Wikipedia talk:Short description/Archive 14

Archive 10

←

Archive 12

Archive 13

Archive 14

Archive 15

Archive 16

→

Archive 18

Short description length limit

I notice that over the last year a length limit has been imposed on short descriptions, apparently without broad community consensus. Each short description should: be short – no more than about 40 characters (but this can be slightly exceeded if necessary), and there seem to be editors attempting to enforce this limit, which goes against policy and guidance, in that the limit appears to have been arbitrarily imposed, has no grounding in policy or guidance, and is counterproductive to the goals of the encyclopedia by being arbitrary rule creep with no demonstrated value. It is an unnecessary restriction on the freedom of editors to develop an optimum short description for any given article. A number of other suggestions seem to also have become regarded as rules, contrary to the spirit in which they were originally written. To the best of my knowledge, the only solid broad consensus for matters regarding short descriptions was produced in the debates at the time that that short descriptions were forced upon us by WMF. It is likely that none of the more recent, prescriptive recommendations now published on the page have no standing more substantive than personal opinions and preferences of the people who wrote them. I intend to revise some of these changes to bring them back in line with policy, guidance and in some cases, reality. Rules limiting article content must fit in with the needs of the encyclopedia, and limitations should be restricted to what is necessary to achieve the needs of the encyclopedia. Bear in mind that local consensus does not overrule policy and general guidance, and evidence of necessity or benefit is required before a rule can be imposed on editors. Everyone is welcome to participate in this revision and associated discussion.

I will open by challenging the claim that Each short description should: be short – no more than about 40 characters as inappropriate. A short description should be brief within the more important constraints of being useful to the reader, valid, and not misleading. Where reasonable practicable it should be less than about 100 characters, or it is not short, but if a useful, valid and not misleading short description takes more than 100 characters, it must be as long as necessary.

The short description should also preferably be in simple, non-technical language to best serve the largest audience, but if special terms are needed to keep the length reasonable, that should be considered on a case by case basis like any other content, and discussed on the talk page. If you are reverted for a bold edit, explain your reasoning. Keeping the length to below 40 characters should not be used as a reason. Whether removing valid, useful and not misleading information from a short description in the pursuit of brevity is a valid argument is basically up to talk page consensus, like other content debates.

Notice that the constraints I have suggested are fully compliant with the goals of the encyclopedia. Anything more restrictive must be proven necessary. Cheers, · · · Peter Southwood ^(talk): 15:59, 26 August 2022 (UTC)

The "not more than about 40 characters" guidance evolved, if I recall correctly, as an informal consensus following several discussions on this page and its predecessors, though there has been no formal RFC. If by "I intend to revise some of these changes" you mean that you intend to unilaterally replace it with your own proposal Where reasonable practicable it should be less than about 100 characters, or it is not short, but if a useful, valid and not misleading short description takes more than 100 characters, it must be as long as necessary – then you should get agreement here first. That would definitely be introducing text that has no consensus as yet.

What you refer to as 'rules' are, I hope we can agree, actually suggestions or guidance on an Information page, which, according to the template, "may reflect varying levels of consensus and vetting". The suggestions are of course not formal guidelines. As the page itself says in bold "The short description is part of the article content, and is subject to the normal rules on content". It's perfectly normal for editors to refer to this and other information pages as representing what they consider to be good practice, but in every case a contrary consensus on an article talk page will take precedence. MichaelMaggs (talk) 19:00, 26 August 2022 (UTC)

As you will have noticed, I opened this discussion without making any unilateral edits, precisely because I disagree with the way changes were made and wish to follow correct procedure. I noticed a gradual change in the wording, becoming more restrictive and prescriptionary as it changed. I am not aware of the informal consensus to which you refer, (and would have objected at the time if I had noticed such discussions), and invite you to point it out to me.
Whether or not an informal consensus exists within this page's earlier discussions, the current wording of the project page is beyond the authority of local consensus, informal or otherwise, as it applies to every article in the Wikipedia, and is being used to attempt to persuade editors that 40 characters is a formal and official limit, which it is not.
Regarding existing consensus: I think there is existing local consensus that a short description that is longer than 100 characters is not short, but I am not fussy about a specific limit, or specific wording. I also think that requiring a short description to be valid is covered by policy, and requiring it not to be misleading is covered by policy. Compliance with policy takes precedence over any subsidiary guidance regarding length, therefore if a short description needs to be longer to comply with policy, it needs to be longer. Guidance cannot put a limit on the allowable length needed to comply with policy. Perhaps you disagree? Whether a short description is required to be useful is more open to debate. Personally, I think that a short description that is not useful may just as well be omitted altogether, which is what English Wikipedia originally counterproposed to WMF during the original objections regarding the use of WikiData descriptions and the related RfC, but WMF in the person of User:DannyH (WMF) insisted that we would have short descriptions, and we have had to make the best of it.
I agree that these are, or should be considered, suggestions, but they have been expressed as if they are rules, which is misleading, whether intentional or not. Wikipedia avoids unnecessary rules as this allows necessary flexibility. I have seen editors arguing about reducing a short description by two characters because it exceeded the 40 character limit, using the shortcut WP:SDSHORT as their reasoning.

If you have a suggestion for wording that is not excessively restrictive, allows full compliance with all policy, and cannot be used to bludgeon less knowledgeable editors into accepting procrustean edits to existing policy-compliant short descriptions, please make a proposal.

Until we have sufficient consensus for a policy compliant guidance statement, I propose that we remove non-compliant suggestions, guidance, or whatever description fits best, or at least flag them as being under dispute. Cheers, · · · Peter Southwood ^(talk): 09:17, 27 August 2022 (UTC)

I don't have a particularly strong opinion on this, but short descriptions are displayed in a few places (such as the Vector 2022 search results where it is truncated if too long. — Qwerfjkltalk 09:33, 27 August 2022 (UTC)

I feel that 40 chars is an appropriate number to quickly describe an article, but if others feel that the guidance requires stronger consensus, re-wording it may be helpful. There have been several cases where I have shortened descriptions, but not specifically to go below 40 chars, rather only in cases where I feel I can make it shorter without sacrificing too many details. I think that's a good thing to strive for. ASUKITE 13:11, 27 August 2022 (UTC)

From personal experience I know that there are topics that are very difficult to usefully, validly and not-misleadingly describe in under 100 characters. If short descriptions are worth having, they are worth having in a useful form. An arbitrary length limit goes against the spirit of Wikipedia. 40 characters is often enough, but it is sufficiently often not enough that is is entirely inappropriate to try to enforce it as an arbitrary limit.

If Vector 2022 does not handle short descriptions appropriately, it should be fixed so that it does. That is a simple matter of programming, and the WMF is paying people to do that programming. It is not an obstacle. The WMF does not dictate editorial decisions to the projects, and short description length is an editorial decision.

Shortening descriptions is fine provided the shortened description is not worse. I have often done it myself, and it is often possible to write a better and shorter description. No-one is arguing against that.· · · Peter Southwood ^(talk): 13:31, 27 August 2022 (UTC)

As this seems to be the same discussion addressed in the #Too-short descriptions section of this page, maybe I have to repeat exactly what I said there:

I've done many thousands of these, and in my experience long SDs are almost always the result of an editor attempting to define the subject matter, contrary to WP:SDNOTDEF. That causes problems particularly in mathematics and science where fully accurate definitions are frequently and necessarily long and technical. Such definitions, breaching both WP:SDSHORT and WP:SDJARGON, need to be shortened and simplified to achieve the purpose of the SD, namely A very brief indication of the field covered by the article; a short descriptive annotation; a disambiguation in searches, especially to distinguish the subject from similarly titled subjects in different fields.

Sometimes that can't be done by starting with a full definition and chopping out parts of it, as that results in an SD that is inaccurate. The way to go then is to use simple expressions to give the reader, in broad terms, A very brief indication of the field. The idea is to give the non-technical reader enough information to decide whether the article is or is not the one being sought. For example String theory can't reasonably be defined, but uses the broad expression "Theoretical framework in physics" to make it clear that it's something to do with physics; and Platform-independent model has "Software engineering model" to make it clear it's software-related. Baseband should be something like "Frequencies in a telecommunication signal" (41 characters), indicating that has to do with telecommunications as opposed to, say, a type of rubber band. While WP:SDSHORT does envisage going beyond 40 characters "if necessary", it's really not necessary to go beyond 45, even in the most difficult cases

I've no objection to some other recommended maximum character value provided that the recommendation is flexible and envisages longer descriptions where necessary. I don't think it's sensible, though, to have no length recommendation at all, as that doesn't give the desired gentle push to new users who very commonly try to include as much detail as they can, as if the SD is just a slightly shorter version of the lead. It also strongly encourages jargon-heavy highly-specialised definitions contrary to WP:SDNOTDEF, particularly in science and maths articles. As Asukite says, it's a good thing to have a value to strive for.

I don't know what's intended by the term "valid" in the comments above, but if it means not contrary to policy and guidelines, that's already stated on the page. An SD that briefly and non-misleadingly describes the field is pretty well what is meant by "useful". If it needs greater emphasis in different words, fine. MichaelMaggs (talk) 17:22, 27 August 2022 (UTC)

A gentle push towards brevity is acceptable, A recommended maximum length must be expressed in a way that does not allow editors to suggest that it is a hard limit and that anyone is justified in shortening it just to make it shorter, without regard to loss of usefulness.

I have also written thousands of short descriptions, some of which were trivially easy, and some of which taxed my understanding and creativity beyond their limits, and I have given up after failing to work out what the article is supposed to be about. There have also been a few occasions where the topic was sufficiently complicated that even though I could make some sense of the scope, I did not feel competent to describe it briefly and remain confident that the description would be useful, valid or not-misleading. These topics I leave to people with more appropriate knowledge. On other occasions my best efforts were over 100 characters long, and I left them like that in the hope that someone else with a deeper understanding of the topic might be able to write something that is shorter, while still being useful. What I fear is that someone will shorten them in a way that makes them useless or misleading just so they can fit a recommendation for length.

Valid is used with its normal meaning of having a sound basis in logic or fact, and implies that the short description is supported by the content of the article to which it is applied, in the same way that the lead is expected to be supported by the rest of the content. A short description which is not supported by article content would not be valid. In most cases the short description will be supported by the content of the lead, but not always, as some articles have extremely poor lead sections which give the reader little or no idea what the topic is about. · · · Peter Southwood ^(talk): 16:27, 29 August 2022 (UTC)

References

Summary

Is someone able to summarize this discussion or let me know if there is anything new here WRT § Too-short descriptions? The two discussions seem to start with the same observation. Sorry for not keeping up. ~Kvng (talk) 17:02, 28 August 2022 (UTC)

It is unclear to me why the OP started a new discussion without a reference to the detailed information in the above discussion, the phab link, and the statistics that followed the discussion. It is unclear why the OP would suggest that a 100-plus-character short description could be necessary without providing any examples, and when ZERO of the 4,600,000 articles with short descriptions have SDs of that length. 100-plus is beyond an edge case.

As discussed in the statistics section above, just 3% of articles have SDs with a length over 60 characters, and many of them have been found to be useful as well as not necessary to shorten. I think that softening the language with some descriptive rather than proscriptive language might be helpful. Something like: be short – preferably no more than about 40 characters (but this can be slightly exceeded if necessary; just 3% of short descriptions are longer than 60 characters). In the section immediately below WP:SDSHORT, the language is preferably no more than about 40 characters, which I think is better phrasing than no more than about 40 characters from SDSHORT. – Jonesey95 (talk) 11:50, 29 August 2022 (UTC)

Jonesey95, I am against prescriptive limits mainly because there is no way of knowing that there cannot be an article which needs a longer short description than those which already exist, and someone will insist that anything that can be construed as a fixed limit must be enforced as a fixed limit. Already there are shortcuts being used as reasons why short description lengths must be limited and descriptions which are entirely reasonable must be shortened. As soon as a shortcut exists, it tends to be used as a substitute for a logical explanation and rational discussion. I have no objection to recommending brevity, but not at the cost of functionality, and not in a way that can be used to intimidate inexperienced editors. Cheers, · · · Peter Southwood ^(talk): 15:48, 29 August 2022 (UTC)

Kvng, What is a too-short description? Do you mean a short description that fails to provide useful, valid, unbiased, and not-misleading information to the reader, but could provide those functions by being replaced by a longer short description? Cheers, · · · Peter Southwood ^(talk): 15:48, 29 August 2022 (UTC)

The language suggested by @Jonesey95 is good. It makes a recommendation for the usual case without being prescriptive. It cannot in good faith be misinterpreted as a fixed or rigid limit, and it gives no support to anyone attempting to intimitate by arguing the contrary. It suggests a norm – extremely useful for inexperienced editors – while allowing the flexibility that may be needed in more difficult cases. MichaelMaggs (talk) 17:24, 29 August 2022 (UTC)

I think just adding "preferably" is enough.
For reasons for the 40 character limit, see § Reasons why SDs over 40 characters are a problem. — Qwerfjkltalk 18:53, 29 August 2022 (UTC)

Pbsouthwood, at this point in the discussion, it would be helpful if you would respond to points in the actual discussion, both this one and the detailed one above, with details, links, examples, and concrete evidence rather than opining, spinning possible futures, and stating your general philosophy. I think we all understand your general philosophy; if you have specific objections to specific statements on this information page, please state what they are along with your proposed revisions. I have proposed a specific change; would it satisfy your concerns? If you have found an article, amid the two million or so remaining article-space pages without an SD, that requires a 100-plus-character short description, please link to it so that editors here can help craft a short description that is actually short. – Jonesey95 (talk) 18:55, 29 August 2022 (UTC)

Do you really expect a serious response to this request? · · · Peter Southwood ^(talk): 08:23, 30 August 2022 (UTC)

Yes, please. If you are unsatisfied with the language on this page, please propose a specific change to the page's text. – Jonesey95 (talk) 13:22, 30 August 2022 (UTC)

Ok. I will have a go. I will start out by mentioning again my point that this discussion is too limited in participation to impose a binding requirement on the format of the encyclopedia, and therefore any suggestion made here by myself or anyone else has no more value than an opinion based on inadequate evidence. The encyclopedia remains incomplete. We do not know what other articles will be added in the future, and therefore do not know what their titles and appropriate short descriptions might be. We should not be trying to make arbitrary limits where they are not necessary, and it has not been established that such limits are necessary.

I have been able to reduce the length of short descriptions longer than 100 characters, and therefore have no examples to give you. This does not prove that longer short descriptions cannot be necessary in the future. I consider it unlikely but not impossible. Providing links, examples, and concrete evidence of things that have not happened yet is not an actionable request. That they have not happened yet is not proof that they will not happen.

For a specific suggestion, see my response below.· · · Peter Southwood ^(talk): 05:14, 1 September 2022 (UTC)

@Pbsouthwood: If we're only talking about something that's unlikely but theoretically not impossible – something we can be fairly confident will be very rare, if it exists at all – would that not be a case for WP:IAR? Graham (talk) 00:33, 6 September 2022 (UTC)

Graham11, There are two aspects of length limits under discussion, Only one of them is only theoretically not impossible, the other is frequently exceeded. A counter-argument would be: why make a rule that would possibly have to be ignored, when it is not necessary to have the rule in the first place? Wikipedia traditionally avoids constraining editors by unnecessary rules, partly because Wikipedians traditionally squabble about the interpretation of rules, and some tend to weaponise them during disputes, which can be a huge time-sink and can drive away less obsessive editors, leading to a more unpleasant and unproductive editing environment overall.· · · Peter Southwood ^(talk): 03:54, 6 September 2022 (UTC)

If there is a way to misinterpret a rule, some Wikipedian will find that way. If there is no way to misinterpret a rule, some Wikipedian will nevertheless find a way. · · · Peter Southwood ^(talk): 04:02, 6 September 2022 (UTC)

• I'm wondering if instead of trying to put all these normative parameters devoid of context, why not just lay out the issues? Something like "Short descriptions exceeding 40 characters may be not be displayed in full by some tools, and SDs longer than 60 characters are quite rare. Please try to keep SDs as brief as possible as longer, more specific SDs can be less useful than a short, general one." VanIsaac, LLE ^cont_WpWS 19:26, 29 August 2022 (UTC)

  It looks like the answer to is there is anything new here is no. I do like Vanisaac's suggestion for explaining the range of workable lengths. If we put the salient points and presented evidence of these discussions in the recommendations, we're likely to repeat the discussion less often. ~Kvng (talk) 21:02, 29 August 2022 (UTC)

  I agree with Vanisaac that we should avoid normative parameters, That is basically my point, this small group of editors here, myself included, should not be setting normative parameters that affect the entire encyclopedia. The data currently available suggest that in most cases short descriptions are shorter than 40 characters, but as far as I can tell there is no statistical analysis of their quality and usefulness in comparison with length, so using that data to support a length limit is invalid. Setting out the facts is acceptable, lets just be sure that they are facts, and are set out neutrally. I have no problem with requesting editors to keep short descriptions as brief as possible providing they fulfill their purpose, but I question the utility of the claim that longer, more specific SDs can be less useful than a short, general one, as the opposite can probably also be said with equal truth, ie: longer, more specific SDs can be more useful than a short, general one. In some cases this claim may be true, in other cases the other. · · · Peter Southwood ^(talk): 08:06, 30 August 2022 (UTC)

  Jonesey95, Which part of my opening statement is not clear? Cheers, · · · Peter Southwood ^(talk): 08:15, 30 August 2022 (UTC)

  Your statement is clear, but it does not suggest a specific "change X to Y" modification to the language on this page. Perhaps I misunderstand your reason for posting. – Jonesey95 (talk) 13:22, 30 August 2022 (UTC)

  I suggest that all mention of a specific character count be omitted from the recommendations, for example:.

  Each short description should:

  • be short – no longer than is needed to fulfill its functions effectively

    The Shortdesc helper gadget displays a character count as the description is changed.

  If it is considered desirable to provide context, I suggest following Vanisaacs suggestion of laying out the issues, neutrally presenting the unbiased facts in a way that does not suggest any specific length limitation of the short description itself.

  Jonesey95's statistics can also provide context, but should include the caveats appropriate to their scope. I.e. They do not include any analysis of quality, and that many short descriptions provided through infoboxes and other such automated methods are not optimised for the specific topic. I have no analysis to support this opinion, but personal experience suggests that custom short descriptions will generally be a bit longer than template generated short descriptions, and may well be significantly more useful to the reader. This does not mean that I disapprove of template-generated short descriptions as a stop-gap, they are almost always better than nothing, but I suspect that the average length of short descriptions will become a bit longer over time as they are improved to better suit their articles, and we should not be providing advice that can be misused to hinder genuine improvements.· · · Peter Southwood ^(talk): 05:14, 1 September 2022 (UTC)

No further discussion?· · · Peter Southwood ^(talk): 05:02, 11 September 2022 (UTC)

It seems that no-one has any more to say. I will revert the advice back to what it was before the instruction creep began as there is no broad based community consensus for any hard limit or recommendations for a hard limit. Before changing to anything that implies a limit, please conduct a wider RfC announced to the wider community. · · · Peter Southwood ^(talk): 10:58, 18 September 2022 (UTC)

No, please do not do that. Having opened this section with an over-bold statement that you “intend” to make changes, you now effectively repeat the statement you started with and purport to grant yourself permission to do what you wanted to do in the first place. You appear to have overlooked the four thousand words of discussion in this section alone, in which four editors supported mention of some numerical value (perhaps with softer wording) and precisely zero supported the complete removal you are advocating. You also persist in referring to a “hard limit”, when you know perfectly well that the current recommendation says no such thing. MichaelMaggs (talk) 21:16, 18 September 2022 (UTC)

MichaelMaggs, I did and do intend to make changes, unless someone else does them first. Either by reverting to what had been the status quo before the contested changes, which I could have reasonably done before opening this discussion, or to another version which is yet to be determined, and which omits mention of a numerical limit until such limit is approved by the general community. You are free to open a RfC any time you like. I gave you the opportunity to propose a version which does not limit the community without general consensus, and am still waiting for it. If 40 characters is not a limit, it need not be mentioned, and should not be used as an argument for changing better short descriptions which exceed 40 characters, which is implied by one of the shortcuts. People are misinterpreting the current version and misusing it, therefore it needs clarification or reversion. I prefer fewer rules, less instruction creep. This discussion has had very little input from the broader community and does not constitute consensus on an issue which affects the whole encyclopedia, so it does not really matter that a small number of editors like a numerical value and wish to impose it one everone. · · · Peter Southwood ^(talk): 07:48, 20 September 2022 (UTC)

You're not helping your case by demanding that the 40 character recommended target (it has never been a "hard limit") be entirely removed as "arbitrary rule creep". It has in fact been an integral part of this WikiProject page ever since it was first created – by you – in February 2018. The exact wording has changed slightly, and is open to improvement, but 40 has always been there. It has helped editors draft around 4 million new SD texts and countless SD improvements for more than 4 ½ years. MichaelMaggs (talk) 07:39, 21 September 2022 (UTC)

It is not "my case" that needs help.

Am I correct in assuming that The short description should be as short as is reasonably practicable while serving its purpose. A target of 40 characters has been suggested, but there can not be a hard and fast limit. sometimes more words will be necessary to make sense. is the integral part that you refer to? Can I draw your attention to a) A target of 40 characters has been suggested, and b}, directly following, and clearly referring to a), but there can not be a hard and fast limit. sometimes more words will be necessary to make sense, which was not an instruction or a recognised recommendation, but a description of the situation at the time, having no standing as guidance or even as a "recommended target" just a statement of reality – it was suggested. There has been no broad community based consensus to change that in the interim to the best of my knowledge. Perhaps you can show evidence of such consensus? Perhaps not?

I recommend reverting to that original wording as an improvement on the current wording· · · Peter Southwood ^(talk): 17:53, 22 September 2022 (UTC)

The current wording in Formatting is:

• be short – no more than about 40 characters (but this can be slightly exceeded if necessary)

I think we all agree that the current wording is not optimal. Above, I proposed:

• be short – preferably no more than about 40 characters (but this can be slightly exceeded if necessary; just 3% of short descriptions are longer than 60 characters)

Pbsouthwood proposed:

• be short – no longer than is needed to fulfill its functions effectively

If we do not provide some sort of numerical information, people will forever argue about the length. I think that we need an explanation somewhere, perhaps immediately following the bullets in the Formatting section. How about moving the last bullet to the end and following it with an explanation:

• be short

and then an explanation below:

More than 80% of short descriptions fulfill the purposes listed above using fewer than 40 characters. Less than 3% of short descriptions are longer than 60 characters, and short descriptions longer than 100 characters will be flagged for editing. Short descriptions exceeding 40 characters may be truncated by some Wikipedia tools. Because they are intended to be scanned quickly, longer, more specific descriptions can be less useful.

I tend to get a little wordy, so maybe that is too long. Input is welcome. – Jonesey95 (talk) 16:16, 20 September 2022 (UTC)

Jonesey95, If you want the numerical information, we can revert to the original wording, (see above, and via the link conveniently provided by MichaelMaggs), which states that the 40 characters was a suggestion, not a recommendation, followed by the explanation of why it cannot be rule, and for that matter, should not be a recommendation.

The statistics are descriptive, and have not been analysed sufficiently to justify using them to propose recommendations, or to claim that they even fulfill the purposes listed above, as most short descriptions are in their first iteration, and a large number are automated, via infobox, and in my experience, are good enough as a start, but far from ideal, so we cannot assume how many of them might approach the ideal to the extent of making recommendations based on current lengths. I commend your work on the statistics, but they are not useful for extrapolation of trends. They are only evidence of what exists now, not of what should exist, or even of what might exist in the future. For that it would be more informative to analyse short descriptions which have gone through several changes, to see how the length trends over number of edits, preferably using examples which have had several changes, and which have stabilised as descriptions recognised as good by a panel of topic experts. I hypothesize that very short descriptions will tend to get longer, and longer ones shorter. Whether the average will end up above or below 40 characters is beyond my capacity to accurately predict, and how much variance we can expect is another unknown. A wild-assed guess might predict the mode at around 40 characters, with the 25th and 75th percentiles near 30 and 60 characters, but I would only expect the reality to start to become apparent in another 5 to 10 years. · · · Peter Southwood ^(talk): 17:53, 22 September 2022 (UTC)

Above, Pbsouthwood said If it is considered desirable to provide context, I suggest following Vanisaacs suggestion of laying out the issues, neutrally presenting the unbiased facts in a way that does not suggest any specific length limitation of the short description itself. I think that I have done that. As far as I can tell, the original wording has support from only one person in this discussion, so there is no consensus to return to it, or something like it. I am supportive of change, and I have proposed changes multiple times above. I have provided actual analysis of actual millions of actual short descriptions, and descriptive text for the page based on that analysis. Instead of helpful suggestions to guide us to a new consensus, Pbsouthwood's responses have unfortunately contained speculation and guesses such as the above. Speculation and guessing has a place, but using them to craft helpful consensus guidance for editors is unlikely to lead to positive outcomes. If you want the wording of the guidance changed, please make a suggestion that builds on the current, but suboptimal, consensus wording. Please take into account my feedback on your proposed wording change that omitted the numbers entirely; leaving out any mention of numerical length will just lead to more conflict and arguing here and elsewhere. Use facts and data, please. – Jonesey95 (talk) 19:29, 22 September 2022 (UTC)

Jonesey95,

1. I find one change that you have proposed above, at 11:50, 29 August 2022 (UTC), It is directly above the reasons I gave for why it is unsuitable. Another change is at 16:16, 20 September 2022 (UTC), which, as you have pointed out, I have not previously adequately addressed. Are there more than two multiple changes that you proposed? I may have missed something.
2. More than 80% of short descriptions fulfill the purposes listed above using fewer than 40 characters. Do they? How well? This is unlikely to have been analysed sufficiently to be able to make that claim with confidence. Perhaps More than 80% of short descriptions use fewer than 40 characters (assuming the analysis is accurate, which seems likely), and it is likely that most of them fulfill the purposes listed above (speculation, but perhaps not unreasonable, and based on having read a large number of them and more likely than the claim that they do fulfill the purposes – I am fairly sure that some of them do not, and it does not necessarily have anything to do with their length) This would be good enough and can be corrected if the numbers change.
3. Less than 3% of short descriptions are longer than 60 characters Probably true, but may not remain true. Can be changed if later analysis shows the percentage to have changed, so OK as it stands.
4. and short descriptions longer than 100 characters will be flagged for editing. Short descriptions exceeding 40 characters may be truncated by some Wikipedia tools. Does not allow for the possibility of actually needing more than 100 characters. I suggest a slight amendment: and short descriptions longer than 100 characters will be flagged for attention. Short descriptions exceeding 40 characters may be truncated by some Wikipedia tools. If the tools change the explanation can be amended to suit without controversy
5. Because they are intended to be scanned quickly, longer, more specific descriptions can be less useful. Fair enough.
6. I do think that be short – no longer than is needed to fulfill its functions effectively is a useful qualification. It should be implied by other content on the page, but some people may not notice the context when their attention is drawn by a shortcut. If nobody else wants it, I am not going to push the point. It can be added later if confusion is common enough.
7. How do you propose making it clear that the explanation is not a rule? I would like to see a mock-up of the whole section to see how it reads, but we may be getting there.
8. The original wording was a description of the situation at the time, it does not require support from anyone in the discussion, A recommendation that is in effect a change to the Manual of style requires support from more people than have been involved in this discussion (about 7 if I haven't missed anyone). The comparison is irrelevant.
9. The actual analysis of actual millions of actual short descriptions has no analysis of the actual quality of those short descriptions, quality being how well they fulfill the described purposes, and is therefore of little relevance in making recommendations which are intended for improving quality of short descriptions.
10. One of the places for speculation and guessing is when considering future trends for which little relevant data is available, Speculation is one of the early stages of the scientific method. Without speculation one will not come up with a testable hypothesis, and may end up assuming that available data is relevant and sufficient because the problem has not been sufficiently analysed and that is all you have. It is a common error.
11. Speculating or guessing that the current distribution of short description lengths is a necessary or desirable model for recommending future short description lengths is not supported by logic, necessity, or policy.
12. Leaving out all mention of numerical length in recommendations removes the arguments about how the numerical length was proved to be relevant and valid, which would be a good thing, since so far the relevance is known to be sketchy, and the validity is unproven, and is based on speculation and guesses about the quality of existing short descriptions and the effects it might have on the quality of future short descriptions. Mentioning the current distribution of lengths without claiming it to be of any relevance for guidance is just stating the facts as we know them, therefore acceptable. Instead of arguing about the length, provide a better short description which is shorter, If you cannot, maybe the length is necessary. If there is disagreement as to which version is better, that is a content matter and should be sorted out by talk page discussion.
13. Label and dismiss is not a valid form of logic, particularly when the labeling is vague, inaccurate, and poorly targeted.

Cheers, · · · Peter Southwood ^(talk): 16:51, 23 September 2022 (UTC)

As requested immediately above, here's a mockup of the Formatting section, with comments incorporated:

Formatting

Each short description should:

• be written in plain text – without HTML tags or wiki markup
• start with a capital letter
• avoid initial articles (A, An, The) except when required for correct grammar and meaning
• avoid a final full stop
• be short – no longer than is needed to fulfill its functions effectively

  More than 80% of short descriptions use fewer than 40 characters. Less than 3% of short descriptions are longer than 60 characters, and short descriptions longer than 100 characters will be flagged for attention. Short descriptions exceeding 40 characters may be truncated by some Wikipedia tools. Because they are intended to be scanned quickly, longer, more specific descriptions can be less useful.

My apologies if I missed any feedback above. – Jonesey95 (talk) 17:29, 23 September 2022 (UTC)

Looks good as it stands in isolation. I need to take a look to see how it integrates with the rest of the page. Will get back to you soon, but it may take up to a day due to other stuff happening. I don't think you missed anything. Cheers, · · · Peter Southwood ^(talk): 16:33, 24 September 2022 (UTC)

Jonesey95, I am happy with this version. We will need to remove the redundant and soon to be inaccurate repetition of the 40 character advice from the Examples section just below the Formatting section, Here are some examples, though they can all be varied if the context allows something better. See also § Inclusion of dates. Remember to ensure your description is short: preferably no more than about 40 characters, as it will be adequately covered in the formatting section, it could be replaced by (See above), and the shortcut WP:SD40 that mentions 40 characters should be deleted in keeping with the descriptive character of the new mention. Cheers, · · · Peter Southwood ^(talk): 09:30, 25 September 2022 (UTC)

Updated. I took the liberty of formatting the "be short" notes as bullets instead of what looked to me like a wall of text. – Jonesey95 (talk) 17:35, 26 September 2022 (UTC)

Thanks Jonesey95, that is fine, though as Wikipedia walls of text go that was an unusually low wall;-) · · · Peter Southwood ^(talk): 07:34, 27 September 2022 (UTC)

Thank you, thank you, thank you !!! — ALL of you. I’m a relative newby, only a copy editor, and spend ~92.53% of my extremely limited WP bandwidth on SD’s & Intros. I’ve been bitten a few times on SD length (both directions); I really appreciate this clean solution set and all the time y’all put into it. (I also have some concerns on format, consistency, and quality…. largely irrelevant to the many happy hours that you folks put into this fix. They will keep, for now. And I hasten to add that I’ve made plenty of mistakes, and am trying to learn as I go.) This helps me a lot. All blessings!! Left Central (talk) 03:36, 24 October 2022 (UTC)

Former and defunct

I've noticed a number of SDs starting with "Former ..." or "Defunct ...", or ending with "now defunct". Those are unlikely to be useful for disambiguation or clarification. In general, our SDs are timeless, but if that information is useful, I'd think that using precise dates ("Computer hardware manufacturer, 1960-1980") would be better, leaving the detail to the end (where it might be truncated).

I suggest that we explicitly say that starting with "former", "defunct", "late", etc. is not recommended. Thoughts? --Macrakis (talk) 16:49, 25 October 2022 (UTC)

It's always a good idea to check the archives. See this discussion from December 2021. – Jonesey95 (talk) 18:36, 25 October 2022 (UTC)

Thanks. Based on that discussion, I've updated the Project Page to say that those adjectives should be avoided. --Macrakis (talk) 21:46, 25 October 2022 (UTC)

@Macrakis, I agree with the text of the update you made, but I think it belongs in the "Content" sub-section (there is a "Content" section and a "Content" sub-section"; I am referring to the sub-section) rather than the "Formatting" sub-section. I also think it could be helpful to add an anchor SDAVOID and shortcut WP:SDAVOID to link to that content sub-section. – Archer1234 (talk) 22:12, 25 October 2022 (UTC)

@Archer1234: OK (although there is "avoid" content in the next section, too). --Macrakis (talk) 22:42, 25 October 2022 (UTC)

Protected edit request on 29 October 2022

Replace: {{#ifeq:{{{pagetype}}}|Disambiguation pages|s}}

with: {{#ifeq:{{{pagetype}}}|Disambiguation pages|s}}{{#ifeq:{{{pagetype}}}|Disambiguation page|s}}

i.e. insert the second clause, making the plural usage "descriptions" active for the type "disambiguation page" (singular).

Reason:

This is the actual usage, and there's an avoidable category redirect as a result.

E.g. Template:Place name disambiguation includes Template:Disambiguation page short description (singular "page" & "description"), which adds Category:Disambiguation pages with short description (plural "pages", singular "description"), which is a redirect to Category:Disambiguation pages with short descriptions (plural both).

This is e.g. visible at Template:Short description/doc#Pagetype parameter, which shows the disambiguation category with zero members (because it's the singular form, which is a redirect). Sai ¿?✍ 14:13, 29 October 2022 (UTC)

@Saizai, this looks like it would add pages to Category:Disambiguation pages with short descriptionss. — Qwerfjkltalk 19:44, 29 October 2022 (UTC)

 Not done for now: Per Qwerfjkl's comment. Izno (talk) 07:52, 10 November 2022 (UTC)

Category:Disambiguation pages with short descriptions has been nominated for discussion

Category:Disambiguation pages with short descriptions has been nominated for possible deletion, merging, or renaming. A discussion is taking place to decide whether this proposal complies with the categorization guidelines. If you would like to participate in the discussion, you are invited to add your comments at the category's entry on the categories for discussion page. Thank you.— Qwerfjkltalk 18:46, 12 November 2022 (UTC)

Religion in short description; specificity more broadly

An edit recently revised "American Mormon leader" to "American religious leader". I can see that more general description can be better under many circumstances, but I'm wondering if there has been discussion of this type of distinction. For instance "footballer" vs. "sports person", "governor" vs. "politician", "surgeon" vs. "doctor". This is related in some ways to the inclusion of dates in the short description in that a guideline could help to reduce churn of values between camps with different points of view on "proper" content. --User:Ceyockey (talk to me) 16:28, 20 November 2022 (UTC)

Although there was no edit summary, that was probably done because the LDS church advised that members be called "Latter-day Saints" instead of Mormons. MB 17:17, 20 November 2022 (UTC)

The specific should be preferred until it is too specific. If Mormon is truly contentious then maybe it should be avoided (not a debate I can weigh in on), but unless the person was also a leader of other religions then Mormon is better; if a player is truly best known for multiple codes, or if the football/soccer label is too much, then use sportsperson (though I think sportsman and sportswoman are probably preferable); if a governor could be well known for different periods of their political career then use politician; etc (I’m sure readers here can infer the doctor/surgeon case). — HTGS (talk) 01:16, 21 November 2022 (UTC)

I usually try to copy whatever description is in the lead, assuming that the text there is the consensus. If the lead says "Joe Smith was a Turkish religious leader", then I put "Turkish religious leader" in the SD, even if the person was Muslim, Christian, or Wiccan. – Jonesey95 (talk) 15:57, 21 November 2022 (UTC)

In that example I recommend changing the lead. — HTGS (talk) 19:35, 21 November 2022 (UTC)

> The specific should be preferred until it is too specific.

I agree. Spel-Punc-Gram (talk) 14:09, 28 November 2022 (UTC)

Text of template – the actual short description – should displayed in Visual Editor

When a user is editing an article in the Visual Editor, this template displays as "Short description": this is only normal for templates that have no text element; templates with a text element show as that text.

Viewing the description text is a hassle: it is two clicks away, only visible when actually editing the template.

Therefore I'd like to request that this template be changed so that it actually shows the short description text.

Spel-Punc-Gram (talk) 14:17, 28 November 2022 (UTC)

Screenshot of Abraham Lincoln

The screenshot of the Abraham Lincoln article in this project page has a short description that no longer matches the short description on the Abraham Lincoln article. Jamplevia (talk) 16:06, 2 December 2022 (UTC)

Protected edit request on 27 November 2022

{{Annotated link}} apparently produces erroneous results when {{short description}} has a missing or blank first unnamed parameter. I have attempted to check for this situation with this change to the sandbox. I tested the change here (no parameters) and here (multiple empty parameters), and it appears to work. If other editors think that this change will work with no side effects, I believe that it should be implemented. There is no reason that I can think of to have an empty first parameter. As of earlier this month, there were five articles with empty |1=, so this is an actual (although rare) problem.

We may also want to check parameter 2 to see if it contains a value other than "noreplace". – Jonesey95 (talk) 02:13, 27 November 2022 (UTC)

Would it be better to fix Template:Template parameter value so that blank parameters do not cause confusion? — Martin (MSGJ · talk) 13:21, 29 November 2022 (UTC)

The problem of empty short description templates would remain. The first parameter is required, but there is currently no code to enforce that requirement. – Jonesey95 (talk) 15:14, 29 November 2022 (UTC)

Okay, done that. Did you consider a warning message as well? — Martin (MSGJ · talk) 23:04, 7 December 2022 (UTC)

I did not, since the short description is not normally displayed. I think the crew of gnomes here will be able to monitor the category. – Jonesey95 (talk) 00:46, 8 December 2022 (UTC)

Auto-generated SD for languages

Would it make sense to generate SDs for the ~9,400 articles with {{Infobox language}}? Could be "(Extinct or Endangered) (family) language (of country if one or two countries are mentioned in "Native to" / otherwise could be continent or region). Examples:

• English language: "West Germanic language"
• Nanerigé language: "Senufo language of Burkina Faso"
• Gogodala language: "Papuan language of Papua New Guinea"
• Shira language: "Bantu language of Gabon"
• Daju Mongo language: "Eastern Sudanic language of Chad"
• Nyah Kur language: "Austroasiatic language of Thailand"
• Pa-Hng language: "Hmong–Mien language of China and Vietnam"
• Tamien language: "Extinct indigenous language of California"
• Trumai language: "Endangered language isolate of Brazil"

a455bcd9 (Antoine) (talk) 10:32, 19 December 2022 (UTC)

Less than 1.5 million missing…

Hi folks. Looks like we just reached somewhat of a milestone with the number of articles lacking a short description having just dropped below 1.5 million for the first time since the feature was added? Link That would mean less than a quarter of English Wikipedia's about 6.5 million articles don't have a short description. Progress! Kind regards, Robby.is.on (talk) 20:15, 23 December 2022 (UTC)

Removing categories for more namespaces

Following the deletion decisions at CFD here and here, more namespaces are to be de-categorised following Wikipedia:Categories for discussion/Log/2022 December 15#Pages with short description.

Qwerfjkl or other interested editors: I have implemented this in Template:Short description by extending the previous code, but please review the code again, as it could no doubt be streamlined to work only in article space (using {{Main other}}?) rather than excluding a list of other namespaces. – Fayenatic London 09:55, 24 December 2022 (UTC)

Opportunity for a MediaWiki tag?

This seems like a kludge that could be productized into a cached tag extension for the MediaWiki ecosystem (mediawiki-extensions-MetaDescriptionTag). The current solution relies on .nomobile, which was recognised as an antipattern. IIRC the tech stack was changed to expunge the class. 82.14.9.253 (talk) 17:40, 24 December 2022 (UTC)

Pages with most views.

I requested a very useful query at WP:QUERY that was just fulfilled at User:Uhai/Pages without short descriptions by view count. It contains the top 1000 viewed articles without short descriptions. Would anyone like to contribute on this? — PerfectSoundWhatever (t; c) 00:30, 5 January 2023 (UTC)

I recommend simply removing lines from the report once they are done rather than taking the time to mark them as "done". – Jonesey95 (talk) 01:23, 5 January 2023 (UTC)

While true, all I have to do is copy and paste {{ya}} to each article that I finish. At least in source editor, I feel deleting each row is more effort. You have to make sure you delete the right amount of lines and stuff. It's very simple to delete rows in vis editor though, so maybe it's better to switch over... — PerfectSoundWhatever (t; c) 02:16, 5 January 2023 (UTC)

Only 50 remaining now. MichaelMaggs (talk) 19:11, 2 February 2023 (UTC)

Now all done. MichaelMaggs (talk) 21:31, 2 February 2023 (UTC)

Module:AnnotatedLink listed at Requested moves

A requested move discussion has been initiated for Module:AnnotatedLink to be moved to Module:Annotated link. This page is of interest to this WikiProject and interested members may want to participate in the discussion here. —RMCD bot 23:33, 2 February 2023 (UTC)

To opt out of RM notifications on this page, transclude {{bots|deny=RMCD bot}}, or set up Article alerts for this WikiProject.

Dispute on short description at Marshfield station

Opinions on the utility of two alternative short deacriptions are requested. · · · Peter Southwood ^(talk): 08:06, 8 February 2023 (UTC)

Short description for Marshfield station

There is a discussion on the appropriate short description for Marshfield station at Talk:Marshfield station which would affect several articles on the mass transit system of Chicago and other pages. The discussion concerns the purpose of short descriptions and whether local consensus or precedence overrides broader agreements on the purpose and style recommended for short descriptions in the encyclopedia as described in this project. Members of this project and any other interested parties are invited to consider the implications and provide input. · · · Peter Southwood ^(talk): 05:53, 12 February 2023 (UTC)

Proposed change to WP:Short description#Content

I propose to change current text
A short description is not a definition and should not attempt to define the article's subject nor to summarise the lead. to:
A short description is not a definition and need not attempt to define the article's subject nor to summarise the lead.

Current text implies that a short description that is incidentally a definition would be wrong, but if a definition serves the purpose adequately it should be quite acceptable. This is a minor point, and I am not expecting any objection, but you never know, and it is polite to ask, Cheers, · · · Peter Southwood ^(talk): 09:49, 13 February 2023 (UTC)

I do object to that, as the wording "need not attempt" clearly implies – as I read it, and in conflict with the sentence's opening words – that an editor drafting a short description can, if they wish, positively attempt to define the article's subject or to summarise the lead, but that that's not essential. While I agree that an incidental definition is perfectly OK, we should not encourage editors to start thinking in terms of definitions, as they are very often too complex, jargon-filled, and unnecessarily long.

I suggest: A short description is not a definition, and editors should not attempt to define the article's subject nor to summarise the lead. Of course there is no objection to an otherwise-suitable short description that also happens to work as a definition. MichaelMaggs (talk) 11:42, 13 February 2023 (UTC)

I don't read it that way, but it seems it could be confusing for some people as you do read it that way. Your alternative basically looks fine to me. However I would leave out the "of course" as it does not add anything useful.
Would you accept A short description is not a definition, and editors should not attempt to define the article's subject nor to summarise the lead. There is no objection to an otherwise-suitable short description that also happens to work as a definition.? Not a crisis either way. If you think the "of course" is useful I can live with it. I do agree that clarity is important, and that in most cases tying to define the topic would not work for the reasons you give. Cheers, · · · Peter Southwood ^(talk): 15:54, 13 February 2023 (UTC)

Happy to omit "of course". MichaelMaggs (talk) 16:38, 13 February 2023 (UTC)

DOB

Is it really necessary to include the date of birth in a short description, when the description has already made it obvious who the person is, for disambiguation and other purposes, especially for BLPs? I see that quite a lot of them do, and it just doesn't sit right with me. IMO it should be discouraged. Laterthanyouthink (talk) 01:28, 1 January 2023 (UTC)

Check the archives for one or more pretty good discussions that led to the current guidance. – Jonesey95 (talk) 05:20, 1 January 2023 (UTC)

Dates are recommended if and only if they are useful. WP:SDDATES says "sometimes there may be more important information that needs to take precedence", which is often the case for articles on living people. Even where useful, the guidance also notes that "Care should be taken where the biographies of living persons (BLP) policy applies: birthdates for living people should not be included unless sourced within the article." But generally they are indeed encouraged. MichaelMaggs (talk) 09:29, 1 January 2023 (UTC)

Thanks, and I read the hits on DOB in the archives. I'm constantly on the lookout for unsourced DOBs in articles, which is partly what led me to ask the question. It just seems redundant on SDs in all but a few cases, to me - considering how I see SDs trimmed so often, including dates of films (which IMO is more useful than living people's birthdates, which are for the most part unimportant). Anyway, I'll go with the consensus, if that's been decided. Laterthanyouthink (talk) 10:02, 1 January 2023 (UTC)

@Laterthanyouthink: DOBs aside, years of films should almost never be excluded. — HTGS (talk) 02:27, 15 January 2023 (UTC)

Yes, I agree about films, and was doing so regularly until sometime last year when an editor started changing some I'd added, reducing the SD to "British film" or "Australian film". I haven't done any recently, I don't think, but will return to adding the dates. Laterthanyouthink (talk) 03:58, 15 January 2023 (UTC)

There is unlikely to be a good reason to remove the date of a film unless the short description is long. "British film (date)" or similar is not long, and the editor truncating useful disambiguation information should probably be advised to stop. · · · Peter Southwood ^(talk): 06:05, 14 February 2023 (UTC)

Abbreviations, initialisms and acronyms

These may be unfamiliar and therefore confusing to the average reader. Should we specifically discourage their use if reasonably practicable? The counterargument is that they are usually a way to make the description shorter. · · · Peter Southwood ^(talk): 09:10, 18 February 2023 (UTC)

@Pbsouthwood, I think this is written in the MOS somewhere, but common abbreviations e.g. c. are okay, whereas more obscure ones like fl. are not. — Qwerfjkltalk 10:25, 18 February 2023 (UTC)

From WP:SDDATES:

Where a date is not known exactly, "c. " may be used for "circa". Other examples are given at WP:APPROXDATE, although "fl. " for "floruit" should be avoided as it is not universally understood. Centuries should not be abbreviated "c. " due to the potential for confusion with "circa".

— Archer1234 (t·c) 10:34, 18 February 2023 (UTC)

I think it's already covered well enough by avoid jargon, and use simple, readily comprehensible terms. So Television -> TV and United Kingdom -> UK are fine, but I often find myself expanding or otherwise avoiding acronyms that will be familiar only to readers who already know about the subject. Sports, TV and transport articles are the worst for jargon acronyms. MichaelMaggs (talk) 11:26, 18 February 2023 (UTC)

Thanks everyone, I think we are all on the same page here. We are relying on the editor having a reasonable knowledge of MoS and a modicum of common sense.· · · Peter Southwood ^(talk): 13:56, 18 February 2023 (UTC)

Some ridiculous possibilities that should probably be handled better

I've just been writing test cases for Module:GetShortDescription and had to create a bunch of pages under my sandbox/ to store the various wacky possible configurations the module has to handle. Other than the various mixes of numbered and unnumbered params including |pagetype= and |2=noreplace; {{short description|         }} will produce an eight character short desc entirely of whitespace (see the example), and {{short description|||}} will produce an empty short desc. Even {{short description}} renders is treated as existing. I see empty descs are added to Category:Pages with empty short description, but why render treat them as existing at all? I also see that my lovely string of whitespace isn't considered to be empty. Fred Gandt · talk · contribs 19:49, 26 January 2023 (UTC)

Fred Gandt, does all this matter? Cheers, · · · Peter Southwood ^(talk): 09:53, 13 February 2023 (UTC)

Detection of nothing useful as something useful is just wrong. All cases of empty short descriptions should be treated as not being set, and although slightly more difficult to measure (due to there being a large variety of whitespace characters with legitimate uses), all strings of only whitespace should also be treated as not being set. These cases are not explicitly none, but they're also not short descriptions. Does it matter? If the system detects that a short description is set, and that description has no actual content, its page is miscategorised as having a short description, and that description displayed in e.g. mobile searches, even though there's nothing to see. Fred Gandt · talk · contribs 15:06, 13 February 2023 (UTC)

I think I see now. What you are doing is more complicated than I expected. Unfortunately it is also too complicated for me to suggest anything helpful, but at least I can offer my sympathy;- Cheers, · · · Peter Southwood ^(talk): 15:37, 13 February 2023 (UTC)

I was flat out wrong an aspect of the handling of empty short descriptions; they are not only categorised; even though the HTML is created for them, the servers do understand that the page has no short description (bit weird but okay). Strings of white-space though are treated as meaningful, not categorised and are shown as descriptions where descriptions are shown (mobile searches etc). It's not the end of the world, but the creation of HTML for an empty short description is, as I say, weird, and a description consisting of nothing but white-space is clearly no use to anyone. Fred Gandt · talk · contribs 16:11, 13 February 2023 (UTC)

Are they parsed any differently than if they were part of body text?· · · Peter Southwood ^(talk): 05:43, 14 February 2023 (UTC)

See https://phabricator.wikimedia.org/T326898#8653326 as well -- I'd like to use `{{SHORTDESC:}}` to handle the `none` case, and have the template handle `{{Short description|}}` (and possible other all-whitespace values) by putting the page in the category "empty short descriptions" but *not* using the `{{SHORTDESC}}` magic word in that case. Does that sound reasonable? C. Scott Ananian (talk) 16:38, 28 February 2023 (UTC)