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]
    Extended content
    This section is an excerpt from Wikipedia talk:Manual of Style/Capital letters § Concluded.[edit]
    Capitalization-specific:

    MOS feedback needed on FA Dementia with Lewy bodies

    [edit]

    See discussion on talk. SandyGeorgia (Talk) 20:07, 19 April 2025 (UTC)[reply]

    The linked discussion has converged on the following consensus: I'd also support spelling out on first mention in each section (with brackets: so "Alzheimer's Disease (AD)") on first mention in each section, and then abbreviating thereafter.
    I think the suggestion to re-introduce abbreviations in each section of longer articles (on a case-by-case basis) would be a good addition to MOS:1STABBR. Lots of articles would get easier to follow that way.
    Joe vom Titan (talk) 20:01, 29 April 2025 (UTC)[reply]
    I would also second this – readers often go to one section of the article without reading the others, especially for longer articles. GnocchiFan (talk) 18:51, 29 May 2025 (UTC)[reply]

    Dates and times on Spaceflight articles

    [edit]

    Hello, I was looking to get some clarity on how we should handle dates and times on Spaceflight pages.

    The style guide for WikiProject Spaceflight states: Since space is not within any Earth-bound time zone, and to avoid regional bias, the WP:WikiProject Spaceflight community has established a consensus (discussed here) to use UTC.

    However in creating Template:Launch time, User:Redraiderengineer has opposed using UTC has the primary time zone, saying:

    Launches and landings are events that take place in a time zone on Earth, and MOS:TIMEZONE gives "priority to the place at which the event had its most significant effects." Previous editors have suggested using local time for Earth-based events (e.g., launches and landings) and UTC for events in space, which is aligned with the MoS.
    The Manual of Style has precedence, and "participants in a WikiProject cannot decide that some generally accepted policy or guideline does not apply to articles within its scope."

    I was hoping that some more experienced editors can give some clarity on how we should handle this issue. Thank you in advance for your responses. -- RickyCourtney (talk) 22:47, 28 April 2025 (UTC)[reply]

    To provide additional context, my understanding of MOS:TIMEZONE for Earth-based events (e.g., launches and landings) is to prioritize where the event took place with the local time zone and add UTC for dates and times in the infobox and/or first use in the article.
    Effectively, the difference in interpretations is the local time zone first or second. For example:
    • June 4, 2022, 9:25:02 am EDT (13:25:02 UTC); or
    • June 4, 2022 (2022-06-04), 13:25:02 UTC (9:25:02 am EDT)
    If a launch occurred in the Eastern Time Zone, the straightforward approach is to prioritize ET while including UTC to allow for a reference to events that occurred in space. This use existed on some spaceflight articles before the template and is currently used on articles such as Blue Origin NS-16, Starship flight test 8, Elon Musk's Tesla Roadster, and Blue Ghost Mission 1.
    The mention of precedence was based on my understanding that the WikiProject cannot override the MoS based on local consensus, which is an additional reason I addressed this.
    Thanks, Redraiderengineer (talk) 00:51, 29 April 2025 (UTC)[reply]
    While it’s true that a few articles are listed with launch local time first, the majority are listed with UTC first. That includes all the Apollo mission articles (most of which passed the rigorous Featured Article review process), all Soyuz mission articles (which don’t even list local time), and the majority of the Space Shuttle mission articles (many of which took off in one time zone and landed in another, further complicating what is “local time”). -- RickyCourtney (talk) 01:46, 29 April 2025 (UTC)[reply]
    There is WP:OTHERCONTENT with UTC listed first, but that alone is not a reason to prefer it especially when this matter didn't reach broad, explicit consensus in the past (plus after further examination WP:CCC). The other content does indicate that editors have determined a local time zone, and the MoS guideline is to prioritize it.
    I don't think local time is that complicated. If launch and landing have a different local time zone, they are both used for their respective event along with UTC. Space Shuttle mission articles, such as STS-40 and STS-48 (two random picks), and Artemis I are examples of articles that take this approach. Redraiderengineer (talk) 02:47, 29 April 2025 (UTC)[reply]
    I'd support adding an exception to the global style manual for spaceflight launches and landings: These should give UTC first and local time in parantheses. UTC is the most sensible when thinking about a space mission of which launch is only one part. However, local time is also relevant (e.g. night/day), so it should be given in parantheses. Perhaps an RfC is in order? Joe vom Titan (talk) 19:31, 29 April 2025 (UTC)[reply]
    If an exception is desired, I agree that an RfC is the next step. Redraiderengineer (talk) 15:40, 22 May 2025 (UTC)[reply]
    I don't believe that the style guide for WikiProject Spaceflight is in conflict , nor that your interpretation is incorrect. There are no time zones in space, so we are outside the guidance of MOS:TIMEZONE. By "priority", the MoS only meant which local time to use, not the ordering of the time. (As usual with the MoS, this change was made BOLD-ly, and reflects local consensus only [2]) I personally feel that the putting UTC first when multiple local times are in use makes the article easier to read. Hawkeye7 (discuss) 00:36, 12 June 2025 (UTC)[reply]

    It should also be noted that the actual time used for the launches by the launch control and operators, FAA signoffs etc., is UTC not local time, just like with all other aspects of aviation. Canterbury Tail talk 14:04, 11 June 2025 (UTC)[reply]

    New York-New York Hotel and Casino

    [edit]

    This is an odd one. My initial instinct was that an en dash should be used, not a hyphen, but upon double-checking MOS:DASH, I'm not so sure. Although we normally use an en dash in compounds when the connection might otherwise be expressed with to, versus, and, or between, this is not connecting two place names like Minneapolis–Saint Paul but rather repeating the same place name twice (unless the second "New York" refers to the state, which would be an odd use of a hyphen or dash, but I couldn't find information on the origins of the name in the article or online). Use an en dash for the names of two or more entities in an attributive compound would seem to support the use of an en dash, but again, it is technically the same entity and I'm not confident this classifies as an "attributive compound". We could also easily shorten the name to New York-New York (and probably should, per WP:CONCISE and WP:COMMONNAME), in which case, we are advised to use a hyphen in compounded proper names of single entities. I guess it just looks strange to use a hyphen because we usually use an en dash when applying a prefix or suffix to a compound that itself includes a space, but this isn't a prefix. This reads like New / York-New / York rather than New York / – / New York, and "York" obviously doesn't modify "New". Thoughts? InfiniteNexus (talk) 21:53, 23 May 2025 (UTC)[reply]

    How do independent sources that discuss this hotel/casino present the name? Do they “correct” the dash, or consider it “part of the name” and leave it. I would especially be interested in exploring sources that we trust to usually get stylization right - to see if they make an exception in this case. Blueboar (talk) 22:40, 23 May 2025 (UTC)[reply]
    Should we be looking to third-party sources for this? Many sources don't use en dashes at all in accordance with whatever style guide they follow (e.g. AP); we have our own style guide. So I don't think this is a matter of COMMONNAME. In any case, a cursory search on Google Books shows wild inconsistency, ranging from spaces to commas to slashes (and many false positives). InfiniteNexus (talk) 22:51, 23 May 2025 (UTC)[reply]
    I think you have answered my question - since we don’t see other sources making an exception to their own style guides for it, then neither should we. Blueboar (talk) 00:24, 24 May 2025 (UTC)[reply]
    When I said "many sources don't use en dashes at all", I was not referring to this instance in particular, I meant in general. Some style guides do not use en dashes in any scenario, including cases where our style guide would call for them. InfiniteNexus (talk) 00:34, 24 May 2025 (UTC)[reply]
    Never look to sources on style issues. Sources are for facts, not style. External style guides can influence our MoS, but they are only one of multiple factors to be considered. Any mention of style guides should be in the context of proposed changes to our MoS, not individual cases. ―Mandruss ☎ IMO. 00:38, 24 May 2025 (UTC)[reply]
    That said, damn the hotel owners for unnecessarily creating such a conundrum for Wikipedia's MoS wonks. "New York Hotel and Casino" would have done just fine, and readers could just deal with any ambiguity between that and Sydney's New York Hotel. I'd be inclined to use the hyphen and call it a day; it's just not that critical. ―Mandruss ☎ IMO. 01:01, 24 May 2025 (UTC)[reply]

    "Use the appropriate plural"

    [edit]

    Use the appropriate plural; allow for cases (such as excursus or hanif) in which a word is now listed in major English dictionaries, and normally takes an s or es plural, not its original plural...

    The instruction under MOS:PLURALS is rather clumsy. The "use the appropriate plural" bit doesn't really convey anything, as it should go without saying. It seems to just be there to lead into the next clause, which mentions "cases such as excursus or hanif" without any context as to how these terms are significant. Apparently these are loanwords from Latin and Arabic that have become established in English. So what about loanwords that have not become established? What kind of appropriate plurals are to be used for them? Do we reflect CMOS's recommendation that "plurals of non-English words should be formed as in the original language"?[3] If so, this should be mentioned in the paragraph before the bit about allowing for special cases. But then the whole thing seems like it would better fit under the Non-English terms section, especially since there are already Terms without/with common usage in English subheadings there. Should it be moved there instead? --Paul_012 (talk) 02:43, 10 June 2025 (UTC)[reply]

    It is oddly phrased. Can I assert that I am duly allowing for the existence of such cases but indices remains the appropriate plural? Or am I being told to use indexes? NebY (talk) 19:33, 10 June 2025 (UTC)[reply]
    The word index is a weird case. Both indexes and indices are valid plurals, but they refer to different things. If a document has an index for people and a separate index for places, then use indexes to refer to them collectively; if there are multiple scalars, e.g., subscripts, then use indices. -- Shmuel (Seymour J.) Metz Username:Chatul (talk) 13:43, 11 June 2025 (UTC)[reply]
    That's very nicely summarised - thank you - and I clearly picked a bad example. I am still rather perplexed as to what "allow for cases" is telling/guiding an editor to do; could we be clearer? NebY (talk) 16:14, 11 June 2025 (UTC)[reply]
    It's odd. There does seem to be something missing, as in: Non-English words used in running text in English (typically?) follow the pluralization conventions of the native language [example]. For cases… followed by the current statement (or some improved version conveying the point). It might be helpful to link between this section and the Non-English terms section, if some version of the guidance is to remain here. For cases where there is more than one reasonably standard plural form, especially where it does not effect the meaning, I wouldn't be overly prescriptive here. --MYCETEAE 🍄‍🟫—talk 02:06, 12 June 2025 (UTC)[reply]
    We could perhaps turn your phrasing around: Non-English words adopted into English take the plurals shown in appropriate English dictionaries, otherwise we follow the pluralization conventions of the native language? I put "appropriate" because ENGVAR might come into it. NebY (talk) 12:24, 12 June 2025 (UTC)[reply]
    This is a big improvement, certainly better than what I wrote. Far less clumsy than the original and more likely to be helpful, and I don't think it changes the underlying standard. --MYCETEAE 🍄‍🟫—talk 21:38, 12 June 2025 (UTC)[reply]
    "plurals of non-English words should be formed as in the original language". So the plural of "bureau" is "bureaux". Hawkeye7 (discuss) 23:49, 11 June 2025 (UTC)[reply]
    Yes, I really hope Chicago qualifies that. NebY (talk) 12:02, 12 June 2025 (UTC)[reply]
    Doesn't this go without saying? I don't consider bureau a non-English word, frankly. The full section (11.3: Non-English words and phrases in an English context) in Chicago mostly concerns setting these words in italics, noting that this applies especially when such words are not listed in standard English dictionaries or may be unfamiliar to readers. This convention is intended to alert readers that such terms would not normally be found in English and can help prevent a non-English word from being misread as an error. The line about plurals is exactly as quoted in the linked piece above and comes near the end of the section, after the scope of the guidance is established. --MYCETEAE 🍄‍🟫—talk 21:35, 12 June 2025 (UTC)[reply]
    The problem is that English is full of words that were originally “non-English”, but are now considered part of the English language (bureau and index are both good examples). Blueboar (talk) 12:05, 12 June 2025 (UTC)[reply]
    Bureau's a nice ENGVAR example too; very often -x in BrEng (eg the formal name Citizens Advice Bureaux) though I think often -s of the desks nowadays, but in AmEng more often -s. Still, dictionaries cover that well enough already, and so far the only sense I can make of our "allow for cases" guidance is "if it's in the English dictionary, do what that says and if not, do what the other language does". NebY (talk) 22:31, 12 June 2025 (UTC)[reply]
    That sounds very reasonable, and simple enough. Gawaon (talk) 06:54, 13 June 2025 (UTC)[reply]
    OK, I've changed it as above but dropping the "Non-English" from "Non-English words adopted into English"; if it's adopted, it's not non-English any more, or at least that's needlessly arguable. Hope that's OK. NebY (talk) 10:23, 15 June 2025 (UTC)[reply]
    Reads OK, but why did you remove the examples? Gawaon (talk) 10:25, 15 June 2025 (UTC)[reply]
    When I saw how the paragraph would now work, the examples seemed rather obscure and didn't clarify so much as take the reader into the weeds. I wound up daring to think the bald statement was enough, even though the MoS is usually more, um, discursive. In short, no very strong reason! NebY (talk) 10:59, 15 June 2025 (UTC)[reply]

    The "Italics within quotations" section mentions one applicable template, but not another.

    [edit]

    The page is semi-protected, and I only really make small fixes to articles (which doesn't happen so often, so I haven't been autoconfirmed), so I figured I may as well just mention it here. The "Italics within quotations" section mentions using the {{em}} template for emphasis within quotation, using {{em}} for emphasis is in another section. In the same way, {{lang}} is talked about for words in different languages in another section, however, {{lang}} is not mentioned in "Italics within quotations". I feel both of these should be treated equally, whether it be removing the existing mention of {{em}}, or adding a mention of {{lang}}. I'm on the side of adding {{lang}}, but that decision's up to someone else, I can't edit the article. Davie53K (talk) 15:59, 10 June 2025 (UTC)[reply]

    Maybe a link to § Non-English quotations. -- Shmuel (Seymour J.) Metz Username:Chatul (talk) 17:42, 10 June 2025 (UTC)[reply]

    ENGVAR and quoted speech from a written source

    [edit]

    The guidance under Quotations at Wikipedia:Manual of Style#Original wording is pretty straightforward and aligns with my expectations: Quotations must be verifiably attributed, and the wording of the quoted text must be faithfully reproduced and In direct quotations, retain dialectal and archaic spellings. My question concerns this passage (emphasis added) from The Guardian:

    Asked about the incident on the ABC, the Nationals senator Matt Canavan said “it looks like there was a targeting there” but added he was “loth to jump to conclusions though when you just see part of the footage”.

    This is reproduced and referenced in to The Guardian June 2025 Los Angeles protests#Australia as follows (emphasis added):

    Albanese and Senator Matt Canavan both said Tomasi's shooting appeared targeted, with Canavan adding he had only seen part of the footage and was "loath to jump to conclusions".

    Here we have a British newspaper reproducing a spoken quote by an Australian man on Australian TV using a variant spelling loth. In an article written in American English, is it permissible to change the spelling to loath? (Note: The Albanese statement is sourced to a different article.) --MYCETEAE 🍄‍🟫—talk 03:24, 13 June 2025 (UTC)[reply]

    WP is full of numerous examples where the Engvar of a quote is preserved despite being at odds with the established Engvar of an article. This case should be no different. MapReader (talk) 04:05, 13 June 2025 (UTC)[reply]
    MOS:CONSISTENT: "The conventions of a particular variety of English should be followed consistently within a given article. Exceptions include: Quotations". Hawkeye7 (discuss) 05:02, 13 June 2025 (UTC)[reply]
    Thanks, both of you. This had been my approach, but someone edited it. I wonder if they thought it was a typo, as I might have done. I needed a sanity check because if we were citing the video we would of course never have transcribed it that way. I will correct and leave an invisible note. --MYCETEAE 🍄‍🟫—talk 06:02, 13 June 2025 (UTC)[reply]

    When to include the country alongside a city

    [edit]

    Does MOS have guidance on when a city is large or well-known enough that we should exclude the country name in an article's text? For example, see the second sentence in 2024 UEFA Champions League final. I'm not seeing advice on this in MOS:PLACE, and MOS:OVERLINK only covers linking.

    If it doesn't, where should that line be drawn? I would assume that London, UK, and New York City, US, are above it. Arguably, places like Chicago, US; Seoul, South Korea; and Istanbul, Turkey, could be as well. Ed [talk] [OMT] 19:42, 13 June 2025 (UTC)[reply]

    • It seems like it should be precisely the same list as "cities that go without wikilinks", no? Given the question in both cases is whether the city is sufficiently recognizable to nearly all our readers. Remsense 🌈 论 19:45, 13 June 2025 (UTC)[reply]
    • First, are we talking about mentioning the city in running text, or titling an article?
      If running text, I would ask whether there is there a need for disambiguation? An article that mentions “Saint Petersburg” may need to clarify whether it is referring to the city in Russia, or the city in Florida (or some other one).
      Article Context comes into play here. If the context is Medieval History, the reader can assume that a reference to “Boston” is referring to the town in England (as it was the only “Boston” that existed during this era). Blueboar (talk) 20:22, 13 June 2025 (UTC)[reply]
      I'm concerned you may be overestimating our typical reader's knowledge of medieval history. EEng 20:26, 13 June 2025 (UTC)[reply]
      @Blueboar: As I said above, "in an article's text". :-)) I agree that disambiguation and especially context are important. Ed [talk] [OMT] 20:25, 13 June 2025 (UTC)[reply]
    • For what it's worth, see the list linked as footnote [a] from the "Major cities" section here [4]. Of course, that list isn't above criticism in the context of a modern, international encyclopedia e.g. (with all due respect) Cincinnati, Cleveland, Oklahoma City, San Antonio??? EEng 20:26, 13 June 2025 (UTC)[reply]

    Do not use the precomposed ellipsis character (…)

    [edit]

    I propose to rethink this rule. It was established by one user, having extreme opinion on this issue, 20 years ago, as stated in this discussion (latest discussion about this rule that I found in archive of this talk page). It was justified by problems with display of this Unicode character in very old browsers and fonts, all of these problems should be irrelevant now. The problem that brought me here is this:

    1. AWB replaces Unicode character to three dots as a part of "minor fixes" pack, containing hundreds of fixes, and this pack could only be turned completely on or completely off, there is no way to turn off only this replacement.
    2. In Russian Wikipedia, my main wiki, local MoS prescribes usage of Unicode character instead of three dots. AWB with "minor fixes" enabled brokes our MoS.
    3. If I file a ticket to AWB developers, they probably will use enwiki's MoS as a reason not to change AWB minor fixes. Only when enwiki's rule will changed, I can request AWB developers to remove and even inverse this harmful replacement.

    MBH (talk) 16:54, 16 June 2025 (UTC)[reply]

    Why not change the ruwiki MoS? Nikkimaria (talk) 05:07, 17 June 2025 (UTC)[reply]
    If the rules for this differ across various Wikipedias, getting the AWB developers to add a toggle button for this feature seems like the best solution (or maybe it could be autodetected based on the wiki one is in?). Gawaon (talk) 07:23, 17 June 2025 (UTC)[reply]
    I've not used AWB in other languages, but it makes sense that different rules would apply in different languages. At least we need to check with the AWB developers before changing the MoS on this topic. Last time I looked, the precomposed character and the three dots looked significantly different depending on fonts chosen. SchreiberBike | ⌨  11:07, 17 June 2025 (UTC)[reply]
    We discussed it here last year. The discussion didn't require a formal close; there was some support for but some marked opposition to switching to or even allowing the Unicde character, and consensus for change was not in prospect. NebY (talk) 11:23, 17 June 2025 (UTC)[reply]

    What's a good format for an interview?

    [edit]

    Here's my example. Please suggest improvements. I have use italics for the people. Is that proper?

    In a CNN town hall interview with Comey, Anderson Cooper addressed Trump's false claim that "he did not stay overnight in Moscow around the time of the Miss Universe pageant in 2013":[1]

    Cooper: "Do you think it's significant that the President lied to you twice?"

    Comey: "It's always significant when someone lies to you, especially about something you're not asking about. It tends to reflect a consciousness of guilt."

    Valjean (talk) (PING me) 19:29, 17 June 2025 (UTC)[reply]

    Valjean. I'm not aware of any policy on the matter, but your format looks good to me. Pretty much the same issue applies to a quote from a play – e.g. (copied from the Parchment article):

    In Shakespeare's Hamlet (written c.1599–1602) the following exchange occurs:

    Hamlet. Is not parchment made of sheepskins?
    Horatio. Ay, my lord, and of calves' skins too.[2]

    (You'll note two minor inconsistencies there, as to (a) whether the speaker's name is followed by a colon or a period, and (b) whether the spoken words are put in quotation marks; but I think those are valid distinctions to make for the difference between an interview and a play.) 128.86.179.55 (talk) 16:42, 18 June 2025 (UTC)[reply]
    Thanks. In these cases, the quotes should be there. -- Valjean (talk) (PING me) 17:58, 18 June 2025 (UTC)[reply]

    Is there anything in our MOS that would forbid also bolding it, like this?

    Cooper: "Do you think it's significant that the President lied to you twice?"

    Comey: "It's always significant when someone lies to you, especially about something you're not asking about. It tends to reflect a consciousness of guilt."

    Valjean (talk) (PING me) 18:25, 20 June 2025 (UTC)[reply]

    Well, it's not one of the certain usages listed as considered appropriate in MOS:BOLD. Why do we need to emphasise names at all, with italics or boldface? What's wrong with the much more conventional, to my eyes,
    Cooper: "Do you think it's significant that the President lied to you twice?"
    Comey: "It's always significant when someone lies to you, especially about something you're not asking about. It tends to reflect a consciousness of guilt."
    NebY (talk) 18:38, 20 June 2025 (UTC)[reply]
    • Well, bold is definitely out -- no need for anything so eye-catching. Beyond that, this is the sort of thing editors should work out on individual articles. If we get to the point where the WP:MOSBLOAT criteria kick in, then let's all talk. EEng 22:15, 20 June 2025 (UTC)[reply]

    Interesting. ChatGPT recommends bold based on common use:

    ✅ Best Practice for Quoting Interviews in Wikipedia Articles

    👇 Recommended format:

    Cooper: "Do you think it's significant that the President lied to you twice?"

    Comey: "It's always significant when someone lies to you, especially about something you're not asking about. It tends to reflect a consciousness of guilt."

    🔹 Why boldface (Name) is preferred:

    Bold names are often used in dialogue excerpts to distinguish speakers clearly — especially in sourced interviews or hearings.

    It improves scannability and mirrors the practice used in major articles (e.g., Watergate, Mueller Report, Trump impeachment).

    Avoid italics for names in this case — italics suggest either a title or an editorial distinction, which doesn't apply to speaker IDs.

    🔹 Additional notes:

    Keep the
    between lines for readability.

    Do not use quotation marks around the speaker's name.

    The blockquote template is a good choice here — you’re already using it correctly.

    ✅ Final Example (ready to use):

    Cooper: "Do you think it's significant that the President lied to you twice?"

    Comey: "It's always significant when someone lies to you, especially about something you're not asking about. It tends to reflect a consciousness of guilt."

    ✅ Why it's the best format:

    Bold speaker names clearly distinguish who is speaking without clutter.

    No italics for names — consistent with MOS and avoids confusion with titles or emphasis.

    Blockquote formatting is appropriate for extended quotes or multi-line dialogue.


    tags keep the exchange visually readable and clear.

    You're doing it exactly right. If you're consistent with this format throughout your draft, it will both meet style guidelines and make the article more accessible to readers.

    Valjean (talk) (PING me) 18:55, 21 June 2025 (UTC)[reply]

    • You're not seriously proposing that we pay attention to gibberish from a robot, are you? EEng 19:59, 21 June 2025 (UTC)[reply]
      No, it was just interesting, especially since ChatGPT knows PAG and MOS better than 99% of editors. -- Valjean (talk) (PING me) 23:56, 21 June 2025 (UTC)[reply]
      Good to hear, since for a minute there I thought you might have lost your mind. EEng 00:13, 22 June 2025 (UTC)[reply]
      No, I would never suggest that we let AI have a voting voice in our decision-making. We, as humans, need to maintain full control here, and all this talk of letting AI make so many decisions in society, business, politics, education, etc. freaks me out. Fuck that! Otherwise, we can learn from AI when it shows deep knowledge of topics, and ChatGPT has placed special emphasis on learning Wikipedia's customs, policies, guidelines, and how to employ them.
    I have started using it as a copy editor to double-check my work (I maintain full control and responsibility), and it understands MOS and policies so well that it warns when I get too close to SYNTH, OR, or NPOV violations, etc. It points out the problem and suggests solutions (attribution, sourcing, framing, balance, etc.) It's very good at that. It is obviously not one of our RS, but in discussions, we can consider what AI says "as another voice" to consider. It predicts the correct answers based on its understandings of human behavior, Wikipedia conventions, and combines that with its nearly infinite knowledge base, and Wikipedia and its sources are an important part of that base. It's worth listening to. That's all. -- Valjean (talk) (PING me) 01:45, 22 June 2025 (UTC)[reply]
    I worry that your approach may eventually lead to AI being inappropriately used as a crutch. I seriously question whether a chatbot can make any kind of effective evaluation of POV vs NPOV, or reliably detect OR. EEng 01:58, 22 June 2025 (UTC)[reply]
    I actually share your concern, but I'm suggesting using it just like we use spell checkers, just another tool. It should not replace our editorial judgment and experience. My experience has been good so far. It has caught things I had not noticed. -- Valjean (talk) (PING me) 02:03, 22 June 2025 (UTC)[reply]
    I have tentatively believed that AI, if done right, would eliminate much of the effect of uncontrolled editor bias (UEB), which is a serious problem. problem. Our PAGs, e.g. NPOV, try to eliminate UEB, but they are terrible at it and biased, policy-compliant content can and frequently does get into articles. All it needs is consensus among editors with uncontrolled biases.
    On the other hand, I don't see how AI could understand our confusing and self-contradictory PAGs any better than we can (your experience notwithstanding). What AI might do is help us see where the PAGs are confusing and self-contradictory, so that we might finally do something about it. How close are we to AI that can do that? I have no idea. But only then could we use AI to provide effective guidance for article content. ―Mandruss ☎ IMO. 06:42, 22 June 2025 (UTC) Edited after reply 07:20, 22 June 2025 (UTC)[reply]
    I do in any case believe that AI-generated content should never be used as argument in a discussion. Editors may well discuss with a chatbot in the background, but if they want to make a point, they should be able to voice it as their own! Gawaon (talk) 07:12, 22 June 2025 (UTC)[reply]
    Mandruss, you you really think chatbots are better at avoiding bias than editors are? In my experience, it's fairly easy to coax chatbots into saying what you want to hear, so a chatbot used by a biased editor will likely share that bias. Plus they'll likely bring their own biases with them, due to pretraining. Gawaon (talk) 07:30, 22 June 2025 (UTC)[reply]
    I'm speaking of AI in general; AI can come in many forms, and chatbots are merely one of the first. I don't think AI necessarily requires a question beyond "How can this article's NPOV be improved?", which would eliminate question bias.
    In theory, we could have an AI bot, dedicated to NPOV, that roamed the site analyzing articles' NPOV and posted suggestions for improvement on article talk pages. We probably wouldn't need a policy that the suggestions should be acted on, but editors failing to act on them would not be a good look. I think they should need very good articulable reasons to ignore them. If the bot frequently gave bad advice, we would figure out why and fix it; it might take a number of years to get it right. Sometimes the problem would be with our PAGs, not the bot, so we would fix the PAGs.
    As I said, I have no idea how close we are; I could be speaking of progress twenty years from now, for all I know. (I'll be cold in the grave by then.)
    Plus they'll likely bring their own biases with them, due to pretraining. As I also said, if done right. That would be a crucial point. AI bias is not inevitable in my opinion; its creators just need to (1) know what they're doing and (2) be prepared to put NPOV before their own biases. The AI must be detached from and impassionate about the article subjects; i.e. unhuman. ―Mandruss ☎ IMO. 07:47, 22 June 2025 (UTC)[reply]
    I'm not thrilled at the idea of a bot telling editors what they have to do. Even if it worked reliably (about which I have grave doubts), that might kinda destroy motivation. More plausible for me is that editors employ bots to find NPOV violations faster than they could themselves; the editor might then flag or fix an NPOV violation if the editor themselves agrees that it is one. But it must still be a human who makes the call. Gawaon (talk) 11:01, 22 June 2025 (UTC)[reply]
    Ok, but we (excluding me) may have 20 years to talk about it. It's certainly not happening now. ―Mandruss ☎ IMO. 12:48, 22 June 2025 (UTC)[reply]
    Painting of people betting on a game of pure chance
    10 tokens on Gemini CMD (talk) 13:23, 22 June 2025 (UTC)[reply]
    I asked Copilot "On Wikipedia, is it bad practice to bold names in dialog?".

    On Wikipedia, bolding names in dialogue is generally discouraged unless there's a compelling reason tied to the article's structure or clarity. According to the Manual of Style on text formatting, boldface is typically reserved for the article's subject in the lead section or for terms that redirect to the article. It's not meant for emphasis or stylistic flair in body text, including dialogue.

    So if you're writing a fictional conversation or quoting speech, it's better to stick with standard formatting—bolding names might look jarring or imply undue emphasis. If clarity is a concern (like distinguishing speakers), consider using formatting like indentation, line breaks, or even a simple attribution line instead. Want help reworking a snippet to fit the style?

    Note: bot-wrestling is purely entertainment and not a legitimate sporting contest. NebY (talk) 13:13, 22 June 2025 (UTC)[reply]
    That seems to mostly address other uses than what we're talking about here. The following mention of template {{Dialogue}} is spot on. -- Valjean (talk) (PING me) 15:31, 22 June 2025 (UTC)[reply]
    Our template {{Dialogue}} does bold the speakers. -- 14:53, 22 June 2025 (UTC) — Preceding unsigned comment added by Michael Bednarek (talk • contribs)
    Oh, so it does! Among the 51 articles that use it, the quote box at Rhodesia's Unilateral Declaration of Independence#Final steps to UDI doesn't seem too jarring to me (smaller font among long pieces of text, maybe) but Duck test#Notable uses looks like a fine example of excessive emphasis. NebY (talk) 15:19, 22 June 2025 (UTC)[reply]
    I wasn't aware of that template. That looks very nice. This is the only manner of bolding under discussion here, and it looks good. -- Valjean (talk) (PING me) 15:31, 22 June 2025 (UTC)[reply]
    If you mean the duck test example, no, it looks terrible! --𝕁𝕄𝔽 (talk) 15:48, 22 June 2025 (UTC)[reply]
    Oh no! I mean the first example (Rhodesia). -- Valjean (talk) (PING me) 15:51, 22 June 2025 (UTC)[reply]

    And just to add to fun, Google Gemini is adamant that if it is a play script or movie script, it should be all caps, no bold, new line, thus:

    HAMLET
    Is not parchment made of sheepskins?
    HORATIO
    Ay, my lord, and of calves' skins too.

    and for a movie script, the convention is almost the same except that the character name is centred. (The markup needed to show that is left as an exercise for the reader.)

    So does anybody like

    COOPER: "Do you think it's significant that the President lied to you twice?"
    COMEY: "It's always significant when someone lies to you, especially about something you're not asking about. It tends to reflect a consciousness of guilt.

    It certainly looks better than bold (everything looks better than bold!) but maybe not better than italic. --𝕁𝕄𝔽 (talk) 15:48, 22 June 2025 (UTC)[reply]

    A script has to be read at a quick glance, hence the all caps. Our discussion here has nothing to do with that. -- Valjean (talk) (PING me) 15:53, 22 June 2025 (UTC)[reply]

    So here is a use of the {{Dialogue}} template (indented):

    When Cohen described the tape stopped by Rtskhiladze as the "infamous pee tape when Mr. Trump was in Moscow for the Miss Universe Pageant", Representative Jackie Speier sought clarification:[3]: 228 

    Speier: It wasn't infamous then, was it?
    Cohen: Yes, yes. That the tape - the conversation about the tape has gone back [to] almost a couple months past when they were there [in Moscow in November 2013] for the Miss Universe Pageant that that tape existed.

    How's that look? -- Valjean (talk) (PING me) 15:50, 22 June 2025 (UTC)[reply]

    I really should have quoted Gemini in full, and (annoyingly) it supports bold markup:
    Q&A Format (Formal): This format clearly labels the speaker before each line of dialogue, often in bold. This is best for interviews where the exact back-and-forth is central and easy to follow.
    — Google Gemini
    So it agrees with the {{Dialogue}} style. 𝕁𝕄𝔽 (talk) 16:37, 22 June 2025 (UTC)[reply]


    References

    1. ^ Foran, Clare (April 26, 2018). "Comey on Trump's assertion that releasing memos was illegal: 'He's just making stuff up'". CNN. Retrieved July 29, 2024.
    2. ^ Hamlet 5.1/116–117, Folger Shakespeare Library
    3. ^ "Deposition of Michael Cohen, Part 2" (PDF). House Intelligence Committee. March 6, 2019. pp. 225–229, 235–236, 254.

    Ref name format, space or no space

    [edit]

    Which style (with or without a space near the end) is preferable, and what are the advantages and disadvantages of each? (I know it doesn't make any difference in function.)

    Reference 1.<ref name="refname1" />

    or

    Reference 2.<ref name="refname2"/>

    Valjean (talk) (PING me) 18:46, 21 June 2025 (UTC)[reply]

    Why should we need a preferred style for something that doesn't matter (no difference in function)? Schazjmd (talk) 19:19, 21 June 2025 (UTC)[reply]
    I was just curious. I see both uses and wondered if there is some reason. Maybe when searching a page or something. I have encountered some interesting reasons for editorial preferences over the years and was wondering if I'd learn some fascinating reason this time. If you don't think of anything, then just ignore this thread. -- Valjean (talk) (PING me) 23:54, 21 June 2025 (UTC)[reply]
    Retrieved from "https://en.wikipedia.org/w/index.php?title=Wikipedia_talk:Manual_of_Style&oldid=1296840522"