Wikipedia talk:Short description

Ethnic groups in short description

Recently, I've massively changed to Ethnic group that contradicts with WP:SDNOTDEF. Most ethnic groups, such as Russians, Jews, Bengalis, etc. of country/region is redundant per WP:SDNOTDEF. But in French people, which People of France is set to none, to avoid duplicating information that is already in the title. Absolutiva 21:29, 21 July 2025 (UTC)

I don't mind using ethnic groups in the short description. However, in the case of "East Slavic ethnic group" for Russians, this descriptions requires someone to know what East Slavic means (unless you are counting on the article name providing context). While this will not work for all, another possibility is to describe people by their general geographic area instead of OR in addition to their ethnicity. For example, "Slavic group from Eastern Europe", "Eastern European Slavic group", "Ethnic group from Eastern Europe", "Cultural group from Eastern Europe" or "Slavic people from Eastern Europe"--assuming we want to avoid "People from Russia". That being said, East Slavs has the short description "Subgroup of Slavic peoples" which would also apply to Russians. Rublamb (talk) 22:19, 21 July 2025 (UTC)

Ethnic groups also referred to "Indigenous peoples" for example:

• Alaskan Natives – Indigenous people of the United StatesPages displaying short descriptions of redirect targets
• Quechua people – Indigenous people of South America
• Tupi people – Indigenous people of Brazil

Absolutiva 01:02, 22 July 2025 (UTC)

Indigenous has a different meaning than ethnic, indicating that the group originates from that location and still lives there. An ethnic group can have a shared culture, history, and language but does not need to live in their place of origin. For example, Russians and Jews who live in the US are part of an ethnic group but are not indigenous to the US. (There is probably cultural bias in the usage of indigenous that might be worth considering. That is, we seem to acknowledge mobility for some groups, but others seemed fixed to a specific location.) Another difference is that Alaskan Natives is an artificial grouping based on geography that includes several different tribal cultures and ethnic groups; thus, it is the opposite of an ethnic group. Rublamb (talk) 01:38, 22 July 2025 (UTC)

Notice that I made a mistake as mentioned in my talk page. Absolutiva 03:03, 22 July 2025 (UTC)

Going back to the original question, "East Slavic ethnic group" is seriously POV. It is an argument for "ethnic purity": a person who is not "East Slavic" ipso facto can't be Russian. No. A Russian is a person who is a citizen of Russia. (Or arguably, if a first or second generation immigrant family who could be described as "Russian-American", someone whose ancestors were Russian citizens.) 𝕁𝕄𝔽 (talk) 11:49, 3 October 2025 (UTC)

Yes, a Russian can be "a citizen of Russia". That would be a modern geo-political viewpoint. However, the article in question, Russians, is about "an East Slavic ethnic group native to Eastern Europe. Their mother tongue is Russian, the most spoken Slavic language", according to its lede. This makes sense because the country Russia has not always existed, for example during the era of the USSR, but the Russian people still existed as a cultural group. Rublamb (talk) 17:27, 3 October 2025 (UTC)

Russia has not always existed... someone should have told Napoleon, saved himself a lot of bother. And obviously Peter the Great and Catherine the Great must have been playing Risk – pity Putin hasn't done likewise rather than trying to recreate Greater Russia. And Rus existed in modern Ukraine when Moscow was a village in a swamp – another of Putin's insane rationales.

for example during the era of the USSR Eh? See Russian Soviet Federative Socialist Republic (1917 to 1991). 𝕁𝕄𝔽 (talk) 19:36, 3 October 2025 (UTC)

You're still thinking of modern geopolitical divisions. The Russian cultural and language group dates to around the 9th century, before there was nation called Russia. Rublamb (talk) 20:06, 3 October 2025 (UTC)

Migrating the "Recent progress" graph to Charts?

After the Graphs extension was disabled, the "Recent progress" graph at Wikipedia:WikiProject Short descriptions#State of the project has been unavailable, although GhostInTheMachine has still been diligently keeping the graph data up to date with the table. Now that mw:Extension:Chart is available and is more or less usable, I wonder if we should switch over to the new extension?

I've already drafted a chart using the data on the WikiProject page (I think I got everything right, though I used an LLM to help parse the data into json). The data table is at c:Data:Sandbox/Liu1126/SdescProgress.tab and the chart is at c:Data:Sandbox/Liu1126/SdescProgress.chart. This is what it would look like in the same image frame as it is on the WikiProject page:

Extended content

Recent progress

Thoughts? Liu1126 (talk) 10:59, 29 September 2025 (UTC)

diligently? Diligent would be the same day every month. Moving on ...

My first attempt to use the chart thingy back in June was not very pretty, but recognise the data? See User:GhostInTheMachine/TestChart. The Image frame would seem to help with setting a chart width, but sadly we still cannot set the chart height. There are also display issues when the chart is combined with {{cot}}/{{cob}}. I have met some gossip that talks about a way to get the new chart system to collect data from the article as it used to, but the method still seems rather dirty. I think that we need to wait until the chart thingy is genuinely "production ready" — GhostInTheMachine ^{talk to me} 14:47, 29 September 2025 (UTC)

"Crime labels" in a short description...a quandary

I came here looking for guidance on a crime label used in a short description. Perhaps this guidance can be revised to better help users on the issue. There is an essay WP: Crime labels that has influenced changes to guidance on biographies. Here is the specific problem at the moment: An editor, @Sebbog13, recently changed the short descriptions for Sara Jane Moore and Squeaky Fromme to read: "American attempted presidential assassin". The description has some logical confusions, but also it uses the crime label that the subject is a "presidential assassin" - e.g., it could be taken that that's their identity and permanent occupation, c.f., the essay noted above.

Biographical article guidance is that rather than employ a crime label for a subject, just a statement as to the nature of the crime is to be given. Such a statement, while more objective, NPOV, and accurate, is necessarily longer than a label. In this present case, a short description for Moore (who recently died): "American who tried to assassinate President Ford (1956-2025)" might be preferred, though this description is 61 characters; perhaps others can suggest a shorter form. Per this article, this length is longer than 98% of Short descriptions. I write this to request two considerations for the guidance of Short descriptions: (1) that as tempting as they may be to use, "crime labels" are to be avoided, and (2) perhaps allowances can be made in such cases for longer descriptions, or examples given as to how a Short description can avoid the label, be ca. 40-character brief, yet still accurately reflect the subject. Bdushaw (talk) 11:47, 2 October 2025 (UTC)

Don't worry about the 40-character recommendation. Sixty-one characters is fine if they are all useful and help us comply with policy and guidelines. We are currently hunting 95-character descriptions, so 61 is way down the priority list. Please do use a dash in the year range, though: "(1956–2025)". – Jonesey95 (talk) 12:55, 2 October 2025 (UTC)

The 40 character limit is arbitrary and arose because of a technical limitation in the app for iPhones. Yes, SDs should be short, even terse, but they must be useful to someone who needs to have some clue as to the content of the article. So no shorter than they need to be – but equally no longer than they need to be either. Perhaps originally SDs were only written with the app in mind but nowadays {{annotated link}} is widely used to annotate articles listed in WP:See also sections, to help readers identify which topics they might want to explore further. 𝕁𝕄𝔽 (talk) 13:55, 2 October 2025 (UTC)

I am glad the SD lengths are not quite as strict as I thought. Bdushaw (talk) 14:57, 2 October 2025 (UTC)

I don't think we can issue general guidance against crime labels in SDs. There are many assassins, mobsters, murderers, pirates, etc. whose SDs would be much less informative if they didn't focus on the main aspect of notability. Firefangledfeathers (talk / contribs) 14:04, 2 October 2025 (UTC)

Sure, there is no hard and fast rule, to be sure. Al Capone was "a mobster", Bonnie and Clyde were "bank robbers"; no one would disagree. My concern is that crime labels in many subjects are inappropriate, while the length requirements of a SD beg for the use of a label. Elizabeth Holmes, the article where I first encountered this issue, is NOT "a fraudster"; rather, she was "convicted of fraud". Perhaps the request here is to note that an SD should respect the label usage, or absence, of the article it is to represent. In the biographical articles noted above, the "assassin" label is not used, hence I should conclude the SD should not use that label. Bdushaw (talk) 14:57, 2 October 2025 (UTC)

With Elizabeth Holmes, the short description is currently "American businesswoman". That seems to skip over why she is notable. I don't have a issue calling her a fraudster, after all that simply means "a person who commits fraud, especially in business or commercial dealings". Thus, someone who is convicted of a fraud, is correctly called a "fraudster", if that is why they are notable. Given the discussion here, "American businesswoman convicted of fraud" seems like a good solution. Rublamb (talk) 19:04, 2 October 2025 (UTC)

Either one is fine, because either one briefly describes her and distinguishes her from Elizabeth Holmes (writer) and Beth Holmes and Elisabeth Holmes Moore. – Jonesey95 (talk) 19:48, 2 October 2025 (UTC)

There was an RFC on the label "fraudster" and Holmes that clearly decided the label was a bad idea. There have been a multitude of other similar RFCs for other subjects and similar labels with the same result, and so on, hence the motivation for writing the essay on the subject noted above. One of my secondary aims for this inquiry here is to continue to raise awareness and sensitivity to the issue that often such labels are a bad idea. The definition argument you give has been worked over time and again. (And Holmes was a notable, high-profile businesswoman well before the fraud became apparent, but that's beside the point here; and not that I agree with that SD...) I similarly think it is a bad idea to have an article that studiously avoids such labels, and then have such a label creatively and carelessly employed for the Short description. As noted in the essay, the discussions on the crime label issue have been voluminous - we should avoid repeating them! Bdushaw (talk) 20:13, 2 October 2025 (UTC)

Remember that all the SD is to be used for us to help disambiugate when search on a list particular via the mobile. The SD need not be precise, just enough to distinguish. So when including crime labels, we should be careful about going too far. So using the word "assassin" seems very strong, and in a case like Moore, something like "American convicted of presidential assassination attempt" Masem (t) 15:47, 2 October 2025 (UTC)

Where the SD tends to be on the longer side, as here, dates can be omitted. WP:SDDATES says that "biographies of non-living people ... generally benefit from dating, but since the description should be kept short, other information may need to take precedence." In these examples, if the consensus is that a slightly longer text is needed to avoid what you call a 'crime label', I'd leave off the dates. MichaelMaggs (talk) 17:34, 2 October 2025 (UTC)

There is no good reason to omit the dates. The reason for the 40-character advice is that in some edge cases, SDs are truncated. Truncating the dates in those edge cases while providing them in the majority of cases is user-friendly behavior. – Jonesey95 (talk) 18:52, 2 October 2025 (UTC)

There is no 40 character limit. It is not even a target. · · · Peter Southwood ^(talk): 16:20, 1 December 2025 (UTC)

Thanks to you all for the discussion - I've gone with "American convicted of presidential assassination attempt" with the date. For me it has been useful, I hope also for the guidance for Short descriptions. Bdushaw (talk) 09:02, 3 October 2025 (UTC)

Incorrect short descriptions

Brad Meador, Paul Friedman (announcer), and Jeff Fair all use Template:Infobox baseball biography that is generating short descriptions of 'Baseball player' - which are incorrect in these cases. My bot's import of these to Wikidata was reverted, but I'm not sure how to fix that here. Can someone help please? Thanks. Mike Peel (talk) 17:18, 8 October 2025 (UTC)

According to Template:Infobox baseball biography, you can override the auto-generated SD by manually adding a normal SD template with a suitable description. I'm not sure if your bot would then port the right one over to Wikidata. Firefangledfeathers (talk / contribs) 17:21, 8 October 2025 (UTC)

Hmm, OK, I tried changing the short description at Brad Meador to use a new one on Wikidata per the template instructions, but it now just displays 'wikidata'? Mike Peel (talk) 17:51, 8 October 2025 (UTC)

I think you'd want something like {{Short description|American baseball executive (born 1975)}}. Firefangledfeathers (talk / contribs) 17:56, 8 October 2025 (UTC)

Enabling the gadget called "Shortdesc helper" in your Preferences makes it very easy to override the default short description generated by the infobox. I recommend it. – Jonesey95 (talk) 18:24, 8 October 2025 (UTC)

It's fixed on Wikidata, how do I get that to display here? Thanks. Mike Peel (talk) 18:52, 8 October 2025 (UTC)

GhostInTheMachine already fixed it :) Putting "wikidata" into the template won't transfer the description on wikidata over here. You have to manually set the description here yourself, like in the suggestion of Firefangledfeathers above. YuniToumei (talk) 18:57, 8 October 2025 (UTC)

Mike Peel: The English Wikipedia does not display descriptions from Wikidata. This is on purpose and by design. In order to get short descriptions to display on the English Wikipedia, the {{short description}} template is required, either in the article itself or in a template that is transcluded in the article (almost always an infobox). Putting a {{short description}} template in the article will override any automatic short description created by transcluded templates. – Jonesey95 (talk) 23:28, 9 October 2025 (UTC)

SDFORMAT: "plain text"

The guideline currently states:

Each short description should:
•   be written in plain text – without any type of markup or pictorial elements

This does not define "plain text" as ASCII. It seems, however, that there is a bot that is flagging anything else as an error.

So I want to propose that the guideline [sic] be clarified so that it does not preclude

• letters with simple diacritics typically encountered in English language texts such as ó, ò, ö, â, ā, ñ (French, German, Irish, Māori, Scots Gaelic, Spanish).
• extensively used typographic symbols like @, ©, #, ¶, § and so on.
• Mathematical signs widely seen outside mathematical text books and journals, like +, −, ±, /, ÷, ×, ~, ∴
• Various brackets and dashes like (, [, {, ⟨ and their mirrors, endash and emdash

Discuss. 𝕁𝕄𝔽 (talk) 11:49, 2 November 2025 (UTC) Added ñ to letters-with-diacritics samples --𝕁𝕄𝔽 (talk) 13:16, 2 November 2025 (UTC)

JMF, what bot? — Qwerfjkltalk 12:07, 2 November 2025 (UTC)

Sorry, some error report. @GhostInTheMachine tells me that Pilcrow (SD:Symbol ¶ used to identify a new paragraph) and Plus and minus signs (SD: Mathematical symbols (+ and −)) "are being flagged in an error report". My perspective is that Symbol used to identify a new paragraph and Mathematical symbols are uselessly uninformative at best. 𝕁𝕄𝔽 (talk) 13:12, 2 November 2025 (UTC)

JMF, I suppose you mean Wikipedia:Database reports/Short descriptions containing possibly invalid characters. — Qwerfjkltalk 12:25, 3 November 2025 (UTC)

Yes, thanks, that's obviously the one. And I see it does invite reviewers to regard its output as potentially infringing. But clearly some editors regard it as definitive. Most of what it flags make sense but just taking the first three that caught my eye,

• Breve Diacritical mark, ◌̆
• Bullet (typography) Typographical symbol (•)
• Cedilla Diacritical mark (¸)

These are all valid SDs that are sensible and helpful to readers. The symbol shown should be left in place.

It is not obvious why endash and emdash should be declared ineligible. 𝕁𝕄𝔽 (talk) 12:43, 3 November 2025 (UTC)

Most of these have specialist or non-English uses. Since short descriptions are intended to be widely comprehensible to English-language readers worldwide, it's not evident to me that many of those could be validly used. Do you have any apecific examples in mind where these characters would be needed? MichaelMaggs (talk) 21:49, 2 November 2025 (UTC)

The specific cases that caused me to raise this are pilcrow and section sign because to return "typographic symbol" is uselessly uninformative and even insulting to the reader's intelligence. For the SD to have any meaningful value, it must include the symbol – if only so the the reader can tell this is indeed the article they wanted. Same goes for division, multiplication (hybrid sign).

Words that contain letters with diacritic are not usual in English but not exceptional. (Perhaps less true in en.us? except ñ.)

I fail to see any rational basis to exclude characters that are in routine use. ASCII is a very long time ago, I can't believe that there is any debate about accepting The ISO Latin-1 set. I guess the Unicode basic plane will have to wait another 50 years.

Finally and to be clear, I am not bidding to add another rule to SDSTATUS, but just have whatever it is that checks for errors to lighten up. 𝕁𝕄𝔽 (talk) 00:12, 3 November 2025 (UTC)

I disagree that the SD for pilcrow needs to include the actual symbol. If I am wondering whether a "pilcrow" is some sort of bird, or a made-up word from Lewis Carroll, or a punk band from Omaha, the SD of "typographical symbol" is enough to set me straight. If I'm wondering which specific symbol it is, I can go to the article. The same goes for all of the symbols currently included in short descriptions – I think they can all be removed without making the SDs worse. – Jonesey95 (talk) 16:25, 3 November 2025 (UTC)

I suspect that you are thinking only of searches, where the user already knows the word pilcrow and there your assessment is correct. But equally important, if not more so, are WP:annotated links in See Also lists. These are immensely helpful to readers who don't know what the symbol is called, but recognise it when they see it and can see where to go next. This kind of serendipity is what makes Wikipedia great. 𝕁𝕄𝔽 (talk) 17:19, 3 November 2025 (UTC)

Formally...

• So to state this formally, I am requesting that the term "plain text" be taken to mean U+0020 thru U+007E and U+00A0 thru U+00FF. (Same as is illustrated here, 20–7E and A0–FF). I exclude non-standard use of codepoints in the U+0080 thru U+009F range (used by Windows-1252 and MacRoman). --𝕁𝕄𝔽 (talk) 08:57, 3 November 2025 (UTC)

• Now that I have seen Wikipedia:Database reports/Short descriptions containing possibly invalid characters and the exclusion list it contains, another option would be to add to the exclusion list all the symbols given at List of typographical symbols and punctuation marks?
  • Does context matter? For example, would we want to accept U+2122 ™ TRADE MARK SIGN anywhere except in the SD for Trademark symbol, given its sarcastic usage in other contexts?
  • Maybe we just need to strengthen the rubric at 'Short descriptions containing possibly invalid characters' to emphasise potentially to assert that context matters? --𝕁𝕄𝔽 (talk) 12:56, 3 November 2025 (UTC)

Context does indeed matter. The SD for Pilcrow, for example, seems fine, but I doubt that the symbol would be acceptable in many other articles. Some of the potential issues listed at Wikipedia:Database reports/Short descriptions containing possibly invalid characters are similar explanations of symbols, but I don't think we can just amend that list to accept everything you mention. The list is useful in flagging up chemistry and mathematics SDs that attempt to define the topic or that use technical terms (WP:SDNOTDEF / WP:SDJARGON). Better to clarify the rubric if need be. None of the symbols are blocked, just flagged for attention. MichaelMaggs (talk) 13:12, 3 November 2025 (UTC)

Agreed, I think that this can be the easiest way to resolve the issue without introducing additional bureaucracy. (But see Pilcrow: Revision history for yesterday: the conflict of perspectives has been a real issue and source of friction for some time.)

How about this additional sentence in the rubric in Wikipedia:Database reports/Short descriptions containing possibly invalid characters:

In an article about a specific non-ASCII symbol, it is helpful to readers for the SD to include that symbol (in parentheses). In such cases, the SD should be allowed to stand, but not otherwise. (For example, the SD for the Trademark symbol reads Symbol (™): an unregistered trademark and this is a valid use – but it is not valid to use the ™ symbol in other SDs, whether alongside a trade name, used sarcastically or otherwise.) Similarly, correct spelling of the names of some localities requires use of diacritics and these should not be discarded.

Does that hit the spot? 𝕁𝕄𝔽 (talk) 16:15, 3 November 2025 (UTC)

I created this report and have made all or most of the adjustments to its documentation and exceptions. I welcome better documentation if there is evidence that reasonable people do not understand the caveat in the emphasized word "potentially" that is currently the fourth word on the page, or the word "possibly" that appears in the page title. (It is much harder to help unreasonable people or those who simply will not read documentation.)

I think it works well to highlight real problems; it originally contained about 2,000 SDs and catches many minor problems such as misused en-dash-like characters. I think that including all Unicode characters would cause the report to be a lot less useful. Including characters like the em dash, precomposed fractions, guillemets, and other characters at list of typographical symbols and punctuation marks would make the report a lot less useful. – Jonesey95 (talk) 16:25, 3 November 2025 (UTC)

Accepted. I recognise that my first proposal was overkill and I am abandoning it. I found the comment by @MichaelMaggs persuasive and am now content that it will be sufficient to improve the rubric. I will do that now: further tweaks can be resolved there, 𝕁𝕄𝔽 (talk) 10:10, 4 November 2025 (UTC)

There is no need for additional language about diacritics; the report documentation already explains that "Other non-ASCII letters, such as Roman letters with accents" are accepted by the report. If you find a specific character that is needed for a place name and that the report does not accept, please note it on the report's talk page. – Jonesey95 (talk) 01:01, 5 November 2025 (UTC)

Announcing new userscript: FormatSeeAlso

Hi everyone. Since this page discusses using short descriptions in "See also" sections via {{Annotated link}}, I think editors here might be interested in a new userscript I've developed: FormatSeeAlso.

It's designed to help editors format "See also" sections per MOS:ALSO. The script converts plain wikilinks to use {{anl}}, pulling in the articles' short descriptions. It also formats section links with {{slink}} and optionally sorts the list alphabetically, while preserving formatting like italics or quotes, and handling redirects.

The script includes an initial step that scans all the "See also" links and presents a dialog with the articles that are missing a short description. This makes it easy to identify and populate missing SDs before adding the annotation templates.

Thanks to JMF for coming up with ideas and beta testing the script - couldn't have done it without you! If anyone has questions or thoughts, I'd love to hear them. Anne drew (talk · contribs) 14:51, 7 November 2025 (UTC)

I would like to endorse this tool, it works brilliantly. If you have been using {{AnnotatedListOfLinks}}, this is a gear change in convenience and performance. It identifies many of the usual errors and encourages action to fix them. And, like AnnotatedListOfLinks, it identifies the (still too many) cases of articles that lack an SD and does a great job of encouraging the user to contribute such. And yes, from time to time, to identify that the generic SD is not going to be good enough in a particular article and stop to give it a hand-tailored annotation instead. 𝕁𝕄𝔽 (talk) 11:16, 8 November 2025 (UTC)

Short descriptions on redirection pages

I'm afraid that I don't understand what this is intended to convey:

where the redirect page is the subject of an {{Annotated link}} (which is increasingly the case)

• The redirect target page or the redirecting page?
• the advice would make sense if the redirection is to a section at the target article, but it doesn't say that?

What have I missed? 𝕁𝕄𝔽 (talk) 10:03, 22 November 2025 (UTC)

I have rewritten that line so that it now reads

3.   where the redirection page is to a section of the target page and thus a more specific SD is required;.

I leave it others to decide whether the line The short description of a Redirect to section should refer to the section content and should not generally be the same as that of the article itself. should continue to be at the end or whether it can be appended to to item 3. --𝕁𝕄𝔽 (talk) 12:31, 25 November 2025 (UTC)

Probably close enough for most cases. A different short description is typically needed for a redirect to section or a redirect from subtopic (often both apply at the same time). There are a lot of redirects where the target article's SD is sub-optimal to just plain wrong. Cheers · · · Peter Southwood ^(talk): 16:45, 1 December 2025 (UTC)

Clarification on birth years for BLPs

I propose rephrasing this sentence:

Care should be taken when the biographies of living persons (BLP) policy applies: birthdates for living people should not be included unless sourced within the article.

To this:

Care should be taken when the biographies of living persons (BLP) policy applies: birth years for living people should not be included if not sourced within the article.

The reason for replacing "unless" with "if" is that if I'm not mistaken, this sentence sets a necessary but not sufficient condition (i.e. "You may or may not add the birth date, but if you do, it has to be sourced within the article"). But some contributors seem to interpret it as "if sourced in the article, birth years should be added to the SD of BLPs" and pointed to WP:SDDATES to justify adding birth years (e.g. " (born 1976)") to BLPs. The proposition also changes "birthdates" to "birth years" for extra clarity (no one is arguing for adding day and month anyway). I'm open to other phrasings if you think my proposition isn't ideal. Alenoach (talk) 15:59, 22 November 2025 (UTC)

It’s astonishing how people will interpret what seems fairly plain language. But yes, by all means clarify the wording if need be. MichaelMaggs (talk) 18:33, 22 November 2025 (UTC)

Putting one "if" statement and two "not"s in this wording will make it more confusing, not less. How about: "birth years for living people should be included only if they are sourced within the article"? (we could also write "may" instead of "should") – Jonesey95 (talk) 02:34, 23 November 2025 (UTC)

"should be included only if" leaves room for misinterpretation (it's close to "should be included if"). Using "may be included only if" is an option. Alenoach (talk) 03:04, 23 November 2025 (UTC)

That sounds reasonable. I'd suggest:

Care should be taken when the biographies of living persons (BLP) policy applies: birth years for living people may be included only if reliably sourced within the article."

Adding "reliably" provides a useful link to WP:RS, for beginners who might think that a date in the short description can be 'sourced' to any date in the text. MichaelMaggs (talk) 12:29, 23 November 2025 (UTC)

Looks good. Alenoach (talk) 14:15, 23 November 2025 (UTC)

Done. MichaelMaggs (talk) 14:37, 23 November 2025 (UTC)

SD on the page "List of S&P 600 companies"

I added a short description to the page here but it was reverted under the criteria SDNONE. However, just below this is the criteria SDLIST which states that articles that that are lists should never have the short description "None". So I'm a bit confused. Can I get some clarification? Thanks for any help Pearsejward (talk) 15:55, 24 November 2025 (UTC)

Don't stop reading in the middle of the sentence. – Jonesey95 (talk) 17:22, 24 November 2025 (UTC)

SDLIST does not say that articles that that are lists should never have the short description "None": it says they should not do so solely because they are lists. Not all list articles are the same, and not all should use "None"; they should be looked at individually. In this case, the wording "S&P 600" is not "sufficiently self-explanatory to English language speakers worldwide" (WP:SDNONE), and a proper short description is required, I've suggested "US stocks with small market capitalization". MichaelMaggs (talk) 17:22, 24 November 2025 (UTC)

Thanks. Very helpful Pearsejward (talk) 09:21, 25 November 2025 (UTC)

"Concept in" considered harmful

I keep finding short descriptions like "Concept in mathematics", despite the warning in WP:SDEXAMPLES to "Try to be more specific". When short descriptions are used to disambiguate mobile searches, these only have any value as disambiguators from other search results that happen to be from remote topics rather than others in the same field (consider for instance someone trying to distinguish Blaschke sum and Blaschke product, two very different topics named after the same mathematician, one of which had a short description of this form until today). When short descriptions are used in lists of annotated links, such as Wilhelm Blaschke § Mathematics or many see also sections, it is worse, as the context of the list usually provides more context than "concept in mathematics" would. And from the point of view of keeping short descriptions short, "Concept in" is actively harmful, as it adds 11 characters (counting the space at the end) without adding any meaning. I think our wording in SDEXAMPLES should actively discourage this sort of wording rather than providing it as a positive example with a vague admonition to be more specific. —David Eppstein (talk) 22:35, 26 November 2025 (UTC)

I disagree. The guidance tells you how to write a short description, including what to avoid, and provides a list of examples, ending with a final long-standing option of "Concept in [academic field]" – wording which is discouraged and allowed only in cases where something more specific isn't possible due to the complexity of the subject. In such cases, "Concept in" is by no means content-free, serving as it does as a means of specifying the broad field in terms that are clear to non-specialist English-speaking readers worldwide. "Concept in mathematics" will clearly be understood as "Something [that's too complicated to explain] in the field of mathematics". "Concept in" is a useful last-choice fallback when more specific terms aren't available or would require specialist knowledge to interpret. The disambiguation purpose of short descriptions according to WP:SDPURPOSE is to distinguish for the benefit of general readers "especially between similarly titled subjects in different fields" (my italics, eg mathematics, art), and not to make fine distinctions for the benefit of specialists between two highly technical articles within the same general field.

Let's talk about a specific example, Star coloring, in which yesterday you reverted my short description of "Concept in mathematics" to "Graph coloring avoiding 2-colored paths" with an edit summary starting "Please do not EVER use "concept in" in your short descriptions". No. A general reader might expect "Star coloring" have something to do with art, or perhaps astronomy. The title doesn't in itself suggest mathematics, and "Graph coloring avoiding 2-colored paths" does not achieve that as it relies on specialist knowledge of the technical mathematical meaning of "graph" contrary to WP:SDJARGON. A non-specialist reader looking at that short description will be none the wiser, perhaps guessing based on the common usage of "graph" that the article is about some sort of artistic technique to do with the colours used to display bar or line charts. There may be something better than "Concept in mathematics", but your proposed wording is not it.

When an editor works through the list of short description examples and finds that the article they're looking at really can't be captured in a just a few words, the correct response is to go broad and mention just the overall field, not to give up and feel free to ignore the fundamentals WP:SDPURPOSE, WP:SDJARGON and WP:SDNOTDEF with a short description addressed to mathematicians that will be incomprehensible to the vast majority of readers. When used within an article, the {{Annotated link}} template provides an override specifically designed for more appropriate language to be used where it is needed, for example in the See also section of a mathematics article. MichaelMaggs (talk) 13:06, 27 November 2025 (UTC)

I thought first that this was another case of clash of perspectives: "top down" (what is presented to a search) -v- "bottom up" (the list of relevant topics presented in an annotated See Also list). But now I wonder if it is actually something deeper, a world view that any topic in Maths, Science and Technology ("MST") is ipso facto jargon. If the SD of every MST topic is to be reduced to "concept in Mathematics" or "mathematical concept" [and this is by no means the first case of its kind] then the whole principle of Short Descriptions is fatally undermined.

So, although it is true that editors are advised to supplement the SD presented by a {{anl}} if the context merits it [or even do a complete local variation], that is for the 'long tail' cases. The base SD should be meaningful and if that means breaking the arbitrary 40 character limit, let it be broken. To put "Try to be more specific" another way, WP:Think of the reader (and not just those readers who discarded MST studies at age 16). 𝕁𝕄𝔽 (talk) 14:19, 27 November 2025 (UTC)

Your first thought, that this is a "top down" -v- "bottom up" discussion, is probably right. This isn't to do with anything mathematical being ipso facto jargon – I can't speak for others but that is absolutely not my world view. But in this example expecting more than a very small proportion of readers who see Star coloring in a search result to recognise, wihout more, that "graph" is being used in a very specific technical sense fails the fundamental WP:SDJARGON rule "use simple, readily comprehensible terms that do not require pre-existing detailed knowledge of the subject". Better might be something like "Type of mathematical graph coloring". MichaelMaggs (talk) 14:38, 27 November 2025 (UTC)

"Type of mathematical graph coloring" conveys, beyond the article title "star coloring", only the clarification that this is part of mathematics (rather than, as one might guess, astronomy). That is, it is almost completely content-free and unspecific. It is a bad short description. It does not disambiguate any search that is looking for a specific type of mathematical graph coloring and where the searcher wants to pick this one out of all the other search results. It is useless as an annotated link in any mathematics article, where one already knows that the topic is mathematical. In short, it has no value as a short description. Your argument about "a small proportion of readers" is pointless: only a small proportion of readers will ever see this short description, and it is those readers, not all readers, that the short description should target. Some of them will see it because of the word "star" and those are the ones who need to know that it is mathematics, or at least not astronomy, which any mathematical terminology in the short description can convey. The rest of them will be coming to this from a context of mathematics and (if they need a short description) need one that is specific enough to distinguish this from similarly-named mathematics articles. If this is the way you think short descriptions should be written, I can only conclude that you either do not understand what short descriptions are used for or are deliberately trying to undermine their value by making them as useless as possible.

To put it a different way: our categories group items on similar topics together. Our short descriptions have the opposite function, to differentiate items on similar topics or with similar names, to ungroup them. Your vague short descriptions appear to come from a position of categorization, not of differentiation, and that makes them bad as short descriptions. —David Eppstein (talk) 21:45, 28 November 2025 (UTC)

Your "I can only conclude" deduction might have been an offensive personal attack in some other area, but in the context of short descriptions it is just hilarious. MichaelMaggs (talk) 22:42, 28 November 2025 (UTC)

Well, ok, then, maybe it is me that doesn't understand short descriptions. I think their function on Wikipedia is to be displayed in search results to help differentiate those results, and to make it easier to write short blurbs in lists of links such as see-also sections by allowing the annotated link template to copy the short description rather than hand-crafting it. Your short descriptions are almost entirely useless for that purpose: they cannot be used at all in see-also sections, and they only disambiguate searches whose results are already far apart in topic, not the common case where a specific technical word leads to many closely-related results.

So tell me: what other function am I missing? What function do you think short descriptions serve, for which vagueness rather than specificity is actually preferable? And if vagueness is so preferable, why do you still use short descriptions that are not as vague as possible, when "Wikipedia article" or "Concept" would be even vaguer? —David Eppstein (talk) 23:16, 28 November 2025 (UTC)

And that is what it means to me too. But clearly not to some. How about this for a helpful, informative, useful SD - not!

• Natural computing – Academic field

By comparison, "concept in mathematics" is dangerously close to being helpful. 𝕁𝕄𝔽 (talk) 16:00, 30 November 2025 (UTC)

To apply my position in my post below, something like "Coloring in graph theory avoiding 2-colored paths" is 49 characters with "graph theory" in the first 40 to clue readers into seeing it is not about graphs in general. "Mathematical graph coloring avoiding 2-colored paths" is slightly longer at 52 characters, but is clearer about it being about math. -- Patar knight - ^chat/_{contributions} 04:37, 1 December 2025 (UTC)

Yes, if two similarly named things are in the same general field, then a SD like "Concept/term in X" is probably not the optimal SD, and some additional information is often helpful to include. However, WP:SDJARGON, WP:SDNOTDEF, and/or WP:SD40 all still apply.

However, I'm not sure your examples are good examples of this.

The SD for Blaschke sum ("Polytope combining two smaller polytopes") is useless if you don't know what polytopes are. The average reader with some familiarity with STEM is more likely to incorrectly guess it has something to do with isotopes, polymers, and combination reactions than think of geometry. The previous SD "Aspect of convex geometry" is better even it is more vague because geometry is something most people learn about in elementary school. There also doesn't appear to be an issue of distinguishing it from Blaschke product, since that as far as I can tell, is not about geometry. However, something like "Polytope in convex geometry" is a better fusion of indicating a general field while also providing further details for those more familiar. Even something like "Polytope in convex geometry combining two smaller ones" would only be 54 characters and still have the core information in the first 40 characters.

The SD for Blaschke product is not as bad, but "Analytic function with prescribed zeros" leaves room for confusion with general data analysis techniques that involve zeros whereas the previous SD "Concept in complex analysis" at least uses "complex analysis" which while still somewhat jargony, is at least a discrete term for a mathematical subfield and is again, not really confusable with the sum article, which is about geometry. "[Mathematical f/F]unction in complex analysis bounded by zeros" or something similar would probably be a better balance while not being overlong.

My position, even in cases where there aren't similar named things in the same general field, is that a field that will be recognizable to most readers should be included in the first 40 characters, and the overall SD can then be as long as 60 characters (which IIRC fits most common use cases) for some more specific/differentiating details. -- Patar knight - ^chat/_{contributions} 04:28, 1 December 2025 (UTC)

Let's go back to first principles. The major misunderstanding that seems to have prompted this section is that it's permissible to write short descriptions to prioritise their use within {{Annotated link}} templates at the bottom of technical pages, where they are seen and used by specialists, at the expense of basic fundamental requirements such as WP:SDJARGON, WP:SDNOTDEF, and WP:SDSHORT. Template use is not and has never been one of the primary purposes of an SD, as set out in WP:SDPURPOSE. Right from the start, when the template was first created (by Pbsouthwood I believe), it was always provided as a convenience for editors who might find it useful. It was always recognised that in many cases the SD as written would not be ideal for 'See also' sections, where more specificity is often useful, and that editors would need an override functionality.

That remains the position: WP:SDPURPOSE sets out the primary purposes, which do not include template use, while simply noting as a fact that "Short descriptions are also used in lists of links" and that "The template {{Annotated link}} is available for this purpose". Wikipedia:Short description#Annotated links notes that the template "may be used", and "may be useful". Template use is simply not a defining feature of SDs. Of course there's no objection to writing a SD that covers all possible use cases, where that's feasible, but there is never any excuse for ignoring the clear admonition of WP:SDAVOID to "avoid jargon, and use simple, readily comprehensible terms that do not require pre-existing detailed knowledge of the subject".

In the example we have here, "Coloring in graph theory avoiding 2-colored paths" as suggested by Patar knight works well, as 'graph theory' makes it clear that it's something to do with mathematics (not art or astronomy) without wasting characters on 'mathematics' or 'mathematical'. MichaelMaggs (talk) 11:36, 1 December 2025 (UTC)

Some short descriptions are easy, some are not. There is no rule that could cleanly distinguish a good SD from a bad one for all cases, so we have to rely on editor discretion. Sometimes our editors possess discretion. Sometimes the best we can do is not make it worse. When all else fails, we have talk page discussion and consensus to fall back on, like any other content.

Also, if a short description is too difficult, move on and do an easier one. If you can't make it better, leave it alone. · · · Peter Southwood ^(talk): 16:11, 1 December 2025 (UTC)