrine of "prior restraints" on speech: the notion that certain speech is so inherently wrong that it can be prohibited in advance, without the need to consider the circumstances or justifications that might be made for it. This doctrine has fallen considerably out of favour in American jurisprudence over the last half century. But here the proposal does something similar: it broadly prohibits the citation of relevant and important sources without regard to the circumstances or justifications, or whether they are in accord with mainstream scholarship, or whether very recent scholarship has anything to say about the particular point for which they are being cited.
That result runs counter to the basic process of verification on Wikipedia: statements may be challenged when they are controversial, removed if they cannot be verified; but editors are strongly encouraged to seek out sources that could verify or refute challenged statements and cite them. When an article lacks sufficient sources, or needs better ones than those it has, the remedy is rarely to delete it or the unsourced contents. The remedy is nearly always to find sources that address the points being made, and revise the article accordingly, adding the missing or replacing the incorrect material. And as AfD makes clear, there are no deadlines for improving articles. So if the only source cited for a claim is say, Livy, the citation to Livy should not be deleted simply because it lacks recent corroboration.
If a source addressing the point can be found, then by all means, add it. If no source can be found verifying or refuting what Livy says, then the fact that the claim is made should still be cited to Livy if it is relevant to the article. But the statement should not be deleted merely because no recent authority has yet been cited alongside it, or because the authority cited for it is from older scholarship, or because nothing is found addressing the point for which it is cited. That would be consistent with general Wikipedia policy: primary sources—in the non-historical sense—are reliable for their own contents, provided that no attempt is made to analyze the text; that is exclusively the province of secondary sources; and that the source is not cited misleadingly, or presented as inarguable fact in the sense that Wikipedia is somehow endorsing anything beyond the fact that the source says a particular thing.
With respect to the argument that citing historical writings is itself a form of synthesis or analysis: that is a flawed argument, because it applies equally to modern sources. All articles on Wikipedia are cobbled together using various sources to discuss different aspects of a topic; that is not what is meant by "synthesis", and properly cited facts presented as nothing more than what the sources have to say are not analysis. We should always look for secondary sources to provide or explain what the original sources for anything say; but we should not seek out and destroy citations to those sources because a corroborating source is absent. Rather, find secondary sources that address the point and add them. And there should be a better reason to challenge and remove cited content than that the source is old, or primary—in the historical sense. P Aculeius (talk) 03:15, 17 May 2025 (UTC)
I can agree with some of this. WP:Attribution esentially places wieght on the source (ancient and/or modern scholarly) instead of putting something in the wikivoice. There are tags to place where secondary sources may be needed or preferred. And again sometimes primary sources like from Loeb Classical Library have footnotes from modern scholars as you read the classical texts - thus providing modern interprtation as you read the text - and that modern footnote can be mentioned in wikipedia. Ramos1990 (talk) 06:01, 17 May 2025 (UTC)

Proposed second version

Using Ifly's above adaption of RSPSCRIPTURE's text as a starting point, we've discussed a few parts which seemed as though they didn't quite work for our purposes, and I think we agree that the following passage could represent a step forward:

Ancient literary sources like Cicero or Plutarch are primary sources. Such primary sources may be cited alongside a reliable secondary source which cites the ancient source, or used for attributed and relevant quotes. In general, do not add statements based on such primary sources (including summaries of those sources) without a reliable secondary source establishing their relevance and veracity. Analysis or synthesis of such primary sources by Wikipedia editors is prohibited by WP:NOR. Content that interprets such primary sources must be cited to appropriate scholarly sources (such as those in classical studies). See also WP:CLPRIM.

See that discussion for the full reasoning behind these changes. To briefly summarise, the old wording at least sounded somewhat as though primary sources could only be cited when they were being quoted (whereas I think there's agreement here that parallel primary source citations, for example, are acceptable), and we've avoided mentioning "interprets" and "summarises" alongside one another (as these aren't really the same thing, at least in this context). This has led to an explicit acknowledgement that using parallel primary source citations is acceptable (may be cited alongside ...), and the addition of a further sentence (In general, do not add ...) – based mostly upon a passage from WP:CLPRIM's lead – which touches on summarising ancient sources. There have also been general changes to the previous passage's wording. – Michael Aurel (talk) 05:50, 17 May 2025 (UTC)

I dig this adjusted version. Seems nuanced, which is how these things should be. Ramos1990 (talk) 05:54, 17 May 2025 (UTC)
Perhaps move Analysis or synthesis of such primary sources by Wikipedia editors is prohibited by WP:NOR. to before Content that interprets...? Ifly6 (talk) 06:08, 17 May 2025 (UTC)
Sure, done. – Michael Aurel (talk) 11:32, 17 May 2025 (UTC)
This is still problematic for multiple reasons. The premise remains that every detail that can be cited to the original source must be duplicated, discussed, and confirmed or refuted by a secondary source, or it may be subject to deletion without further inquiry. This is backwards: if a statement in a primary source is not clearly corroborated by a secondary source, then an attempt to find such a source should be made; and for some statements we may not find any direct comment in recent scholarship: that alone should not result in deletion.
Further, the relevance of a claim made by any source should generally be self-evident; while we may and ideally would cite Jones alongside Suetonius in discussing the murder of Caesar, we would not ordinarily expect or require Jones to state that Suetonius' description of the murder is relevant to it, nor should the inclusion of Suetonius depend on whether Jones agrees or disagrees with his account.
The inclusion of "summaries" is vague and unhelpful: if Suetonius spends two paragraphs discussing the attack and mentioning each significant individual and their part in it, is the wording "Suetonius mentions several of the assassins, including Brutus, whom he says struck the fatal blow" a prohibited "summary"? That result would be absurd; it is a plain description of what the source says, and is in no way analysis of it. Again, a secondary source may be useful, desirable; the lack of one—or perhaps the placement of the citation to a secondary source that describes the same event later in the same paragraph—should not lead to the conclusion that the citation to Suetonius should be excised as a prohibited use of a primary source!
That analysis or synthesis of sources is not permitted as original research is a restatement of a general Wikipedia policy, and in essence says the same thing as the following sentence. Including the "see also" pointing to that essay implies that the essay (written by one of the participants here, whose opinions are much stronger and less accepting of the use of various sources than this discussion suggests we should be) implies that that essay is being incorporated into the policy or elevated to policy status, which should not be the case. P Aculeius (talk) 11:55, 17 May 2025 (UTC)
This seems to assume a working method of starting with a primary source, then checking it against secondary sources. Isn't that quite contrary to the fundamental idea that Wikipedia should be based on secondary sources? It's certainly not an approach that we want to encourage in new editors, even if we have a long-standing (in Wikipedia years) legacy of articles developed that way. NebY (talk) 14:58, 17 May 2025 (UTC)
This gets to the heart of the matter. Aculeius' claim of harm (or, in my mind, really a mere inconvenience) is essentially only realised if an editor started writing an article by paraphrasing Suetonius or some other source. I think the consensus of participants here is that we don't want someone to do that. That's all: the spectre of "purging" older secondary sources like they're the four olds and trampling over editors' First Amendment rights is an impressive exercise in alarmism. Ifly6 (talk) 19:45, 17 May 2025 (UTC)
That's not the situation I'm describing: what I'm saying is that a missing or deficient secondary source for some particular detail should not result in the deletion of the fact cited to a primary source, and as currently written the proposal seems to recommend that instead of attempting to cure the deficiency by locating an appropriate secondary source. Let's suppose that a paragraph about a battle was based on "Williams, 1917", citing Livy and Dionysius of Halicarnassus, and that the article is written with the specific details from Livy and Dionysius cited to the passages in each, as both mention certain things but each also mentions details that the other does not. Williams describes the account as given by Livy and Dionysius, whom he mentions, but he says nothing specifically about whether either is "relevant" or "truthful". Does this mean that the relevance and veracity of the passages has not been established, and therefore the citations should be deleted?
At some point the description of the battle is expanded, and the citation to Williams is no longer obviously connected to the details cited to Livy and Dionysius. Or somebody comes along and decides that Williams is too old a source, or disagrees with Williams' expertise for reasons having nothing to do with the account of this battle, and replaces him with some other secondary source that may not mention all of the details in Livy or Dionysius, though it does not contradict any of them, or at least does not contradict how they are presented in the article. The proposed language now seems to say that whatever is cited to them must be deleted to whatever extent the secondary source is either disconnected from the text it formerly supported, or the details are no longer corroborated by a secondary source, even though no secondary source disputes them either.
To repeat: I'm not recommending building articles beginning with primary sources and only then finding secondary sources. But the remedy for inadequate sourcing is to find more sources. I'll also clarify that my objection has nothing to do with the First Amendment; it has to do with the notion of placing preconditions on all citations to historical sources that could be used as justification to delete them, without considering what they've been cited for, how they're cited, or whether they can be corroborated using secondary sources. P Aculeius (talk) 20:11, 17 May 2025 (UTC)
I'm all for citations to classical and primary sources in general, while they may not always satisfy editors in checking the WP:V box for the content i think they are of significant benefit to readers. But also i don't think we should downplay the difficulties here. Take for example Hemiauchenia's example article Ptolemy IV Philopator and the claim: so Magas was scalded to death in his bath.
This is easily cited under the PAG's with for instance JSTOR 41289743 p. 243:

...he had his mother poisoned and his brother Magas scalded in his bath (Polyb. 15.25.2).

This would probably satisfy editors looking for "secondary" sources to satisfy V, but it is a pretty useless citation for the reader for a few reasons. When i read a passages such as that outside of WP which is cited to a classical source i read into it an implied attribution. According to Polybius' Histories Magas was scalded to death in his bath. But by asserting that content to a WP article and providing such a citation i think naive editors and readers are given an unwarranted assurance that this is a "fact".
Polybius 15.25.2 does not say anything about scalding in the bath (an unlikely sounding assassination to modern readers). Near as i can tell the story is Magas had cauldrons of boiling water poured over him and this is known merely from two fragments:
Well, that seems like pretty limited information to assert something in a WP, especially considering Phylarchus might have been something of a sensationalist writer, Polybius and Plutarch both accusing him of carelessness and inaccuracy, unnecessary description of horrors, even insustainable, more suited to the dramatist than the historian to cause violent sensations in the reader.fn 10
I think creating a rule for everything here which naively satisfies WP:V will generate some pretty ugly looking articles: lots of "according to X..." and a great deal of over-citing. This doesn't really benefit the reader at all. But how would you word this bit of content and how would you cite it? fiveby(zero) 16:05, 17 May 2025 (UTC)
That's not ideal, no. But the story remains relevant even if it's not strongly supported; and if a secondary source mentions it, it should be included as something that some sources report, and cited to them, albeit with whatever the secondary source has to say about whether it is likely to have occurred, and whether other sources agree with it or fail to mention it at all. And any source that is mentioned in the article ought to be cited.
How would you word this content? Let's assume that a secondary source (Fooson) mentions all of what you state: he notes the anecdote from Phylarchus and the unknown proverbs, but states that it isn't mentioned by Polybius, and that Polybius and Plutarch describes Phylarchus as sensationalistic and frequently inaccurate.
Say all of that: "Phylarchus and a source of unknown authorship mention an anecdote that Magas was scalded to death by having cauldrons of boiling water poured over him.<Phylarchus, Proverbiis> This detail is not mentioned by any other sources, including Polybius, who along with Plutarch describes Phylarchus' accounts as sensationalistic and frequenly inaccurate.<Fooson>" P Aculeius (talk) 20:29, 17 May 2025 (UTC)
Sounds like a lot of work to get there, i think diligent editors are probably willing to do that—and know how to do that. As a reader i wouldn't like to see such editors hamstrung by a rule which requires source shrubbery, useless citations to "secondary" sources such as that JSTOR article which only mention the subject in passing.
On the other hand look at the citations provided: #9 Polybius 15.25.2; Plutarch, Life of Cleomenes 33; Pseudo-Plutarch Proverb. Alexandr. 13 is obviously simply cut-and-pasted from #4. This doesn't show a great deal of diligence. I see that quite a bit, references to classical works just copied from another source w/o examining them or attempting to clean up the citations. These are the type of editors that need some rules to help them out.
I would take a different tack here: the "Legacy and reception" section already has some discussion of the accounts of Polybius and Phylarchus. Move that up front to alert the novice reader as to the basis and trustworthiness of the sources used to reconstruct the life of Ptolemy. If that is done i don't think it really necessary to say "according to Polybius" or always include additional citations for all the statements in the article. fiveby(zero) 16:34, 18 May 2025 (UTC)
Passing mentions are usually not helpful to readers, though one might make the case to keep one if it is the only corroborating material one can find. That wouldn't be a good result, but it might be better than deleting it! I don't think that anyone here is advocating that sources that are repetitive or cumulative need to be added; but two varying accounts in different historians (or the fact that a sensationalistic one describes something that a more conscientious one omits) may be relevant.
Presumably only one secondary source is required to support the discussion, if it covers the subject well, but two or three that do so in differing ways or with different authority might be cited together. Additional sources that merely allude to it but add nothing of significance can be safely omitted, or pruned if the citations are already cumulative (I note that one way of dealing with the appearance of excessive numbers of inline citations is to group the sources, which there are various ways of doing).
I am not particularly worried about copypasta citations, since they are easily checked and pruned if the sources are available online. For Polybius or Plutarch that should be fairly straightforward; locating English translations of the letters of Cicero or the complete works of Valerius Maximus can be trickier, but once someone is able to do so—or read the citated material in Latin or another language they have been translated into—then they can also be cleaned up.
I don't see any objection to a "legacy and reception" section or something similar discussing the trustworthiness of various sources. My concern in this discussion has been about creating a policy that actively discourages citations to historical sources, even when they might be desirable, or which leads to existing citations being deleted irrespective of their usefulness, or the context in which they're used, based entirely on whether they're "primary" in the historical sense, or potentially dated among modern sources. P Aculeius (talk) 18:26, 18 May 2025 (UTC)

Suggested rewrite

After leaving the matter for the night and turning over a few ideas in my mind, I decided to try writing something from scratch that I hope will address the main concerns here. I've tried to avoid proscriptive language ("do not"), and incorporating or elevating essays into policy, but I have tried to make clear that modern secondary sources are expected whenever primary sources are cited. I've also included a cautionary note concerning older modern historical writing. Here's what I've come up with:

In a historical context, primary sources include the works of ancient, medieval, and other pre-modern writers, even when they were not participants in or witnesses to the events they were chronicling. They are described as "primary" because their writings form the body of texts from which modern historians write about the past. These writings can, and in many cases should be cited alongside modern secondary sources that discuss the same topics or events. However, editors must take care not to present the claims of historical sources uncritically. Statements made by these sources should be clearly attributed to them, rather than presented as objective fact. Whenever possible, authoritative scholarship from the modern era should address the same subjects.

As with all other articles, editors should not attempt to analyze or comment on the topic (original research), nor should articles include novel theories or conclusions not clearly stated in and attributed to the cited sources (synthesis). Editors should also remember that modern scholarship is the product of continuing research and development. In some cases, recent sources will incorporate new discoveries that were unavailable to earlier generations of historians; and opinions concerning matters such as politics, social conditions, race, sex, or religion have changed over time. As a result, older modern writing, while still acceptable to cite for many purposes, may not always be the best choice; or in the case of evolving views, it should be cited alongside more recent works.

I'm aware that to some participants in this discussion, this may seem too wordy or too permissive. But because all articles are still subject to the basic editorial standards of Wikipedia regarding verifiability and neutrality, I think we should avoid a tone that is too proscriptive, and which would encourage the overly-aggressive pruning of sources. P Aculeius (talk) 14:49, 18 May 2025 (UTC)

Statements made by these sources should be clearly attributed to them. Would that always require attribution inside a sentence of material to its source, in addition to an inline citation. Take for instance: Battle of Thermopylae#Prelude

Leonidas took with him the 300 men of the royal bodyguard, the Hippeis. [Herodotus VII, 205]

Doesn't the citation itself clearly attribute the statement to Herodotus? Must this sentence begin with "According to Herodotus..."? fiveby(zero) 17:36, 18 May 2025 (UTC)
I don't think it's always necessary to name the authority in addition to citing it, but it may be desirable to do so at or near the beginning of a passage that cites the source, especially if it's being cited multiple times. The important things about attribution are (1) that the reader is made aware of the source, and (2) the reader isn't misled into thinking that Wikipedia is endorsing something about which there could be some doubt. That won't necessarily mean that the words "according to ..." or similar words are always required, or that they have to come at the beginning of a sentence. Context will help determine the best way to attribute something, and I expect that editors will sometimes disagree on whether an attribution is explicit enough, but that kind of debate is normal in Wikipedia. P Aculeius (talk) 18:06, 18 May 2025 (UTC)
I'd say the leading "Sources" section along with the citation to Herodotus would be sufficient here, though could always be improved. But whatever you add here as a policy or guideline needs must serve a variety of readers and control a horde of editors all with differing levels of competence. Just thinking about those editors who will take this as a "rule" and that there always must be in-text attribution, or those which take it to mean anything goes for inclusion as long as it is attributed. More often i see the debates which take more of a facile and overly literal view of the P&G's rather than those of reasonable editors trying to exercise good editorial judgment. fiveby(zero) 18:33, 18 May 2025 (UTC)
That's literally what I'm trying to avoid here: I see the other proposals as too proscriptively worded. I suppose if you think it should be even looser, maybe I've hit the sweet spot in the middle. But I'm under no illusion that my suggestion will be universally lauded—if anyone thinks it's good, or that it just needs a few tweaks, I should probably be happy. P Aculeius (talk) 21:05, 18 May 2025 (UTC)
I think this describes what we do now, when we are doing good work, and therefore I like it. The earlier proposal seemed intended to change how we work generally and I don't think that's warranted. Furius (talk) 16:35, 19 May 2025 (UTC)
One of the core issues here is whether such a thing is or isn't good work. There are some WP:GA-rated articles which are essentially paraphrases of Livy, Plutarch, or Appian. A lot of editors, myself included, are of the view that this is not good work and have expressed this regularly in good article reviews. Simple questions like "what year did Veii fall?" are not really answerable with Livy alone. (According to Livy... 396 BC would be an anachronistic fiction probably inserted by editors with overconfidence in the number of the edition before them; the "correct" In the year of the consular tribunate of Lucius Titinius, Publius Licinius, Publius Maelius, Quintus Manlius, Gnaeus Genucius, and Lucius Atilius, per Livy is not really helpful for readers.) I think serious editors would agree that "good work" under such circumstances would require reading the secondary literature on chronological problems and a modern determination of absolute BC.
Among other examples, what that means for this discussion is that there are places where editors cannot trust what appears before them. This is why some editors want primary sources to be used in WikiVoice™ with secondary sources only. In WikiVoice is a distinction here which I think is somewhat confused. WP:PRIMARY says A primary source may be used on Wikipedia only to make (1) straightforward, descriptive statements of facts that (2) can be verified by any educated person with access to the primary source but (3) without further, specialized knowledge.
  • An unattributed claim in WikiVoice that Veii fell in 396, cited to Livy, would fall on its verifiability: in this instance it's certainly straightforward but does not meet either of the latter since Livy can't use a dating system that didn't exist yet.
  • An unattributed claim in WikiVoice that Leonidas brought just 300 men (from above) falls on whether Hdt said it and whether (when he said it) it is reliable. The troop numbers that Hdt reports are, frankly, not reliable. Both number and precision indicate against it. And in just this case, it is the consensus that Hdt is deliberately suppressing the existence of the c. 700 Lacedaemonian perioikoi.
  • An attributed claim, which would put the claim outside WikiVoice, that Herodotus says that Leonidas brought just 300 men falls just on whether Hdt said it. (Setting aside the fact that Hdt later implies that there were 1000 Lacedaemonians in the epitaph. Roel Konijnendijk has done some excellent public history on this topic.)
These are issues that most commonly, almost only, crop up when editors write articles by paraphrasing primary sources (or bleating their uncritical WP:FRUIT) with a careless disregard towards modern scholarship. fiveby raises a reasonable concern above that anything goes for inclusion as long as it is attributed: I think that's a question relating to WP:NOR rather than our here reliable sources noticeboard. But I doubt anyone wants that would stretch PRIMARY into a ban on the claim that the XYZ edition of ABC says what it says.
Aculeius' suggestion seemingly differs from Michael Aurel's in only minor ways. The core is Statements made by these sources should be clearly attributed to them, rather than presented as objective fact. Whenever possible, authoritative scholarship from the modern era should address the same subjects. This seems to direct essentially the same things as Michael Aurel's: (1) no WikiVoice claims to primary sources; (2) no primary-source-only passages. Two substantive notes, however: a note on the primary sources is not the place to address the question of obsolete sources; if it is intended that this wording would create only optional recommendations, that would be in contravention of WP:PRIMARY's "may be used on Wikipedia only" wording. Ifly6 (talk) 20:45, 19 May 2025 (UTC)
By this logic, an article about Leonidas or Thermopylae could not cite Herodotus because modern scholarship concludes that his description is inaccurate, and Livy could not be cited for when Veii fell, even if the article cites secondary sources clarifying what parts of Herodotus' account are unreliable and why, or why the chronology employed by Livy—at least how we render it in modern terms—is probably off by about four years. That would be an absurd result; citing a source and reporting what it says is not the same as endorsing it uncritically. No reputable history would cover these events without citing Herodotus or Livy; if that's how our policies are to be interpreted, then we should give up the pretense of being an encyclopedia.
The wording I proposed is not intended to suggest that editors are free to disregard the need for secondary sources; but it is intended to prevent the purging of sources already in articles simply because secondary sources are deficient or absent (or as suggested above, because primary sources contain biases or inaccuracies that could be and perhaps already are addressed by secondary sources); elsewhere on Wikipedia the preferred remedy for inadequate or missing sources is to locate and cite them. The remedy for a source incorrectly being presented as authoritative when it is not, is to edit the article to change how it is presented, and cite reliable sources explaining why it's unreliable in this respect or that.
Lastly, with regard to the contention that sources should be avoided or deleted because they are "obsolete": the contention that sources become "obsolete" when they reach a certain age, irrespective of what they say about a topic or whether it is accurate, or that it is "obsolete" because its author held views that are no longer deemed acceptable about matters that don't affect the thing for which the source is being cited, or because the author was unaware of some detail that also doesn't affect the point for which the source is cited, should be discarded altogether.
Classical scholarship has advanced considerably since the days of Gibbon, Niebuhr, or Mommsen, but that doesn't mean they should automatically be regarded as uncitable, irrespective of the purpose for which they're being cited. If a specific point for which sources such as these have been cited is called into question, or shown to be erroneous, then add or replace the source with another; if it's still accurate, or the veracity of the opinion remains an open question, then the citation shouldn't be removed just because the source is old. P Aculeius (talk) 23:44, 19 May 2025 (UTC)
I have no idea from where you've conjured the idea that Herodotus cannot be cited under, seemingly, any circumstances. A notice on the reliability of Herodotus and his bookshelf companions is not the place to discuss the reliability of obsolete sources. Ifly6 (talk) 00:12, 20 May 2025 (UTC)
All of the cited cases are dealt with by WP:Primary without a further policy. The facts you mention are not in fact straightforward. Good article reviews are already catching articles (new or old) that are misusing primary sources. Everything's working. So, why is new verbiage required? Furius (talk) 09:34, 21 May 2025 (UTC)
I'm not convinced that it is, but if there is a rewrite, then I want to make certain that it doesn't imply that primary sources (in the historical sense) can't be cited appropriately, or that they should be removed either automatically, or because the secondary sourcing in an article is inadequate. I also read some of the discussion as taking issue with the use of modern sources that aren't very recent (i.e. 19th, 20th century scholarship), but which may be perfectly accurate with respect to the basic facts for which they're being cited, and the implication that citing to them might also be prohibited or strongly discouraged without regard to context. But it sounds to me like the present wording is satisfactory and no new restrictions are imminent, and if so then I'm happy to leave things as they are. P Aculeius (talk) 13:04, 21 May 2025 (UTC)
Unlike what has been implied by some editors, nobody is changing, adducing, or furthering policy. WP:RSPIS:
  • checkY a list of sources whose general suitability for most purposes has been discussed repeatedly
  • checkY a very brief and simple summary of the consensus found in such discussions
This is meant to summarise the consensus as to the general suitability for most purposes of ancient primary sources. That is intended so that editors have somewhere to point to when other editors demand proof of consensus for such things as "Herodotus isn't a primary source" and "You can't use Herodotus to say only 300 Spartans went". If your contention is you can back the conclusions that are reached here out of WP:PRIMARY so there's no need to state the conclusions, I think that's more a challenge against RSP's existence than it is relevant to this discussion. Ifly6 (talk) 15:21, 21 May 2025 (UTC)
Very few articles are ever assessed as Good Articles. Wikipedia has and needs a constant influx of new editors; they need accessible guidance, not text buried in footnote [d]. Even editors who have been around a while venture into fresh areas of the encyclopedia without knowing everything already; they need accessible guidance, not text buried in footnote [d]. My watchlist is sad and depressing; I want accessible guidance I can cite, not a footnote. This is not the best of all possible worlds; not everything works and the encyclopedia that everyone can edit will always be in flux. NebY (talk) 09:52, 22 May 2025 (UTC)

Asia

Would this apply to works like the Twenty-Four Histories and Veritable Records of the Joseon Dynasty as well? I admit I have not read all of the arguments and counter-arguments presented above, but I have mixed feelings on disallowing the use of these sources for anything but directly attributed quotations. For one, they are pretty clearly secondary sources and, despite having clear biases, were often compiled with fairly rigorous scholarly standards. For many basic facts, they are (in my view) reliable sources. Also, a time-period based definition like "classical" or "pre-modern" might not adequately include these, if that is the intention. Toadspike [Talk] 15:14, 19 May 2025 (UTC)

I don't know enough about these histories, or the nature of historiography in China and Korea to give a definitive answer, but given that all but the last of the Twenty-Four Histories predate the 18th century, which I would say is on the cusp of modern historical writing—at least in the European tradition—I expect that the entirety would probably be considered a "primary source" for Wikipedia purposes, along with the majority of the Veritable Records, most of which (if I understand correctly) significantly predate the last portion, from 1865. I just don't feel competent to judge the reliability of Chinese or Korean historical writing from the 18th and 19th century, so with caution I would still probably treat them as "primary sources" in the historical sense.
That said, I don't believe that the policy as written limits the use of such sources to quotations. As long as editors make clear what the source is, and it's not being accepted uncritically—that is, spoken in the voice of Wikipedia, as if the article simply adopted whatever they said—it should be fine to cite them for what they say. And they should still be backed up by secondary sources that describe the same topics, and include context or any commentary or analysis of those histories, indicating things such as where they might be inaccurate or incomplete, or what might have influenced the tone of their writing. P Aculeius (talk) 17:53, 19 May 2025 (UTC)
I agree that for Wikipedia's purposes those are also probably primary sources. Re their reliability, I have a bad feeling in my gut, but that is based only on generalisations of pre-modern histories' loose treatment of numbers, melodramatic details, and teleologies. Between the possibility of differing interpretations of what "basic facts" are and my complete ignorance of their transmission, I wouldn't want to comment. Ifly6 (talk) 20:51, 19 May 2025 (UTC)
I think I agree in principle with both of you, P Aculeius and Ifly6. The Veritable Records of Joseon and other dynasties are primary, being compiled by their own governments. Most of the 24 Histories, dealing with the regime immediately preceding the author's, could be considered primary too. However, some are technically not primary. Take, for example, the Shiji, which does (in part) cover the author's own dynasty and the dynasty immediately preceding it, for which it could be considered primary or at least potentially biased. But it is hard to imagine any reasonable definition of "primary" that applies to the rest of the work, which cover events hundreds or thousands of years before the author's time and is based on primary or secondary sources itself (see Shiji#Source_materials). This puts it far outside the bounds of WP:PRIMARY as written. If the intent is that it should only be used for attributed quotes (which I think I'd agree with), we're gonna need a new guideline. Toadspike [Talk] 05:28, 20 May 2025 (UTC)
In terms of Wikipedia, I think such sources would be considered primary sources: Further examples of primary sources include: ... medieval and ancient works, even if they cite earlier known or lost writings. Such a thing is not unprecedented: one of the "primary sources" for the return of the Heracleidae is Thucydides, writing maybe 700 years after the events are said to have taken place. (Notably, almost nobody believes this "return" or the associated Dorian invasion are real.) Separately, I don't think anyone is seeking to restrict usage of primary sources to attributed quotes: the most maximally restrictive position seriously advocated is merely requiring people to cite the secondary literature when citing primary sources in WikiVoice™. Ifly6 (talk) 05:42, 20 May 2025 (UTC)
This was a question I was eventually going to ask on Asian, indigenous, Mayan, Aztec sources etc. Many of the sources from archaeology globally are also primary sources. Would the rewrite account fo all of these? I think that the rewrite, if to be used at all should be very generic and universal across cultures and time. Ramos1990 (talk) 06:10, 20 May 2025 (UTC)
I don't think anyone is suggesting we rewrite WP:PRIMARY. Just in the same way WP:RSPSCRIPTURE concerns itself only with scriptural texts, rather than every archaeological report to come out of the Levant, there's no reason to suppose that a notice on classical sources like Cicero or Plutarch also needs to scope-creep into every archaeological report on the classical world (let alone globally). While Aculeius' version does drop the examples that suggest a limitation to those ancient primary sources commonly cited in the field of classical studies, I don't read into his comments above any intent to expand the scope in such a manner. But, of course, for final clarification you'd have to wait for his reply. Ifly6 (talk) 12:38, 20 May 2025 (UTC)
Similarly, Wikipedia:Reliable sources/Perennial sources#What this page is not. Ifly6 (talk) 12:42, 20 May 2025 (UTC)
In my opinion, all historical sources should receive a similar treatment, in the sense that historians refer to them as "primary sources", and in that they were not subject to the same standards, expectations, or approach that modern history is. But because they are the foundation of modern historical writing on the topics that they cover—and often the only substantial sources for the details of history—citing them, in the sense of naming them and providing references that readers can use to locate relevant passages, is essential to articles about the topics they describe.
As far as quotations go, I don't read anything in Wikipedia policy requiring passages to be quoted word-for-word, as long as what's cited to them accurately represents what they say. Obviously some passages will be ambiguous, and others may be cited erroneously. These are things that can be addressed by finding discussions in secondary sources (which should be cited anyway, even if there's no obvious ambiguity), or through the normal editing process.
I tend to quote sources when the language is relatively brief, strongly or colourfully worded, or particularly lucid and memorable (bearing in mind that sometimes this is at least partly the work of a translator, which may also be subject to criticism). But in general, a summary or paraphrase that conveys the meaning of the original text with reasonable fidelity is consistent with Wikipedia policy and should be fine. P Aculeius (talk) 13:10, 20 May 2025 (UTC)
From the linked page: Sima Qian was a methodical, skeptical historian well, that certainly needs some attention! I would go as far as to say the Shiji is an *unreliable* source. Many of the biographies are basically folk legends, and there's supernatural stuff reported as fact in there, which to my mind is an automatic disqualifier for anything other than bare attribution. Also for a more sober note see this lower down on the page: It was commonly maintained that Ssuma Chhien [Sima Qian] could not have adequate historical materials for his account of what had happened more than a thousand years earlier
Outside of East Asia and the Mediterranean I'm not aware of any other places where this would apply though. There aren't surviving contemporary histories of India for most of its history, for example. Similar with Maya, Aztec, other indigenous sources; we're generally dealing with oral traditions, not texts. Psychastes (talk) 01:06, 21 May 2025 (UTC)
  • Not on the basis of their post-medieval date. So far as these sources were written after 1500 (and their WP articles say that at least some parts of them were), they are not medieval, let alone ancient, within the meaning of the policy. RSN cannot rewrite NOR. James500 (talk) 02:19, 21 May 2025 (UTC)
    yeah, agreed. the post-1500 ones don't apply here at all. Given that they're official imperial histories and also not academic scholarship from the past century, that's going to be a WP:NPOV judgment. Psychastes (talk) 03:13, 21 May 2025 (UTC)
    In fact, there was a huge scandal in the academic history world just recently about using the Twenty-Four Histories uncritically, that largely revolved around treating the Qing imperial histories (as in, the 1600s-1900s) as fact. So I would say "these are not reliable sources" and "this is totally unrelated to the issue of using classical sources" Psychastes (talk) 03:23, 21 May 2025 (UTC)
    It largely revolved around, amongst other things, cherry picking and misrepresenting sources, as you are cherry picking Qiao's article. James500 (talk) 16:14, 21 May 2025 (UTC)
    I agree that histories written prior to the adoption of modern historiographic standards aren't reliable sources for their claims, but I would still maintain that they can be cited for the fact that they make those claims. To borrow from the example above, without any additional background: "Much of what is known of this period comes from the accounts of Sima Quian. Sima's writing contains numerous instances of folk legends and supernatural occurrences, and modern historians are skeptical that he would have had access to sufficient records to compile an accurate account of events occurring a thousand years earlier.<Smith 2011>"
    As for New World records, it's true that we don't have indigenous narrative histories of whole nations or empires. But there are fragmentary narratives from codices and carvings in native languages—at least, that's my understanding as it relates to the Mayan civilization. Their contents should be citable, provided that it's clear that what's being reported is just what a particular source says, and that these accounts aren't necessarily accurate.
    That's why we want secondary sources: to provide context for and analysis of what surviving writings say. On the one hand, there's not much reason to doubt "here lies buried the Lord of Z, who died on the 15th day of the Month of the Jaguar in the 24th year of his reign." But context would be needed for claims about his rule, deeds, and legacy, irrespective of what was written about him during or shortly after his life. That context requires a source with more distance and modern analytical methods. P Aculeius (talk) 04:05, 21 May 2025 (UTC)
To address the specific question asked (Would this apply to works ...), that's essentially our choice, but I would prefer we restricted ourselves to Greco-Roman literary sources. I think this was the original intention of the discussion (Classical sources (Herodotus, Plutarch etc)), and I think that most of the editors who've been involved here edit mostly on topics related to classical history. Looking at other entries at WP:RSP, most of them are quite specific (many relate to a particular website or news outlet), and the category of all ancient (or pre-modern) literary sources seems very broad in comparison. While I can see that some considerations would indeed apply across the board, trying to write anything substantial which applies universally seems considerably more difficult, and would prevent us from writing something tailored more specifically to classical sources. There would also be quite some increase in the number of editors and WikiProjects we would need to consult. Plus, I think the category of all ancient literary sources would include the kinds of texts covered by WP:RSPSCRIPTURE (and presumably there shouldn't exist entries at WP:RSP which "override" other entries). As to archaeological sources, I would exclude them, as at that point (to echo what Ifly said above) I think we're getting closer to a rewrite of WP:PRIMARY. – Michael Aurel (talk) 04:46, 21 May 2025 (UTC)
  • Having given the matter consideration, I oppose treating ancient literary sources (or Greco-Roman sources, or medieval sources, or early modern sources, or any similar grouping) as I group. I think they should be evaluated one at time, on a case by case basis. If I wanted to, I could use all the credulous news stories about the alleged existence of weapons of mass distruction in Iraq, or the alleged link between the MMR vaccine and autism, to argue that all 21st century newspapers are unreliable; and I could use people like Jan Hendrik Schön and Andrew Wakefield to prove that all 21st century academics are unreliable. That is the problem with selectively cherry picking particular errors from particular sources, and then applying them to them to an entire class of sources. James500 (talk) 16:16, 21 May 2025 (UTC)
    I think this suggestion is borne largely from ignorance of approaches common in ancient history. Yes, historians do analyse things source by source, but they also generalise: ancient literary sources as a class require critical scrutiny because of shared structural features in a culture where invention, rhetoric, and ideology shaped how events were described.
    Insufficiently critical reading of ancient literary sources is a common criticism in reviews. There books on the topic, eg Grant Greek and Roman historians (1995), Morley Writing ancient history (1999), or, more provocatively, Gill and Wiseman (eds) Lies and fiction in the ancient world (1993). There is an entire literature, stemming from Wiseman Clio's coemstics (1979) and Woodman Rhetoric in classical historiography (1988) which takes the view that they are as a class relatively unconstrained literary inventors. Indeed, that literature is extensive enough to warrant polemic – eg Lendon "Against Roman historiography" in Feldherr (ed) Cambridge Companion to the Roman historians (2009) pp 41ff – accusing the Wiseman-Woodman-following scholars of undermining ancient history as a project with compelling counterargument.
    Even among defenders it is well established that scepticism of ancient literary sources as a class is warranted. This is a basic lesson for undergraduates: eg ex review Morley 1999 there is no harm in having written a published version of the sorts of comments one might make on a student essay. Scholars defending their reliability still must justify their acceptance of detail against a default reference of unreliability. If we turned up new material from the Herculaneum papyri such an approach that would still be the default. The idea no generalisations could at all be made is fringery posing as scruple. Ifly6 (talk) 19:03, 21 May 2025 (UTC)
    I did not say that no generalizations could be made. I said I do not trust you to make them. I would not have any objection to RSP quoting Michael Grant et al verbatim and in context. James500 (talk) 20:58, 21 May 2025 (UTC)
    You said [you] oppose treating ancient literary sources... as I [one?] [a?] group ... they should be evaluated one at time, on a case by case basis. This precludes generalisations. I think there's little else to be said inasmuch as your response is essentially disconnected from the self-evident record of this discussion. Ifly6 (talk) 22:49, 21 May 2025 (UTC)
    I am sorry if you did not understand what I said. I admit that, with hindsight, it probably could have been more clearly worded. I only intended to object to "unreferenced" generalizations advanced by Wikipedians (not to generalizations made by Michael Grant et al). James500 (talk) 23:17, 21 May 2025 (UTC)
    Wikipedia articles need citations. Wikipedia policies and guidelines are different. They are formed in discussion with reference to appropriate scholarship, style guides or whatever else may be pertinent but also considering what is appropriate for Wikipedia. We document the outcomes in our policy and guideline pages, on some pages such as WP:RSP footnoting the discussions but not citing the scholarship etc. employed in the discussions. In short, our policies and guidelines consist almost entirely of "unreferenced" generalisations. NebY (talk) 10:30, 22 May 2025 (UTC)
    For the avoidance of doubt only: It is not normally in the interests of Wikipedia's readers for editors to include ideas they have invented themselves in Wikipedia's content policies and guidelines (or essays). I can see that there is no policy or guideline forbidding the inclusion of ideas invented by Wikipedians in policies or guidelines (or essays). But there is equally no policy or guideline that says that Wikipedia's policies or guidelines (or essays) must include ideas invented by Wikipedians, so I am entitled to !vote against the inclusion of such ideas. James500 (talk) 23:18, 22 May 2025 (UTC)
    Have you read Michael Grant's Greek and Roman Historians: Information and Misinformation? As well as examining the most prominent ancient historians in a number of ways, he introduces and summarises the overall situation; for example
    "A tendency to deviate from strict truth in the interests of a good story";
    "methods and devices ... in complete disagreements with these modern requirements" such as accuracy and objectivity;
    "used their sources in manners which we find inadequate and surprising";
    "the borders between fact and fiction are overstepped in many ways which seem to us inexcusable";
    "they do not tell us enough, and some of what they tell us is wrong"
    "it is necessary to repeat, once again, that ancient history was understood not as history, according to our meaning of the word, but as literature. There is no doubt that this is detrimental to their value as historians";
    "Mommsen was not far wrong when he classified historians among artists rather than scholars" and quoting Mommsen "'A historian had to entertain, and for that purpose he did not need truth as much as wit' 'In the end, the demands of artistry gained precedence over those of science'"
    Grant spends some time, as you might guess, asserting that the ancient historians are still well worth reading. For example,
    "That is precisely what makes the task of today's historians, attempting to reconstruct ancient Greece and Rome, so interesting. There is no need to repeat here all of the arguments in favour of our learning about the classical world. We ought to learn about it, and yet the task of doing so is difficult and must be recognised as such."
    For the avoidance of doubt: these are not ideas invented by Wikipedians or in any sense Fringe theory; all the above was published in 1995 and Professor Grant died, at age 89, in 2004. NebY (talk) 23:55, 22 May 2025 (UTC)
    This view is itself the fringery posing as scruple. Scepticism of primary sources as discussed here is a common lesson for undergraduates. The epistemology of Wikipedia policies is well settled in the manner NebY describes. You are entitled to your opinions: you may need new ones if they are not to be discarded in the consensus process. Ifly6 (talk) 00:10, 23 May 2025 (UTC)
    To be clear, footnote (d) to WP:PRIMARY, part of our policy document Wikipedia:No original research, says Further examples of primary sources include:... medieval and ancient works, even if they cite earlier known or lost writings. It has included ancient works since 2011 at the latest. NebY (talk) 00:40, 23 May 2025 (UTC)
    Please stop twisting my words, and putting words into my mouth, on purpose. You know perfectly well that I did not say any of that. You know perfectly well that posted my original comment before Grant's book was mentioned; that I was not aware of Grant's book at the time I posted that comment; and that I would not have posted that comment if I had been aware of Grant's book. Further, you know perfectly well that this comment did not refer to anything said in this discussion, but was merely a description of what I would do in a hypothetical scenario not connected with this discussion, prompted by this comment, which was so misleadingly worded (probably on purpose with the intent of entrapping me into saying something that NebY would be able to twist), that it appeared to me to be refering to a hypothetical scenario not connected with this discussion. You know perfectly well that I did not suggest that anything in Grant's book was a fringe theory. You know perfectly well that if I had meant to refer to anything in Grant's book I would have done so in express and unambiguous words naming Grant by his name. You know perfectly well that this comment does not mention ancient or medieval sources at all, that the comment is not talking about ancient or medieval sources, and that if I had meant to refer to ancient or medieval sources in that comment, I would have done so in express and unambiguous words. You know perfectly well that I did not suggest that there was anything objectionable about "scepticism of primary sources", and you know that I have in fact expressed skepticism of ancient and medieval sources above in this discussion. You know perfectly well that I did not at any time suggest that WP:PRIMARY is not policy. Now please stop. James500 (talk) 01:35, 23 May 2025 (UTC)
    May I respectfully suggest to both sides at this point that it is pointless to keep arguing over who misunderstood whose words. I don't think we have any substantial disagreement at this point: primary sources may still be cited, but with care to distinguish between what those sources say and what modern historians have to say about them. That's how it should always have been, and if we occasionally err by failing to provide a secondary source to back up, explain, or dispute the accounts of the ancients, we should look for and supply those sources when we find anything that seems to require them. Surely nobody here can object to that. And if we're all on the same side, it doesn't do anyone any good to keep arguing about who's mischaracterizing whose remarks. P Aculeius (talk) 04:48, 23 May 2025 (UTC)
    Yes please. Paul August ☎ 12:42, 23 May 2025 (UTC)
Noting that we've had a number of other discussions on the Korean Veritable Records and currently list its reliability as no consensus on WP:KO/RS. See WP:SILLOK for a listing of discussion on the Veritable Records. grapesurgeon (seefooddiet) (talk) 03:12, 25 May 2025 (UTC)

Another rewrite

This is based on Aculeius' text earlier:

Greco-Roman literary sources,[a] such as the writings of Aristotle, Cicero, Herodotus, Livy, and Plutarch,[b] must be used in accordance with WP:PRIMARY: editors may not analyze or synthesize material found in them.[c] Editors must take care not to present the claims of such sources uncritically.[d] Statements made by such sources should be clearly attributed to them, rather than presented as objective fact.[e] An entire article, or large parts thereof, may not be based only on such sources.[f] Whenever possible, editors should consult and cite modern published scholarship on the same subjects.[g]

These notes explain each text choice and are not intended as part of the final text.

Notes

  1. ^ More specific and avoids "primary sources"
  2. ^ Listed a few more to ensure that the scope is not thought too narrow or wide.
  3. ^ Everyone agrees on this and PRIMARY itself is clear that such sources are covered by it. The portion after the colon is also from PRIMARY.
  4. ^ This is verbatim from Aculeius' suggestion, except that "historical sources" is replaced with "such sources".
  5. ^ This is verbatim from Aculeius' suggestion with a link to WP:INTEXT added.
  6. ^ This is almost verbatim from WP:PRIMARY.
  7. ^ This is from Aculeius' suggestion, replacing "authoritative scholarship from the modern era should address" with "editors should consult and cite modern high quality reliable sources on", to be more consistent with Wikipedia terminology. I read Aculeius' text to imply that HQRS should be consulted and cited; I strengthened it to make that explicit. Edited. Per Prefer secondary sources – Articles should rely on secondary sources whenever possible in WP:RS, strengthened to WP:SCHOLARSHIP.

This is meant to avoid the red herrings related to defining "primary sources", the suitability of secondary sources, and the scope of the entry. The entry would be on the page WP:RSP. This is a long quote therefrom:

What this page is
  • checkY a list of sources whose general suitability for most purposes has been discussed repeatedly
  • checkY a very brief and simple summary of the consensus found in such discussions
What this page is not
  • ☒N a policy or guideline
  • ☒N a list of pre-approved sources that can always be used without regard for the ordinary rules of editing
  • ☒N a list of banned sources that can never be used or should be removed on sight
  • ☒N a list of biased or unbiased sources
  • ☒N a list of sources that are guaranteed to be correct regardless of context
  • ☒N a list of every source that has been discussed
  • ☒N a list of sources that have never been discussed, or whose reliability should be obvious to most editors
  • ☒N a list of primary, secondary, or tertiary sources
  • ☒N a list of independent or affiliated sources
  • ☒N a list of self-published or traditionally published sources
  • ☒N a representative sample of all sources used on Wikipedia or all sources in existence

It is my view that nothing in what I suggested directly above is anything beyond a summary of the WP:CONSENSUS of repeated discussions on primary sources, including those here. The only live issue that remains, I think, is about in text attribution. While I do not like Aculeius' example, it does show how someone can write an acceptably lucid narrative while also scrupulously doing in text attributions. Indeed, if we accept Furius' contention that a statement that the Lacedaemonians sent only 300 men, cited to Herodotus, violates PRIMARY because that claim is itself not straightforwardly derived from Herotodus, PRIMARY would require that in text attribution. Ifly6 (talk) 23:28, 21 May 2025 (UTC)

@P Aculeius and Furius: Pinging you to ensure that I did not misstate your positions. Ifly6 (talk) 23:28, 21 May 2025 (UTC)

I would also add the following:

Some editors believe that it would be undue weight to cite one of these sources without a corresponding secondary source demonstrating its relevance. Some editors also emphasise that old or public domain editions and translations of these sources may not reflect current scholarly readings or views.

I think the problems of undue weight and obsolete translations should warrant mention. Even if you disagree with these position, I think this discussion demonstrates that some editors hold these beliefs. Ifly6 (talk) 00:02, 22 May 2025 (UTC)

(edit conflict) This comment was removed in an edit conflict. I restored it. Due to the time it was made, it must have been directed towards the upper portion only. The edit conflict created a longer OP. Clarification may be in order. Ifly6 (talk) 00:05, 22 May 2025 (UTC)
I am striking my !vote because the proposed text has changed. No comment on the new version, and I am not going to !vote again until I am satisfied the proposal is stable. James500 (talk) 00:42, 22 May 2025 (UTC)
I don't find anything particularly objectionable in this proposal, but I think the language should be clarified in a couple of places:
  • In the first line, insert "such as" before the list of authors, to make clear that it's not intended to be comprehensive, or treat the named writers differently from other classical sources.
  • I might substitute a prominent Greek writer (Homer, Hesiod, Polybius, perhaps a philosopher) for Appian, who seems less likely to come up than the others named (I've cited Appian many times, but I think the others are much more likely to be encountered).
  • "May not be based only on these sources": substitute "such" for "these", since again we mean all such sources, not the ones listed at the beginning specifically.
  • Also in the same clause I strongly recommend substituting "should" for "may", not because it's a "suggestion", but because we know that this often happens, even though it shouldn't. This choice of wording points to the solution: look at what is said, determine whether a particular instance is really problematic, and then find a secondary source to back it up, add context, or refute it.
  • Delete the words "high quality" before "reliable sources"; the phrase links to "reliable sources", which provides a great deal of guidance in choosing sources of high quality, but does not define the phrase or have a section about it. Including the phrase might encourage quibbling over which otherwise reliable sources are "high quality" and which aren't, without necessarily addressing what they say in a specific instance and whether it's accurate. I'm not arguing that we don't want high-quality sources; I'm saying that we don't want to treat something that isn't a term of art as though it were one. We should just supplement or replace inferior sources when possible, instead of having lengthy debates over whether a particular source meets the definition of "high-quality". P Aculeius (talk) 01:08, 22 May 2025 (UTC)
Such as: checkY. Appian → Aristotle. These → such. I'm not sure what changing "should" to "may" does other than weaken something which we are essentially all on agreement on. I would strengthen directly to WP:SCHOLARSHIP regardless, which is on that page: Prefer secondary sources – Articles should rely on secondary sources whenever possible. Classical studies is fortunate enough to be a field with scholarship. Ifly6 (talk) 01:20, 22 May 2025 (UTC)
Suggest deleting the words "those of" in the first line, or else change it to "the writings of". Possibly change "Graeco-Roman" to "Greek and Roman" (isn't the usual spelling "Greco-Roman", or is "Graeco-Roman the UK variant?). Readers might associate the phrase "Greco-Roman" chiefly with wrestling, not classical scholarship. I still think that "should not" is better than "may not"; "should" does not make it optional, but ought to make overzealous correction (i.e. deleting sources that have been cited, instead of adding those that ought to be cited) less probable. P Aculeius (talk) 03:06, 22 May 2025 (UTC)
The spelling Graeco-Roman is BrE. See one of our articles. Spelling correction, however, is well taken. Much of the RSP list is in AmE, so I'll adopt that spelling. Are you discussing "may not" in editors may not analyse or synthesise material found in them? Ifly6 (talk) 03:18, 22 May 2025 (UTC)
Those of → the writings of. I think "Greco-Roman" is sufficiently common that one wouldn't confuse it with wrestling (especially so in the contexts where this would be linked). Ifly6 (talk) 03:27, 22 May 2025 (UTC)
We could link Greco-Roman to Greco-Roman world, for clarity and even to avoid having to explain later that this does include Alexandrians, Josephus, Martial .... NebY (talk) 10:43, 22 May 2025 (UTC)
Sure, let's do that. Ifly6 (talk) 16:37, 22 May 2025 (UTC)
(edit conflict) I think this is better than previous versions. It makes use of some of the toned down language from the "Suggested rewrite" version, it stays directly on topic, it makes the scope of the guidance clear (Graeco-Roman literary sources ...), and I think it accurately reflects the views of editors in this discussion. One specific point on wording: with the addition of "such as" (which I agree with), the first sentence isn't a complete sentence. Maybe "Graeco-Roman literary sources include writings by figures such as ..."? We could also link "synthesise" to WP:SYNTH if we wanted. I would argue we should strengthen our language in the Editors must take care ... sentence (eg. "cautious", or even "very cautious"; cf. WP:SCHOLARSHIP, which says When relying on primary sources, extreme caution is advised.), though I understand others may disagree.
I'd agree with the inclusion of the Some editors believe that ... sentence (with the missing word fixed). We could also try to find a less strong version of the statement that can be agreed upon: there isn't agreement (and there won't be) that the presence of an accompanying citation to a secondary source is necessary to indicate the relevance of a claim from a primary source, though I do think our existing policies (specifically, WP:DUE and WP:PROPORTION) already require that the weighting (and perhaps, depending on your reading, the inclusion) of claims from primary sources should be at least based upon secondary sources in some way (even if those secondary sources aren't necessarily cited after each statement). To put it another way, what might be missing from the Whenever possible, editors ... sentence is the awareness that claims for which it isn't possible to find secondary sourcing are often those that shouldn't be mentioned at all. – Michael Aurel (talk) 03:16, 22 May 2025 (UTC)
Thanks for noticing the fragment. Altered to such as → include, but are not limited to, which maybe will satisfy everyone. I agree with you on both points (due weight needs to be based on secondary sources; secondary sources ignoring or dismissing a primary source define how we should frame our coverage of that same source). A broad consensus notably not unanimity, however, would probably be necessary to note something like that latter element down as the "orthodox" interpretation of PRIMARY point #2 do not put undue weight on [a primary source's] contents. Ifly6 (talk) 03:27, 22 May 2025 (UTC)
I meant to note the sentence fragment, but wasn't sure whether it followed from something else, and forgot to check. However, instead of changing "such as" to "include, but are not limited to", I suggest merging this with the following sentence by deleting "use of such sources". The key verb here is "be limited", not "include"; our chief concern is the manner in which sources are cited, not which sources are included among the examples.
With respect to the "may not", I mean the second instance: "may not be based on" should read "should not be based on". The change in wording does not make the use of secondary sources optional, but it ought to leave the remedy less open to misinterpretation: editors running across a section or passage that is missing secondary sources for whatever reason might interpret "may not" to require that the text in question be deleted, without attempting to locate and cite the needed secondary sources. P Aculeius (talk) 04:50, 22 May 2025 (UTC)
I've gone ahead and merged the first two sentences, along the lines of your suggestion. I've also added a link to WP:SYNTH (not sure if we're agreed on this, but if anyone disagrees they can remove it). No real opinion on the "may" vs "should" dispute. – Michael Aurel (talk) 07:21, 22 May 2025 (UTC)
Yes, that's better. I may have mistyped what I meant by the "key verb", but you understood what I meant anyway. Thank you! P Aculeius (talk) 11:45, 22 May 2025 (UTC)
No objection to your changes here. Ifly6 (talk) 16:36, 22 May 2025 (UTC)
I would like to refocus on to the latter portions I proposed above. Are there any meaningful objections to it as a summary of views held? Ifly6 (talk) 00:12, 23 May 2025 (UTC)
If the text we were just discussing and refining is what you're referring to, I think it's a formulation we can all live with. P Aculeius (talk) 01:55, 23 May 2025 (UTC)
To echo what I wrote above, I'd support the addition of the Some editors believe that it ... sentence, though I'd prefer a (less strong) version which doesn't rely on the "some editors" wording. Something which states that the weighting of ancient sources should be determined by secondary sourcing, without imposing new (or at least new, overly tight) restrictions. For example, something along these lines could work:

The weight given to one of these sources should be roughly proportional to its treatment in modern, published scholarship on the same subject, and this scholarship should be cited when possible.

This language is mostly taken from WP:PROPORTION, and the statement hopefully doesn't imply anything not already implied by that policy. The last part is intended as a simple rephrasing of the Whenever possible, ... sentence, connecting the two for the sake of brevity (if anyone thinks the phrase is saying anything beyond the original sentence, it can be reworded).
A stronger version of this would also cover the inclusion of ancient literary sources. To be clear, I'm not suggesting that secondary sources need to be included alongside every citation to a primary source, that content cited to a primary source should be deleted if editors can't see a relevant secondary source in the article, or that it's necessary for scholars to spend paragraphs on a primary source to justify its inclusion (for example, citing the source would be perfectly fine). In my view (and others may disagree), this is already implied by policy: if there is zero coverage of an ancient source in reliable, published scholarship on a certain topic, then it would be "proportionate" for us to have zero coverage of it in the article on that topic.
I wouldn't say I have strong feelings on the second sentence. If we include it, I'd suggest removing the "Some editors also emphasise" part (as I don't think anyone disagrees with the unqualified statement?). – Michael Aurel (talk) 06:13, 25 May 2025 (UTC)
I think it needs some qualification, because differences in translation will rarely affect the meaning of anything cited, except when very fine minutiae are being considered. I'm sure the experts here can think of a dozen examples where that's the case, but those will still be a very small number in historical texts. Differences in viewpoint may also occur, but I doubt that obviously outdated views (as opposed to changes of opinion that could as easily swing back and forth over time) are likely to be cited as authority.
For instance, secondary sources may differ as to whether Vespasian was even-handed or despotic, but that's not something that can be proved either way, because it's a matter of perspective. On the other hand, if an old writing suggests that the people of Sardinia were easily subjugated because of their inferior culture, that would be an outdated viewpoint—no recent scholarship would say that, although it also seems unlikely that editors would add and cite that. That doesn't mean that a source containing outdated viewpoints can't be cited for anything else.
As for how best to qualify the statement, may I suggest, "Editors should be aware that in some instances..." or "Please take note that in some instances..." In these two lead-ins, we're not implying that editors disagree about the rest of the sentence, but "in some instances" emphasizes that this is only occasionally an issue. We can all agree that views change over time, sometimes due to new discoveries (or translations) or evolving social standards, but the extent to which this affects individual citations needs to be determined on a case-by-case basis, not resolved by a blanket statement. P Aculeius (talk) 13:55, 25 May 2025 (UTC)
I would say that the "in some instances" part is essentially already covered by the word "may" (which indicates that this is only sometimes the case), but "in some instances" is more precise, and I'd have no issue with your suggested change (I would probably omit the words in front of "in some instances", though, for brevity). My main concern, as you've identified, was that the "some editors" part could potentially lead someone to think there are editors who disagree with the statement itself. – Michael Aurel (talk) 23:23, 25 May 2025 (UTC)
That's fair, I don't think it's necessary to lead in to it. But "in some instances" does seem more precise, as you say, which as I see it means it's less likely to be treated as an across-the-board statement, and that would have been my concern. P Aculeius (talk) 02:30, 26 May 2025 (UTC)
Yes, better without the lead-in. A straight "In some instances" looks to me like an effective warning to editors new to this area or these issues - "Here's another trap you could run into". With a lead-in, "Editors should be aware that in some instances" looks like "We should probably mention that if you're very unlucky ... but you'll probably be fine." NebY (talk) 12:39, 1 June 2025 (UTC)
I would broadly deny that this is rare. The old translations which render "democratic party" out of words like populares are wrong. The old editions which call Marius' parents farmers are wrong. There are lots of cases where this happens. It's really not that rare. We have access to current Loeb translations. There's really no need at all to be consulting Dryden's translation 1683 of Plutarch's Lives (something I have actually seen). As to the language, I agree that may already covers in some instances. I also agree with Michael as to inclusion of the language around WP:PROPORTION. Ifly6 (talk) 07:03, 27 May 2025 (UTC)
Taking into account the above comments, here's an attempt at a version which includes the two further sentences:

Greco-Roman literary sources, such as the writings of Aristotle, Cicero, Herodotus, Livy, and Plutarch, must be used in accordance with WP:PRIMARY: editors may not analyze or synthesize material found in them. Editors must take care not to present the claims of such sources uncritically. Statements made by such sources should be clearly attributed to them, rather than presented as objective fact. An entire article, or large parts thereof, may not be based only on such sources. The weight given to one of these sources should be roughly proportional to its treatment in modern, published scholarship on the same subject, and this scholarship should be cited when possible. In some instances, old or public domain editions and translations of these sources will not reflect current scholarly readings or views.

– Michael Aurel (talk) 05:29, 1 June 2025 (UTC)

Note: I've unarchived this discussion, as I think there was at least general agreement that something should result from this, and because I think we were fairly close to agreeing upon a paragraph to be added to WP:RSP. – Michael Aurel (talk) 05:23, 1 June 2025 (UTC)

The board is very overloaded at the moment, and this is a very large discussion, have you thought of just being WP:BOLD. -- LCU ActivelyDisinterested «@» °∆t° 07:21, 1 June 2025 (UTC)
Hmm, I suppose I could have, though I'm an active participant in the above discussion and I wouldn't want to be seen as inserting my preferred version. – Michael Aurel (talk) 10:02, 1 June 2025 (UTC)
I think there's a clear consensus to add something. Any addition can be reworked or reverted if anyone disagrees strongly, per normal WP:CONSENSUS. -- LCU ActivelyDisinterested «@» °∆t° 10:50, 1 June 2025 (UTC)
I would support this. Paul August ☎ 09:57, 1 June 2025 (UTC)
I support this being added to WP:RSP and I'm certainly seeing more than enough consensus for WP:BOLD. Also, I don't think any tweak I could suggest would be worth delaying for. Thanks for unarchiving it and to all for developing it. NebY (talk) 12:51, 1 June 2025 (UTC)
Sure, do it. Ifly6 (talk) 16:11, 1 June 2025 (UTC)
As there seems to be at least a general consensus in favour of adding the above paragraph, I've done so. I've added it with a status of "No consensus", as this seems as though it most accurately reflects the above discussion, which has focused more on when and how it's appropriate to use such sources (this can of course be changed with further discussion, though). Any wording tweaks or suggested rephrasings can dealt with separately, as ActivelyDisinterested says. – Michael Aurel (talk) 01:22, 2 June 2025 (UTC)
Retrieved from "https://en.wikipedia.org/w/index.php?title=Wikipedia:Reliable_sources/Noticeboard/Archive_479&oldid=1297938008"