Notice: file_put_contents(): Write of 1060530 bytes failed with errno=28 No space left on device in /opt/frankenphp/design.onmedianet.com/app/src/Arsae/CacheManager.php on line 36

Warning: http_response_code(): Cannot set response code - headers already sent (output started at /opt/frankenphp/design.onmedianet.com/app/src/Arsae/CacheManager.php:36) in /opt/frankenphp/design.onmedianet.com/app/src/Models/Response.php on line 17

Warning: Cannot modify header information - headers already sent by (output started at /opt/frankenphp/design.onmedianet.com/app/src/Arsae/CacheManager.php:36) in /opt/frankenphp/design.onmedianet.com/app/src/Models/Response.php on line 20
Wikipedia talk:Manual of Style - Wikipedia Jump to content

Wikipedia talk:Manual of Style

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia

Welcome to the MOS pit[Humor]


    Style discussions elsewhere

    [edit]

    Add a link to new discussions at top of list and indicate what kind of discussion it is (move request, RfC, open discussion, deletion discussion, etc.). Follow the links to participate, if interested. Move to Concluded when decided, and summarize conclusion. Please keep this section at the top of the page.

    Current

    [edit]

    (newest on top)

    Pretty stale but not "concluded":

    Capitalization-specific:

    This section is an excerpt from Wikipedia talk:Manual of Style/Capital letters § Current.[edit]


    Move requests:

    Other discussions:

    Concluded

    [edit]

    Use of main article template

    [edit]

    Can you please tell me which is preferred?

    1. A section heading followed by a {{main}} template, followed by prose that might wikilink to that main article.
    2. A section heading containing prose with a wikilink to the main article.

    My feeling is that a {{main}} template follows Wikipedia style and is easiest for a reader who may wish more information. The opposing view thinks {{main}} is redundant when the prose wikilinks to more. I looked at a featured article, and found an example. Earth §Size and shape, first has a {{main}} and then in prose, a wikilink to the same article. I chose #1 because it avoids us sending readers on a scavenger hunt for more information. MOS:LAYOUT, WP:SUMMARYHATNOTE, and WP:HATNOTERULES all appear to agree. Thank you. -SusanLesch (talk) 19:35, 10 September 2025 (UTC)[reply]

    I prefer #1 and I agree with your reasoning. Hatnotes are occasionally overkill but they serve a distinct purpose and should be used accordingly. {{main}} signals something important to readers that is not immediately obvious from the mere inclusion of wikilinks in prose. --MYCETEAE 🍄‍🟫—talk 22:00, 10 September 2025 (UTC)[reply]
    I believe the use of {{main}} means that we don't wikilink that article in prose. I prefer that over no {{main}} and linking it in prose. – Muboshgu (talk) 17:00, 11 September 2025 (UTC)[reply]
    When, instead of having the other article, its content could logically have been included directly in the section, and the reason for having it as a freestanding article instead is a combination of size and level of detail, then the other article is the main article for the section, and it can be seen as a direct offshoot of the article containing the section. Indeed, it may have been created by splitting the longer, more detailed text from the higher-level article. In that case the {{main}} template is virtually necessary, the other article is a component of the one containing the section. This is irrespective of whether the section also links to it. Or maybe it means the link is somewhat contrived and shouldn't be there. Largoplazo (talk) 15:26, 17 September 2025 (UTC)[reply]

    Myceteae, the inline link was suggested in two rounds of discussions on Talk:Donald Trump involving SusanLesch and Muboshgu. The links were then discussed in Talk:Donald_Trump#Follow-up to Trump-Epstein which would have been the place to discuss a hatnote. As for the section itself, it's not a summary of the Relationship of Donald Trump and Jeffrey Epstein, so IMO WP:SUMMARYHATNOTE doesn't apply. The two sentences are the lowest common denominator the editors could agree upon after more than a month of discussion and two rounds of !voting. Option 2 (section heading with a wikilink) wasn't proposed by anyone. Editors agreed on an inline link in the first sentence of the text. Space4TCatHerder🖖 18:05, 11 September 2025 (UTC)[reply]

    Thanks for the background. As always, context matters for these discussions and there will be reasonable editorial decisions to take a different approach based on the article. The example given in the discussion here was related to Earth#Size and shape. The usage there makes sense and any concern about redundancy is overridden by the fact that this is likely very useful to readers, based on the use of piped links and numerous wikilinks in hatnotes and in the section body. I haven't read through the discussion you linked, but it appears the situation was sufficiently different and was subject to a lengthy discussion, so a different approach is appropriate there based on the particulars. --MYCETEAE 🍄‍🟫—talk 18:41, 11 September 2025 (UTC)[reply]
    Space4TCatHerder, thank you for accepting my invitation to post here. Your version of facts are false and stymied the discussion.
    Myceteae, thank you for your replies. Multiple {{main}}s can cause a log jam and be overkill as you say. Nevertheless, may I remind everyone.
    A controversial context is exactly where editors should follow Wikipedia precedent and execute within the MOS shared wisdom we are lucky to have. Wikipedia has hundreds, maybe thousands, of contentious articles on wars and world conflicts. Each one observes MOS—otherwise they'd have splintered into diverging encyclopedias. We should treat this subject like every other to help our readers. -SusanLesch (talk) 19:17, 12 September 2025 (UTC)[reply]

    Will anyone stand up for MOS?

    "Editors should structure articles with consistent, reader-friendly layouts and formatting"

    . -SusanLesch (talk) 14:38, 17 September 2025 (UTC)[reply]

    MOS is a Wikipedia guideline. Why isn't this discussion board stepping up to support it? Four readers have added {{main}} since this thread started. Each was reverted because one editor decided WP:IDONTLIKEIT. -SusanLesch (talk) 22:52, 4 October 2025 (UTC)[reply]

    Superseded by topic below Should MOS:COLOUR permit multiple colors in some situations?

    [edit]

    Please see discussion about MOS:COLOUR at Wikipedia_talk:WikiProject_Military_history#Using_colors_in_military_maps_to_convey_information:_MOS:COLOUR_issues? Noleander (talk) 22:24, 15 September 2025 (UTC)[reply]

    This discussion in WikiProject Military History has been moved to the MOS Talk page, in the topic immediately below. Noleander (talk) 13:58, 16 September 2025 (UTC)[reply]

    Should MOS:COLOR be updated to clarify if and when color hues may be used to convey information in maps?

    [edit]
    Red and blue distinguishing armies
    Red, blue and green distinguishing voyages
    Blue, green, grey distinguishing range of an animal
    Many colors distinguishing railways

    I have seen a lot of confusion in WP:Featured Articles and WP:Good Articles about how to apply the MOS:COLOR accessibility guideline to maps. That guideline says:

    "When using color, editors should keep accessibility for users with low vision impairments and color blindness in mind:... Color should not be used as the sole visual means of conveying information, or for distinguishing elements such as links, templates, or table rows. Always provide an alternative method [such as icons, symbols, etc]" [emphasis added].

    On the right are four maps that illustrate the confusion I observed in FA and GA. The confusion revolves around the question: When does MOS:COLOR permit the use of hue (red/blue/green/yellow/orange) as the sole means of conveying information in maps?

    Notably, everyone agrees that color lightness  – e.g dark blue/ medium blue/ light blue/white – is consistent with MOS:COLOR. Likewise, using dotted/dashed/solid line styles is acceptable for conveying information in maps.

    WP:MOS says "New content added to [WP:MOS] should directly address a persistently recurring style issue." I think this color issue is "persistently recurring" because I've seen it four times in the past six weeks, so I suspect that hundreds of editors are facing this issue monthly as they add or review maps within articles. Therefore, keeping MOS:COLOR unchanged is, in my opinion, not wise.

    Personally, I have no opinion on how to apply MOS:COLOR to maps. I raise the question here only because it is clear that many editors are confused about it.

    Therefore, I pose the question:

    Should MOS:COLOR be updated to clarify if and when color hues (such as red/blue/green/yellow/orange) may be used to convey information in maps?

    Noleander (talk) 13:56, 16 September 2025 (UTC)[reply]

    • I don't think the content of images (for example maps) falls under the purview of the MOS. —Kusma (talk) 13:59, 16 September 2025 (UTC)[reply]
      In the interest of minimizing future confusion, what do you think of adding "The MOS:COLOR guidelines do not apply to the content of images, including maps." to MOS:COLOR, perhaps as a footnote? Noleander (talk) 14:07, 16 September 2025 (UTC)[reply]
      I disagree that images should be excluded from the MOS. Images (and other media) are an important part of how we present information to our readers and we should be putting as much effort into them as we put into text. I think we too often fall back on "Not my problem, that's the best I could find in commons", which is a shame. I understand that not everybody is skilled at manipulating images, but we have projects like Wikipedia:Graphics Lab which can help. RoySmith (talk) 14:34, 16 September 2025 (UTC)[reply]
      Agreed with RoySmith -- once an image becomes part of an article, it's subject to the same standards as anything else in it. Of course, the MoS is only ever a guideline, so maps which use colour to convey information aren't strictly verboten, but we acknowledge that, in theory, it's better if some other supporting method can be used, or at least best not to use pairs of colours that are commonly the subject of vision problems (e.g. red and green) where they can be easily confused. If it really is the best in Commons, then we can discuss whether it adds net value being in, even given the accessibility issues, or whether it's better taken out. It's a similar situation with citation -- information in a map isn't exempt from WP:V because it's in a map, and reviewers at FAC, in particular, will often ask for proper verification on the image's page. UndercoverClassicist T·C 15:21, 16 September 2025 (UTC)[reply]
      I am worried about raising the standards too high (if the MOS says every map needs to be colourblind accessible we will have people going around to remove all others). Maps are extremely helpful to understand many articles. If there is a high quality map that violates the accessibility guideline, we should include it in the article because it helps many of our readers. Of course we should continue to try to get a better map. Just like we will use low resolution photographs of living people until we get a better one. —Kusma (talk) 16:05, 16 September 2025 (UTC)[reply]
      What about adding wording to MOS:COLOR that focuses on recommendations, rather than mandates. Such as: "When colors are used as the sole means of conveying information in a map, the colors should be chosen with color-blindness in mind: Using color lightness or line-patterns is recommended. Relying on hue alone is discouraged, but if hue must be used, pairs of colors subject to color-blindness (such as red/green, or orange/red) should not both be used. Consider using this external tool to evaluate the colors.". Noleander (talk) 16:14, 16 September 2025 (UTC)[reply]
    • MOS:IMAGES. EEng 14:23, 16 September 2025 (UTC)[reply]
      Could you elaborate? I could see no mention of maps or color hue in that page. Are you agreeing with user Kusma above that MOS:COLOR has no bearing on the content or colors used in maps? Do accessibility concerns impose any guidance on colors used for graphic overlays drawn in maps? Noleander (talk) 14:28, 16 September 2025 (UTC)[reply]
      I'm just responding to the assertion above that images are outside MOS's purview. EEng 14:46, 16 September 2025 (UTC)[reply]
      Okay, thanks. The indentation confused me. Noleander (talk) 14:56, 16 September 2025 (UTC)[reply]
      The MOS does not apply to outside the articles space. WP:MOS: Changing "articles" to "pages" (or any change broadening MOS's scope of applicability) would require a widely advertised RfC
      However, provisions related to accessibility apply across the entire project, not just to articles. Hawkeye7 (discuss) 19:27, 16 September 2025 (UTC)[reply]
      Thanks for that clarification ... I was unaware of those details. To restate: Although the MOS as whole does not apply to image and map files (outside article space), WP:ACCESSIBILITY (aka MOS:ACCESS) does apply to images and maps; and MOS:COLOR is part of WP:ACCESSIBILITY. Therefore MOS:COLOR does apply to map files. Noleander (talk) 19:38, 16 September 2025 (UTC)[reply]
      Huh? Images included in articles are obviously part of the articles, in "article namespace", to which the manual of style self evidently applies. The images themselves might be hosted on Wikimedia Commons, where there's (as far as I know) no restriction against hosting as many atrociously bad images as anyone feels like uploading. But there's a whole sub-page Wikipedia:Manual of Style/Images pertaining to the use of images in articles. –jacobolus (t) 19:45, 16 September 2025 (UTC)[reply]
    • Where are the discussions in question where editors are confused? ~~ AirshipJungleman29 (talk) 14:33, 16 September 2025 (UTC)[reply]
      I've seen at least four. Here are three I can find now (search for "MOS:COLOUR" in the following):
      User:Nikkimaria can probably supply others, since they are an expert in this area, and raised accessibility concerns in all of the above. And I should mention that I'm one of the confused editors: in my nomination for James Cook I felt obligated to spend a lot of time revising maps to change the colors; only later to find other editors that stated MOS:COLOR did not require the revisions. Again, I have no opinion on the matter, but I am confused :-) Noleander (talk) 14:48, 16 September 2025 (UTC)[reply]
      @AirshipJungleman29 - I found dozens, perhaps hundreds examples of the confusion by looking at links to MOS:COLOUR, here: https://en.wikipedia.org/w/index.php?title=Special:WhatLinksHere/MOS:COLOUR&limit=500 I only scanned a few of them, but it appears there are scores of cases, going back a decade, were FA nominators thought their images were acceptable, and someone (usually a reviewer?) suggested that the image failed to comply with MOS:COLOUR. Noleander (talk) 15:03, 16 September 2025 (UTC)[reply]
    Two relatvely recent discussions/comments at FAC are here (re. a graph) and in this review (re. a diagram). —Fortuna, imperatrix 15:12, 16 September 2025 (UTC)[reply]
    • There's MOS:ACCESS/MOS:COLORS (Note: to be distinguished from MOS:COLOR, mentioned above) to consider, perhaps, per WP:FA?#1F. —Fortuna, imperatrix 14:46, 16 September 2025 (UTC)[reply]
    • Only somewhat related to the original question, but I do find it frustrating that while MOS:ALT is part of the MOS, folks at WP:FAC seem to interpret It follows the style guidelines as including a carve out making MOS:ALT optional. In general, I think we give accessibility issues less emphasis than we should; we should not just be the encyclopedia anybody can edit, we should also be the encyclopedia anybody can read. And I'm still doing a bit of a slow burn that when I recently pointed out to somebody that their user page was formatted with so much fancy styling that it difficult for me to read, they blew me off with (paraphrasing) "I can read it just fine, you must have vision problems". To be sure, my vision isn't as sharp as it was 10 (or 20, 30, 40, 50, or even 60) years ago, but you'll all get here too if you survive long enough. And while I still see well enough to qualify for a driver's license, I guess I'm quickly reaching the point where I can't read an encyclopedia. RoySmith (talk) 15:18, 16 September 2025 (UTC)[reply]
      folks at WP:FAC seem to interpret It follows the style guidelines as including a carve out making MOS:ALT optional. I share your frustration here. The usual point that gets made is that we don't have good guidance on what makes good alt text, and external sources of wisdom often disagree -- but then I still think there ought to be general agreement that you can't follow MOS:ALT without at least having a go at writing some alt text, and therefore that an article with no alt text at all can't pass FAC. But that may be a discussion for another venue. UndercoverClassicist T·C 15:24, 16 September 2025 (UTC)[reply]
      I'd appreciate a ping if this gets raised elsewhere. I hadn't encountered this issue before, and it boggles the mind. Firefangledfeathers (talk / contribs) 15:25, 16 September 2025 (UTC)[reply]
      +1 Absolutely fundamental. We're effectively excluding a chunk of our readers from "seeing" the same article as others. —Fortuna, imperatrix 15:42, 16 September 2025 (UTC)[reply]
      MOS:ALT sucks and should be replaced by a useful guideline about writing helpful ALT text. (I raised this at Wikipedia talk:Manual of Style/Accessibility/Alternative text for images#Rewrite but we didn't get anywhere). —Kusma (talk) 16:08, 16 September 2025 (UTC)[reply]
      This is somewhat off topic here, but what would be a good practice for the alt text in Lexell's theorem? I just set the alt for every diagram to alt=Refer to adjacent text on the recommendation of MOS:ALT because I can't think up a way of describing these images in a way that is independently valuable, but would it be better to leave alt out or blank? In general, a diagram of a geometric proof conveys roughly the same formal information as the text of the proof, but the text is often pretty hard to follow without the image as an overview and mental anchor. I frankly don't know how to make an article like this accessible to someone who can't see (especially since most screen readers mangle mathematical formulas, which are the primary non-image content; I wouldn't necessarily mind providing explicit alt text for every formula, but we don't currently have a way of accomplishing that). –jacobolus (t) 19:54, 16 September 2025 (UTC)[reply]
    • I am colour blind and I often find maps useless because I cannot tell apart the colours. I do not have a problem if they are poster colours, that is pure red, blue, etc, with no shade of another colour. I do not know how typical I am. Dudley Miles (talk) 16:57, 16 September 2025 (UTC)[reply]
    • user:GeogSage said at WT:MIL § Using colors ... "The average person making maps for Wikipedia is not qualified." And pointed to the work of Cynthia Brewer, especially her ColorBrewer tool, as a quick shortcut to getting a better result than whatever a non-expert might naïvely first try. Quoting Brewer, GeogSage pointed out that "Hue variations are used to allow more contrast between categories, but lightness should still be the primary differentiating characteristic."
      Brewer (and GeogSage) is right here, but that's not the end of the story.
      Lightness contrast is completely essential for making anything that will be placed directly side-by-side clearly distinguishable, whether in a photograph, user interface, diagram, map, or whatever. If you place two large filled objects with very limited lightness contrast next to each-other, the result will be that they blend together; if the two are of similar color, then they will tend to just look like a uniform blob, and if they are both highly colorful but of substantially different hue they will be extremely distracting and "clashing", while also simultaneously being hard to distinguish. It is completely essential that any colors representing different data values be split by lightness (say in a choropleth map or the like).
      But in some contexts the lightness contrast between objects of one type and objects of some other type is more important than contrast between various objects of the same type, so we sometimes have a relatively limited range of contrast to work with for, e.g., several fill colors that need to all contrast with foreground text, or for several line colors or text colors that need to contrast with the same fill colors.
      Just like writing decisions, color design takes some taste, knowledge, and careful trade-offs, based on the content, output medium, intended viewership and their expectations, etc. It's hard to boil these decisions down to any kind of ironclad rules, but only general guidelines, like: Try to make the most contrast you can, while also being non-distracting in the context of a text article people are trying to read (this limits how colorful, busy, or high-contrast any image should be, and argues against very active animation), keeping labels and symbols legible, trying to support readers with non-standard color vision as best as possible, etc.
      I think more useful than trying to force any kind of universal requirements (or e.g. blocking articles from being featured if the images have issues) would be to have a place where image authors could consult some more experienced Wikipedians who have thought about these aspects of design and offer advice, or maybe some kind of team effort to produce some examples (perhaps drawn from existing images on Wikipedia/Commons) showing specific problems and possible ways of improving them. –jacobolus (t) 06:58, 16 September 2025 (UTC)[reply]
      That is an excellent suggestion: "... have a place where image authors could consult some more experienced Wikipedians who have thought about these aspects of design and offer advice" If that were implemented, those advisors would inevitably create a FAQ that summarizes the common issues that arise. Such an FAQ would be required, for example, when the advisors are busy IRL, or retire from WP, etc. Can you sketch-out the advice that hypothetical FAQ would provide to editors regarding the use of color in maps (in the context of accessibility)? Noleander (talk) 18:46, 16 September 2025 (UTC)[reply]
      Is this the kind of request that could be added to the Wikipedia:Graphics Lab/Map workshop? Sariel Xilo (talk) 19:26, 16 September 2025 (UTC)[reply]
      As an example and to follow up on my statement quoted above, I would like to use the opportunity to point out the category on Wikicommons Category:Choropleth maps using absolute numbers.
      I can list several sources if anyone is interested, but this entire category is filled with maps that the literature would consider misleading. Attempts to outright delete them have been blocked, and instead we've been told to just replace them as we go to avoid disrupting the project. Bad maps are worse then no maps, there is no defense for the maps in this category (I've done the literature review). This is ONE example of how bad the problem is with Wikipedia's cartography. Many maps have no sources, including many of the boundaries on country pages. The projections are often not documented, but when they are are often inappropriate to the map type. Color is a huge issue, but we have a lot of other glaring problems that need to be addressed. GeogSage (⚔Chat?⚔) 20:03, 16 September 2025 (UTC)[reply]
      Next thing you will be saying that bad images are worse than no images. Hawkeye7 (discuss) 21:21, 16 September 2025 (UTC)[reply]
      Maps are not the same thing as an image. Maps are models of reality, bad maps should be viewed the same as an inaccurate or misleading piece of text. All maps lie, however there are conventions and norms that help us to ensure they are white lies. If a bar chart or line graph is improperly used, it is worse then no graph at all as it might paint an inaccurate view of the underlying data. GeogSage (⚔Chat?⚔) 03:52, 17 September 2025 (UTC)[reply]
      To be fair to the Wikipedian mapmakers, most other aspects of Wikipedia have the same kinds of problems: poor sources or no sources, poor organization, inaccurate factual claims, logical fallacies, misleading interpretations, claims not reflective of the scholarly mainstream consensus, poor writing, bad API design of templates, etc. etc. But despite all its problems, it remains a marvel, and generally better than I would hope it to be when just imagining the concept. –jacobolus (t) 04:43, 17 September 2025 (UTC)[reply]
      The issue is that if I point out an article is making inaccurate claim, or if material is poorly sourced, I can remove it without much of a fuss. When it comes to maps, people demand that they must be replaced with something better, rather then deleted. That entire category I linked is filled with maps that are worse then no maps at all, not only do they present data with inappropriate symbolization leading to misleading representations, they set an example that other amateur cartographers follow. Bad maps out number good maps online, and we are part of the problem. GeogSage (⚔Chat?⚔) 04:56, 17 September 2025 (UTC)[reply]
      Bringing this thread back to the topic of this section :-) .... I'm perceiving that there is no appetite to make the MOS:COLOR hue guidelines mandatory for maps. As I read the comments above, there appears to be a strong feeling that some maps can and should use hues to convey information. Of course, there are many, many variables, and every map must be considered on a case-by-case basis. But I don't see any editor that says "maps must use colors in a in a manner that permits visually impaired readers to clearly grasp the data". So perhaps we could add text or a footnote to MOS:COLOR that clarifies that the color guidelines are recommended as an aspirational goal for maps; but may be waived if there is a valid reason. That should put an end to the frequent confusion that has been observed in FA nominations over the past decade about how MOS:COLOR applies to maps Noleander (talk) 19:34, 17 September 2025 (UTC)[reply]
      "Guidelines may be waived" also seems like the wrong framing (and our current guidelines are pretty weak). More like: people should try to do the best they can within available constraints, and most of the criteria involved should not be interpreted in a black-and-white fashion.
      If there are obvious improvements that can be made to particular maps folks should feel free to make those or offer specific critiques. In many cases a bad map is still better than no map, and in my opinion a map being ugly or amateur isn't, by itself, sufficient reason to throw it out.
      But sometimes a bad map is worse than nothing: If maps have more serious issues violating core Wikipedia policies (e.g. are inaccurate, non-neutral, not supported by reliable sources, link to sources but mischaracterize their content, etc. they should probably be removed from articles.
      There are many resources available (including free resources online) for learning about graphical design, data visualization, cartography, human vision, etc. Wikipedia's articles about these topics are often mediocre, so if someone wants to work on this topic, such articles are an obvious target for incremental (or radical) improvements. If anyone feels like making more specific advice or guidelines for Wikipedia maps per se, that would also be a good project. –jacobolus (t) 20:48, 17 September 2025 (UTC)[reply]
      Thanks, that text you suggest is useful: "People should try to do the best they can within available constraints, and most of the criteria involved should not be interpreted in a black-and-white fashion." That looks like it would address the primary goal of this topic (helping to minimize confusion about usage of colors in maps); and it appears to be consistent with the comments other editors have made above. What do you think of this wording: "Applying these color guidelines to maps is recommended, but they should be balanced with other cartographic goals and constraints. [This would be added to MOS:COLOR as either text or a footnote]. Does that sound sensible? Noleander (talk) 22:33, 17 September 2025 (UTC)[reply]
      a bad map is still better than no map I can not stress enough how much I disagree with this statement, although my definition of "bad map" might be a bit different then yours, as being ugly or amateur does not inherently make a map "bad." Bad maps break cartographic conventions that misrepresent the underlying information to the user, and the situation on Wikipedia is REALLY bad. Another example, outside color, can be seen in our locator maps, like the one of South Sudan I'm including.
      South Sudan (orthographic projection) highlighted
      Try digging into the sources for the boundary files on Wikimedia. I've found that most of the time, it is completely impossible to find out where a user is getting these from, and several list the source as "Own work," or that they are derivatives of other maps in the commons. This is the exact same thing as original research, and would not be tolerated anywhere else on the project.
      In terms of improving the state of cartography pages on Wikipedia, I can not agree more. My general mission on Wikipedia is to improve the geography literature where ever possible; I've nominated several for Wikipedia:Articles for improvement/Nominations, and Cartography is high on my list. I've done some work on Computer cartography, GIS, Distributed GIS, and created pages for Web GIS and Internet GIS, and want to take on the page for Map, but it really needs a lot of editors eyes on the problem. If you're interested in helping, I'd personally appreciate others involvement. My comments here have mostly focused on the broader situation involving maps as a whole because I want attention called to the problem, it is difficult to address the finer points of the MOS like color choice when we have so many maps that aren't following basic conventions like proper projection, thematic map type, and citing sources. Some fun pages to add to the list you have are Cartographic design, Cartographic generalization, and Map communication model. GeogSage (⚔Chat?⚔) 00:26, 18 September 2025 (UTC)[reply]
      Your example map of south sudan is a very helpful basic locator, and still would be with moderately inaccurate borders for any reader who is not trying to carefully parse every pixel of the border. It is in my opinion clearly better than having no map at all. –jacobolus (t) 00:49, 18 September 2025 (UTC)[reply]
      If I write a sentence that says, "North and South Korea are divided 39th Parallel," it would be immediately deleted. If I make a map that shows North and South Korea divided on the 39th Parallel, we would keep it until someone made something better. This is stuff wars are fought over, and different groups have different sets of boundaries. If you do not provide the source for the border you are using, it is as useful as an uncited textual claim describing the the land holdings of Genovia in the article History of Europe. We are endorsing disputed territorial claims in Wikivoice, and people with political agendas are actively exploiting this to push versions of the world map that fits their world view. GeogSage (⚔Chat?⚔) 03:48, 18 September 2025 (UTC)[reply]
      You are far over-interpreting the map here, and fundamentally misunderstanding/mischaracterizing its purpose. There are a variety of detailed maps later in the article which show various more precise claims about territory and explain more about their sources (I have no idea if those are accurate; I didn't investigate), but the basic country locator thumbnail is showing the globe-scale context, and does not make any of the precise border details even visible; we're talking about differences of 1 or 2 pixels on someone's screen, which are not even going to be distinguishable unless someone is extremely knowledgeable and peeping the thumbnail with a magnifying glass.
      A better analogy would be the sentences that say: "South Korea [...] is a country in East Asia. It constitutes the southern half of the Korean Peninsula and borders North Korea along the Korean Demilitarized Zone, with the Yellow Sea to the west and the Sea of Japan to the east." While it's true that this description is not incredibly precise, it's also still very useful to someone who is trying to figure out where in the world the place is that we are talking about. If the South Korea locator map accidentally drew the border in the wrong place by 50 kilometers (2 pixels at the scale of the locator map in South Korea), it would be a regrettable error and worth fixing, but it would basically not matter in any practical sense. –jacobolus (t) 05:02, 18 September 2025 (UTC)[reply]
      You are far over-interpreting the map here, and fundamentally misunderstanding/mischaracterizing its purpose. From the perspective of the outside literature involving cartography, I do not believe I am. Look at the map of South Sudan again, in that there are boundaries showing Western Sahara as a separate country without any indication of border disputes. This IS a political stance, and without a source it is Wikipedia taking a stance on that issue, without a citation, the Wiki IS the source. Different maps have different levels of cartographic generalization, which is why it is important to document the source. It will tell you the level of accuracy you can expect from the boundary files. Another way to look at the problem, if I drew the border of Egypt 50 kilometers in the wrong place, I could move ALL of Gaza into Egypt. GeogSage (⚔Chat?⚔) 16:56, 18 September 2025 (UTC)[reply]
      At the scale of this map, Gaza isn't even visible at all. It's like 1 pixel or something. –jacobolus (t) 17:21, 18 September 2025 (UTC)[reply]
      But Western Sahara and Morocco are visible. If a set of borders are not cited, it is original research, plagiarism, or in some cases a combination of both. Borders should be viewed as an exact quote, the South Sudan example, like many, omits the person or organization that is responsible for the base map. I suspect there are several maps in the commons that are using copyrighted material inappropriately. GeogSage (⚔Chat?⚔) 20:24, 18 September 2025 (UTC)[reply]
      I don't see what Western Sahara has to do with anything. In any case, map borders are not protected creative works, and copying them is not "plagiarism". The person responsible for the base map (User:Martin23230, who originated the file File:Africa (orthographic projection).svg) was a Wikipedian in 2009. They didn't cite a source and haven't made any contributions here since 2015.
      If someone wants to make (and keep updated) better base maps with better explained sources, that would be fine, and I'm sure widely appreciated. I'd recommend using https://www.naturalearthdata.com/ as a source. –jacobolus (t) 20:33, 18 September 2025 (UTC)[reply]
      Western Sahara is a disputed territory and the United States now recognizes Morocco has sovereignty over it, which is disputed by several other countries. The choice to include the line dividing Morocco and Western Sahara is a political stance, which is one reason for requiring citations. As I've said, for some reason, things that would not be acceptable in text, like uncited direct quotes and misinformation, are kept in map form unless someone takes the time to make something better. Per Wikipedia:Verifiability and Wikipedia:No original research, unsourced or user generated map boundaries should be removed, not kept until we find something better. This is a problem across the project. GeogSage (⚔Chat?⚔) 21:04, 18 September 2025 (UTC)[reply]
      That sounds extremely off-topic. This discussion is about colours, not borders. Gawaon (talk) 21:21, 18 September 2025 (UTC)[reply]
      I stated above, Color is a huge issue, but we have a lot of other glaring problems that need to be addressed. The discussion is about cartography on Wikipedia. On Wikipedia, we have a lot of maps that are violating basic cartographic conventions, that are missing sources for their data, that are complete user fabrications, and that are spreading misinformation either intentionally or by the ignorance of the map makers. Discussing color choices is great, but we have absolutely no consistently upheld standards when it comes to maps. Whenever these issues are brought up, it is dismissed. The current discussion is kind of like an HOA discussing how to paint houses that are violating every possible housing regulation imaginable, on fire, and actively spreading that fire to other parts of the city. GeogSage (⚔Chat?⚔) 01:05, 19 September 2025 (UTC)[reply]
      Sorry to get a bit off topic in the above.
      I meant to mention before, but one of the basic problems here is that MOS:COLOR is (a) not a very complete guideline, and (b) doesn't really try to make itself applicable to graphical content like images or maps. Its focus is on content consisting of primarily text, e.g. in lists, tables, image captions, warning banners, or the like. It has a few useful advice tidbits in this context, such as: don't rely on color alone to convey information, make sure that links are obviously clickable, and don't go overboard with colors. But it isn't an adequate replacement for serious resources on this topic (I second GeogSage's recommendation to look up Maureen Stone's advice). It leans heavily on the WCAG 2.0 text contrast recommendations, which are unfortunately fundamentally broken because they are not based on an accurate model of human perception, and as a result provide extremely misleading guidance to designers. (User:Myndex was so dissatisfied with these that he spent years making up a better text contrast metric.) I don't think anyone should be over-interpreting MOS:COLOR; I don't know how it achieved its current form, but I doubt there was too much expert effort put into crafting it. –jacobolus (t) 01:27, 18 September 2025 (UTC)[reply]
    • I think that MOS ACCESS is overly restrictive for maps. It ignores the role of alt text to remedy some visual issues. --Guerillero Parlez Moi 08:27, 17 September 2025 (UTC)[reply]
      Unfortunately writing alt text that could reasonably substitute for a complex map is not easy to do. Nikkimaria (talk) 23:48, 17 September 2025 (UTC)[reply]
      A personal anecdote to illustrate this, in a map use class I took the final assignment was to write a 10 page paper that could communicate everything contained on a topographic map, with the main point was to prove how futile such a task is. Even in a simple map, how do you write all of the spatial relationships between the objects of interest out? GeogSage (⚔Chat?⚔) 00:28, 18 September 2025 (UTC)[reply]
      I'm very doubtful that any wording about maps would work, as it would need to be so hedged about with ifs and buts that it would be worthless.
      I have instead a small suggestion: add a guideline that use of colour on maps and diagrams should be redundant with a different cue, such as text labels, tone (black/grey/white), type of shading, etc. So, you don't just colour an area blue, you label or mark it as well. Chiswick Chap (talk) 15:23, 18 September 2025 (UTC)[reply]
      @Chiswick Chap - Thanks for the suggestion " ... use of colour on maps and diagrams should be redundant with a different cue, such as text labels, tone (black/grey/white), type of shading, etc..." That wording seems to be consistent with the existing guidance in MOS:COLOR, which says "Color should not be used as the sole visual means of conveying information.... Always provide an alternative method—such as an accessible symbol and/or text [or icons or labels, etc]" [emphasis added to "sole"]. Can you clarify: are you suggesting that maps that solely rely on hue (red/blue/green/yellow/orange) are never acceptable? or are merely discouraged? What do you think about notion, mentioned above by other editors, that there are there some situations where hues may be used as the sole means of conveying information in maps (after balancing accessibility goals with other map requirements)? Noleander (talk) 16:19, 18 September 2025 (UTC)[reply]
      My own recommendation would be for authors to try to add a moderate amount of lightness contrast to lines or fills of different colors (not always feasible), try to choose colors such that the map can still be accurately read when various "color-blindness simulator" type filters are applied, depending on the context consider also using alternate graphical differences in addition to color, and ask for help if they are having trouble making a map accessible. –jacobolus (t) 16:33, 18 September 2025 (UTC)[reply]
      A web search turned up this paper about the topic of map colors specifically:
      Jenny, B.; Kelso, N. V. (2007). "Color Design for the Color Vision Impaired" (PDF). Cartographic Perspectives (58): 61–67.
      –jacobolus (t) 16:40, 18 September 2025 (UTC)[reply]
      I don't think, as I implied already, that we can use a "shall" for any of this. Hue alone is unlikely to be any better; given that light, eyesight, and computer monitors all vary, it must be a dead duck as a sole method. In particular, when there is some sort of scale, say from warm brown for mountain tops down to dark blue for ocean depths, distinguishing particular intermediate levels can be quite difficult even if there are contour lines to assist, and is pretty much hopeless on its own. Chiswick Chap (talk) 17:32, 18 September 2025 (UTC)[reply]
      @Chiswick Chap - You wrote: "Hue alone is unlikely to be any better; given that light, eyesight, and computer monitors all vary, it must be a dead duck as a sole method." That sounds like it applies primarily to gradient-type maps such as File:Alicella Projected Distribution.jpg (third map at top of this topic). Referring to the two topmost maps at the top of this topic (File:Vicksburg Campaign April-July 1863.pdf and File:Cook Three Voyages 59.png) those use red and blue to distinguish graphical overlay lines. Thousands of maps in Wikipedia use hues for graphical overlay lines that way. For those sorts of overlay situations, it is not clear to me what you are suggesting. Do you suggest that those kinds of red/blue overlays be prohibited? Permitted? Discouraged? Avoided (but okay as a last resort when balanced against other map-design factors)? I apologize for pestering you, maybe I'm too dense to grasp your intention :-) Noleander (talk) 18:41, 18 September 2025 (UTC)[reply]
      Hue, Saturation, Lightness: I supposed you were suggesting hue as the key variable, and I said it wouldn't do. Maps do indeed often use such scales: I'm saying the scales are very difficult without additional cues. Please stop pinging me. Thanks. Chiswick Chap (talk) 19:00, 18 September 2025 (UTC)[reply]
      It applies to all kinds of map elements including lines, discrete symbols, etc.
      • File:Alicella Projected Distribution.jpg has a badly chosen depth gradient because it does not have enough lightness contrast. The legend explaining the depths is extremely confusing, and the distance scale is distracting, unnecessary, and somewhat misleading for this projection. The map uses a not very effective map projection, the tectonic plate boundaries are confusing, the gray color has way too little contrast with the green/blue, etc. This map is certainly better than nothing, but was made by cartographic amateurs (biologists) and leaves a lot to be desired.
      • File:Vicksburg Campaign April-July 1863.pdf is very hard to read because the background shading is too dark and the text and pattern are much too busy. Much of the text is nearly illegible, the foreground elements are easily confused with background elements, the symbols for battles are not effectively shaped or colored, etc. The description page says it's the creator's own work, but it's not clear what that means (how was the background map generated, from what sources?) The red and blue colors for lines are okay though; they have sufficient lightness contrast and well enough chosen colors to be pretty clearly distinguishable by color blind readers.
      • File:Cook Three Voyages 59.png has very problematically chosen colors. The red and green are almost completely indistinguishable to deuteranopes and hard to distinguish for protanope. This is a significant accessibility problem, and should make this map a priority for fixing. A basic fix (choosing better colors) would be easy. A more complete fix might also involve using different line styles for the three voyages. (While we're at it, the Eckert IV projection cut near the prime meridian is probably not the best choice for this map, since a lot of the action of the map takes place at the edges where there is a lot of distortion.)
      –jacobolus (t) 19:07, 18 September 2025 (UTC)[reply]

    Tools

    [edit]

    I found what looks like a really nice tool for visualizing what an image looks like to people with various sorts of color vision deficiencies: https://www.terrific.tools/color/color-blindness-simulator. If people know of other similar tools, please list them here. RoySmith (talk) 15:48, 16 September 2025 (UTC)[reply]

    If that tool (and its website) is reliable, accurate, and stable, it might be worth mentioning it in a footnote within MOS:COLOR ... especially if the outcome of this discussion encourages maps to use hues in a manner that promotes accessibility. Noleander (talk) 16:02, 16 September 2025 (UTC)[reply]
    Thanks for this. Brb after I go through all the articles I've worked on. ~~ AirshipJungleman29 (talk) 16:30, 16 September 2025 (UTC)[reply]
    If you use Firefox, hit Ctrl-Shift-I - the accessibility menu allows you to simulate various kinds of colour blindness as well as contrast loss. Nikkimaria (talk) 03:36, 17 September 2025 (UTC)[reply]
    @Nikkimaria: Thank you for pointing that out. That's a nice tool. Along the top bar of that screen, select "Accessibility", then there's a drop-down list next to "Simulate:" where various vision patterns can be examined. It's explained here as "Color vision simulation". There are some other useful tools too. Do other browsers have that?  SchreiberBike | ⌨  13:21, 17 September 2025 (UTC)[reply]
    • Remember that this and every MOS page starts with a disclaimer that says “exceptions may apply”. A map depicting a complicated situation can probably be justified as one of those “exceptions”. Blueboar (talk) 00:25, 18 September 2025 (UTC)[reply]
      Thanks for pointing out that "exception" statement .. I was not aware of it. It looks like most editors commenting above agree that many maps need an exception to the MOS:COLOR guidelines. WP:MOS says "New content added to this page should directly address a persistently recurring style issue." There have been scores of examples in FA nominations over the past decade where confusion ensued after editors suggested that maps were not in compliance with MOS:COLOR ... so this may be a "persistently recurring" situation. What do you think about the addition of a brief sentence or footnote to MOS:COLOR? Perhaps a footnote that references this Talk page topic? Or summarizes its themes? Noleander (talk) 01:01, 18 September 2025 (UTC)[reply]
    • In addition to this, I notice people making maps tend to use graphic design software, which has limitations. With GIS software freely available that is capable of doing almost anything you can do with the paid stuff, there really ins't an excuse for not using it. To add to the list a bit:
    1. ColorBrewer is a great tool to help with picking map colors that considers color vision impairment. You can see the website here.
    2. I recommend QGIS which is free and opensource to actually make maps, you can download here.
    3. Another free software that can be used to make maps is GeoDa, which you can download here. GeogSage (⚔Chat?⚔) 04:31, 18 September 2025 (UTC)[reply]

    Is this a "persistently recurring" issue?

    [edit]

    To gauge the magnitude of this issue, I looked in the "What links here" lists for MOS:COLOUR and MOS:COLOR. Limiting the search to the year 2022, and only in the FA nominations, I found about twenty (20) discussions where MOS:COLOR was applied to maps: Some examples (not limited to 2022): 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16.

    The typical pattern is: an editor improves an article and nominates it for FA; the article includes a map with color usage that is not consistent with MOS:COLOR. A reviewer points it out. The editor is puzzled, because the map was in WikiCommons, and there does not seem to be any remedy available. Sometimes the map is removed from the article. Sometimes a new map is created. Most often, the map is left in the article and the nomination continues.

    Contrasting maps with tables: Roughly the same number of FA nominations in 2022 (about 20) uncovered MOS:COLOR issues in tables, but those reviews took a different path: nearly always, the tables were immediately rectified. So, MOS:COLOR - on its face - is understandable. It is only when applied to maps that confusion arises.

    Extrapolating: if there were 20 COLOR-map issues in FAs in 2022, then there were roughly 200 in the past decade in FA. Extrapolating to the entirely of English Wikipedia, as a wild guess, there were 2,000. And that is only the discussions that involved multiple editors ... certainly there are also a large number of situations where isolated editors consulted MOS:COLOR for map guidance and were confused. Noleander (talk) 22:07, 18 September 2025 (UTC)[reply]

    Tables are usually trivial to fix, because the they are rendered by the browser and any text editor can be used to create or edit them (sometimes smarter software or small programs can be helpful, but is not necessary). They are usually graphically pretty simple, and their styling and assessment involves relatively straight-forward criteria. They are also hosted either in the page or in a template, here on English Wikipedia.
    Maps and diagrams are typically at least an order of magnitude harder to fix, because they are made using a variety of software running on various hardware platforms, usually involving at least some steps using (often expensive) commercial software, and the published version on Wikimedia commons is just a rendered output file, not an editable version which preserves the full chain of steps/materials for creation/modification; even when a fully editable source is provided, it is often nontrivial for someone else to work with it. Various map/diagram authors have different preferred tools and incompatible workflows. There has been limited effort building broadly accessible tools for creating maps and diagrams for Wikipedia per se. Making better ones would be a very valuable project, but would take a lot of effort and probably some funding from Wikimedia. –jacobolus (t) 22:28, 18 September 2025 (UTC)[reply]
    No one will fix the problem if we continue to allow the bad maps to exist. We need to purge them all, and then spend the next few years carefully monitoring new maps added to fill the void. GeogSage (⚔Chat?⚔) 01:08, 19 September 2025 (UTC)[reply]
    For your sake I hope you aren't serious. You are definitely not going to find any consensus for a position of «let's delete most of the maps from English Wikipedia and then reject any new ones unless they are made by professionals». If someone else made this post, I would say they were writing an reductio ad absurdum parody to discredit you. –jacobolus (t) 01:15, 19 September 2025 (UTC)[reply]

    "Anybody can publish any kind of a map, however bad, and get away with it. Ordinarily a field is subject to the law of natural selection the things that are bad or inadequate fail to survive. But in cartography this law does not operate effectively because the ability to discriminate among maps is not widespread in this country."

    — Richard Edes Harrison, Cartography in Art and Advertvertising, in The Professional Geographer 1950

    "If the wine is sour, throw it out!"

    — The Agony and the Ecstasy, 1965

    I'm absolutely serious, as someone who teaches cartography I believe it is my ethical responsibility to advocate for good maps. To clarify, I think we should absolutely delete maps that violate basic cartographic standards, that spread misinformation, and that are not cited in the same way we would delete text/articles that are plagiarized or spread misinformation. This not not necessitate that we only have professional cartographers in the same way insisting on proper grammar and factual accuracy does not necessitate English professors and professional researchers. Anybody can edit Wikipedia and publish anything if others don't check it. Misinformation is not excusable because it is contained in a map. POV pushing is not acceptable when done using cartography. We don't leave bad information until someone replaces it with something better. GeogSage (⚔Chat?⚔) 01:46, 19 September 2025 (UTC)[reply]

    Proposed update to MOS:COLOR

    [edit]

    Summarizing the discussion above:

    1. There were about 20 instances of confusion related to applying MOS:COLOR to maps in FA nominations during year 2022. Extrapolating: 200 in FA the past decade, and more than 1,000 across the encyclopedia.
    2. The number of incidents of confusion meets the WP:MOS requirement: "New content added to this page [MOS] should directly address a persistently recurring style issue." (emphasis added).
    3. Review of the FA nominations shows that MOS:COLOR is applied to tables without confusion, which suggests that MOS:COLOR, by itself, is understood by most editors.
    4. Ideally, all WP map data would be understandable to visually impaired readers. Where possible, maps should use the recommended MOS:COLOR techniques (lightness shading, labels, icons, dashed lines etc) to make map data accessible to those readers.
    5. There were no editors (in this Talk page discussion above) who suggested that MOS:COLOR guidelines must be strictly applied to all maps
    6. There are many competing design goals that a map must satisfy, including sophisticated cartographic design standards. Good map design is a complex process and must consider many factors.
    7. There are some maps which may not be able to meet the MOS:COLOR guidelines without adversely impacting other map design goals.

    Therefore, I propose adding the following to MOS:COLOR:

    a) "These color accessibility guidelines apply to maps, but they should be balanced with other cartographic goals and constraints."

    or

    b) ""Applying these color accessibility guidelines to maps is recommended, but they should be balanced with other cartographic goals and constraints."

    This could be text within MOS:COLOR or, less obtrusively, included as a footnote. The new text could include a wikilink pointing to this Talk page discussion. Thoughts? Noleander (talk) 17:19, 21 September 2025 (UTC)[reply]

    As written, I think both of those run the risk of creating an exception that ends up un-writing a rule. Everything in the MoS applies to everything, but should be balanced with other goals and constraints. But then I also see that this is a persistent issue with maps but not (as you note) with tables, so perhaps there's some room to draw attention to the fact that some uses of colour are more difficult to substitute than others. UndercoverClassicist T·C 20:50, 23 September 2025 (UTC)[reply]
    @UndercoverClassicist Good point. Do any of these options move in the right direction? (boldface emphasis added to indicate key differences between the proposals):
    c) Applying these color accessibility guidelines to images and maps sometimes requires judgement when balancing them with other design principles.
    d) Applying these color accessibility guidelines to maps is sometimes difficult due to other cartographic design principles.
    e) When applying these color accessibility guidelines to maps, balancing them with other cartographic design principles may be difficult in some cases.
    f) Applying these color accessibility guidelines to maps sometimes requires judgement when balancing them with other cartographic design principles.
    All of these would include a footnote referring the reader to this Talk page discussion, so they can find additional insight, examples, and background. Noleander (talk) 21:26, 23 September 2025 (UTC)[reply]
    I think c) is the best of those: being difficult isn't normally a problem for this kind of thing (you don't get a pass on MOS:PEOPLETITLES because the rules are complicated and confusing), and it's not just maps where this could be an issue: there are other types of images where colour is useful and can't easily be done away with. UndercoverClassicist T·C 06:30, 24 September 2025 (UTC)[reply]
    Okay, thanks for the feedback. I'll wait a couple of days, and if there are no objections, I'll implement the (c) proposal. Noleander (talk) 15:59, 24 September 2025 (UTC)[reply]
    It may be worth mentioning the technical difficulty of making significant changes. Just exchanging colors in an SVG file is usually not too hard (search/replace one set of color coordinates for another), but trying to change line styles, add patterns to fills, move labels around, etc., requires some skill, access to tools, and time to do the work. –jacobolus (t) 16:42, 24 September 2025 (UTC)[reply]
    That's definitely useful information. There are three places we can convey info to the reader:
    1. Text within MOS:COLOR (probably a single sentence, such as proposal (c) above)
    2. In a (new) footnote attached to the new text in MOS:COLOR; footnote would appear at bottom of MOS:COLOR
    3. Here in this Talk page discussion (new MOS:COLOR text's footnote will link to this discussion)
    @Jacobolus Would you be willing write a couple of paragraphs here in this Talk discussion? Describing "Best practices applying MOS:COLOR to maps". It would be readily available for future editors, and may even turn into something more significant (essay?) in the future. Noleander (talk) 17:06, 24 September 2025 (UTC)[reply]
    I think would be hard/time consuming enough to do well (especially with possible discussions around it) that I don't want to commit to doing it in the short term. I'd recommend again:
    Jenny, B.; Kelso, N. V. (2007). "Color Design for the Color Vision Impaired" (PDF). Cartographic Perspectives (58): 61–67.
    Feel free to ping me in a month or two and I might feel more motivated. –jacobolus (t) 17:14, 24 September 2025 (UTC)[reply]
    I understand. It is clear from the discussions above that is a very complex and subtle topic, with many nuances. That document you cite looks excellent. I propose to include a link to it in the new footnote (at bottom of the MOS:COLOR page). Noleander (talk) 18:01, 24 September 2025 (UTC)[reply]
    As discussed above, I added option (c) to MOS:COLOR. The change was made in this diff. Noleander (talk) 18:41, 27 September 2025 (UTC)[reply]
    [edit]

    Greetings, all. Have there been discussions about tolerating wikilinks more in articles whose subject is scientific? I searched but it appears there aren't any; and if that's indeed the case, I'd start a discussion about it. -The Gnome (talk) 09:09, 28 September 2025 (UTC)[reply]

    • Wikipedia:Manual_of_Style/Linking: Articles on highly technical subjects might demand a higher density of links than general-interest articles, because they are likely to contain terminology unfamiliar to the average reader. However, do not use links as a substitute for explanation; if a technical term can be simply explained in a few words, do so. EEng 09:15, 28 September 2025 (UTC)[reply]
    That's what I was looking for. Thanks! -The Gnome (talk) 11:30, 28 September 2025 (UTC)[reply]

    Readability

    [edit]

    I came to the MOS looking for guidance on paragraph size for readability and could find none. Viriditas (talk) 19:41, 3 October 2025 (UTC)[reply]

    Try MOS:PARA (in MOS:LAYOUT): Sections usually consist of paragraphs of running prose, each dealing with a particular point or idea. Single-sentence paragraphs can inhibit the flow of the text; by the same token, long paragraphs become hard to read. That's not massively prescriptive, but then the MoS generally isn't on matters of style, and it would be difficult to pin down a general rule that applies to all cases, sentence lengths, subject matters etc. UndercoverClassicist T·C 20:09, 3 October 2025 (UTC)[reply]
    Thanks. Viriditas (talk) 20:19, 3 October 2025 (UTC)[reply]
    Added this.[2] Let's see how many seconds it lasts. I don't want to mention their name, but I know who is going to revert it. Viriditas (talk) 20:38, 3 October 2025 (UTC)[reply]
    Please, my blushes! EEng 03:37, 4 October 2025 (UTC)[reply]
    I'm not going to revert it, but I think it's too schematic expressed like this. I'd prefer a more cautious "rule of thumb"-style expression. Gawaon (talk) 21:22, 3 October 2025 (UTC)[reply]
    Can you modify what I added? Viriditas (talk) 21:24, 3 October 2025 (UTC)[reply]
    Well I tried, but of course I can't prevent the reverting if it happens. Gawaon (talk) 21:31, 3 October 2025 (UTC)[reply]
    Ok, looks good. Now, using the Hawaii article linked below as a guinea pig, you see the excessive paragraphs. What do you do? Cut them back first, or leave a message on the talk page? I'm asking because going straight to cutting them back could upset the primary editor. Viriditas (talk) 21:37, 3 October 2025 (UTC)[reply]
    I thought you were the primary editor, so ... do you mind? Gawaon (talk) 21:45, 3 October 2025 (UTC)[reply]
    Oh, I explicitly invited you below! I was talking in the abstract, in terms of approaching this elsewhere. Viriditas (talk) 21:46, 3 October 2025 (UTC)[reply]
    Ah, you got me a bit confused there! No, I didn't, and wouldn't ask for permission. Paragraphs aren't owned by anyone, and this is basic copyediting work. Gawaon (talk) 21:49, 3 October 2025 (UTC)[reply]
    Got it. Viriditas (talk) 21:50, 3 October 2025 (UTC)[reply]
    My personal rule of thumb: if a paragraph has more than 210 words, it's probably too long and should be spit. Conversely, if two adjacent paragraphs are very short (no more than three to five sentences together, say), they might be suitable for merging. But that's just me, you won't something quite like this in the MOS. Gawaon (talk) 20:13, 3 October 2025 (UTC)[reply]
    Thanks. Two questions: the 200 word rule appears to be quite common when you go looking for it off-wiki. 25 years ago, 150 words was considered the maximum in academic journal articles (not sure which countries, but assuming US). Newspaper articles are much smaller, around 40-70 words or so. But with the rise of social media, we've had a major change in the expectations of paragraph size, with younger readers expecting smaller, not larger paragraphs. A discussion from last year on the village pump asked people to check their mobile phones as a rule of thumb (most of our readers are mobile). Is there a historical reason as to why we can't seem to nail this down? It appears that 200 words is the upper limit outside of Wikipedia, yet I'm dealing with a 242 word paragraph on wiki at the moment, and this is more common than not. Second question: is adding a general range of words to the MOS out of the question? We do that elsewhere for many different things. For example, adding 30-200 words seems reasonable for a standard paragraph on Wikpiedia. Viriditas (talk) 20:21, 3 October 2025 (UTC)[reply]
    It might very well be a good idea for Wikpiedia, but here at Wikipedia we're better out without such one-size-fits-all misguidance. ;P EEng 03:57, 4 October 2025 (UTC)[reply]
    Like I said, I'd split such a paragraph if I encountered it. I don't think I've ever been reverted for splitting a long paragraph (but of course, one never knows). Gawaon (talk) 20:23, 3 October 2025 (UTC)[reply]
    I don't think I've ever been reverted for splitting a long paragraph – That's the most obvious reason (though there are other compelling reasons as well) that any kind of rule on the subject is a bad idea -- see WP:MOSBLOAT. EEng 03:57, 4 October 2025 (UTC)[reply]
    Oh, today is your lucky day! I wrote Hawaii series by Georgia O'Keeffe and need to be admitted to rehab. What's the longest paragraph there you ask? I hesitate to tell you. Send help immediately. Viriditas (talk) 21:11, 3 October 2025 (UTC)[reply]
    So why, uh, did you write it like that? Gawaon (talk) 21:25, 3 October 2025 (UTC)[reply]
    Experimenting with different styles, mostly. I do that a lot. I wrote that article in May of 2024; it passed GAN in early 2025. I only got on the paragraph bandwagon several months ago when I was informed by three younger editors that their generation does not read long paragraphs like that, mostly due to their experience growing up with social media. Since then, I've been trying to cut my paragraphs back to accommodate them, but haven't yet touched that one. Viriditas (talk) 21:29, 3 October 2025 (UTC)[reply]
    I've fixed some of it already. Gawaon (talk) 21:42, 3 October 2025 (UTC)[reply]
    Looks good. Viriditas (talk) 21:47, 3 October 2025 (UTC)[reply]
    I haven't touched the part from "Departure and return", if you or anyone else wants to continue the work. Gawaon (talk) 21:47, 3 October 2025 (UTC)[reply]
    Glad to be a guinea pig. Viriditas (talk) 21:47, 3 October 2025 (UTC)[reply]
    Done.[3] Viriditas (talk) 21:56, 3 October 2025 (UTC)[reply]
    I found a few more. Think it's fine now. Gawaon (talk) 22:33, 3 October 2025 (UTC)[reply]
    Much obliged. Viriditas (talk) 22:37, 3 October 2025 (UTC)[reply]
    younger editors ... their generation does not read long paragraphs ... mostly due to ... social media. ... I've been trying to cut my paragraphs back to accommodate them — As an editor of an encyclopedia, I find it disturbing that social media might be setting the standard for our writing. I know that we write for our readers, but surely the paragraph length should be dictated by the content, not by the apparently limited reading skills of social media users. Mitch Ames (talk) 03:17, 4 October 2025 (UTC)[reply]
    Second question: is adding a general range of words to the MOS out of the question? We do that elsewhere for many different things: is this true? Off the top of my head, I'm struggling to think of guidelines that make a quantifiable ruling on a matter of style, rather than general recommendations. We have MOS:LEADLENGTH, but that's been whittled down from a much more prescriptive length recommendation which proved more frustrating than useful in practice. UndercoverClassicist T·C 20:40, 3 October 2025 (UTC)[reply]
    I was referring to WP:SIZERULE. Viriditas (talk) 20:43, 3 October 2025 (UTC)[reply]
    Ah, yes, of course. UndercoverClassicist T·C 20:45, 3 October 2025 (UTC)[reply]
    We also have MOS:MILSECTLEN: Overly lengthy continuous blocks of text should be avoided; sections which are so long as to impede reader understanding should be broken down into subsections. There remains some disagreement regarding the precise point at which a section becomes too long, so editors are encouraged to use their own judgment on the matter. Hawkeye7 (discuss) 21:31, 3 October 2025 (UTC)[reply]
    I'm glad you brought this up because the lengthy Carl Borgmann#Academic career section has been bugging me. I want to leave it as it is, but subsections might be required. What do you think? Is the section too long? Viriditas (talk) 21:40, 3 October 2025 (UTC)[reply]
    • I know! Let's have a bot that roves through article space, calling on ChatGPT to merge and split paragraphs throughout thousands of articles per day. Problem solved -- with no thinking, consideration of the material (or of anything else for that matter) needed! EEng 03:51, 4 October 2025 (UTC)[reply]

    Counting words in a paragraph

    [edit]

    What is the easiest way to count words in a paragraph? Do you have to copy the text to an external app, like a word processor? I know about and use User:Shubinator/DYKcheck, but I think that only counts words in a whole article. Nurg (talk) 22:20, 3 October 2025 (UTC)[reply]

    I don't know about easy, but I copy "suspicious" paragraphs into a word processor to check. That also makes it easy to find a suitable split. Gawaon (talk) 22:28, 3 October 2025 (UTC)[reply]
    The easiest way is to install an extension in your browser. Viriditas (talk) 22:31, 3 October 2025 (UTC)[reply]
    Good answer. That got me on the track. Thank you. Nurg (talk) 00:20, 4 October 2025 (UTC)[reply]

    RfC: commas and ENGVAR

    [edit]

    May editors revert edits that add commas or remove existing commas from articles written in British English on MOS:ENGVAR grounds? Yours, &c. RGloucester — ☎ 01:54, 4 October 2025 (UTC)[reply]

    ♬ Comma comma, comma comma, comma chameleon! They come and go! They come and go-uh-uh-oh! ♬
    Note: Joke recycled from the last comma knock-down-drag-out. See also: The Tragedy of the Commas

    Introduction

    [edit]

    It has recently come to my attention that some of our colleagues who congregate at featured article review (FAC) have formed a local consensus to use MOS:ENGVAR as a pretext to remove commas from articles with the British English tag. I have opened this RfC to confirm whether the community supports this practice.

    One thing that I want to make clear is that I believe that the omission of commas in these cases can be legitimate, based on the guidance found at MOS:COMMA, which indicates a preference for the modern style of light punctuation. MOS:COMMA, however, allows for discretion; whether a comma is required in a given place is often a subjective judgement. In my understanding of our present guidelines, consensus through editing and discussion should determine whether a comma is actually needed in any given piece of text.

    My objection is to the use of ENGVAR to justify the removal of commas from articles tagged for British English, and to the reference by certain editors to 'American' commas. My understanding of ENGVAR is that it explicitly excludes punctuation from its purview, mentioning only vocabulary (elevator vs. lift ), spelling (center vs. centre), and occasionally grammar. Our MoS proscribes the use of single inverted commas for quotations (commonly used in British contexts) and prescribes logical quotation style (uncommon in everyday American use) on the basis of MOS:COMMONALITY. To make punctuation an ENGVAR issue is to set the stage for endless disputes, all of which will be fruitless and distracting from the process of building an encyclopedia.

    The crux of the issue, however, is that I can find no clear difference between standard British and American comma use in reliable sources. I must first disclose that I am a native speaker of British English. I am familiar with the major style guides. I have never once encountered the idea that there is some rigid distinction between American and British comma use. As all writers know, comma use is often subjective. Some British and American style guides favour serial commas, others castigate them; however, even style guides that proscribe the serial comma as a general rule allow for their use when they are needed to avoid confusion. Indeed, discretion is what both British and American style guides tend to advise. Yours, &c. RGloucester — ☎ 01:54, 4 October 2025 (UTC)[reply]

    Style guide review

    [edit]
    Succinct summary:
    Reliable British and American style guides are generally in accordance on the issue of comma use.
    There is no evidence of a clear transatlantic divide. The use of ENGVAR to enforce any kind of comma use is baseless.

    Let us take up the case of commas after an introductory phrase, the locus of the dispute at the Liz Truss article (see below for details), which brought about this RfC. A British style book, Plain Words, states: Some writers put a comma here as a matter of course. But others do it only if a comma is needed to emphasise a contrast or to prevent the reader from going off on a wrong scent...On the principle that stops should not be used unless they are needed, this discrimination is to be commended. Next, I grant you a quotation from a renowned American style guide, The Chicago Manual of Style (6:34): Although an introductory adverbial phrase can usually be followed by a comma, it need not be unless misreading is likely. Shorter adverbial phrases are less likely to merit a comma than longer ones. My reading of these two passages is that they both advise the same thing: commas may be placed after an introductory phrase, some writers do so, some do not. However, in line with the modern trend towards light punctuation, commas may be removed unless confusion will result. If my reading is correct, there is no evidence of a transatlantic divide in this instance.

    Next, please allow me to cite a passage from the style guide I tend to use in my own work: Grammar and Style in British English. The guidance reads: Introductory adverbs are traditionally followed by commas...But with the current fashion of minimal punctuation, they are now often omitted. Again, please note that no indication is made of an Americo-British distinction. The real distinction, as noted by our own MOS:COMMA, is between traditional and modern use; modern style on both sides of the Atlantic favours light punctuation. Plain Words concurs: Present practice is markedly different from that of the past in using commas much less freely.

    Next, I will cite the Oxford Style Manual (1st ed). Firstly, on page 117 it states: The modern tendency is towards use of rather fewer commas. Too few commas can cause confusion, however, just as too many can cause distraction. Since there is a great deal of acceptable variation in their use, they are perhaps the most abused type of punctuation. Note again, the mention of a 'modern' style; also note the point about 'acceptable variation'. It seems that Oxford are slightly more tolerant than the Wikipedia editors in question! Then, about adverbial phrases specifically (pg. 119): Adverbial material, whether clauses, phrases, or single adverbs, obeys no single rule regarding commas, though the length of the material and what it modifies in the sentence regulates where commas are placed. Again, 'no single rule', and no evidence of any transatlantic divide.

    Finally, allow me to refer to the MLA Handbook, another American style guide. This is information taken from the MLA Style Centre, rather than Handbook itself: A comma may generally be omitted from an introductory phrase of two or three words, but consider using a comma when you wish to emphasize the phrase. Again, the omission of commas is deemed acceptable, at the discretion of the writer. No mention is made of a transatlantic divide, nor is there any clear conflict with the British guidance above. Yours, &c. RGloucester — ☎ 01:54, 4 October 2025 (UTC)[reply]

    Details of the edits that brought about this RfC

    [edit]

    The edits that brought my attention to this issue demonstrate the problem. One gnomish editor with a penchant for copyediting added some commas to the Liz Truss article. This editor, one may note, is almost certainly a native speaker of British English, based on his editing history. Then, one of the FAC group of editors reverted his change, with the edit summary: BrE tag. Now, this edit summary would not be sufficiently clear in even the best of cases; the gnomish editor was rightly vexed, and reverted with the edit summary: 'No reason given for reverting'. He was then reverted again, this time by another member of the FAC group. The edit summary used this time was more clear, if no less vexing: rv good-faith change per previous edit summary: these commas are used in American English but are not standard in British English, which is the established variant of the article (WP:TIES). In this instance, we have the absurd situation of an editor-cum-administrator rejecting an edit made by a probable native speaker of British English for being 'American'.

    I challenged the editor who wrote this edit summary to provide a basis for his claim; despite many gyrations, he was unable to do so. As we have seen above, comma use cannot be rigidly delineated as British or American – it is a matter of discretion on both sides of the Atlantic. Kindly note again that I have no objection to the reverts made in and of themselves. The text as it is was formed on the basis of consensus, and reviewed at FAC. MOS:COMMA and the above style guides favour the modern preference for light punctuation, and can be used to justify the style used at the article in question. My objection is to the use of a baseless ENGVAR straitjacket to gatekeep British English writing, without any grounding in Wikipedia policy or guidelines, or indeed, in reliable sources and actual usage. Yours, &c. RGloucester — ☎ 01:54, 4 October 2025 (UTC)[reply]

    Survey (commas and ENGVAR)

    [edit]
    • No – Commas are outwith the purview of ENGVAR; even if they were, there is no evidence of a clear-cut distinction between allowable commas in British English and American English. Comma use in a given article is subject to editorial consensus and MOS:COMMA. If editors seek to remove unnecessary commas, they are welcome to invoke MOS:COMMA. Yours, &c. RGloucester — ☎ 01:54, 4 October 2025 (UTC)[reply]
    • No – per RGloucester and because I, too, am not aware of any clear-cut, general ENGVAR distinction in this regard. Also because I find that, when in doubt, adding commas is usually better for readability than leaving them out. Gawaon (talk) 07:13, 4 October 2025 (UTC)[reply]
    • My perception is that American English more strictly/readily uses commas after introductory phrases, such as "However,…" or "Last year,…" or "In 2023,…", whereas in British usage these are sometimes omitted so that formulations such as "However I disagree" or "In 2023 he died", or indeed longer such constructions, are more commonly encountered. Whether this reflects anything that has been formalised in any style guidelines, I do not know. MapReader (talk) 08:13, 4 October 2025 (UTC)[reply]
    • Reluctant yes. The question this RfC asks depends on a more fundamental one: are there BrEng and AmEng styles of comma use? (And I think that should have been the question the RfC asked.) If there are, clearly ENGVAR can be cited for reverts. The problem with that question is that ENGVARs are not absolute -- there are some very clear differences such as colour/color, but I've seen reverts (though I can't lay my hands on one) where the distinction was less clear cut. I don't think that's caused a problem in the past and I don't think we should start legislating this sort of thing -- instead, if someone reverts and you don't agree, debate the issue on the talk page. So I'm at yes, largely because of WP:CREEP and the inherent fuzziness of ENGVAR; reluctant, because I don't want to see people arguing that anything they say is BrEng has to be believed. But overlegislating this would be more harmful. Mike Christie (talk - contribs - library) 02:31, 5 October 2025 (UTC)[reply]
    • There appears to be no clear-cut distinction between British and American English in this regard. Graham11 (talk) 03:00, 6 October 2025 (UTC)[reply]

    Discussion (commas and ENGVAR)

    [edit]
    • Comment – Unfortunately, I don't have access to many style guides at the moment and have only snippets of some online versions. But some that I have seen do have some fairly fixed rules on commas after introductory phrases. For example, the University of Oxford Style Guide (for staff, not students) says, 'Do not use a comma after a time-based adverbial phrase.'[1] For example, 'In 2010 the most popular game among children was hopscotch.' Whereas my understanding was that the Chicago style guide would have it, 'In 2010, the most popular game among children was hopscotch.'
    There do seem to be a lot of British editors who remove commas after a time-based adverbial phrase. So the practice must be pretty widespread. One editor, Keeper of Albion, did say to me once that I wouldn't find a single British style guide that gave instructions to use a comma after a time based adverbial phrase. And to be fair to them, I never have. Dgp4004 (talk) 17:59, 4 October 2025 (UTC)[reply]
    The University of Nottingham in England is also very clear: 'However, even though it's tempting, don't use a comma after a time-based adverbial phrase'.[2] Dgp4004 (talk) 18:23, 4 October 2025 (UTC)[reply]
    The 'University of Oxford Style Guide' and the 'University of Nottingham Style Guide' are both corporate manuals for use within each respective university's promotional material, and should not be confused with authoritative sources like the Oxford Manual of Style. That some British style guides say that it is acceptable to omit the comma in this case has already been established; that does not change that some American style guides allow for the same thing, and that other British style guides allow for the comma. Chicago allows for 'In 2010 the most popular game among children was hopscotch': it clearly specifies that the comma can be omitted as long as it will not cause confusion. This is not an ENGVAR issue, but an issue of style guides. Consider similarly that most British style guides prefer -ise spelling, but on Wikipedia we allow Oxford spelling as well because it is an acceptable variant within British English. Yours, &c. RGloucester — ☎ 21:33, 4 October 2025 (UTC)[reply]
    You have access to many more British style guides than I do I understand. Is there nothing in any of the British guides that mention commas after a time-based adverbial phrases? It seems odd to me that it would be explicit in two corporate style guides and then left either vague or unmentioned in the academic ones. Dgp4004 (talk) 21:50, 4 October 2025 (UTC)[reply]
    Nevermind, there's a copy of the New Oxford Style Manual (2012) on archive.org. Just sharing here so that we've all got a copy of the text:

    When a sentence is introduced by an adverb, adverbial phrase, or subordinate clause, this is often separated from the main clause with a comma...This is not necessary, however, if the introductory clause or phrase is a short one specifying time or location: 'In 2000 the hospital took part in a trial involving alternative therapy for babies'...Indeed, the comma is best avoided here so as to prevent the text from appearing cluttered.[3]

    Now of course, shouldn't is not the same as mustn't. And Oxford, of course, is not every style guide. But it doesn't seem all that unreasonable to me for those editors you cited to claim that such a comma shouldn't be added to British English articles.
    Perhaps what we should really be discussing is altering MOS:COMMA to make Wikipedia practice clearer. Dgp4004 (talk) 22:21, 4 October 2025 (UTC)[reply]
    The argument you are making does not logically make sense. That a comma may be omitted in these circumstances in British English is something that no one contests. However, as you have seen above, American style guides also permit the omission of the comma. Take for instance the MLA, which I linked above. It explicitly gives a date-related example, and allows for omission. If omission is permitted in certain American and British style guides, it cannot logically follow that commas should not be added because an article is written in British English. This does not even take into account the fact that the use of commas is explicitly mentioned as an acceptable variant in numerous style guides, such as the ones I cited above. Your excerpt from the New Oxford Style Manual was selective. The full text reads: Indeed, the comma is best avoided here so as to prevent the text from appearing cluttered. Whichever style is adopted should be implemented consistently throughout. Clearly, if either style may be used, one cannot say that the style that uses a comma is either 'non-British' or 'American'. Yours, &c. RGloucester — ☎ 00:50, 5 October 2025 (UTC)[reply]
    Let me also cite New Hart's Rules (2014), since I think it is germane here. This guidance is more or less the same as that found in the 2012 New Oxford Style Manual, but it adds the following caveat: If commas are omitted, be vigilant for ambiguities: In 2000 deaths involving MRSA in males increased by 66 per cent – Prefer In 2000, … or recast the sentence. Please note again that the omission of commas is governed by discretion, as with all comma use. Yours, &c. RGloucester — ☎ 01:01, 5 October 2025 (UTC)[reply]
    Could you clarify why this matters? If everyone agrees that some British style guides support omitting commas in the circumstances under discussion, why would it be any more "fruitless and distracting" for an editor to dispute a comma addition "per British English" vs "per British Style Guide X"? If we're not looking to change MOS:COMMA, what is the practical outcome of this discussion? Nikkimaria (talk) 01:28, 5 October 2025 (UTC)[reply]
    As you can see above, these style guides permit the omission of the comma in these circumstances, but do not offer any prohibition on comma use. Indeed, there are circumstances where ambiguity demands the comma. Disputing a comma addition 'per British English' is unfounded when the use of a comma is permitted within British English, at the discretion of the writer. The practical outcome of this discussion will be to determine if ENGVAR can be used to justify the blanket removal of commas from a given article merely because the article is tagged as written in British English, effectively eliminating the discretion that has heretofore been allowed to the writer within that variant. I would argue that the principle of discretion should be maintained, in line with the guidance found in the style guides above.
    If we were to follow your argument, I could edit an American English article to remove commas in these circumstances 'per American English' because the MLA and Chicago specify that commas may be omitted. In actual fact, this is an oversimplification of the guidance provided. Yours, &c. RGloucester — ☎ 02:18, 5 October 2025 (UTC)[reply]
    I haven't made any argument here, other than that I don't think this discussion is worth the words being spent on it. If the discussion concludes that comma use is entirely discretionary, that is likely to lead to more disputes rather than fewer (see the previous commas discussion linked above). Nikkimaria (talk) 02:26, 5 October 2025 (UTC)[reply]
    I do not remember having said that. Comma use on Wikipedia is governed by MOS:COMMA, not 'entirely discretionary'. I simply do not see any basis anywhere for a blanket prohibition of certain kinds of comma use from British English articles on ENGVAR grounds, when this is not in accordance with MOS:COMMA or British style guides. What these editors are trying to do is impose an editorial straitjacket where one does not exist in actual practice. This will negatively impact the readability of our encyclopaedia, and raise the barrier for British English contributors who have a different style. Yours, &c. RGloucester — ☎ 02:32, 5 October 2025 (UTC)[reply]
    • Comment: Our Australian Style Guide says: Use commas after adverbial phrases and adverbial clauses.[4] Hawkeye7 (discuss) 18:28, 4 October 2025 (UTC)[reply]
      No, that's not "Our Australian style guide"—it's this ridiculous Australian Government style guide, widely recognised as shyte. I know you've latched onto it as laying down the law; please don't. Tony (talk) 02:11, 5 October 2025 (UTC)[reply]
    • @Mike Christie: I did not ask that question because 1), other forms of specifically British and American forms of punctuation are already prohibited by the MOS on COMMONALITY grounds (see my example about single inverted commas and logical quotation above), 2) editors do not have the ability to determine whether there are British and American styles – this is something that needs to be determined by reliable authorities on style. There are some mistaken Britons who believe for instance that the '-ize' spelling is American, when in fact this is an acceptable and recognised variant within British English. Do you really think it is a good idea to allow editors to go around removing certain commas from articles written by British writers on ENGVAR grounds, when this practice is not actually rooted in consensus? Yours, &c. RGloucester — ☎ 02:43, 5 October 2025 (UTC)[reply]
      We have established above that there does appear to be some difference in usage, across the Atlantic, and there are some instances where punctuation markings are part of ENGVAR, such as the treatment of abbreviations, contractions and acronyms. But the punctuation of a sentence is clearly covered by MoS guidelines common to all varieties of English, and therefore the answer to the question as you asked it is clearly ‘No’. Nevertheless I don’t see anything in COMMA that directs as to whether "Nevertheless, he went and did it" or "Nevertheless he went and did it" is preferred, and hence it surely comes down to editor preference and consensus? When I have created content, as British I have tended to leave out the comma, and have sometimes noticed another editor coming along later and adding those introductory commas into the article; as I don’t feel particularly strongly about it, I usually just let these edits run, even though to my eye as a reader they sometimes seem to break the flow unnecessarily. MapReader (talk) 06:04, 5 October 2025 (UTC)[reply]
    Putting aside the question of acronyms and initialisms, and the clearly unquantifiable difference – i.e. what you called a 'tendency' above – in Americo-British usage...I think I am in agreement with you. I hope you will acknowledge that under the present guidelines a British English tag alone cannot possibly justify comma removal. While I am reluctant to enter the realm of primary source analysis, a random sampling of the BBC and The Guardian provides the following examples:
    • BBC – On Friday, the Met wrote to Defend Our Juries to raise concerns about the amount of police resources the protest would divert at a time when "visible reassurance and protective security" was needed for communities.
    • The Guardian – Last month, the publisher of the Nottingham Post warned of an “increasingly Trumpian approach” to the media creeping into British politics.
    Neither of these publications is a paragon of English, but neither is their use of commas 'American'. To be clear, I would omit these commas or rewrite the sentences. Yours, &c. RGloucester — ☎ 06:47, 5 October 2025 (UTC)[reply]
    Yes. And I don’t think I would suggest that not having these commas is standard British usage; it’s just that British usage seems more relaxed about whether they are there or not, whereas my perception is that American English treats them as mandatory. The difference is most notable when it comes to dates - “In 2023 he got married” seems fine to me, but I suspect an American editor would write “In 2023, he got married”. MapReader (talk) 14:50, 5 October 2025 (UTC)[reply]
    I think "mandatory" may be too strong a term here. I have edited academic manuscripts for publishers in both the US and the UK, and in my experience not all American writers would consider a comma necessary in that example. It depends partly on register and partly on the length of the clause that follows: "In 2023 he got married" might not trigger a comma, but a need for one would be felt more strongly in a sentence like "In 2023, he retired from the advertising business and opened a small hotel in southern Italy, where he and his wife had met thirty years earlier during a cycling trip". But you are absolutely correct about the bigger picture: American writers and publishers are generally much more likely than their British eqivalents to use a comma after such phrases at the beginning of a sentence, and therefore this can reasonably be considered an aspect of ENGVAR. In any case, it's pretty clear that this RFC is not really about comma usage at all: it's about the OP's indignation that another editor dared to describe some aspect of his English as "American", a description which the OP considers "slanderous". His intemperate reaction to that innocent comment in the edit summary for a revert which even he concedes was proper demonstrates the very opposite of good will and collegial editing (not to mention common courtesy). 87.7.87.236 (talk) 17:43, 5 October 2025 (UTC)[reply]
    Having read the exchange, and RGloucester's complaint that "nothing can be more reprehensible than to have one's actions labelled 'American'. The gravity of this disgrace demands one's utmost attention", it does seem that this whole RfC is merely ammunition to 'win' an extremely petty argument elsewhere. So I'll not waste any more time on it. Dgp4004 (talk) 18:18, 5 October 2025 (UTC)[reply]
    No, this RfC is an attempt to determine if there is broader community consensus for the interpretation of guidelines used by certain editors elsewhere. That comment was in reference to the British editor who was confused when his edits were rejected on ENGVAR grounds, not me. I may have an ornate personal style, for which I can beg your pardon, but I am perfectly capable of separating personal and professional. Here is the real question. Which editor who was a party to that discussion decided to log out to and make a comment here to avoid WP:SCRUTINY? Yours, &c. RGloucester — ☎ 21:53, 5 October 2025 (UTC)[reply]
    It really doesn't matter. Stop going about metaphorically slapping people with a glove like an 18th century dandy and WP:Grow a thick skin. Dgp4004 (talk) 23:00, 5 October 2025 (UTC)[reply]
    And to be clear: no going about literally slapping people with a glove, either. EEng 15:38, 6 October 2025 (UTC)[reply]
    I thought you're supposed to slap with a trout instead of a glove. Gawaon (talk) 17:28, 6 October 2025 (UTC)[reply]
    I was not a part of that original discussion, but I did read it, and that's precisely why I logged out to comment. People who get into such a lather over perceived slights on the internet can't be trusted to behave rationally, and I don't want you following me around and filling my talk page with hectoring comments over a trivial disagreement about punctuation that offended your amour-propre. Dgp4004's advice is good; you should follow it. 87.7.87.236 (talk) 15:35, 6 October 2025 (UTC)[reply]
    I don't feel any need to comment on what are clearly personal attacks by someone who refuses to disclose their identity in contravention of established Wikipedia practice. This is not a matter of 'perceived slights'; I have behaved rationally at every step of this discussion, supporting my statements with reference to Wikipedia policy, guidelines, and external sources. My concern is really very simple, and it has very little to do with glove-slapping. It is that Wikipedia is meant to be an egalitarian space, where actions are governed by policy, guidelines, verifiability, and accountability. When editors in positions of authority use their power to subvert policy and guidelines without broader consensus, they must be challenged; otherwise, one ends up with the disaster that was the recent arbitration case. What started as trivial gatekeeping at FAC could very well end up spreading like a plague across the encyclopaedia. I have every right to ask the broader community whether there is consensus for this practice. That you seek to avoid scrutiny speaks for itself. I hope we can allow uninvolved editors to come to their own conclusion. Yours, &c. RGloucester — ☎ 00:11, 7 October 2025 (UTC)[reply]

    References

    1. ^ "University of Oxford Style Guide" (PDF). University of Oxford. p. 12. Retrieved 4 October 2025.
    2. ^ "Commas". University of Nottingham. Retrieved 4 October 2025.
    3. ^ New Oxford Style Manual. Oxford: Oxford University Press. 2012. p. 69. Retrieved 4 October 2025.

    edit request, contractions' apostrophes

    [edit]

    Should Wikipedia:Manual of Style#Apostrophes link to (or maybe transclude) Wikipedia:Manual of Style#Contractions?

    Wishing everyone safe, happy, productive editing. --70.22.1.45 (talk) 10:19, 4 October 2025 (UTC)[reply]

    A short mention and link to Wikipedia:Manual of Style/Abbreviations § Contractions would probably be appropriate, e.g. (as the penultimate point, i.e. before "For further treatment ... see the article Apostrophe"):
    Mitch Ames (talk) 10:30, 4 October 2025 (UTC)[reply]

    ENGVAR and gaol / jail ?

    [edit]
    Discussants were mostly divided between those favouring "jail" for clarity and consistency with MOS:COMMONALITY, and those defending "gaol" as a valid British spelling tied to historical institutions. Even though several editors argued that "jail" is now the dominant spelling across all varieties of English and should therefore be preferred in general prose, others pointed out that "gaol" remains a correct British form still used in official and historical contexts, and that wholesale changes would conflict with MOS:ENGVAR and MOS:TIES. The discussion clearly shows strong policy-based support for using "jail" in ordinary descriptions, but also clear consensus that "gaol" should be preserved in proper names, quotations, and historically accurate contexts. In short, there is no consensus for a blanket replacement of "gaol" with "jail"; editors should please avoid mass, indiscriminate changes and instead apply case‑by‑case judgment. The outcome is therefore a compromise: apply "jail" as the default modern spelling but retain "gaol" where it is part of a name or historically appropriate. (non-admin closure) Vanderwaalforces (talk) 22:52, 11 October 2025 (UTC)[reply]

    The following discussion is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.

    Can I replace uses of 'gaol' with 'jail' per MOS:COMMONALITY because gaol is a dated spelling? This would apply to existing articles like Cork County Gaol or Dic Penderyn.

    See Talk:Beechworth. More eyes on that thread would be useful. Andy Dingley (talk) 23:52, 21 August 2025 (UTC)[reply]

    Here's an article showing the Australian view: https://www.abc.net.au/listen/radionational/archived/booksandarts/jail-or-gaol-how-should-australia-spell-it/7532694
    As a child in school in Australia in the 1970s, I would see "gaol" and be confused by it. As an adult I still have to stop and re-read it. I suspect many Brits are the same with thinking "gaol" is archaic (can you confirm?). And of course, Americans and most non-native English readers will have vast trouble with "gaol".
    Of course, if a building has an official name including "gaol" then we use that.  Stepho  talk  00:09, 22 August 2025 (UTC)[reply]
    Per Garner's Modern English (5th ed., 2022), jail has been more common in British sources since circa 1967 and the current ratio in print in "World Englishes" is 17:1 in favor of jail. The Ngram also shows that jail took off around 1967 and has been the dominant form in BrE ever since. Fowler's Dictionary of Modern English Usage (4th ed., 2015) begins the entry for gaol with: the traditional and treacherous spellings in the UK, are now under severe and probably unstoppable pressure from jail, jailer, which are dominant in most other parts of the English-speaking world. The jail entry also says this is the more common spelling in BrE and is the preferred spelling except in historical contexts in which the gaol- forms might be more appropriate. Learner's dictionaries from Cambridge and Oxford call gaol "old-fashioned". I don't have access to OED but several sites say they label 'gaol' "archaic", including this 2013 post which states all four British dictionaries consulted list jail as standard. --MYCETEAE 🍄‍🟫—talk 14:40, 24 August 2025 (UTC)[reply]
    I don't have access to the full OED content either, but one can already see in the public part that they list "gaol" as a mere "variant of jail". Gawaon (talk) 17:35, 24 August 2025 (UTC)[reply]
    The Shorter OED (6th edition, 2007) says:

    gaol (noun & verb) var. of jail

    and under jail:

    – NOTE: In Britain gaol is used in some official contexts, but otherwise is restricted to literary use, jail being the usual form. In American English jail is the usual spelling.

    Mitch Ames (talk) 11:06, 25 August 2025 (UTC)[reply]
    Just as an aside, as a kid I read 'gaol' before I knew how it was pronounced and thought is was pronounced with a hard 'g' and almost rhymed with 'cowl'. I assumed it was a different type of jail, and that there were three things: gaol, jail, and prison! I also thought Geoff was 'gee-off'. Ah, the dangers of too much reading! TreeReader (talk) 22:34, 27 August 2025 (UTC)[reply]
    I read Tale of Two Cities in high school or so, and when we watched a video (or film, maybe?) in class, I was so confused that he said "are you the jailer's daughter", instead of the line as written. :) I think we should use "jail" as appropriate, since it's also considered proper BE. "they built the gaol to replace the old jail" doesn't work for me, though. Maybe "old facility"? SarekOfVulcan (talk) 17:12, 10 September 2025 (UTC)[reply]
    Not in that article. "Gaol" is the official spelling. Hawkeye7 (discuss) 00:15, 22 August 2025 (UTC)[reply]
    • So how do we proceed here? The Beechworth thread seems to have broken down into useless WP:IDHT retrenchment and demanding proof that tourist attraction flannel doesn't stand as WP:RS. Should we really reword a whole bunch of articles about 18th century Irish prisons on the basis of modern-day usage in Australia? Do a whole bunch of recent changes need to be rolled back? Andy Dingley (talk) 09:16, 22 August 2025 (UTC)[reply]
    • MOS:COMMONALITY is a fairly gentle suggestion -- using vocabulary common to all varieties of English is preferable), not use vocabulary common to all varieties of English or similar. "Gaol" is slightly old-fashioned but perfectly correct in BrE, so while there might be good reasons to opt for "jail" (or "sulfur", "fetus" and similar), it would be inappropriate to insist on a change on the grounds of COMMONALITY in an article with WP:TIES to the Commonwealth. UndercoverClassicist T·C 09:33, 22 August 2025 (UTC)[reply]
      MOS:COMMONALITY also says "When more than one variant spelling exists within a national standard variety of English, the most commonly used current variant should usually be preferred" and there are exceptions listed when there's a different meaning with a different most-common spelling, proper names, and direct quotations. Wikipedia:Manual of Style/Spelling#English spelling comparison chart points out that even in British English, "jail" is the more common modern variant, and so should be the preferred one when it's lowercase. In fact, that chart shows "jail" is the most common modern variant in the UK, Ireland, South Africa, New Zealand, Canada, and the United States. I'm not sure any standard dialect in the world actually uses "gaol" more commonly. -- Beland (talk) 17:03, 22 August 2025 (UTC)[reply]
    • Avoid both, and use "prison" except in names of institutions with "Gaol" (and any that might have "Jail"). Ham II (talk) 09:54, 22 August 2025 (UTC)[reply]
    There is a legal technical difference between prison and { jail, gaol }. Also most of these are using 'Gaol' in the first place because that's some fundamental part of that specific article, like a historical title, e.g. Beaumaris Gaol [5]. Of course if we're writing ab initio we would choose clear, common language. Nobody is looking to write an article, 'Should electric motorbike riders be sent to gaol?' Andy Dingley (talk) 11:07, 22 August 2025 (UTC)[reply]
    How can we tell which institutions are jails and which are prisons? The article Beaumaris Gaol describes it as both a "gaol" and a "prison". -- Beland (talk) 16:55, 22 August 2025 (UTC)[reply]
    The distinction of terms, in the UK (but not in the US), is mostly down to chronology (as I found out during my recent self-revert on the Wilde poem) -- the prison system was nationalised in 1878, after which most prisons were officially called "Her Majesty's Prison Suchandsuch". Beaumaris closed in 1877, so wouldn't ever have officially held that name. Other laws earlier in the C19th used "gaol" for all of them. The Beaumaris article correctly says "gaol, or prison" -- "prison" being provided as an alternative word for "gaol" rather than an alternative function it may have served. UndercoverClassicist T·C 17:09, 22 August 2025 (UTC)[reply]
    I'm a bit confused by this. Words have meanings; if "prison" means something different than "gaol", it should not be presented as a synonym for "gaol". For example, in the US "prison" formally means long-term detention and "jail" means short-term detention. "Jail or prison" would not be a correct formal descriptor of most facilities (perhaps there are some that do both). If "gaol" and "prison" mean the same thing in the UK system, then for the sake of native English speakers who don't have "gaol" in their standard national dialect and who would have a lot of trouble figuring it out, shouldn't we use "prison" instead of "gaol" because that word is universally legible? -- Beland (talk) 17:30, 22 August 2025 (UTC)[reply]
    There's a few different things we might scope out that could also help us with spelling and word choice issues:
    • The words used in the names of institutions (in laws and other proper names in use)
    • The words used in the names of laws
    • The common nouns used to describe facilities in everyday speech
    • The legal classes of things and their names
    To start out, I took a look at the text of Gaols Act 1823. That name does not appear in the text, and that article says the same law is also called the Prisons Act 1823. It appears the names of many laws are applied retroactively and in multitudes, and many early laws may not have had names at the time they were enacted.
    Turns out it defines two classes of prison: gaols and houses of correction, with different legal requirements for each. I started documenting at Prison#UK and Ireland.
    There's a whole bunch of confinement-related laws in the UK and I don't have time to research them all now, but it would be interesting to figure out a beginning and an end time for the legal category of gaol. From what UndercoverClassicist wrote above, the end point may be Prison Act 1877. There are references on Gaols Act 1823 to laws talking about gaols or gaolers going all the way back to the 1300s, and it will take a bit of research to go past that to see if there is any notional beginning of the idea of a building of confinement, and also figure out if legal categories changed at all in those 500 years. Though that far back the statutes are in Middle English, when spelling was unstable. Wiktionary lists the spellings "gayole, gaiol, gaylle, gaille, gayle, gaile".
    I'll also note Category:Defunct prisons in the United Kingdom is a good place to find examples of names, and there is a lot of variation, many without either "Gaol" or "Prison" in the name. -- Beland (talk) 08:08, 23 August 2025 (UTC)[reply]
    In my own poking around, I found Inveraray Jail (built 1820 in Scotland), which looking at Google Books is generally referred to as a "jail", substantiating the claim on Talk:Inveraray Jail by someone who says they currently run the place that it was never known as "Inveraray Gaol". The one exception I found was in the law that created it, which of course was written in England, not Scotland. From this I infer that "jail" and "gaol" were both used historically to refer to the same class of institutions, which seems to undermine the idea that the only way to refer to UK confinement institutions from a certain era is "gaol". (Of course that law also capitalizes all nouns and uses long s and has lots of other archaic spellings we don't use, so it has forms like "Court-Houſe" instead of the modern "courthouse".) Beland (talk) 18:01, 22 August 2025 (UTC)[reply]
    Let anyone be misled: the use of long s does not represent a spelling variation; it's merely a typographical style. Long s is still ess. EEng 00:15, 23 August 2025 (UTC)[reply]

    Short answer? No, you may not replace the official names of places with names you prefer. Cork County Gaol and Cork City Gaol and Kilmainham Gaol are their names. BastunĖġáḍβáś₮ŭŃ! 10:01, 22 August 2025 (UTC)[reply]

    But what about re-writing inside that article so that Cork City Gaol now states the building of a new Cork City Gaol to replace the old jail at the North Gate Bridge (the old jail, [6] ? Andy Dingley (talk) 11:07, 22 August 2025 (UTC)[reply]
    With this diff, I have just boldly changed The Ballad of Reading Gaol to replace the second use of the word gaol to make HM Prison Reading explicit. Let's see if it survives. 𝕁𝕄𝔽 (talk) 12:19, 22 August 2025 (UTC)[reply]
    • What about this revert on Dorothy Smith, Lady Pakington? It just says "...Pakington was forced to appear before the court of high commission, and was committed to gaol." The capitalized word "Gaol" does not appear anywhere in this article, and the name of the facility that this person was sent to is not mentioned in the article. @Roger 8 Roger: on the other thread, you seemed to think "jail" was better in these instances; is "prison" OK? It's the most common variant in both British English and across national standard dialects. @Ham II: It sounds like you'd prefer "prison" here, or is "jail" OK? @Bastun: Unclear what your position is? @Andy Dingley and UndercoverClassicist: Do we need to research whether the thing being referred to was named "Gaol" or "Prison" or was part of a distinct legal class of facility regardless of name to decide between "gaol" and "prison"? There seems to be some notion that context flips the more-common "jail" to the less-common "gaol" and I'm trying to figure out for those favoring "gaol" in some circumstances what the scope of that is. -- Beland (talk) 17:45, 22 August 2025 (UTC)[reply]
      Nope -- in this case, while MOS:COMMONALITY would suggest "prison", it doesn't insist, and the relevant question is whether "gaol" is a valid word for this thing in in British English (which it is). UndercoverClassicist T·C 19:31, 22 August 2025 (UTC)[reply]
      OK, so it sounds like you're ambivalent between "gaol" and "prison" in this case? Or would you object to changing it to "prison" if others object to "gaol"? -- Beland (talk) 19:49, 22 August 2025 (UTC)[reply]
    • Use "Gaol" if it's part of the official name of an institution (hence with a capital letter), "jail" everywhere else. In that way, no commonality is wasted and spelling sanity prevails as far as possible. Gawaon (talk) 19:52, 22 August 2025 (UTC)[reply]
      In the case of the Beechworth article, the article refers to the Old Melbourne Gaol, an official name. There is a semantic difference in meaning between "gaol" and "prison": the former is a facility for the temporary holding of persons, the latter is for long-term detention. Hawkeye7 (discuss) 01:02, 23 August 2025 (UTC)[reply]
      We solved the controversy in that article by not using lowercase "gaol" but the proper names "Melbourne Gaol" and "HM Prison Beechworth". Some of Andy's recent reverts could be resolved in that way, though we'd have to verify that "X Gaol" (or "HM Prison X" or something else) is in fact the correct name. But there would be other reverts where we are not using a proper name and can't because it would be too repetitive or the text doesn't specify which facility. -- Beland (talk) 06:29, 23 August 2025 (UTC)[reply]
      What's the problem with "gaol" anyway? Hawkeye7 (discuss) 08:35, 23 August 2025 (UTC)[reply]
      Mostly that it's highly irregular, so it can be nearly impossible to sound it out and figure out what word it is supposed to be unless you've seen it before and memorized it. (As Prison#cite_note-5 says, soft g does not normally occur in English before "a", and "ao" is not a normal English diphthong.) When it's not coming across as a nonsense word, it's looking like a misspelling of "goal". This is in contrast to typical spelling differences where maybe a silent "e" gets added to the end of a word or a consonant gets doubled or a "u" jammed in. That might be an invalid spelling in one's native dialect, but it doesn't change the sounding out of the word enough to make it unrecognizable. And those changes follow rules, so once you've seen a few examples, new words that follow the same pattern are easily understood.
      Americans probably have the biggest problem because "gaol" is just not a valid spelling of any word in Standard American English, and we don't have historic buildings with "Gaol" in the name or "gaol" in any laws that we still use, as far as I know. In order to have memorized this word you'd need to be reading about jails in a foreign dialect, and that just doesn't happen very often. Even for native users of Standard British English, "jail" is becoming more common so there are fewer chances to learn "gaol" and an even lower probability that documents with "gaol" will get exported.
      We expect readers to know words in order to be able to read, but only the words in the national dialect they've learned. We can bridge this irregularity with a gloss in the same way we bridge terminology differences, like "lift (elevator)" and "pants (underwear)" - "gaol (jail)". This is perfectly comprehensible but a little bumpy. To me it seems cleaner to just write "jail" in the first place, so I'm probing the scope in which we have consensus to do that. Given that "jail" is now either the highly dominant or only spelling in all standard English dialects, we never have to gloss in the reverse direction like "jail (gaol)". -- Beland (talk) 10:55, 23 August 2025 (UTC)[reply]
      @Hawkeye7: Would you happen to have a pointer to any reliable sources that say that Australia uses jail for short-term and prison for long term, either now or during the era of the Old Melbourne Gaol? This is asserted a lot for the United States, but I couldn't find anything establishing that for Australia. (What I did find I suspect was AI-generated and might be completely inaccurate.) I'm mostly trying to document this for readers, as Prison#Australia doesn't mention any such distinction at the moment.
      Looking at List of prisons in Australia, it seems none of the modern facilities have "Jail" in the name, mostly either "Prison" or "Correction Centre". Looking at Punishment in Australia, it seems that the term "remand centre" is used for short-term pre-trial detention, and many of these are located within prisons and correction centres. It would be interesting to find out when and why "Gaol" stopped being used. -- Beland (talk) 06:18, 25 August 2025 (UTC)[reply]
      "Gaol" and "prison" fell out of use when they moved to change the names to "correction centre" and "remand centre" many years ago. I presume this happened at different times in different states. I don't think these terms have currency or recognition outside Australia. Hawkeye7 (discuss) 08:31, 25 August 2025 (UTC)[reply]
      For a reference, see History of the Criminal Justice System in Victoria, Chapter 1. Hawkeye7 (discuss) 09:23, 25 August 2025 (UTC)[reply]
    • OK, this has turned into the same mess of minutiae as happened at Talk:Beechworth
    To try and reset it, can anyone please tell me which policy (not specifically about crime and punishment) support bulk changes to an article text (not title) like these [7][8] ? What's the fundamental reason for it, and benefit derived from it?
    Unless we have a broad reason for doing it at all, we don't even need to start poking at the details. Andy Dingley (talk) 10:54, 23 August 2025 (UTC)[reply]
    A proximal reason is compliance with MOS:COMMONALITY. The fundamental reason is desire to convey information seamlessly to readers in the face of the unreadability of "gaol" for speakers of certain standard dialects (which I explain in detail above in reply to Hawkeye7 at 10:55 today).
    I'll note that the existing text of Derry Gaol uses both "gaol" and "jail", and "Derry Gaol" and "Derry Jail" without explanation. This seems untidy. -- Beland (talk) 11:01, 23 August 2025 (UTC)[reply]
    What part of MOS:COMMONALITY supports arbitrary changes to existing articles? In particular, this goes against the underlying principle of WP:ENGVAR which is to avoid churn. Churn itself being seen as harmful, even outside the argument to 'be right'. A recommendation to use accessible language is one thing, but automated bulk changes to existing articles quite another, and one we've always opposed.
    ENGVAR supports the idea of 'ties'. Now I can't think of a stronger tie than an article about one specific named Gaol then using 'gaol' within its own context. If you claim that this term is too 'unfamiliar' to readers, then they're not going to get past the title.
    There is no justification here, or in any of our policies, for these bulk, blanket changes to stamp out one particular word. Andy Dingley (talk) 21:40, 23 August 2025 (UTC)[reply]
    The changes I have made are not arbitrary, they are to implement "most commonly used current variant (across all varieties of Standard English)". Because "gaol" is not the most commonly used current variant in any national variety of Standard English, it logically follows it should not be used on English Wikipedia outside of exceptions like proper names, direct quotations, citations, and when discussing the word itself. That guideline should not result in churn, because no one should be going around changing "jail" to "gaol". A burst of editing that brings a large number of articles into compliance with consensus guidelines is not undesirable churn, it's desirable cleanup. Churn does occur if there is not agreement as to which direction the guidelines are pointing or there is a desire to change them, which is why we have stopped changing these spellings and are discussing the question.
    Based on what has been written so far, it seems some editors want there to be an exception to MOS:COMMONALITY for articles on certain facilities, and others do not. That's fine; it's a normal thing to have a discussion over. It would be helpful to the discussion to articulate the scope of that exception. For example, does it only apply to jails in the UK up to 1877? MOS:TIES only discusses "strong ties to a particular English-speaking nation". Other editors have objected to adding any other entities to this rule, such as states or provinces or regions with their own dialects. If you want to expand this rule to have something beyond ties to a "nation" influence spelling, that's fine; specific language would be helpful for the purposes of discussion. But this rule currently only concerns choice of dialect, and we all agree that articles about British jails should be written in British English. If you want to introduce the notion of ties to choice of spelling within a dialect, it's possible MOS:COMMONALITY is the right place for that, since it would effectively be an exception to the "the most commonly used current variant (across all varieties of Standard English)" rule. Specific language for discussion purposes would be helpful. Is the thing you want to see happen something like "articles with strong ties to a specific historic period should adopt the spellings of that period if those spellings are still considered correct in the national variety of modern standard English, even if they are considered old-fashioned or are less clear to some readers"?
    Nothing in MOS:ENGVAR prevents edits to "existing articles". MOS:RETAIN prevents changing an existing choice of dialect, but it does not apply to articles with MOS:TIES, which are the ones we are discussing. There are plenty of reasons to change spellings in bulk across all articles with ties to the UK, for example to change "defense" to "defence" because the latter is the British spelling.
    I think it's a bit overboard to say that the fact that "Gaol" is in the title of an article means that readers who don't know what "gaol" means can't find out by reading the article. We can just tell them what it means. We already do this implicitly in Old Melbourne Gaol which starts out "The Old Melbourne Gaol is a former jail"; I think that's fine. We could make it more explicit for readers who don't pick up on the correspondence between the two words by adding '("gaol" is an old-fashioned spelling of "jail")' but I'm not sure that's necessary. You also reverted instances of "gaol" where a specific prison is never named and it's just being used as a general concept, one time in the entire article. Cases like that seem to be single instances of an unclear word that could be improved, uncomplicated by worries about consistency with the rest of the article or confusion that could result from proper nouns with "Gaol" in them. -- Beland (talk) 01:46, 24 August 2025 (UTC)[reply]
    (What I wrote is not quote right; the clarified version of the "most common" rule not only means "use jail because it is more common in British English than gaol" but also "use jail because it is the only valid spelling in American English and also exists in British English".) -- Beland (talk) 04:12, 24 August 2025 (UTC)[reply]
    I agree. Despite the title of this topic, "gaol" vs. "jail" is not really an ENGVAR issue at all, since the latter spelling is preferred in all varieties of English (for BE, according to OED). So changing "gaol" to "jail" is (outside of proper nouns) always fine and improves the article, if in a minor way. Gawaon (talk) 07:01, 24 August 2025 (UTC)[reply]
    While I think jail should generally be used in preference to gaol per MOS:COMMONALITY, I think it's unnecessary and even confusing to require it where the article title uses Gaol (because that's it's the actual name of the institution as in the examples above). I looked at the "mass" changes to the 2 articles that Andy Dingley linked to and they really seem unnecessary. Clearly someone starting to read both those articles will know what a "gaol" is from the article text itself - there's even a reference it to being "jail". There's no need to then replace gaol thereafter - it doesn't add any extra understanding for the (say) American reader. In fact, I can see readers being puzzled why there's one spelling in the title and one in the body. I think the principle would be use jail in preference to gaol except where it would lead to inconsistency with the article title. DeCausa (talk) 09:13, 25 August 2025 (UTC)[reply]
    This seems backwards to me. Oftentimes I will only be reading part of an article, because I have skipped down to the section I'm interested in, because I've followed a link that points to a specific section, or because I'm reading a snippet in search results or an AI answer. If I encounter a word I can't read in the body of a section, it's not going to help me that word is explained in the intro to the article (which I haven't read, and I don't know to go there to decode this word). As a result, I might just not know what certain sentences mean.
    Compare that to the negative consequences of having "Gaol" and "jail" in the same article. Maybe it looks a little weird? Seems a bit incongruous? But everyone knows old spellings and foreign names are weird sometimes, so is it that out of place? If that difference is explained in the intro ("gaol" is an old-fashioned spelling of "jail") that puzzlement is cleared up for everyone reading the article from the beginning. But what if someone skips the intro and goes right to a section? They might still be a little puzzled, but able to understand everything that's written. That seems really minor compared to not being able to read some words at all.
    If an incomprehensible word only shows up inside a proper noun, then e.g. American readers who miss the intro clarification don't really need to worry about decoding it to understand the meanings of sentences - it's at least clear it's the name of a thing, if not the name of a jail. Knowing that "Gaol" in a proper name means "jail" makes that weird name make a lot more sense, but for example I don't need to know that "Lacroix" means "the cross" to know that "Thierry Lacroix" is the name of a person. I don't even need to know how Western names work - that the last part is a family name - in order to understand more about this person's life. If I continue reading I will eventually notice that the same word appears in the names of their family members, but I don't have to wait for this realization to understand other facts. -- Beland (talk) 10:07, 25 August 2025 (UTC)[reply]
    Well, you certainly need to infer a heck of a lot about our readers and the way they read articles to make your argument work. I guess if we (ever) have a reader that matches your template exactly, you'll be right! DeCausa (talk) 11:30, 25 August 2025 (UTC)[reply]
    My impression is rather that you make a lot of odd assumptions about our readers if you seriously think people would be confused by encountering the spelling "jail" in any article. Gawaon (talk) 14:57, 25 August 2025 (UTC)[reply]
    I agree with Beland and Gawaon. Readers who recognize gaol likely recognize it as old fashioned. It's an extraordinary claim that they would be "confused" by the standard spelling. Readers unfamiliar with gaol, who represent the majority, will encounter at least some friction every time they read it, even if it's been defined for them. Some will encounter gaol as a significant barrier to understanding. The behavior Beland describes is common—many readers skim and skip around. We have many practices that anticipate this behavior and aim to improve the experience for such readers. For example, linking to article sections directly, guidance bolding selected terms in sections following the lead, and guidance on redefining and repeating wikilinks to key terms in subsequent sections. --MYCETEAE 🍄‍🟫—talk 16:45, 25 August 2025 (UTC)[reply]
    On that basis we can dispense with the whole MOS principle of internal consistency within an article. No need - they'll work it out! DeCausa (talk) 17:27, 25 August 2025 (UTC)[reply]
    Just because comprehension gets priority over consistency when there's a substantial conflict doesn't mean that consistency should be abandoned. For example, choosing British-style spellings for an article without strong ties might be a good idea if it mentions a proper noun like "Foo Correction Centre". Both "center" and "centre" are comprehensible regardless of the reader's national dialect. -- Beland (talk) 17:53, 25 August 2025 (UTC)[reply]
    I think everyone knows where you stand on his and what your arguments are. DeCausa (talk) 18:01, 25 August 2025 (UTC)[reply]
    OK, that was not clear from your comment that implied we think consistency should be abandoned. -- Beland (talk) 18:14, 25 August 2025 (UTC)[reply]
    You might want to consider whether it's wise to post a response in this thread to every comment that's adverse to your view. You are an admin, after all. DeCausa (talk) 20:42, 25 August 2025 (UTC)[reply]
    Well, I don't see how we can reach consensus without conversing with people who we disagree with; we just need to keep it respectful. I am also frustrated that this conversation lacks focus on comparing rationales and reaching actionable compromise, and there are a lot of easily refuted objections coming out of left field. (You're not at fault for that, and of course helpful points have also been made on both sides.) Instead of complaining about personal jabs or the conversational style of other editors, I've been ignoring personalities and trying to ask questions that get us back to the substance of the question.
    If you truly feel I've engaged in misbehavior here, please report me at WP:AN/I. I don't think people who care about the substance of this question are interested in personality clash drama. -- Beland (talk) 01:32, 26 August 2025 (UTC)[reply]
    Whoa! Steady there. I was just making a suggestion, not saying you've misbehaved! I didn't even link to the B-word. Oops I just did. DeCausa (talk) 07:49, 26 August 2025 (UTC)[reply]
    It's a misapplication of "internal consistency" to use unusual words or spellings in the article body just because they appear in the title. We use "cemetery" in Pet Sematary and Pet Sematary (1989 film); we use "vampire" throughout Nosferatu and Nosferatu the Vampyre; we use "the" in Ye Olde Curiosity Shop and it would be bizarre to use "olde" outside of the store's name, if the adjective were needed in running text. Even a strong reading of internal consistency must be balanced with COMMONALITY and comprehension. --MYCETEAE 🍄‍🟫—talk 20:56, 25 August 2025 (UTC)[reply]
    Oh hang on. I wasn't going to bother posting any further on this thread but that's just bollocks. There's no way in British English gaol is remotely comparable to "Pet Sematary" or "olde". It's standard British English even if less used than jail. DeCausa (talk) 21:47, 25 August 2025 (UTC)[reply]
    Alright, those were on the extreme. These are on a continuum and I maintain that gaol is too nonstandard for internal consistency to overcome the objections. Perhaps examples like The Vitamin Shoppe, The Pop Shoppe, Grand Canal Shoppes are more reasonable. The articles use "shop", "store", etc. in running text. --MYCETEAE 🍄‍🟫—talk 01:12, 26 August 2025 (UTC)[reply]
    That's equally ridiculous. If you think "shoppe" is comparable to "gaol" in British English then you've miscalibrated this badly. Gaol will appear in standard British English dictionaries. DeCausa (talk) 07:15, 26 August 2025 (UTC)[reply]
    They only list is as a less preferred variant, though. So why should we prefer it in some articles just because it happens to show up elsewhere in those article, just in the name of "consistency"? What about consistency with the rest of Wikipedia, and with what the reader is used to? Gawaon (talk) 10:13, 26 August 2025 (UTC)[reply]
    • "There will be no churn because My Way Is Right" is why we have policies against churn. Andy Dingley (talk) 11:14, 25 August 2025 (UTC)[reply]
      Of course. That's why I say people need to make edits in the direction the MOS points, not their own direction, and if there is any disagreement about which direction the MOS points (as there is in this case), that needs to be resolved with discussion, as we are doing. -- Beland (talk) 16:55, 25 August 2025 (UTC)[reply]
    • MOS:RETAIN prevents changing an existing choice of dialect, but it does not apply to articles with MOS:TIES
    You are literally trying to quote a policy supporting stability to support a change in spelling against another policy supporting stability. On the basis of using a web snippet about modern Australian use to change descriptions of 18th century Ireland. This makes no sense, it's just text wallpapering until you've bludgeoned away anyone who disagrees. Andy Dingley (talk) 11:19, 25 August 2025 (UTC)[reply]
    Well, sorry for the long posts, but sometimes it takes more words to explain why an idea is wrong than it does to state the idea.
    I don't follow your logic about stability. Are you saying we're supposed to leave "defense" in articles with UK ties, for example, or are we supposed to change existing text that violates MOS:TIES? -- Beland (talk) 16:54, 25 August 2025 (UTC)[reply]
    1. If the article has a {{Use British English}} tag, then you change "defense" to "defence" without further ado.
    2. If yours is the first non-stub edit, you can tag it, and then change it.
    3. Otherwise, you should establish consensus on the talk page to change the English variant.
    Hawkeye7 (discuss) 00:53, 26 August 2025 (UTC)[reply]
    I agree with (1), but I'm asking Andy if that kind of change violates the sort of stability they are advocating for, if that spelling which apparently violates ENGVAR is long-standing. -- Beland (talk) 03:21, 26 August 2025 (UTC)[reply]
    • Use "jail" throughout the article body. Use "gaol" only where it occurs in proper names or, sparingly, in quotes or other instances where it is beneficial to highlight this usage. In an article about "X Gaol", call it the jail, etc. in prose. Explain or gloss "gaol" on first usage or early in articles that use this word in the title. Obviously, use "prison" instead of "jail" if this is the more appropriate term. --MYCETEAE 🍄‍🟫—talk 14:03, 24 August 2025 (UTC)[reply]
      Also: the (partial) proper name should not be overused. Cork County Gaol, for example, contains phrases like: The Gaol was designed in the Greek Revival style and The County Gaol was the scene of executions by hanging, which took place in public outside the Gaol until the 1860s. We would not typically write the Theatre was built or the School opened in articles about institutions with 'Theatre' or 'School' as part of the name. This isn't an ENGVAR or COMMONALITY issue but it does add to the awkwardness of these articles. --MYCETEAE 🍄‍🟫—talk 18:10, 24 August 2025 (UTC)[reply]
      For the reason I just posted above, I think it will just increase puzzlement to talk about "jail" in an article where there's "Gaol" in the title. DeCausa (talk) 09:19, 25 August 2025 (UTC)[reply]
      I've responded above to keep the thread together. --MYCETEAE 🍄‍🟫—talk 16:46, 25 August 2025 (UTC)[reply]
    What is being argued for by some is that we should have a style guide that applies to British English articles that jail always be used in everything but proper names. But although rarer, gaol remains standard British English, and what is instructive is that when British publications and institutions write about places that are named x gaol, they will use use the same spelling throughout. I presume they do so because this is simply clearer. So should we. Examples: Derby Telegraph [9], or the UK parliament [10], [11], [12]. Here's the National Archive using the term [13]. Sirfurboy🏄 (talk) 07:50, 26 August 2025 (UTC)[reply]
    What's the problem with "gaol" anyway? as Hawkeye7 said above. Both words are current English so both can be used. Just because one is commoner does mean we automatically discard the other. If an article is about a place called a gal then we use gaol throughout the article. It's the same as the UK/US spelling rule about choosing one and sticking with it. The preference for jail should only apply to new article where no actual place name is mentioned. Just because some people don't know what a gaol is doesn't mean we should dumb down. We should assume a certain level of intelligence. Roger 8 Roger (talk) 08:01, 26 August 2025 (UTC)[reply]
    My post above crossed with sirfurboy's. Coincidentally, the same argument. Roger 8 Roger (talk) 08:20, 26 August 2025 (UTC)[reply]
    Yes, I think there's quite a few non-British editors in this thread that have an erroneous impression of how "antiquated" gaol is in British English. One (above) has even equated it with "shoppe" and "olde"! DeCausa (talk) 08:46, 26 August 2025 (UTC)[reply]
    Not if you keep MOS:COMMONALITY in mind, though. Gawaon (talk) 10:19, 26 August 2025 (UTC)[reply]
    Writing "jail" instead of "gaol" is not "dumbing down". It conveys exactly the same information, except with a spelling that more readers can successfully interpret. The point of MOS:COMMONALITY is exactly to avoid readers having to run to the dictionary when simply picking a different spelling would make a sentence comprehensible on its face.
    Readers who don't know how to pronounce "gaol" are not less intelligent; they simply have not been exposed to it because they live in the US or Canada or Australia or NZ, where Wikipedia:Manual of Style/Spelling tells us this spelling is not part of the standard language. (And haven't happened to read either specialized or foreign publications, perhaps because it's never come up, or perhaps because they are in elementary school.)
    I don't think it's reasonable to assume that the Derby Telegraph thinks that "gaol" is clearer for its American readers. It is not writing for an international audience, it is writing for the people of Derbyshire and their neighbors - at most England or the UK. -- Beland (talk) 18:11, 26 August 2025 (UTC)[reply]
    Yes, writing "jail" instead of "gaol" is 'dumbing down' because it to underestimate our readers and assume that they can't cope with the original, still-correct, word. This is Wikipedia, not Simple Wikipedia. Andy Dingley (talk) 00:56, 27 August 2025 (UTC)[reply]
    I'm not sure I would say "can't cope"; nearly anyone who can read the encyclopedia could look a word up in a dictionary. But isn't just factually accurate to estimate that many if not most of our readers would in fact have to look this word up in a dictionary to understand it? -- Beland (talk) 04:03, 27 August 2025 (UTC)[reply]
    Readers from Australia should be familiar familiar with the "gaol" spelling. Hawkeye7 (discuss) 01:15, 27 August 2025 (UTC)[reply]
    That probably depends a lot on how old they are and whether or not they read about certain topics, and we've heard a personal experience above that even with passing familiarity it's problematic. We could research the comprehensibility in Australia in more depth, but it hardly seems to matter as the vast majority of Americans certainly aren't familiar with this spelling. -- Beland (talk) 04:00, 27 August 2025 (UTC)[reply]
    And we can find British, Irish, and Australian sources that use jail in articles about a named Gaol.[14][15][16][17][18] --MYCETEAE 🍄‍🟫—talk 16:27, 26 August 2025 (UTC)[reply]
    And, indeed, American universities publishing papers using gaol [19]. But given that it is clearly a valid current alternative spelling, in use in the British parliament, the British National Archive interpretive information, and academia, as well as newspaper sources, there is simply no reason to impose a style restriction on pages, especially when we note in your examples that when the article uses jail it may find the need to explain itself. It is clearer just to use a consistent spelling, although nothing decided here should over-ride a local consensus relevant for a specific page. Sirfurboy🏄 (talk) 17:34, 26 August 2025 (UTC)[reply]
    Using "gaol" instead of "jail" does not obviate the need for the article to explain itself, as many if not most readers won't know what "gaol" means. The publishers you mention above are for a British national audience, or a specialized audience, not a general international audience.
    When you say "there is simply no reason to impose a style restriction on pages" to avoid using valid spellings, that sounds like an argument for repealing the "most commonly used current variant" of MOS:COMMONALITY. Editors obviously had reasons for adding that in the first place, and I doubt there would be consensus to remove it. But it sounds like you're specifically concerned about the tradeoff between consistency and clarity. Would you support adding an exception like: "except where the article must use a specific spelling, for example in a proper name or direct quotation"? -- Beland (talk) 17:55, 26 August 2025 (UTC)[reply]
    Using "gaol" instead of "jail" does not obviate the need for the article to explain itself - sure, but whereas you can just wikilink gaol, if you start using jail instead, you need to include a sentence of explanation as to why you did so. The publishers mentioned are not just writing to a British audience, because we have seen examples in Australia, Ireland and the USA. You just did not reply on those, nor on the use by Parliament nor TNA. as for "Would you support adding an exception like..." - no. I don't support the multiplication of intricate stylistic rules. Gaol is a variant spelling. It is still in use, including, as we have seen, by academics, parliamentarians and information professionals as well as journalists. I'd support dropping the stick on this one. Sirfurboy🏄 (talk) 20:39, 26 August 2025 (UTC)[reply]
    MOS:COMMONALITY says When more than one variant spelling exists within a national standard variety of English, the most commonly used current variant (across all varieties of Standard English) should usually be preferred, except where the less common spelling has a specific usage in a specialized context, e.g., connexion in Methodist connexionalism. That means if the article is about a proper noun article that includess gaol, stick with gaol throughout. Otherwise jail is 'usually' preferred, ie it's not set in stone. So, why were these mass changes ever made, there's no need?Roger 8 Roger (talk) 22:09, 26 August 2025 (UTC)[reply]
    Roger, if you're saying this falls under the specialist meaning exception, I'm not quite seeing that. The word "gaol" does not refer to a specialized concept or only to jails in one place or period. The most common way to refer to institutions with "Gaol" in the name using that term is "jail", including in British English.
    I do think the specialized meaning applies to something like "gaol fever", where that is actually still the most common spelling (perhaps because this concept is no longer present in modern medicine) and I support retaining that.
    Sirfurboy, sure, some publishers do use "gaol" in the UK, including Parliament and the BNA. As MYCETEAE points out, there are other British publishers that use "jail", and based on the Ngrams graph it appears they are a minority maybe 1/7th of usage) and extrapolating long-term trends, that spelling may just go away entirely in the UK. Wikipedia:Manual of Style/Spelling tells us "gaol" is so little used in the US, NZ, Australia, and Canada, that it's not considered a proper spelling.
    I don't think wikilinking is sufficient; it doesn't help people using print editions or seeing the word in search engine results or AI summaries or excerpts on other web sites. It's just as much of an interruption as going off and looking up the word in a dictionary. If we expect people to go off and do their own research to answer questions raised by the spelling in the article, I don't see why we'd put in a note that "gaol is an old-fashioned spelling of jail", either.
    If your rationale for not changing "gaol" to "jail" is that "gaol is a valid word in British English", that implies to me that we should always ignore the "most commonly used current variant" clause of MOS:COMMONALITY. Would you be in favor of removing that to avoid this sort of issue coming up again, or are there some circumstances in which that advice should be followed? -- Beland (talk) 03:52, 27 August 2025 (UTC)[reply]
    But unlike connexion (as in Connexionalism), gaol has no specialized meaning. It's defined everywhere as a spelling variant of jail. You yourself describe it this way. --MYCETEAE 🍄‍🟫—talk 15:16, 27 August 2025 (UTC)[reply]
    sure, but whereas you can just wikilink gaol, if you start using jail instead, you need to include a sentence of explanation as to why you did so. A wikilink is not a substitute for an explanation (see: MOS:NOFORCELINK) and at a minimum significant ENGVAR differences need to be glossed (MOS:COMMONALITY). Even if we didn't have these guidelines, I'm baffled by the notion that using a less common regional variant is somehow less confusing or requires less explanation than using a universally-recognized variant. --MYCETEAE 🍄‍🟫—talk 14:57, 27 August 2025 (UTC)[reply]

    I mean for example, about specialised meanings, if the article is about melbourne gaol, then use the spelling gaol whenever the word is used in that article, not just when referring to the actual building called melbourne gaol. Gaol is a normal contemporary spelling of the word, not an archaic spelling, even if used less than jail. Also, the mos guidelines say usually use the commoner version, meaning there will be exceptions and allowing for some flexibility and common sense by editors. Roger 8 Roger (talk) 08:25, 27 August 2025 (UTC)[reply]

    This reminds me of when Daniel Lambert was TFA way back in December 2010, we had all manner of people wanting to change "gaol keeper" to either "jail keeper" or "goalkeeper". If the soure says "gaol", we should stick with that spelling. --Redrose64 🌹 (talk) 19:16, 6 September 2025 (UTC)[reply]
    What about articles that use more than one source? (Hopefully, the huge majority!) Gawaon (talk) 20:50, 6 September 2025 (UTC)[reply]
    Well, "gaol keeper" is very hard not to read as "goal keeper", because "gaol" is rare or unknown in a reader's dialect, and "goal keeper" is common. I'm not sure why the right response to complaints about that article was to blame people for not reading carefully, instead of using the common spelling so it's easy to read in a single pass. -- Beland (talk) 23:23, 6 September 2025 (UTC)[reply]
    Good grief. The lead at Daniel Lambert is an example of what not to do. Readers should should not be forced to click a link or read an explanatory note to understand a basic vocabulary word. --MYCETEAE 🍄‍🟫—talk 17:05, 10 September 2025 (UTC)[reply]
    Indeed! Gawaon (talk) 17:50, 10 September 2025 (UTC)[reply]
    Tried to fix it and was swiftly reverted. *sigh* --MYCETEAE 🍄‍🟫—talk 21:41, 10 September 2025 (UTC)[reply]
    I suppose one who would argue to retain this usage might also be described as a gaol keeper, of sorts… --MYCETEAE 🍄‍🟫—talk 21:46, 10 September 2025 (UTC)[reply]
    On the contrary. That FA had a good solution which had gone through our featured article process). You tried "fix" something that didn't need fixing by introducing pointless repetition. Good revert. DeCausa (talk) 21:51, 10 September 2025 (UTC)[reply]

    Adding comment to prevent archiving; has been listed at Wikipedia:Closure requests. -- Beland (talk) 13:52, 11 October 2025 (UTC)[reply]

    The discussion above is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.
    Retrieved from "https://en.wikipedia.org/w/index.php?title=Wikipedia_talk:Manual_of_Style&oldid=1316346935"