Wikipedia talk:Short description

The discussion between me and Marcus Markup at Talk:Influence and legacy of Swami Vivekananda#Short description is transcluded here:

Any input is appreciated, as I feel that the current WP:SDNONE text does not sufficiently address this. it's lio! | talk | work 16:37, 29 December 2025 (UTC)

I think MichaelMaggs is correct in his post (though I would argue that adding a "the" before "Indian" flows better). Swami Vivekananda is definitely not a figure that would be known by English speakers worldwide, so a SD would be helpful in this case. -- Patar knight - ^chat/_{contributions} 07:43, 1 January 2026 (UTC)

I simply don't see it as "helpful" at all for any real world use I can imagine. Who is going to come to Wikipedia, look at a list of articles, and then decide whether it's the article they were wanting based on our adding gratuitous text saying “Impact of Indian monk and philosopher" to the “Influence and legacy of Swami Vivekananda" listing? It does not seem to address any need an actual reader might have. But as I said above, I have thrown in the towel regarding this issue, I accept that my POV on short descriptions is not that of the consensus, and I'm no longer going to edit them except when they can be marked as a "Minor" edit. Marcus Markup (talk) 13:10, 1 January 2026 (UTC)

I think there are a lot of things we do on Wikipedia that might not be helpful to the average reader, but we still do just because there is the potential to be helpful. And then there's the fact that countless niches exist, so there could definitely be someone who reads an article because they see, through the short description, that it has something to do with their interests, no matter how vague the SD is. HKLionel TALK 18:39, 1 January 2026 (UTC)

I think the point of disagreement here is not whether somebody, somewhere might find a thing marginally useful, but whether the benefit justifies the trade-offs. In practice, what makes an encyclopedic tool useful is often not what it allows, but what it does not allow. That something might be "useful" to an edge case (which, in this case, I can't actually imagine) can't alone be the standard, or we’d have no principled way to avoid adding redundant or low-signal metadata everywhere. Deliberate restraint is what preserves signal, reduces noise, and keeps ancillary features from becoming an end in themselves. We could Wikilink every term in an article which someone might find useful in some way, but we don't do that because we show discretion and realize that sometimes, less is more, and enables actually useful information to stand out. The article is not about Vivekananda per se, it is about his influence and the title "Influence and legacy of Swami Vivekananda" is complete unto itself for 99.9% of users. Compelling that 99.9% to also read "Impact of Indian monk and philosopher", so that 0.1% of users don't have to hover over the article is adding unnecessary clutter that solves no demonstrated problem in this case. I will continue to advocate that the field be made mandatory, so that there can be no question of its applicability in debatable cases such as this. Marcus Markup (talk) 14:49, 2 January 2026 (UTC)

As usual at talk:SD, we have the usual competing perspectives of "top-down" via search v "bottom-up" via annotated links. But now, in this particular case, we may be letting that distract us from what is really quite simple: the SD for Swami Vivekananda should say " Indian monk and philosopher". But "Influence and legacy of Swami Vivekananda" shouldn't need an SD because it is self-explanatory. In what context could it arise where Swami Vivekananda has not already been identified? If I were doing a See Also, it would be done like this:

• Swami Vivekananda – Indian monk and philosopher (1863–1902)
  • Influence and legacy of Swami Vivekananda – Impact of Indian monk and philosopher

The second SD is clutter. To me, SD=none is correct (but I wouldn't bother to delete it if I saw it). 𝕁𝕄𝔽 (talk) 19:34, 2 January 2026 (UTC) but if I were building a See Also, I would consciously not use {{anl}} rather than have something so obviously silly. --𝕁𝕄𝔽 (talk) 20:28, 2 January 2026 (UTC)

In what context could it arise where Swami Vivekananda has not already been identified? The most obvious example would be someone who is aware that articles about the impact/legacy of figures exist and is just using the search bar to to see what other articles of that type exist. People go down weird Wikipedia rabbit holes on the time. -- Patar knight - ^chat/_{contributions} 07:02, 3 January 2026 (UTC)

Also in respect to use in See Also sections, using annotated links that would pull the SD isn't mandatory, so it would be fine to leave blank where Swami Vivekananda is already identified. -- Patar knight - ^chat/_{contributions} 07:05, 3 January 2026 (UTC)

Precisely, that's actually how I found the article, because I was curious how many articles of this type existed and how many people have independent legacy articles after reading Legacy of Che Guevara. HKLionel TALK 19:10, 11 January 2026 (UTC)

As Swami Vivekananda isn’t self-explanatory, a title containing this name isn’t either. It could arise in contexts like, for example, focus on the influence and legacy of significant historical figures, in which that of Vivekananda would be treated as an independent topic, rather than a subtopic of his biography.

Precisely, an editor can just not use an ANL if the biography is the main focus, and the subject’s influence and legacy is discussed as secondary to the subject. A lot of SDs seem redundant, but they are there because there are always possible purposes they can serve, so my interpretation of an appropriate blank SD would be only when there is clearly no doubt on how self-explanatory the title is. Better to aim for the chance of being helpful than having none at all, as well as the fact that SDs are by nature meant to merely complement the title, making them much less clutter-like than the SEAOFBLUE example that Marcus has raised. HKLionel TALK 19:08, 11 January 2026 (UTC)

I think there are use cases, as I described above is worth it, but it is a scenario where it definitely doesn't meet WP:SDNONE in that Swami Vivekananda is not self-explanatory to readers, but is also a pretty marginal use case given the niche topic. However, I do agree that the scope of WP:SDNONE should be explicitly drastically reduced instead of vaguely referencing that if one is helpful, since there are almost always additional descriptive/contextual details that are helpful. -- Patar knight - ^chat/_{contributions} 07:16, 3 January 2026 (UTC)

And that's why SDNONE probably needs some form of reevaluation, as I agree that SDs probably shouldn't be indiscriminately added lest they run contrary to their original spirit, yet as long as it seems helpful then I see no reason to not add one whenever I come across an article which title I wouldn't understand at first glance, nor the average reader. I get where SDNONE is coming from, but there should probably be clearer lines drawn to distinguish potentially useful disambiguation from SDs that really wouldn't serve any useful purpose. HKLionel TALK 19:17, 11 January 2026 (UTC)

I believe it's convention to generally exclude articles for simplicity, but I'm not opposed. HKLionel TALK 18:33, 1 January 2026 (UTC)

The current examples for using dates all currently show our MOS-compliant form of "(1668–1735)". Should we also follow the MOS when 'circa' dates are used? Our MOS says that we should have spaces in article text - ie. "(c. 1668 – 1735)" - but how should this be dealt with in a short description? Should is be "(c. 1668 – 1735)" or "(c.1668–1735)" or "(c. 1668–1735)"? Thanks - SchroCat (talk) 09:36, 18 January 2026 (UTC)

"(1668–1735)", without spaces, is long-established usage. Whether "c." should be followed by a space or not seems to vary in practice. Some editors like a space, but I'd prefer to avoid it as making short texts longer without additional information seems pointless. MichaelMaggs (talk) 11:42, 18 January 2026 (UTC)

It's not a question of 'long-established usage' or any user's particular preference: it's what the MOS says to use. The guidelines on spaces in dates for using circa can be found at MOS:CIRCA, which includes spaces after the "c." element and around the hyphen too - the example provided is "John Sayer (c. 1750 – 2 October 1818) ...". The question is whether this also applies for the short description. - SchroCat (talk) 11:49, 18 January 2026 (UTC)

Your question was understood. You asked "how should this be dealt with?". I replied with my view (c.1668–1735). Now, if others agree we could profitably discuss how best to recommend that. MichaelMaggs (talk) 12:18, 18 January 2026 (UTC)

I'm unconvinced we should have a different format to the rest of the MOS, but let's see how others view it. - SchroCat (talk) 12:24, 18 January 2026 (UTC)

I'm unclear how this is even a question. The SD, just like a date range in an infobox or a navbox, is part of the article content and should follow MOS: a space after "c." and a space on either side of the en dash. – Jonesey95 (talk) 15:01, 18 January 2026 (UTC)

Jonesey95, It's only a question because of this edit. I don't tend to get involved with short descriptions, but it seemed an odd statement. The examples on the guideline certainly don't show any spaces, but they don't show any examples with circa either, so I thought it best to ask the question to clear this up, even though I think the MOS would support the use of spaces in these situation. - SchroCat (talk) 15:10, 18 January 2026 (UTC)

but MOS:YEARRANGE says that the ndash should be unspaced... 𝕁𝕄𝔽 (talk) 15:26, 18 January 2026 (UTC)

But not when circa is used - see MOS:CIRCA. - SchroCat (talk) 15:31, 18 January 2026 (UTC)

Well that's just brilliant. So it must be 1750{{ndash}}1766 unless you put a c. first whereupon it must be 1750 {{ndash}} 1766. Whatever happened to consistency? 𝕁𝕄𝔽 (talk) 17:22, 18 January 2026 (UTC)

I think it probably is consistent, if you look at it from the point of view that years without anything else don't have dashes, but if you add anything - another date, circa or flourit, then dashes are added. Consistency, but not necessarily logic... - SchroCat (talk) 17:36, 18 January 2026 (UTC)

The rationale behind this is that c. only applies to one of the years in the range. To compare, you would probably agree that the proper form of linking a hypothetical January 1960 and the year 1965 would be "January 1960 – 1965", and not "January 1960–1965". January is only a property of the first year, not the second. The same applies to circa. This all falls under the "at least one item on either side of the en dash contains a space" logic under bullet point #2 of MOS:DATERANGE. YuniToumei (talk) 17:42, 18 January 2026 (UTC)

The above explanation is a good one. The edit by HKLionel linked above was incorrect per MOS. – Jonesey95 (talk) 17:55, 18 January 2026 (UTC)

Yes, that's a good explanation, thank you.

(fwiw, year1 – year2 looks better more familiar to me in all circumstances but that's an argument for another forum.) 𝕁𝕄𝔽 (talk) 19:10, 18 January 2026 (UTC)

As a guideline, the MOS reflects global consensus. Per WP:SDCONTENT [t]he short description is part of the article content, and is subject to the normal rules on content. One could argue that deviating from it for the purpose of short descriptions would fall under the impermissible category of WP:Local consensus. I do see the argument that a shorter date range style of "(c.1668–1735)" would be desirable to keep short descriptions, well, short. However making that exception might need a WP:PGCHANGE to the MOS itself, not to the short description information page which has not gone through the P&G process. YuniToumei (talk) 13:57, 18 January 2026 (UTC)

I agree; though whether it's actually worth going to the trouble of suggesting changes to the MOS is doubtful. The MOS does contemplate that some things need to be shortened where space is limited, but MOS:CIRCA gets very hung up on whether c. applies to the start of a range, to the end, or both, and the role of spaces in indicating that. There may be better things to do than get into weeks of detailed discussion about the exact meaning of spaces within a short description. MichaelMaggs (talk) 15:42, 18 January 2026 (UTC)

Right. In that case by maintaining the status quo I see no further arguments against applying the MOSCIRCA guideline to short descriptions like the one at Elizabeth Alkin and keeping the spaced endash as SchroCat did. Perhaps Alkin's SD could be added to WP:SDEXAMPLES so that more editors working on short descriptions are aware of this guideline. YuniToumei (talk) 16:23, 18 January 2026 (UTC)

This is a good idea, what does everyone think? HKLionel TALK 09:10, 22 January 2026 (UTC)

I have been bold and added it. If someone reverts or objects we can discuss further. - SchroCat (talk) 09:44, 22 January 2026 (UTC)

Is an ampersand (&) allowed in place of an 'and' in a short description? This isn't allowed in article prose (unless part of a proper noun) but I wonder if that applied to short descriptions as well. Stefen 𝕋ower ^{Huddle • Handiwerk} 23:02, 19 January 2026 (UTC)

As in the discussion above, please follow MOS. Let's not favor shortness over sloppiness. – Jonesey95 (talk) 23:06, 19 January 2026 (UTC)

I concur with that position. It was someone else using an ampersand, and I wondered if I was in the right to change it to 'and'. Stefen 𝕋ower ^{Huddle • Handiwerk} 23:11, 19 January 2026 (UTC)

Short descriptions should still follow general article policies, and I would only recommend an ampersand in places where reliable sources did as well. A hypothetical example off the top of my head would be a tool line that somehow meets GNG might have an SD of "Brand of Black & Decker power tools". VanIsaac, GHTV ^contr_about 00:29, 20 January 2026 (UTC)

 You are invited to join the discussion at Talk:Estonia § RFC: Short description of Baltic states. -- GoodDay (talk) 17:33, 22 January 2026 (UTC)

Would someone please take a look at this discussion, I really have no idea what Thewolfchild is getting at. They have not made a single argument with P&G-based rationale. HKLionel TALK 11:51, 28 January 2026 (UTC)

In the College football article, there is a short description with {{Short description|College version of American-Canadian football}}, yet the page shows "Missing article description". When I saw it first it was {{Short description|College version of American/Canadian football}} and I tried changing the "/" to "-", thinking that maybe the "/" was the problem, but apparently not. Does anyone have an idea what is going on here? rogerd (talk) 02:00, 2 February 2026 (UTC)

The "Local description" on the "Page information" page shows a blank description. {{Infobox sport overview}} includes {short description|none|noreplace}, which seems like it should output a short description that can be overridden locally, but in fact, the "noreplace" is ignored when the short description is "none". That template-driven short description forces a short description of "none" on the article. I moved the short description to work around the issue. There is likely to be a better fix. – Jonesey95 (talk) 04:10, 2 February 2026 (UTC)

Stop embedding short descriptions in infoboxes, where they seem to cause more problems than they solve? 𝕁𝕄𝔽 (talk) 10:55, 2 February 2026 (UTC)

What kinds of problems, other than this edge case? – Jonesey95 (talk) — Preceding undated comment added 17:05, 2 February 2026 (UTC)

A belated reply to Jonesey95's What kinds of problems [do they cause], other than this edge case? This page is on my watch list so I have no idea why I have seen nothing since my post of 10:55, 2 February 2026 (UTC) until now. The main problem is that they are inaccessible to {{annotated link}}, which has caused a lot of hassle with Australian geography articles - see for example Template talk:Annotated link/Archive 2#Failure to return SD from infobox?. No problem for top-down searches when you already know what you want, big problem for bottom-up discovery (via See Also) when you don't. --𝕁𝕄𝔽 (talk) 23:45, 5 February 2026 (UTC)

This problem is due to a bug in the Short description template — {{short description|none|noreplace}} is being treated as {{short description|none}}. This means that an infobox that sets a SD of none is overriding a valid local SD in the article. Looking at the template edit in January 2025, there is a change to use the new behaviour of the magic word: p1=none generates an empty magic word, but the noreplace is being lost.

A while back, I added a workaround using {{Has short description}} for some of the infoboxen, but it really needs fixing in the SD template itself. I think that the opening test needs to be altered from

{{#ifeq:{{lc:{{{1|}}}}}|none|{{SHORTDESC:}}...

to something like

{{#ifeq:{{lc:{{{1|}}}}}|none|{{SHORTDESC:|{{{2|}}}}}...

Sorry, I should have pushed for a fix a while back, but it got lost in RL issues... — GhostInTheMachine ^{talk to me} 17:14, 2 February 2026 (UTC)

@Pppery, Fayenatic london, and Jonesey95: Just in case. You are all busy people — GhostInTheMachine ^{talk to me} 17:45, 4 February 2026 (UTC)

I've implemented that template change as requested, and didn't notice any resulting changes in articles.

As the College football article was in Short description is different from Wikidata, I changed Wikidata to match. Now the article is still in that one but also in Short description matches Wikidata. Anybody got time to chase that one down? – Fayenatic London 19:13, 4 February 2026 (UTC)

I have just undone the workaround in the College football article, by moving the SD template from below the infobox back to the top. The change to the SD template seems to have had the right effect. So life is good. All is well. Thanks Fayenatic london. In the next day or two, I will experiment a little with undoing some of the workarounds I put into infobox templates a while back. — GhostInTheMachine ^{talk to me} 19:29, 4 February 2026 (UTC)

Thanks, GhostInTheMachine and Fayenatic london for looking into this! rogerd (talk) 23:16, 4 February 2026 (UTC)

You're most welcome! From the order of categories on College football, it's now clear that Short description is different from Wikidata is generated by the infobox, not the SD template. – Fayenatic London 22:41, 5 February 2026 (UTC)

The guideline for locations suggest using the country, etc. in the short description. My question is, are schools, colleges, and universities a location? I know there are a few that are virtual, but it seems that most educational institutions have a fixed location. Rublamb (talk) 23:29, 6 February 2026 (UTC)

I certainly find it useful that, for instance, Concordia University has a location: "University in Montreal, Quebec, Canada". There is another Concordia University in the city where I live and therefore this location in the short description helps do exactly what short descriptions should do as their primary function: disambiguate search results beyond what the article title (and any disambiguators in it) would already do. So I think the answer is: yes, locations should be included in short descriptions for schools that have a fixed location. —David Eppstein (talk) 00:01, 7 February 2026 (UTC)

I think something like "Online University founded in 2006" might be even more useful location information for a short description than a purely physical location.

Dates of foundation aren't typically very useful to distinquish. Best practice is to include the country. MichaelMaggs (talk) 09:36, 7 February 2026 (UTC)

Is there a guide/essay/other on short descriptions for films, having the director mentioned, instead of the star? Halbared (talk) 09:40, 6 March 2026 (UTC)

See this discussion from January 2026 at WP:FILM. – Jonesey95 (talk) 14:22, 6 March 2026 (UTC)

Every now and then I do a search for 'hastemplate:"Short description" "Wikimedia list"'. The result is always App suggested edits where someone has replaced {{Short description|none}}. Examples: 1, 2, 3. This feels like an issue because if I edit those back to None then it'd presumably still show up in the App and potentially be replaced once again. Is there a process to complain to the app maintainers to stop letting editors overwrite None? Or is the solution to not use None so we don't have to fight app suggested editors?

I've used the app a few times and there is no effort to distinguish between an article that doesn't have a short description and an article that does have one but is None. Just opened the app to test and the second article it suggested adding a short description to was 2023 in Burkina Faso, which has a {{Short description|none}} //Lollipoplollipoplollipop::talk 20:06, 6 March 2026 (UTC)

See this discussion and T326898, closed as Fixed in April 2025. I see that one of the edits was a couple weeks later, but it probably took that long for the software to be updated in production. Do you have any examples of this sort of edit happening in the last six months? Do you have the latest version of the app? – Jonesey95 (talk) 20:19, 6 March 2026 (UTC)

Also, this search should work better. – Jonesey95 (talk) 20:24, 6 March 2026 (UTC)

Good to know its already being fixed. No idea why it still didn't work, at least for me. I'm using the Wikipedia Beta app: 50569-beta-2026-03-03. //Lollipoplollipoplollipop::talk 22:34, 6 March 2026 (UTC)

Hello, I'm after some advice on behalf of me and another user @Marcin 303.

We have both been busy adding short descriptions to place names but have a slight difference of opinion over how these should be worded. I have been using the names of ethnocultural regions to describe places whereas Marcin 303 believes this is incorrect and only official administrative divisions should be used.

I'll give an example; the town of Bautzen doesn't currently have a short description. If I was to go ahead and add one it would be worded like this: Short description|Town in Lusatia, Germany. However Marcin 303 would phrase it the following way: Short description|Town in Saxony, Germany.

Can anyone advise if either is more correct? I have looked at the examples on the Short description article which specifies that for locations a country should be indicated in the short description, with the "local region" optional and somewhat open to interpretation.

Thank you for any guidance. Wikociewie (talk) 18:07, 10 March 2026 (UTC)

The infobox says that the town is in the state of Saxony. Using official place names is probably the way to go with short descriptions. – Jonesey95 (talk) 19:55, 10 March 2026 (UTC)

I suggest the answer here is to think about how the SD is going to be seen and used. The two most common cases are

1. someone searches Wikipedia for Bautzen and sees the SD as a pop-up
2. Bautzen is listed in a See Also section that uses {{annotated link}}s, so that the name of the town is followed by its SD (as in Berlin – Capital and largest city of Germany)

In both cases, only very local people will recognise Lusatia and they already know it. You need to look at it from a long way away: which variant would you expect for Hellenthal, for example? My advice therefore is the same as Jonesy95's: Saxony. 𝕁𝕄𝔽 (talk) 21:06, 10 March 2026 (UTC)

For articles with titles like "So-and-so's theorem", I have often been using short descriptions in the form of a prepositional phrase, like "On [topic of the theorem]" for the short description. The article itself is not about that topic, per se, but rather about the theorem, and the theorem is on that topic. So it would not work to shorten the description even more, to just "[topic of the theorem]". But you could read the title and short description together as "So-and-so's theorem on [topic of the theorem]". (Another formulation that sometimes works is to give a statement of the theorem as the short description, as Euclid's theorem with short description "Infinitely many prime numbers exist", but in many other cases this would give something too long and technical and it's better just to state the general topic of the theorem than its detailed formulation.)

In a recent edit to one of these articles, User:Btyner objected to this, and changed the short description from "On [topic of the theorem]" to "Theorem about [topic of the theorem]", with the edit summary "short description should generally start with a noun". I cannot find any wording in Wikipedia:Short description stating any such rule, and I can find wording saying that short descriptions should generally avoid words from the title of the article they describe (and that they should be short). Both for reasons of concision and avoiding repetition I prefer the "On [topic of the theorem]" formulation, but I thought maybe this would be something other regulars here might have an opinion about.

So, any opinions? —David Eppstein (talk) 22:38, 13 March 2026 (UTC)

The guidance you are looking for is in the Examples section: "[Article subject] is/was a/an/the ...". This guidance has been the consensus for a long time, and I would rather not see us stray from it. – Jonesey95 (talk) 22:46, 13 March 2026 (UTC)

An example is not guidance. Guidance would say "A short description should be a noun phrase". Examples of short descriptions that happen to be noun phrases could merely be insufficiently imaginative to cover all cases. Also note that the other formulation I gave above, "Infinitely many primes exist", is also not a noun phrase. Also also note that the only example given that covers this sort of topic is the odious "Concept in" formulation that is neither accurate (a theorem is not a concept) nor informative ("concept in" uses ten characters to say nothing). —David Eppstein (talk) 23:06, 13 March 2026 (UTC)

It's not "an example". It's the end of key guidance about the structure of short descriptions, and it serves to explain the table of examples that comes next. My apologies for quoting it only in part. Here's a more formal quotation: "A good way to draft a short description is to consider the words that would naturally follow if you started a sentence like this: '[Article subject] is/was a/an/the ...'" There's a nice discussion about it somewhere in the archives of this talk page, IIRC. As for "Concept in" being odious, that's a pretty strong statement to describe ten characters that mostly do no harm. "Concept in mathematics" would be sufficient to distinguish an article from one about a hip hop album with a similar name or one about a 1980s power pop group. – Jonesey95 (talk) 00:35, 14 March 2026 (UTC)

Logically, there is a big leap from "a good way to..." to "the only way to..."

Many ambiguities in titles, especially in mathematical articles, involve similarities to the titles of other mathematical articles. "Concept in mathematics" is almost as useless a short description as "Wikipedia article" for resolving those ambiguities. That is especially true when the article title already contains the word "theorem", which both makes clear that its topic is very likely mathematics and that it is something different from a concept. (Also your example above of game theory is also not a concept. It is a broad field of study that overlaps both mathematics and economics.)

Or to put it more bluntly, if you think short descriptions should only distinguish the broadest of topic characterizations, and fail to make any finer distinctions within those topics, then why should we have any short description for a biography other than "Person"? After all that would also disambiguate people from music albums and rock bands. —David Eppstein (talk) 06:29, 14 March 2026 (UTC)

I concur strongly with David. Jonesey95, you seem still to be in the mind-rut that SDs are exclusively for searches. They are not. SDs like "concept in physics" are a total waste of space as an annotated link in a See Also list. Yes, in more detailed articles, hand-tailored annotations of the articles listed will certainly be needed but in most cases, a good SD is sufficient to help the reader identify interesting related concepts. 𝕁𝕄𝔽 (talk) 10:26, 14 March 2026 (UTC)

If the guidance that has served us well for so long is inadequate, I recommend that you propose additional guidance text for this page, along with an explanation of when that secondary guidance should be used. It is useful to have SDs that have consistent formatting. – Jonesey95 (talk) 13:13, 14 March 2026 (UTC)

One of the three purposes of the SD, as set out in WP:SDPURPOSE, is to provide "together with the title, a very brief indication of the field covered". So long as the wording chosen focuses on that and on the other two purposes (annotation and disambiguation in searches), and also complies with WP:SDNOTDEF and WP:SDFORMAT, I see no reason why "On [topic of the theorem]" shouldn't be acceptable. While rare, I have occasionally come across abstract topics that can't easily be captured with a noun phrase, because even to get the noun phrase grammatically correct takes too many characters; things called "principle", "rule", "law", "model" and similar in other academic fields come to mind. That's not always the case with theorems, but "So-and-so's theorem on [topic of the theorem]" does achieve what we want. Its only downside is that it might tempt editors to draft an unreasonably long SD, in an effort to define the theorem. But we have WP:SDNOTDEF to cover that. No objection to this being added as a new example. MichaelMaggs (talk) 17:50, 14 March 2026 (UTC)

As well as SDNOTDEF we also have the 40-character target length, which is often too short to provide a precise statement of the theorem (which we should not be trying to provide anyway).

If we're looking for an example to provide, one I just added is Hilbert's lemma, short description "On curvature of surfaces". My aim in formulating this was to be specific enough to distinguish it from the many other things named after Hilbert in different areas of mathematics, but not even trying to formulate a precise statement of the lemma nor using the more technical terminology like "principal curvatures" that would appear in a more precise statement.

Thinking about it some more, I think my real beef with the "concept in mathematics" short descriptions is the attitude they convey of "it's something abstract, you probably don't care, here are some rock bands for you to read about instead". We should not lump all abstract things together as if they are indistinguishable and uninteresting; we can and should do better. —David Eppstein (talk) 18:23, 14 March 2026 (UTC)

I have checked this before. There is no formal 40 character limit, although the app for iPhones has a design error which has that effect (so Apple users should lobby for that error to be fixed asap but the rest of the world should not wait until it is). WP:SDSHORT specifies that the SD should be succinct but not so that the SD is useless to someone coming to it cold. So while

• Hilbert's lemma – On curvature of surfaces

works is necessary and sufficient ,

• Principal curvature – Maximal and minimal curvature at a point of a surface

is no longer than it needs to be. 𝕁𝕄𝔽 (talk) 19:39, 14 March 2026 (UTC)

I concur with this assessment of "concept in mathematics". Pushing ourselves to do better is good exercise. Stepwise Continuous Dysfunction (talk) 19:42, 14 March 2026 (UTC)

For the first example, I'd suggest

• Hilbert's lemma – Theorem on curvature of surfaces

Slightly longer, although still perfectly fine, and adds context given that the word "theorem" is not already in the title. Adding "Theorem" adequately explains the technical term 'lemma' which apart from specialists hardly anyone will understand. In this simple context we don't need anything more to explain 'lemma'. MichaelMaggs (talk) 20:30, 14 March 2026 (UTC)

If I said that the term "concerto" is too technical to be used in a musical composition short description, and that we should replace it with the word "symphony" because it's more widely recognized as a type of musical composition, I think we would rightly expect classical music lovers to object that although both are types of musical compositions they are not at all the same thing. Similarly, lemmas and theorems are both types of proven mathematical statements but they are not the same thing (even though it is not hard to dig up examples where a theorem is used as a lemma or a result treated as a theorem is nevertheless called a lemma). —David Eppstein (talk) 21:01, 14 March 2026 (UTC)

In this example, adding "Theorem" informs readers it's a mathematical topic, which most won't realise from 'lemma'. 'Concerto' is enough on its own to identify the topic as classical music. MichaelMaggs (talk) 21:05, 14 March 2026 (UTC)

You are down in the search-mode gulley again. To use symphony for a concerto or theorem for lemma in an annotated link would be grossly incompetent and have people back here again demanding that {{annotated link}} be banned. And with good reason. 𝕁𝕄𝔽 (talk) 21:15, 14 March 2026 (UTC)

Indeed, which is why {{Annotated link}} provides an override option for such a case. As anyone can see from WP:SDPURPOSE, the annotated link template is "available for" use in See also and index list articles. Use of that template is not one of the three purposes of a SD, and those purposes should not be subverted for ease of template operation. Anyway, let's not go over that ground yet again. It's entirely off the topic of this section. MichaelMaggs (talk) 22:02, 14 March 2026 (UTC)

No! Purpose number 1: a short descriptive annotation to the title

Short descriptions are intended to supplement the title, so as to inform readers unfamiliar with a topic why it may be worthy of further exploration. A misleading SD is a misleading SD, no matter how it is revealed. Under no circumstances should we collude with, let alone actively encourage, creation of SDs that we know to be misleading, incompetent, or just downright false. Annotated link is one way to expose an SD, search is another. The only difference between the two techniques is that [a] search results are transient and the searcher probably won't bother to report crud (though they may have an unprintable reaction to it) whilst [b] SDs in annotated links are there in See Also lists to be revealed in all their glory. In the course of the last few years, I must have corrected many dozens of incompetent, fatuous or false SDs that I just happened to have exposed by converting See Alsos to use annotated links – SDs that have been misleading searchers for years.

My challenge is relevant to this debate because it is the compulsion to be terse at all costs that requires verbal gymnastics and strange proposals to create false SDs. The first priority for an SD must be that it be useful (and for that it must also be accurate). 𝕁𝕄𝔽 (talk) 01:21, 15 March 2026 (UTC)

It is not a WP:SDPURPOSE "to inform readers unfamiliar with a topic why it may be worthy of further exploration". A few editors like to use Annotated links without having to put in the effort to use the override option provided, but that is simply not what the guidance says. We can all agree that SDs should not be incorrect or misleading, but drafting for sophisticated readers who are capable of reading through to the bottom of a highly technical article completely ignores WP:SDJARGON: "avoid jargon, and use simple, readily comprehensible terms that do not require pre-existing detailed knowledge of the subject". It's noteworthy that essentially nobody apart from a very few mathematics specialists have problems sticking to WP:SDJARGON. Not theoretical physicists, chemists or any other highly technical field. Anyway, I think that's enough on this side issue for this thread. Feel free to respond if you wish, but I'll leave it there. MichaelMaggs (talk) 15:06, 15 March 2026 (UTC)

Just one further comment: whilst that may not be an overt SDPURPOSE, it is certainly the purpose of WP:SEEALSO. But for non-obvious article titles, an unannotated wikilink in a See Also list is next to useless. In an ideal world, editors would write context-specific annotations but that is very much the exception rather than the rule. Exposing the SD via {{anl}} gives a "good enough" solution ["the perfect is the enemy of the good"], provided that editors are encouraged to write meaningful SDs. And discouraged from writing vacuous or [mathematically] trivial SDs. --𝕁𝕄𝔽 (talk) 10:56, 16 March 2026 (UTC)

Well, the short description for Ising model uses jargon ("ferromagnetic"), the short description for AdS/CFT correspondence is awful (full of jargon and using an ordinary word in a jargon way), the one for conformal field theory repeats the jargon from the article title, the one for Boltzmann constant uses a term ("particle kinetic energy") that is at least as obscure as "lemma", Stefan–Boltzmann law sneaks in ordinary words overloaded with a jargon meaning ("black body") while using the odd locution "emissive power", Universality class presumes the reader understands "renormalization group flow limit"...

I was going to say that mathematicians are naturally going to have a harder time since their field extends further into pure abstraction, but I honestly don't think the physicists are doing too well either. Stepwise Continuous Dysfunction (talk) 20:34, 15 March 2026 (UTC)

Maybe a corollary of WP:ONEDOWN is that if the lead of an article should be written one educational level less technically than the body, then the short descriptions should be one more level less technical than that. This does not mean eschewing all technicality, though. —David Eppstein (talk) 21:43, 15 March 2026 (UTC)

Aiding search is the most common use case for SDs probably by multiple orders of magnitude. Every single person using the default skin will see it whenever they begin to type anything into the search bar. By comparison, Template:Annotated link is used on 15,000 pages. -- Patar knight - ^chat/_{contributions} 01:26, 17 March 2026 (UTC)

Inserting a word that is factually incorrect—lemmas and theorems are not the same thing, and this is a thing called a lemma that is being used like a lemma—is wrong. Stepwise Continuous Dysfunction (talk) 00:32, 15 March 2026 (UTC)

Our own article (Lemma (mathematics)) describes uses "theorem" to describe it, so it seems fine for an SD, especially since "Lemma" probably fails WP:SDJARGON. We are writing for the average reader, not mathematicians. -- Patar knight - ^chat/_{contributions} 01:22, 17 March 2026 (UTC)

I see no reason not to use a prepositional phrase if that's a way to get a description that is both short and useful. A description being fit for purpose is more important than exactly matching a template in a list of examples. Stepwise Continuous Dysfunction (talk) 19:38, 14 March 2026 (UTC) revised to quote purpose #1 --𝕁𝕄𝔽 (talk) 01:35, 15 March 2026 (UTC)