Proposed notability guidelines

I drafted Wikipedia:Notability (languages) in response to requests by editors policing notability, for verification that we consider all natural languages to be notable. What I drew up is my impression of consensus from years of editing language articles, and where I thought the limits on notability might lie, but it should be reviewed by this project. Once/if we get consensus here, we can add it to {{Notability sidebar}}. — kwami (talk) 20:15, 7 May 2026 (UTC)

I could have sworn some project page somewhere already said every natural language was presumed notable, but I guess not. Thanks for writing this. One question, should/does stability include past stability? So that endangered or long-extinct languages would be presumed notable? Erinius (talk) 23:50, 7 May 2026 (UTC)

Saying "all natural languages are notable" is very open-ended. There are languages for which the historical evidence is a few recorded words, primarily names. There are languages for which the only evidence is someone's comment that a particular people did not speak the same language as their neighbors. There is little to no hope that even a stub article could ever be written about such languages, so there is no benefit in talking about them being notable. Donald Albury 00:50, 8 May 2026 (UTC)

The draft is already more specific than that. You might want to suggest particular changes, or other objections, at the draft or its talk page. — Preceding unsigned comment added by Cnilep (talk • contribs) 01:09, 8 May 2026 (UTC)

"There are languages for which the historical evidence is a few recorded words, primarily names" is not a determinative factor; many of those cases (including some I was reading about via WP just a few days ago!) are subject to a great deal of real-world debate by linguists, in reliable-source publications, so they already clear the WP:GNG hurdle anyway, without any subject-specific notability criterion even coming into play. And "There is little to no hope that even a stub article could ever be written about such languages" is clearly not true with regard to this category. (It could be true of some particular member of that category about which there is little RS material, but having a small stub in such cases is perfectly fine, in the "better than nothing" department). I'm not sure what to make of "There are languages for which the only evidence is someone's comment that a particular people did not speak the same language as their neighbors." Yes, there are notes of this sort in various ancient texts, but that doesn't mean an identifiable, named language exists/existed, in the sense of in-depth coverage in reliable sources, for us to write an article about. So, that would appear to be a self-resolving issue; we can't have encyclopedic coverage of an idea with no RS to use about it. Or another way to look at that: some academic somewhere tentatively hypothesizing that there must have been an Elbonian language, because Brosis of Sparta wrote that the Elbonians did not speak the same language as the neighboring Durkastinians, doesn't equate to "Elbonian constitutes a natural language for encyclopedic purposes"; it's just some person what-iffing. But if such a what-if results in a flurry of debate and hypotheses in various high-end journals and books on linguistics, then it also does in fact pass GNG and is something readers may expect to find an article about here. Much of WP's job is distilling scattershot academic material running through decades of obscure journals, and presenting a summary for the non-expert reader. If a stub about such a [purported] language were too small to bother and likely to stay that way, then we could merge it to the parent language family, if known. That'd be a merge performed as an editorial consensus decision (probably with a categorized redirect left behind). The fact that an subject technically can exist as a stand-alone page due to GNG does not tie our hands and require us to have one in cases where it wouldn't be the best option for readers. Wikipedia is not a bureaucracy or a suicide pact. :-)  — SMcCandlish ☏ ¢ 😼  05:59, 8 May 2026 (UTC)

Per the above: yes, we say that extinction date is irrelevant. "Stability" here is intended to mean that a stable language existed. This is not always the case for pidgins, for example, where RSs distinguish stable, multi-generation languages, which are notable, from people throwing a bunch of words at each other when they meet at the market. For the same reason, we wouldn't have a stub article on every child who ever used home sign [in fact, the ISO code for Rennellese Sign Language was retired for just that reason], but it would be notable once kids are brought together in a school and develop a stable language.

The point of this guideline is to show that we do count distinct languages as inherently notable, because outside editors come along every so often and tag a language article for deletion because it has too little academic coverage to be notable under our general guidelines. But, per SMcCandlish, we do need to have enough info to know that there actually was a distinct language, which means at a minimum having enough data to establish its uniqueness. And of course speech varieties that do not meet that criterion -- and even entirely spurious languages -- may still be notable per the general notability guidelines. — kwami (talk) 10:23, 8 May 2026 (UTC)

This is my personal opinion, but I think if all we know of a language is a few words that are not enough to classify it -- or even to identify it as a separate language -- then that info is best placed in the ethnographic article for the people. And if a taxonomic node is not backed up by reconstruction, so that we can say something about the protolanguage, then it should probably not have a separate article just so we can have a place to park that part of the cladistic tree. I'm guilty of creating quite a few articles like that to break up the list of languages in large families like Austronesian. — kwami (talk) 10:34, 8 May 2026 (UTC)

I believe that any language that is attested should have its own article. Most if not all of these will have been covered by some linguist somewhere who has done some analysis on the language and its classification. Take, for example, Sewee language amnd Amotomanco language, both of them known from a handful of words or phrases, but are analyzed in Zamponi (2023). Because they have been covered in secondary reliable sources, they would have articles. The same goes for unattested languages covered in similar sources, provided they give proper coverage. 🪐Kepler-1229b | talk | contribs🪐 17:50, 8 May 2026 (UTC)

I am thinking of more recent (16th - 17th century) examples, such as Ais people#Name and language, Guale#Language, Uzita and Tocobago (see 5th paragraph in Mocoso), and the possible non-Hitchiti language of Sabacola. We know that each of those peoples spoke a language, we just do not know anything about those languages. Donald Albury 13:28, 8 May 2026 (UTC)

Yes, but if we don't say something, we may get well-intentioned editors creating articles for them anyway because all languages are inherently notable. — kwami (talk) 13:35, 8 May 2026 (UTC)

I would oppose this proposal as it stands. Languages that do not pass WP:GNG are unlikely to have sufficient coverage to establish that they even are genuinely languages. Edge cases like extinct languages only marginally attested in primary sources and have not received WP:SIGCOV in modern sources should not be considered notable enough for standalone articles. A wordlist from a non-linguist 200 years ago is not enough to build an article on.

Obviously all living languages are notable, and so we would only require one decent source that proves this (in practice they will always have this). But "language" is itself an artificial category, and one that has not been, and is not, applied consistently across space and time. We should not be creating articles just because Glottolog notes that some explorer in the 18th century wrote that a language exists. There is no potential for an article there.Boynamedsue (talk) 18:21, 9 May 2026 (UTC)

I am very much in agreement with the proposed guideline. As per User:SMcCandlish's remarks, a lack of reliable sources would regulate any creation of articles on spurious or poorly attested languages. There is a difference between Glottolog noting the existence of a language in passing and Glottolog providing a code for a language. LandLing 01:45, 12 May 2026 (UTC)

..a lack of reliable sources would regulate any creation of articles on spurious or poorly attested languages. That has never stopped editors from creating articles for dubious instances in other topic areas. What benefit would this proposal provide for Wikipedia that is not available from abiding by the GNG? Donald Albury 11:59, 12 May 2026 (UTC)

We occasionally have drive-by tagging of articles as non-notable, with the argument that languages are not notable unless they meet GNG. Rather than simply deleting the tags, as we have been doing for the past 20 years, we'd have a policy to point to that says that languages are notable. — kwami (talk) 19:31, 12 May 2026 (UTC)

There is inherently nothing to write about an unattested language, which serves as a sufficient deterrent to any prospective editor. Just because WP:OTHERSTUFFEXISTS doesn't mean that has happened, or will happen, with languages. 🪐Kepler-1229b | talk | contribs🪐 19:35, 12 May 2026 (UTC)

See my comments in the talk page propoosal. 🪐Kepler-1229b | talk | contribs🪐 18:52, 9 May 2026 (UTC)

Requested move at Talk:Emerillon language#Teko language

There is a requested move discussion at Talk:Emerillon language#Teko language that may be of interest to members of this WikiProject. Robloxguest3 (talk) 20:58, 13 May 2026 (UTC)

Requested move at Talk:Shetland dialect#Requested move 9 May 2026

There is a requested move discussion at Talk:Shetland dialect#Requested move 9 May 2026 that may be of interest to members of this WikiProject. Jeffrey34555 (talk) 18:13, 23 May 2026 (UTC)

Requested move at Talk:Silesian language § Requested move 24 May 2026

An editor has requested that Silesian language be moved to Upper Silesian language, which may be of interest to this WikiProject. You are invited to participate in the move discussion. Fortek67 (talk) 23:52, 24 May 2026 (UTC)