Wikipedia talk:Notability of reliable sources

Previous discussion sought
If anyone has links to any previous discussion on this topic then please share.

If anyone has any memory of this topic ever being discussed, then please share whatever you remember about when and where anyone might be able to track down that discussion.

This topic is difficult to find by search. I have the idea that this must have been discussed in the past, but searching for "notability" and "reliable sources" returns way too much noise. Also this discussion could have happened using many other terms. The most obvious places where previous discussion could have happened are these explanations


 * WikiProject Newspapers/Notability
 * Notability (academic journals)
 * Notability (periodicals)
 * Notability (media)

WikiProject Newspapers began in July 2018. I am participating there which is why I started this page in August 2018. We are considering how Wikipedia can get better coverage of regional and special-interest newspapers as a way to bringing information from those newspapers into Wikipedia. I have been a regular at WikiProject Academic Journals. While people sort of raise this issue, I do not remember anyone ever making the case to confer notability to a source on the basis of that source's importance to Wikipedia. I think there has been the assertion that journals which meet WP:RS often also meet WP:N but I cannot see exactly where this discussion developed.

In the "periodicals" and "media" guides the discussion is about which sources off-wiki cite which other sources off-wiki cite them. I am not immediately seeing discussion about the effects of Wikipedia's content heavily relying on sources.

If anyone has something to share then please do.  Blue Rasberry  (talk)  15:22, 26 August 2018 (UTC)
 * I'm pretty sure that this has come up either at discussions around NJournals or on the WikiProject Academic Journals. --Randykitty (talk) 16:52, 26 August 2018 (UTC)

Bluerasberry's prediction for 2022
I am not aware of any published development plan or roadmap. Neither technology nor social infrastructure exists to make this happen. I am not aware of anyone planning to make this happen. This is just my guess based on my intuition. I could name 100 variables in this which are changing rapidly and I cannot name any individuals or teams which are obviously making anything happen.


 * 1) right now - d:Wikidata:WikiProject Source MetaData collects primary source information about all publishers, companies which own the publishers, and all publications
 * 2) by 2020 - more Wikipedia articles import structured data from Wikidata
 * 3) by 2022 - many topics in many Wikipedias, especially non-English but English also, will have Wikipedia articles based only on Wikidata. This will include topics which fail Wikipedia notability criteria. Maybe these articles will have some tag that they are not typical Wikipedia articles written by a human, or maybe the interplay between human-written text and computer generated text from primary sources will have become the norm.

Source metadata will get imported to Wikipedia in this way, generating Wiki articles for academic journals, newspapers, and databases. There will also be tools to determine how often a source gets cited in Wikipedias, including English, other languages, and in other Wikimedia projects like images pulled from open access journals and put in Commons.  Blue Rasberry  (talk)  15:36, 26 August 2018 (UTC)

Wikidata's WikiProject Source Metadata as a driver for this
WikiProject Source Metadata currently has a focus of collecting metadata for individual articles in academic journals. This project has a few hundred participants. The meta:WikiCite conference which began in 2016 has built community around that project. WikiCite itself is an older idea.
 * d:Wikidata:WikiProject Source MetaData

I am not aware of anyone in WikiCite or WikiProject Source MetaData prioritizing the development of data around publications. Typically project participants are imagining that if they have better information about the publications from a source, then that will create a context from which computation and visualization will report insights about that source.

English Wikipedia has the greater current concern with providing encyclopedia presentations of a source. WikiProject Source MetaData will eventually have all the information necessary to present profiles of all sorts of sources. Since this will mostly be primary information from library databases, Wikidata has capacity to do this quickly with high quality control and with automatic translation into 100s of languages. Wikipedia's current human-centered writing processes will have to align with that information.  Blue Rasberry  (talk)  15:43, 26 August 2018 (UTC)

NJournals
Bad idea. We already had cases where publishers would "seed" WP articles with references to their journals and then subsequently claim during an AfD: this journal is notable, see how many articles link to it. This proposal would open the doors to this kind of abuse. At best, it would lead to a situation where people would go by their subjective impressions of what is a good journal and how much is "many links". I think the current situation is a workable compromise. If a journal is indeed a reliable source, then sooner rather than later it will get into one of Clarivate Analytics databases (apart from the shady ESCI) or Scopus and meet WP:NJournals.

As for the proposal/prediction that journal articles could/should be based on Wikidata, I disagree there, too. WD is rapidly becoming an indiscriminate collection of information and I for one have lost most of my trust in it. --Randykitty (talk) 16:59, 26 August 2018 (UTC)
 * I confirm that the problems you articulate are real. They are as big and bad as you imagine. I am also willing to agree that in the sense that this proposal brings huge problems it is a bad idea.
 * Problems aside, to what extent do you think the wiki community should have sympathy for anyone who sees a source extensively cited in Wikipedia and who consequently questions how they should seek information on that source?
 * Where and how in this world is anyone supposed to get information on the reliability of sources which are themselves not the subject of any WP:RS? Is there any sense in which you think Wikipedia is a concerned party to this difficulty?  Blue Rasberry   (talk)  20:28, 26 August 2018 (UTC)
 * I'm not so sure that the problem is as bad as you think. (Note: I'm only talking about academic journals, I don't know enough about the issues facing newspaper articles). I can't imagine that a journal would be considered a reliable source, be cited in 1000 WP articles (not as the result of a spamming campaign, but real citations) would somehow not be included in, say, Scopus, and therefore meet NJournals. That doesn't mean we would necessarily have an article on it (Scopus has over 25,000 journals, we have articles on about 9,000, not even half), but it does mean that an article could be created on it without us needing a special guideline. As for the next point, yes, it is true that academic journals rarely are the subject of articles discussing them in-depth (except for the more rare case where some scandal gets coverage). NJournals circumvents that by arguing that inclusion in selective databases, which themselves are reliable sources, proves that a journal is notable (inclusion in such databases only comes about after an in-depth review by a commission of specialists). So at worst, we exclude new journals which may be reliable sources but are too young to have been included in said databases. I think that is a small price to pay to keeping out all the trash journals that are out there, too. (And being young, those journals will rarely have many citations in WP; if they already have a lot -not as the result of a spam campaign- then they most ikely get cited a lot in other journals, too, and will very rapidly be picked up by, say, the Science Citation Index Expanded which sometimes accepts journasl that are still in their first year). Do you have an example of an academic journal that is cited extensively in WP but does not meet NJournals? --Randykitty (talk) 23:25, 26 August 2018 (UTC)
 * Scopus only goes back to 2004. If you want to write articles about nineteenth century history, you have to use nineteenth century periodicals, many of which ceased publication long ago and would therefore not be included in Scopus. James500 (talk) 01:58, 25 September 2018 (UTC)

This essay needs to mention the concept of a "reputation for fact-checking and accuracy" as a core concept of RS.
Per WP:RS, the central definition of RS is that a source must have a reputation for fact-checking and accuracy; this essay needs to put that front and center, discussing how and when that can be demonstrated. As written it gives the impression that demonstrating that (which would almost always imply demonstrating notability) isn't necessary, which is obviously contrary to core policy. --Aquillion (talk) 03:28, 10 March 2024 (UTC)