I requested this move as the current title indicates that there is one central election on one day (like 2018 Pakistani and 2019 British election pages); however, there are dozens of different elections in Nigeria throughout 2023 (from February to at least November) making this page more comparable to the 2020 United States elections (especially as they are both presidential systems with a large number of disparate elections throughout the year). Also, as the component elections in this page already have unique pages, it is no longer like the 2019 page where there was no separate election page. In accordance with other like pages, such as the 2022 Nigerian elections, 2023 Nigerian elections is more accurate. When I brought this up first, it was clear that the user that moved this page is not at all familiar with the content of the page; when I requested it be moved others seemingly understood why but stopped short of supporting while the opponent refused to engage until the discussion was closed. On the second move request, an opponent pivoted to a content discussion before also refusing to engage until the discussion was closed. The mover didn't engage on their talk page either, so I am bringing it here as no one has spoken in good faith in months and this needs to be talked out. Watercheetah99 (talk) 04:36, 28 June 2022 (UTC)[reply]
The above is an archive of the move review of the page listed in the heading. Please do not modify it.
The s-variant, -isation, of South Asian English (inherited from non-Oxford British English) has been in use in the Indus Valley Civilisation (IVC) page from its inception in 2001. There may have been times when someone moved it to the z-variant (with -ization), but it was soon reverted back
It is my contention that:
an article that has relied on
MOS:TIES to South Asian Englishes for orthography (which includes the spelling "-isation," in contrast to "-ization") and on
MOS:RETAIN for retaining the spelling for 21 years in the face of monthly attempts to change it to -ization.
must also sample in the corpus of the language of MOS:TIES for capitalization.
In other words, we cannot change midstream in the big river of Standard Englishes from the South Asian Standard English to some other(s) for the purposes of determining capitalization.
I would not employ this argument to change a page name that had existed for 21 years in the lower-case "civilisation" to the capitalized "Civilisation." But I think this statistic is sufficient for retaining a page name in its uppercase "Civilisation" of 21 years.
"plummeted" is not a good word for the relatively modest change of ratio of capitalized to uncapitalized in the sources. And all this was discussed at the RM; it is not a sensible reason to challenge the close. Dicklyon (talk) 19:56, 5 July 2022 (UTC)[reply]
Arbitrary break 1 (Indus Valley civilisation)
Closer's note I told Fowler to file this review without fully engaging with me in my talk over the close because I felt like consulting uninvolved parties over how I closed the RM would be more helpful in resolving these issues. — Ceso femmuin mbolgaig mbung, mellohi! (投稿) 20:03, 25 June 2022 (UTC)[reply]
I have asked the closer specifically below if they were aware at the time of closing that Google ngrams do not sample among the books cited in MOS:CAPS, i.e. "independent, reliable sources," but among everything in the Google database, which includes all sorts of books, including self-published ones. I showed below that in the search for
"IVC/c" with no effort at all 800 self-published books can be rounded up from among half a dozen Vanity Publishers, contrasting with 1300 published by the presses of the dozens and dozens of the world's universities.
I asked because the initiator of the page move had made just this statement:
Statistics from books show tons of lowercase (even valley sometimes): [1] and [2]. Dicklyon (talk) 15:07, 8 June 2022 (UTC)
Another red herring. If you look at the first page of 10 book hits from university presses, you see 5 of 10 lowercase (or at least that's what I see); in the self-published, it's only 3 of 10 lowercase. So if the self-publishers skewed the stats, it's more likely in the opposite direction of what you're suggesting. Self-publishers are the ones more likely to have copied Wikipedia's over-capitalization, as I'm sure you'll agree. Dicklyon (talk) 04:21, 29 June 2022 (UTC)[reply]
These arguments based on "writers may have copied from Wikipedia" are just speculations and imaginations and quite fallacious, and you have yourself admitted that you cannot prove any of it. Please do not keep repeating these un-provable claims. Chaipau (talk) 09:10, 29 June 2022 (UTC)[reply]
I wasn't quoting you directly only calling out a type of argument used often. I could have said writers-may-have-copied-from-Wikipedia or called it differently. But I can see now that it looks like a quote. Thanks for pointing it out. Chaipau (talk) 09:21, 30 June 2022 (UTC)[reply]
Arguments that Wikipedia has influenced writers are not the same as accusations of copying, but yes, there can also be a bit of that. Generally, the "recent trends" in capitalization are positive, partly likely due to Wikipedia influence on topics where Wikipedia caps the title, and partly because self-published books (and wiki-mirror books) have become more common in recent years. So the worst arguments in this whole mess are those that focus on very recent stats (e.g. the blip up in caps in 2017–2019) and call that a trend. The n-gram stats are much more meaningful if you discount the years in which book sales came to be dominated by online sales (i.e. the last 15 years or so); before that, Google had rather few self-published or wiki-based books to scan. Dicklyon (talk) 17:27, 30 June 2022 (UTC)[reply]
I ask the closer again: Were they aware that Google ngrams which they purportedly believed do not sample either independence or reliability, the linchpins, or should I say kingpins, of MOS:CAPS? Fowler&fowler«Talk»08:58, 29 June 2022 (UTC)[reply]
Again: "but ngrams may also sample unreliable sources" line of argumentation is, to put it bluntly, bullshit. We have always used ngrams as a statistical tool that demonstrates general academic usage and there is no reason to believe that reliability of a source would correlates with spelling of "C/civilis/zation". No such user (talk) 09:07, 29 June 2022 (UTC)[reply]
Can you demonstrate that Google ngrams demonstrate "general academic usage"? Google ngrams has a dropdown menu to look for usage in [fictional works Clearly, not all ngrams in that dataset are from serious academic work. That is not the purpose of the dataset either. How many are from self-published books can only be speculated because Google does not provide that information. So use Google Ngrams judiciously. Not like the holy grail. Chaipau (talk) 09:29, 29 June 2022 (UTC)[reply]
Alright, "academic" was a bit strong, but let's say "formal". As for the "self-published" books theory, it's both false and irrelevant
Over 100 sources of metadata information were used by Google to generate a comprehensive catalog of books. Some of these sources are library catalogs (e.g., the list of books in the collections of University of Michigan, or union catalogs such as the collective list of books in Bosnian libraries), some are from retailers (e.g., Decitre, a French bookseller), and some are from commercial aggregators (e.g., Ingram). In addition, Google also receives metadata from its 30,000 partner publishers.[1] That does not leave much room for self-published. Neither you nor Fowler, however, have addressed the other argument of mine: why would you suspect that the results would be significantly different even if a larger ratio of self-published books was included in the corpus? No such user (talk) 10:19, 29 June 2022 (UTC)[reply]
The book sources include fiction (novels, etc.), which is where a lot of less formal usage is found. These are rightly included in the stats for how non-specialist writers treat terms such as Indus Valley civilisation. Dicklyon (talk) 19:15, 29 June 2022 (UTC)[reply]
>>>That demonstrates academic usage.
Google ngrams are an aspect of Corpus linguistics. They track your own little corpus "bluntly" and "bullshit" Has the latter overtaken the former in academic usage in the new millennium?
F&f, for all intents, you WP:BADGERED the closer to close the discussion after they relisted it. One should be careful in what they wish for since they may just get it. Well, it happened and now we are here at your instigation. And it continues. I have already answered what anybody not specifically privy to the n-gram project can know. Cinderella157 (talk) 10:19, 29 June 2022 (UTC)[reply]
I did not badger them. I was attempting to engage them, which they steadfastly refused to do. Please look at the mandatory discussion on the closer's talk page. How insubstantial is that? Not once have they engaged me. Not once. Fowler&fowler«Talk»10:43, 29 June 2022 (UTC)[reply]
Given the post at the closers talk page, I would think that their response was more than adequately engaging and they should be commended. In my view, you didn't really invite a constructive exchange of views. Cinderella157 (talk) 11:46, 29 June 2022 (UTC)[reply]
Endorse close as moved (as nom). It is perfectly common for RM discussions to be closed based on a majority vote count rather than on the strength of arguments based in policies, guidelines, and sources. Thankfully, that's not what happened here. When closer first relisted he encouraged opposers to focus more on such reasons than on their opinions, which there are more than plenty of. But there was really no way to spin the source stats in support of caps, in light of policy as stated at WP:NCCAPS, so no good arguments came about. Closer made it clear that he looked at he arguments carefully. I don't see a problem. Dicklyon (talk) 22:30, 25 June 2022 (UTC)[reply]
Where is the evidence that the closer considered the arguments made by the 70% of the participants? The only evidence we have is that the closer ignored them. Chaipau (talk) 11:53, 26 June 2022 (UTC)[reply]
You mean, evidence-free arguments such as Oppose – Civilization is not a modifier of Indus Valley but is a part of the name (the actual civilization extends well beyond the valley). [RegentsPark] followed by several Oppose per RegentsPark, or Oppose. Generally seen as a proper name. or Oppose. Reliable sources treat this as a proper noun. Or maybe ad-hominem ones such as Oppose for consistency. I distrust proposers who feel they have to keep arguing the toss after every edit., Oppose Any editor so foolishly obsessed on such an unimportant matter that they edit war and get blocked deserves to have their efforts thwarted.. The closer was right to downgrade or outright ignore them. No such user (talk) 09:14, 29 June 2022 (UTC)[reply]
>>>You mean evidence-free arguments such as, "Civilization is not a modifier of Indus Valley but is a part of the name (the actual civilization extends well beyond the valley). [RegentsPark] followed by several Oppose per RegentsPark, or Oppose.
Endorse. Both the P&G and the evidence (usage stats, etc.) support this move, and the closer analyzed it correctly. RM is not a head-count vote, and no amount of opinion-mongering trumps P&G and evidence. Same goes for attempts to capitalize because something is "important" (see MOS:SIGCAPS). — SMcCandlish☏¢ 😼 23:05, 25 June 2022 (UTC)[reply]
Overturn (!voted oppose) The closing editor's rationale relies solely on arguments based on ngram numbers, an argument made by a minority of the !voters and does not reference any of the arguments made by the the opposers, a majority of the !voters. Regardless of whether the opposer arguments are valid, the closer needs to clearly state why they are not. Instead, they have dismissed one side as "emotive" and focused solely on the ngram argument. The implication of this is that arguments based on anything other than ngram numbers are not worthy of consideration, or even of mention. If all we're doing is blindly using google ngrams to decide article titles, we could just, for every article, run a bot that tries every possible combination of titles, counts the ngrams, and then move the article to the winner. We don't do that because the community recognizes that blindly using a number is not the best way to both find accurate titles as well as to leave it to content editors to decide on the best title (Newyorkbrad's comment here expresses the latter view cogently). Because the closing editor has chosen to ignore all arguments other than the ngram ones, this RM close should be overturned. --RegentsPark (comment) 23:52, 25 June 2022 (UTC)[reply]
The biggest problem with F&F's attempts was that he kept focusing on what his favorite experts use as titles, as opposed to in sentences. The first he threw out was Tim Dyson, who has a title-case title, but doesn't cap it in sentences (see my "Bogosity" comment in the RM). And when he wrote What is it you don't understand? We don't capitalize "civilization" everywhere. We do, ideally, only in the page title or section titles. That's exactly what Dyson, Doniger, and the British Museum do. he was both expressing a misunderstanding of our policy and an admission that others often don't cap it sentences. He never acknowledged what our policy is, but kept quoting his own favorite numbers as if they were meaningful. So, yes, his arguments were hardly worthy of consideration, but I'm sure they got considered. You and F&F also both expressed a preference to Indus civilisation at some point, but no serious proposal in that direction was made; a followup RM could do that without objections based on WP:NCCAPS or MOS:CAPS. Your own argument was pretty much nothing but an appeal to "common sense"; I think we did that. And three other opposers were per your argument, whatever that was supposed to mean. Dicklyon (talk) 00:38, 26 June 2022 (UTC)[reply]
It doesn't make any difference what I said in the past. Many of us had not considered the various criteria employed in MOS before. The supporters are all from MOS who have been making their hackneyed arguments for ages. The opposers, many of us experienced Wikipedia editors, some academics, some admins, necessarily, evolved in their thinking as the arguments proceeded. You cannot respond by haggling in fish-market style, saying, "Look but F&f said this, and then they said that but this didn't jibe with that, or with another, so—after I have done my damage to a page of 21 years standing, still stunned how easy it all was, and am now busily employed in removing the curse of "Civilisation," its every vestige, from the brows of Wikipedia, in edit after edit, on page after page—it is they, F&f, who should now reinvent the Indus wheel, write out their arguments in triplicate, drive their cart to post at sites A, B, C, and D, make carbon copies, get a notary to attest their signature, ..." You need to counter my latest argument, that this page move is a slap in the face of MOS:TIES, of MOS:RETAIN, and of the tradition of South Asian Englishes in which the article is written, and that it hurts the claims of Wikipedia to be a global encyclopedia. You should be mindful that by your own admission you know nothing about the Indus Valley Civilisation, have not dipped once in its vast literature, or had made one edit to the large number of Wikipedia pages that have bearing on it. Fowler&fowler«Talk»01:45, 26 June 2022 (UTC)[reply]
It doesn't matter that you call the supporters' arguments "hackneyed". Your ENGVAR arguments have no bearing on the case question. Dicklyon (talk) 03:16, 26 June 2022 (UTC)[reply]
Had you made this change 25 years ago, I wouldn't have uttered a peep. What is the point of making the change now when the whole world is capitalizing in s-variant. It is retrogressive. Fowler&fowler«Talk»10:20, 27 June 2022 (UTC)[reply]
Dicklyon, would you mind holding off changing the capitalisation of uses of the term throughout Wikipedia's articles (like here)? In at least some instances, this is analogous to what was discussed two months ago (where I thought you must have seen there wasn't community agreement to doing that sort of stuff at scale). At the very least, you wouldn't want to waste hundreds of edits when the decision they're based on is currently being challenged. – Uanfala (talk) 01:07, 26 June 2022 (UTC)[reply]
Overturn. I opposed the RM (this is a good time to point out that you're supposed to disclose whether you participated in the RM or not) so it's needless to say that I disagree with the move. My main concern here is that the relist and close comments (by the same editor in the same comment) is a supervote. Both sides have strong arguments in this case -- strength of argument isn't determined by who bludgeons the other side the most or who can be the most verbose -- so it's wrong to dismiss the majority viewpoint in this instance. -- Vaulter03:14, 26 June 2022 (UTC)[reply]
Overturn, as the initiator of this review, and an opposer at the RM. I should like to add in addition to what I have stated at the top initiating the review, and in the link therein, that when editors citing MOS make arbitrary and unannounced interventions on a page, and when closers support their page move in direct opposition to the unanimous views of the content creators of the page, they damage the page, and by consequence, Wikipedia. As an experienced Wikipedian, the major editor of Indus Valley Civilisation, the FA India, British Raj, History of Pakistan, ... I am unhappy that this would be done. I say this as someone who has English grammar books going back to the mid-19th century, and forward to the works of Randolph Quirk, Sidney Greenbaum, Rodney Huddleston, and Ronald Carter, quite a few mentioned in the History of English grammars I wrote on a lark long ago. Although MOS is a phenomenal effort and a valued resource, MOS compulsions in insignificant matters, such as on this page, should not be allowed to create bad blood.
I should also add that on account of the expected, but still momentous, US Supreme Court decision of yesterday (June 24, 2022), and despite us residing in a state in which the women are not substantially affected by the decision, my immediate family has been affected. I will therefore be taking the summer off to support them. Before I do that, I will be spending the remaining couple of days attending to unfinished business at the Darjeeling FAR and quite a few other pages. I will not be around to engage editors on this page or the article's talk page. Best regards, Fowler&fowler«Talk»03:42, 26 June 2022 (UTC)[reply]
Overturn (!voted oppose) , Move request was closed simply based on MOS:CAPS , but MOS:CAPS further links to WP:AT for articles' title issues. There must be a consensus to move such an important article. Northeast heritage (talk) 04:24, 26 June 2022 (UTC)[reply]
WP:AT links WP:NCCAPS for capitalization issues; it says to leave the second and subsequent words in lowercase unless the title phrase is a proper name that would always occur capitalized, even mid-sentence. Both of these policy pages link to MOS:CAPS and MOS:AT which say to use sentence case in titles. WP:NCCAPS was cited and quoted in the RM discussion, so your assertion that the decision was based only on MOS:CAPS is vacuous. Dicklyon (talk) 14:51, 26 June 2022 (UTC)[reply]
Words and phrases in a sentence implicitly carry their grammatical properties. Academics have been using "Indus Valley Civilisation" in sentences. This clearly shows that "Indus Valley Civilisation" is a proper name. Of course, You can other phrases like "Indus Valley civilisation" , "Ancient Indus Valley" to represent the same civilisation. You can't use ngrams to cancel grammatical properties of different phrases. Northeast heritage (talk) 03:24, 27 June 2022 (UTC)[reply]
You may have cited WP:NCCAPS but closer didn't use it. So, My assertion isn't vacuous. Closer didn't consider RM opposers views. That's why there is no concesus. Northeast heritage (talk) 04:07, 27 June 2022 (UTC)[reply]
Comment So far, everyone who has shown up to this move review (except Uanfala) was a participant in the RM. Is it okay to divide this review into two sections, one for comments by original participants and another for uninvolved users' thoughts? — Ceso femmuin mbolgaig mbung, mellohi! (投稿) 04:44, 26 June 2022 (UTC)[reply]
Overturn to no consensus (uninvolved). The closing statement found that "there is no consensus on whether a substantial majority of sources that use a capital C can be found". The earlier relisting statement found the same lack of consensus on that question. I think the close is correct on this finding. (Participants have reasonably disagreed on the weight to give to various ngram queries, and how to analyze and interpret the results; on whether to give more weight to certain classes of sources which participants in their editorial judgment see as more reliable; and whether, to the contrary, citing a certain essay, to give such sources less weight.) Where I disagree with the close is on the conclusion to be drawn from this finding. There are three possibilities. 1. If we have consensus that enough sources capitalise, then the RM should be closed as not moved. 2. If we have consensus that not enough sources capitalise, then the RM should be closed as moved. 3. If we have no consensus on whether enough sources capitalise, then the RM should be closed as no consensus. Under RMCI there are these three outcomes, not two. We are in case (3). Adumbrativus (talk) 06:19, 26 June 2022 (UTC)[reply]
Hi, your statement is: I think the close is correct on this finding and then you make a parenthetic statement which has the appearance of being a paraphrase of the closers statement. However, this does not appear to be the case? Cinderella157 (talk) 10:06, 26 June 2022 (UTC)[reply]
The issue for the closer was "is there a consensus for this move?" That is not what the closer considered. This closure is too heavily compromised. Chaipau (talk) 11:47, 26 June 2022 (UTC)[reply]
The closing statement delineated the types of arguments that the closer felt should be discounted, and did not attempt to recap the other arguments, which presumably weren't discounted. In my parenthetical comment, I mentioned some of the latter type of arguments more explicitly (but non-exhaustively), arguments which I think made it justifiable for a closer to have reached a finding that "there is no consensus on whether a substantial majority of sources that use a capital C can be found". Thank you for the good question – it is not meant as a paraphrase of what the closer wrote, which chose not go into detail in that direction. Adumbrativus (talk) 05:36, 29 June 2022 (UTC)[reply]
Overturn (!voted oppose): The closer relisted the move request with an odd phrasing: The crux of this RM is whether sources use a capital C consistently enough. There is no consensus on whether the sources demonstrate a consistent-enough use of a capital C. At that stage there was a 14-7 vote against the move. In effect the closer had already decided to ignore all the arguments outside an opinionated "crux of this RM". During the relist period evidence was shown that even Ngrams count does not support the move. More importantly, the relist yielded an additional 2-0 vote against the move with the final tally at 16-7. Yet the closer moved the page. It is clear that: (1) the closer had pre-determined the decision with the re-list counting for nothing (2) the closer has shown no evidence that they have considered the arguments made by 70% of the participants in a judicious manner. In effect this closure is no closure at all, but one additional vote for the move (without even an argument, just a per proposer vote). The final tally should be 16-8 against the move awaiting a non-partisan decision. Chaipau (talk) 11:25, 26 June 2022 (UTC)[reply]
Yes, I am aware of WP:DEMOCRACY (and also WP:NOTBURO, which I brought up during the discussion), and I am not making the argument that 16-7 defines the consensus. The argument I am making is that the closer ignored the arguments of more than 2/3rd of the participants. The discussion on Ngrams is still continuing in the talk page (here is the proposer, still making his case, for instance) even after the closure. Clearly there is no consensus on the pivotal evidence the closer used in closing. This closure has been irregular and so deserves to be overturned. Chaipau (talk) 14:28, 26 June 2022 (UTC)[reply]
Overturn (uninvolved). I'm not commenting on what I think should be the correct title for the article, I'm only commenting on how the RM was closed, as don't want this to just turn into another RM discussion. I agree with Adumbrativus' conclusion that the closer should have gone with "no consensus", and hence not moved the article in this instance, and the closing statement did not address enough of the issues raised, as already mentioned above. -Kj cheetham (talk) 11:39, 26 June 2022 (UTC)[reply]
Endorse (involved) F&f would assert that MOS:TIES and MOS:RETAIN are significant additional information not discussed in the page move discussion. F&f would state: ... MOS:TIES for its language of orthography. There is no dispute wrt MOS:TIES that "civilisation" should be spelt as such in consequence in this article. They would then state: In other words, we cannot change midstream in the big river of Standard Englishes from the South Asian Standard English to some other(s) for the purposes of determining capitalization. Where they have linked to orthography, this would state: An orthography is a set of conventions for writing a language, including norms of spelling, hyphenation, capitalization ... F&f has not established that these conventions vary to any significant extent across the corpus of English and that the question of capitalisation should be confined to a particular domain in the same way as spelling (ie "is" v "iz"). While the guidance quoted is novel and new, the evidence relied upon here is not. Before posting this MR, F&f was made aware that: arguments specifically to "is" and collectively ("is" + "iz") have been addressed in the RM and would not fall to the closer being "unaware"here.
F&f would also assert that the closer did not follow the spirit and intent of WP:RMCI. Guidance at multiple places makes it clear that strength of argument and not votes determine consensus (WP:RMNOMIN, WP:NHC, Wikipedia:Advice on closing discussions and Wikipedia:Arguments to avoid on discussion pages). A strong argument is made in respect to acknowledged criteria and particularly WP:P&G, it accurately represents the P&G and it provides evidence (not assertion) to make a cogent case. A strong argument will withstand scrutiny. Most of the comments made, particularly to oppose the RM lack strength. F&f fails to evidence that the closer has failed to follow the spirit and intent of WP:RMCI and particularly that there are any particularly strong arguments made that should have carried the case. While we could individually address the strength of the various arguments, suffice it to say at this juncture that the closer has reasonably followed the spirit and intent of WP:RMCI. There are no reasonable grounds to overturn the close. Cinderella157 (talk) 12:56, 26 June 2022 (UTC)[reply]
It has already been mentioned by a number of commentators here that the close was nothing but a supervote. It is invalid. Chaipau (talk) 04:32, 27 June 2022 (UTC)[reply]
Comment since this is WP:NOTAVOTE, we cannot offer passing opinions without a close analysis and expect our comments to carry weight. To this end, I have made a close analysis at User:Cinderella157/sandbox 1. From this, I would conclude that most of the comments made on both sides of the discusstion were of no significant weight but IMHO, the only comments of substantial weight were made in support of the move. They addressed the criteria established by WP:P&G and drew reasonable and rational conclusions from evidence presented. Many commenters here would suggest that the close has totally ignored comments made by one side. The closer makes two pertinent statements; the first: I am noting this as many voters !voted based on what they personally believed is correct usage (ie not made in consideration of the prevailing WP:P&G) This is a valid observation of the RM discussion. Secondly: This is a classic case of a debate where one side has an emotive majority but the other side has the much stronger arguments. Perhaps some may take umbrage to their comments being described as "emotive". Perhaps "subjective" or "unsubstantiated" opinion might be a better description but the meaning is not so dissimilar and an accurate assessment. The evidence (ie the RM) supports that the closer has weighed the comments in accordance with WP:P&G and the close is sound. Cinderella157 (talk) 04:06, 29 June 2022 (UTC)[reply]
References
^Michel, Jean-Baptiste; Shen, Yuan Kui; Aiden, Aviva Presser; Veres, Adrian; Gray, Matthew K.; Pickett, Joseph P.; Hoiberg, Dale; Clancy, Dan; Norvig, Peter; Orwant, Jon; Pinker, Steven; Nowak, Martin A.; Aiden, Erez Lieberman (January 14, 2011). "Quantitative Analysis of Culture Using Millions of Digitized Books". Science. 331 (6014). American Association for the Advancement of Science (AAAS): 176–182. doi:10.1126/science.1199644. ISSN0036-8075.
Arbitrary break 2 (Indus Valley civilisation)
Endorse. F&F initiation of this MR does not challenge the closer’s compliance with RMCI but merely attempts to rehash the RM. Did or did not the closer follow the intent and spirit of RMCI? I think the closer did. Mike Cline (talk) 16:16, 26 June 2022 (UTC)[reply]
Comment: WP:RMNOMIN in WP:RMCI points to WP:TITLECHANGES which suggests that If an article title has been stable for a long time, and there is no good reason to change it, it should not be changed. The title has been stable for about 21 years on the capitalization issue, as has been pointed out numerous times. Thus the onus was on the proposer to show that there was a good reason for a change, and the closer should have examined whether there was a consensus on the good reason. Instead the closer stated thus: The crux of this RM is whether sources use a capital C consistently enough. There is no consensus on whether the sources demonstrate a consistent-enough use of a capital C. which assumes the opposite - that the title was wrong and the onus was on the defenders to seek consensus for their position. I am sure many would be surprised if if it were to be asserted that there is a consensus on this point. So I cannot agree with you. This closure violates the spirit and intent of WP:RMCI. Chaipau (talk) 17:49, 26 June 2022 (UTC)[reply]
Overturn. The close was such an obvious supervote that it can't have any validity. The whole exercise becomes meaningless if we go with such closes. -- Kautilya3 (talk) 21:52, 26 June 2022 (UTC)[reply]
Overturn (involved, voted against the move; note that several of the above !votes are also involved but did not declare so). Note that I am not saying "overturn to not moved / no consensus" but rather just "Overturn" as this was not a good closure. First off, it's not necessarily a problem for a non-admin to be willing to tackle tough closes, but this is not how you do it. Decrying votes in opposition as being "emotive" might be valid if, say, there was a swarm of new accounts recruited from Reddit that argued along nationalist grounds. But nothing of the sort happened. All of the oppose votes were in good faith and based on valid WP:AT criteria. The ngrams arguably favored the existing title, and ngrams are known to be tricky to interpret anyway. Now, maybe, maybe there's a reason to move this anyway (I think Amakuru had a well-reasoned vote to move, for example, despite disagreeing with it), which is why I am voting for overturn & let someone else reclose it. But it should come from a trusted and experienced closer, and if the result is "move", it should do so on grounds other than merely dismissing opposer's arguments as "emotive", which is not really a way to sell the move when many of the opposers are long-tenured editors with plenty of experience. (It is also worth noting that article titling is something where relying on raw consensus is less dangerous than, say, copyright cleanup... as long as a name isn't directly *wrong* somehow, there usually isn't a harm in letting a long-standing title continue, which is why normally moves expect something more than a 50/50 split. Which didn't even come close here anyway.) SnowFire (talk) 07:01, 27 June 2022 (UTC)[reply]
Overturn (involved). If we allow a closer to effectively ignore what most participants in the debate actually say in favour of their own interpretation of "rules" then we make a nonsense of the whole RM discussion process. Why do we bother discussing anything? -- Necrothesp (talk) 08:45, 27 June 2022 (UTC)[reply]
Endorse (uninvolved) if we don't want to throw WP:NOTAVOTE out of the window. There is no doubt that the current wording of MOS:CAPS and RM practice so far, with the provided body of evidence in form of ngrams and quotations, favor the move: for capitalization the MOS requires substantial majority of reliable soures, and ~50% does not qualify even nearly for "substantial". Trying to invoke MOS:TIES or MOS:RETAIN and plethora of other arguments is just throwing dust. The closing statement was clear and unambiguous: the proposed move is supported by policy. Arguments to overturn it amount to argumentum ad populum. No such user (talk) 12:58, 27 June 2022 (UTC)[reply]
Overturn (!voted oppose) per RegentsPark and in part SnowFire - and without a close analysis my feeling is that no consensus may be a more reasonable close. Doug Wellertalk15:39, 27 June 2022 (UTC)[reply]
Comment Everyone has cited Google ngrams including myself. I eventually went to the Google site to read what ngrams are about. Their main use, apparently, is in cultural and language studies, in studying changes in the use of expressions and collocations over time, Google Books being considered a type of language corpus. Example: Sometime during a substantial stay in England in the early 1990s, I thought to myself, "Hmm, people are using 'kidding' a lot more than they used to," and chalked it up to more people watching American TV shows which seemed to be highly popular. I can view that on Google ngrams in the evolution of "kidding" vs "joking" in both British and American books. It confirms my intuition that sometime in the 1980s, the more common British expression "joking" began to give way to "kidding." The same had happened in American books but earlier. Google doesn't care about reliability. Because for a language corpus to be reliable it is not necessary for the books in its database to be published by reliable publishers. In fact, the less the books are monitored for improper usage, the more reliable is the corpus. So, I am asking, Question: "Does Google ngrams include self-published books?" If it does, then how does their use square with MOS:CAPS: "substantial majority of independent reliable sources?" This can be easily seen when comparing "indus valley civilization" in books published by university presses, which is 1300, and those published by a handful of self-publishers, which is nearly 800. Culture studies folk obviously do want self-published books, Fify Shades of Grey being one. Please don't tell me, "This is not the place for this question." I know it is not, but please humor me this once. I have to start a Wiki break soon. Fowler&fowler«Talk»18:12, 27 June 2022 (UTC)[reply]
I cannot answer your question with any certainty. However, I did follow the links per Kj cheetham to publications here and here. In the latter, it describes that most books in the n-gram corpus were drawn from the libraries of 40 universities and that additional works were provided from publishers. Further, the corpus is a subset of Google Books - ie not all works from a Google Book search for a particular search string will be represented in n-gram corpus and the corresponding n-gram search for the same string. From the information I have, one can neither exclude nor conclude whether self-published works make up the n-gram corpus. Does this invalidate the n-gram evidence for the purpose of determining capitalisation to be used by WP? I would refer you to our discussion during the RM: this is not a case to be be settled by the rules of grammar, but by usage. and that a capitalised form (IVC) can be considered to be a proper name if it becomes institutionalised. Is this not then a study of language and precisely what the n-gram viewer was intended for? It provides access to a statistically significant random sample that allows us to address the question: is it capitalised in a substantial majority of sources? That is far better than the alternative of each side cherry-picking sources until one side runs out or gives up. Though MOS:CAPS would state independent, reliable sources, does the unsubstantiated possibility of the n-gram corpus including sources that are not WP:RSs preclude it from being used as evidence? Independent, reliable sources does not mean academic sources but includes more generalist sources - as much as you might dismiss the idea. And we should also consider WP:SSF. Even if it did rely on some non-WP:RSs it would still serve as good indicator of usage in WP:RSs. It would assume that the usage in non-WP:RSs is similar to that in WP:RSs and or that the proportion of non-WP:RSs is small and not sufficient to significantly affect the results. It is possible to test this assumption and I would hypothesise that non-WP:RSs are more likely to capitalise in much the same way and for much the same reasons as WP:SSF. Indeed, DL reports such a finding, albeit on a small sample. However, that is hypothetical, since I cannot confirm your concerns. Cinderella157 (talk) 05:42, 29 June 2022 (UTC)[reply]
Thanks for being honest. You do realise that a simple search for "indus valley civilisation" in Google Books throws books half of which do not meet WP:RS requirements. Children's books, the self-published, vanity-press published, back-alley published, are mixed with the reliable sources. Fowler&fowler«Talk»12:41, 29 June 2022 (UTC)[reply]
I read "half of which" as a figurative expression, not a quantitative claim. Nevertheless, when I do search for "indus valley civilisation", the first two books are "general knowledge" type of books ([3], [4], and only the third is by a known authority [5]. This result is unsettling to me. The first two may be RS for WP, but we do know that many in Google Books have been identified as not reliable for WP purposes. Here is the [6]] of one such discussion.
The bottom line seems to be that the use of Google Books and, by extension, Google Ngrams require discussion and consensus. For our case, we don't seem to have a consensus on the use of Ngrams data. Chaipau (talk) 10:35, 30 June 2022 (UTC)[reply]
The nature of biases in the n-gram stats has been discussed pretty extensively in this RM, and elsewhere. It's well established that they bias toward over-counting capitalized uses, for multiple reasons, including counting uses in titles, headings, citations, and other title-case contexts, and recently also by amateur and self-published materials that are more likely to have been influenced by over-capitalization in Wikipedia. Nobody has shown any mechanism for, or evidence of, bias in the other direction. Dicklyon (talk) 18:43, 4 July 2022 (UTC)[reply]
Despite the many places you have written many hundreds of words on the topic, there is no consensus that your interpretation of ngrams is correct and/or relevant in every case, specifically including this one. I also do not see consensus that Wikipedia is over-capitalised, let alone that this is some sort of problem we have exported to the world at large. Thryduulf (talk) 19:36, 4 July 2022 (UTC)[reply]
I'm not alleging that "Wikipedia is over-capitalised". Rather that when certain phrases in Wikipedia are over-capitalized in titles, that has an outsize influence on writers who use that phrase. Whether it's true or not, it's a good idea not to peg a theory of proper name status on a very recent blip up in capitalization, which is what the fans of capitalized Civilisation did in the RM and repeated here. Dicklyon (talk) 02:34, 5 July 2022 (UTC)[reply]
Endorse (involved, voted support). This is not really a close case when considering the evidence. The wording of MOS:CAPS is crystal clear that a substantial majority of sources must capitalise for us to follow suit. While the title case version may enjoy a narrow lead with some variants of the title (-ize- vs -ise-, valley vs not valley etc), it does not at all rise to the level of a substantial majority. There really is nothing more to this, no concrete evidence was presented to counter this either here or in the "overturn" !votes above, and the closer had no choice but to close the way they did. — Amakuru (talk) 19:21, 27 June 2022 (UTC)[reply]
Substantial majority was, in fact, shown (70.95% for IVC and 29.05% for IVc). This is the latest 2019 count and the trend was widening further. Chaipau (talk) 20:01, 27 June 2022 (UTC)[reply]
I suggest you apologize for calling someone a liar under WP:CIVIL, WP:AGF etc. This stat was picked with consideration not to conflate the issue with s/z spellings and not to conflate with smoothing artifacts. The result is very clear---from around 2015 there has been a strong trend towards IVC and away from IVc, and by 2019 it has become 2 to 1 head to head. There is no lie here. Chaipau (talk) 23:43, 28 June 2022 (UTC)[reply]
@Amakuru: I hadn’t thought about this myself but MOS:CAPS says, “a substantial majority of independent, reliable, sources." My understanding is that ngrams search all books in the Google database, including the scholarly (the most reliable) and the self-published (the least reliable). A case in point (see my comment and question above) is my being able to whip up 800 self-published sources using the expressions “IVC/c” from a small handful of vanity publishers but only 1300 from among universities' presses of which there are dozens in the US alone. A big question is: Was the closer aware that Google ngrams don’t gauge usage among reliable books, but indiscriminately among all books? If they weren’t aware then this fact constitutes the reason: "the closer was unaware of significant additional information not discussed in the page move discussion” (that grams search among an indiscriminate collection not among a reliable collection) and at the very least "the discussion should be reopened and relisted." A priori, Google has no automated way of determining reliability when experts have a hard time. On the other hand, if the closer was aware, how did they decide that the substantial majority of reliable sources do not favor capitalization? What other evidence was presented by Dicklyon that was convincing? Alternatively still, if the closer knows somehow that Google ngrams do search only among independent and reliable publishers, then please tell us how and why this is the case. Granted I should have asked that question at the top of the page, and a potential closer could disregard this objection for bureaucratic reasons, but the question is serious and it will remain. @Mellohi!, Dicklyon, SMcCandlish, DrVogel, and Cinderella157:Fowler&fowler«Talk» 20:14, 27 June 2022 (UTC) Update Fowler&fowler«Talk»20:24, 27 June 2022 (UTC)[reply]
Ngrams from Google is actually a serious issue. Not just do we have to consider the source of the corpus created by Google, as you F&f have pointed out, there are many other problems. One of the disagreement has been on the use of smoothing. Google seems to be using a simple moving window to smooth it. This is not the best smoothing method one could use because smoothing artifacts show up. For example, the peak in 1900 show up as an asymmetrical plateau after smoothing by window size 10---and even though the red and blue lines are just peaks of different magnitudes, they show up with different features after the smoothing because of distant noise. Some of the inferences that we have been making after smoothing will sound very different if we were to use other smoothing techniques such as wavelet transform, Savitzky-Golay filter or Loess. I submit that none of the inferences, based on uncritical application of the smoothing provided with Ngrams, is believable. There cannot be any reasonable consensus on their use. Chaipau (talk) 22:39, 27 June 2022 (UTC)[reply]
The Ngrams use has both the RS issue and the smoothing issue. These issues are not mutually exclusive. One of them, unfortunately, is a little technical. A close based on Ngrams alone is flawed. Chaipau (talk) 00:25, 28 June 2022 (UTC)[reply]
In other words, my question applies to ngrams without any smoothing. It applies to the raw data collected by Google.
I am suggesting that even there Google might be unable to search in the subset of independent reliable books, because it very likely has no algorithm for determining reliability, and no motivation for finding one, as its primary use is in a large heterogeneous set. In my "kidding" vs. "joking" example, the bigger the data set: street conversations, fiction, fantasy, film, non-fiction, self-published works, ... the better it is for tracking evolving use of those two words in the language. Fowler&fowler«Talk»02:30, 28 June 2022 (UTC)[reply]
Google Ngrams is based on corpora created periodically from the entire scanned Google Books. We have been using the version released in 2019 for this RM. Are there self-published books in Google Books? Are there books in Google Books that are not RS for WP? Yes, there are and you know it. Google Books is the bane of WP because people search for their favorite quotes and cite them here.
The point in this thread is that even if Ngrams were to be acceptable (should not be), the claim of substantive majority fails on statistical grounds. Chaipau (talk) 04:29, 28 June 2022 (UTC)[reply]
@Fowler&fowler: if what I stated (and added to your argument) is not your point what is? I have stated in good faith what looks important to me exactly as I see them. I am OK with your stating your opinion too, though I am a little taken aback by your tone. Chaipau (talk) 06:24, 28 June 2022 (UTC)[reply]
I apologize for the intemperate outburst, but would like to state in earnest that you not interleave your arguments with mine no matter how relevant, supportive, pressing, or complementary they appear. I fear that by entangling the two issues, you have impeded communication, which is fraught as it is. When the arguments are stated independently, they are more easily understood. Fowler&fowler«Talk»09:58, 28 June 2022 (UTC)[reply]
MR is WP:NOTAVOTE At Wikipedia:Move review#Closing reviews: an uninvolved editor will determine whether a consensus exists. Assessing the outcome of this discussion is no different from assessing the consensus of any WP discussion in that (unsubstaniated) opinions carry little weight compared with a strong argument that is evidenced and reasoned on the basis of WP:P&G. I think it ironic to have to point this out in this particular case that most of the comments made essentially fall to WP:ILIKEIT and WP:IDONTLIKEIT style !votes. Cinderella157 (talk) 02:51, 29 June 2022 (UTC)[reply]
But you haven’t answered my question whether you knew at the time of the RfC or during this move review discussion that Google ngrams are not based on sampling either independent or reliable books, the backbone of MOS:CAPS? Fowler&fowler«Talk»04:09, 29 June 2022 (UTC)[reply]
I do not know if the n-gram corpus includes sources that are not considered to be WP:RSs. I am sorry that my priorities may not coincide with yours but you may have guessed I was in the process of providing you with an answer when you made this post. Perhaps you might extend to me the same courtisy and eventually answer my question here. Cinderella157 (talk) 06:17, 29 June 2022 (UTC)[reply]
Frankly, this whole "but ngrams may also contain unreliable sources!" meme is yet another example of throwing a lot of shit in the hope some will stick, so common tactics in this whole sorry episode. We don't really know the ratio of "reliable" and "unreliable" sources in the Google's corpus, and it does not matter at all. What we do know is that the corpus is so large that it covers a substantial majority of reliable sources on whatever subject is being discussed, and thus it has been used in article title discussions since forever, so your raising of this issue is yet another attempt at FUD. Furthermore. we don't have any reason to suspect that "unreliable" sources in the corpus would favor IVC or IVc, so any bias towards any of the versions is just a wild conjecture of yours. No such user (talk) 07:02, 29 June 2022 (UTC)[reply]
Google ngrams are used in culture- and language studies to track usage; they are an aspect of Corpus linguistics
"What we do know is that the corpus is so large that it covers a substantial majority of reliable sources on whatever subject is being discussed, ..."
But then it must cover the substantial majority of unreliable sources as well and they could be larger.
"it has been used in article title discussions since forever"
Overturn While I may have ultimately !voted support had I participated, this appears to be a pretty clear WP:SUPERVOTE of a closing that uses its own interpretation of the guidelines to determine a consensus that very clearly does not exist in the discussion.--Yaksar(let's chat)06:39, 29 June 2022 (UTC)[reply]
Overturn (reclose) was Endorse (uninvolved) The MOS:CAPS arguments are very strong, the regardless of if all the matching items found by ngrams can be used as source, ngrams is usable to show common capital usage is not consistent. I also question the use of 10-year smoothing as counter argument to the close. Looking at this chart without smoothing, the various cycles of ups and downs seem to be about 10 years apart, which means using 10 year smoothing can amplify noise such as the local min at around 2011. That said I do have one concern with the close. While the arguments on if IVC is a proper name seem to be no-consensus, if there was consensus that it was a proper name, that would be an exception to MOS:CAPS. This relisting comment would have been better if it had been clearer on the proper name part of the discussion in reframing the remaining issues for the close, see WP:RMRELIST. PaleAqua (talk) 17:25, 29 June 2022 (UTC)[reply]
As for names, RegentsPark has said it all. If Indus Valley were a pre-modifier for the noun phrase head "civilisation," the urban culture the phrase denotes would not have included sites in the valleys of other major rivers, such as Badakhshan in the valley of the Amu Darya which empties into the Aral Sea, Sutkagan Dor in the valley of Dasht River emptying into the Persian Gulf, Mohenjo-daro on the right bank of the Indus, which empties into the Arabian Sea and Kalibangan near the Ganges, which empties into the Bay of Bengal. It is a proper name.
As for ngrams, there is no proof that they sample trends in the "independent, reliable" sources of the MOS:CAPs exhortation. The Google Books corpus is one of the many of Corpus linguistics. A simple search for "indus valley civilisation" there shows that it has children's books, ruminations by civil servants, by retirees, by nationalists, by sub-nationalists, by scholars, by bloggers, by people searching for aliens, ... the works. Fowler&fowler«Talk»18:30, 29 June 2022 (UTC)[reply]
Fowler&fowler, what RegentsPark actually said at the RM was: Civilization is not a modifier of Indus Valley but is a part of the name. This is true and you and I engaged in a discussion on the semantics of grammar in respect to this. It is not considered civil to misrepresent what has been said by another editor. When they more fully stated: (the actual civilization extends well beyond the valley), you are now asserting that they were stating that there is a semantic difference between IVC and IVc. This was certainly not made clear and definitely not evidenced (ie it is an opinion). Both you and they had the opportunity to substantiate this when we discussed the matter in detail. It is not the purpose of a MR to rehash the RM and give either side an opportunity to make a better case and have a second bite of the cherry. There was plenty of time to get it right the first time and it was you that kept pressuring for a close. However, the line you would now pursue is easily refuted. Ab initio both terms describe a civilisation in (or about) the Indus valley. It is only with further information that the extent of the civilisation is fully known. There is nothing to suggest that this further information applies disproportionately to the terms. It is a philosophical agrument where the premises are not established in a case where linguistics is argued and acknowledged as the prevailing framework (ie that the capitalised form is institutionalised). Cinderella157 (talk) 11:01, 30 June 2022 (UTC)[reply]
The Indus Valley Civilisation is the mature urban part of the "Indus Valley Tradition," which covers the entire sequence from Neolithic, such as Mehrgarh, to Chalcolithic, such as Nausharo, to urban such as Mohenjo-daro. Here is the Google ngram for "Indus Valley Tradition" and "Indus Valley tradition" It has no smoothing. Would you say the substantial majority of books on Google (forget the independent and reliable bit) captalize? If not, why not? Fowler&fowler«Talk»18:57, 29 June 2022 (UTC)[reply]
Would you say that is the substantial majority of sources capitalize? If not, why not? What constitutes substantial? I could find the area under the two curves?
If it is does constitute substantial, then how is it that the larger "Tradition" lasting from 7000BCE to 1300 BCE is capitalized, but the smaller urban in it (IVc) lasting from 2600 BCE to 1900 BCE is not? Fowler&fowler«Talk»19:24, 29 June 2022 (UTC)[reply]
Update: The more I think about the relisting comment, the less I think that I can endorse. The ngrams might be strong, but common usage is only a part of MOS::CAPS, and a better discussion of some of the exceptions might have been useful. In particular there is contention on if the ngram sources might meet independent, reliable sources required by MOS::CAPS. (see some of the replies to my original !vote for examples.) I am not convinced that considering the sourceness is existing practice despite the wording of MOS::CAPS, but it weighs consideration. I'm not specifically !voting for moved, not-moved or no consensus, but ask the closer to weigh the exceptions to MOS::CAPS and discussion of the ngrams both here and in the RM if reclosing. PaleAqua (talk) 17:38, 3 July 2022 (UTC)[reply]
Maybe where it says "where the term is consistently capitalized in sources, or where a whole bunch of editors like caps better for a term that's special to them"? Dicklyon (talk) 18:35, 4 July 2022 (UTC)[reply]
Endorse <uninvolved> per Amakuru and others above. Closure was reasonable, and that's the only reason we're here. So please don't expect me to engage in discussion that re-argues the move request. It's done, and there's an end to it. P.I. Ellsworth , ed.put'r there21:15, 29 June 2022 (UTC)[reply]
Final comment by Fowler&fowler Tomorrow is my last day on Wikipedia for several months.
What have I come away with?
The negatives first: I don't expect any great changes. I don't think they don't know much about ngrams at MOS though they've been using them on faith as a result of watching others whose collegiality they value use them on faith. For no one has been able to answer my simple question: whether they knew ngrams do not sample the "independent, reliable" sources, or that one can easily whip up as many self-published sources as university-presses-published? Apparently, I am the first to raise it. There are also people here the shifting sands under whose feet begin to give way at the slightest intellectual threat. They then brandish glorified Aesop's fables in the form of Wikipedia guidelines.
The positives: I verified my intuition gleaned in Cambridge, England, in the early 1990s about the evolution of "kidding" vs "joking". I verified the widespread capitalization of "Indus Valley Tradition." I had been asked by user:Doug Weller some ten years ago if there was a need for an article on it, and I said no, to my great regret today, for IVC/c (fl. 2600 BCE–1900 BCE) is a subset of the Indus Valley Tradition (ca 7000 BCE–1300 BCE). The capitalization in the latter reflects reliable sources, as no one but academics uses that expression. I even calculated the areas under the two curves Indus Valley Tradition and Indus Valley tradition for my own understanding. I learned something from user:RegentsPark's smart remark that "civilisation" is not the head of a noun phrase pre-modified by "Indus Valley," for there are many IVC sites in other major river valleys, such as Amu Darya or Ganges unconnected to the Indus, and emptying into different bodies of water.
Whether or not I will come back to edit the Indus Valley civilisation article remains to be seen. Very likely other content creators who didn't see the point of this silly sniping excursion will have similar ambivalence. So what then has been the net gain? Fowler&fowler«Talk»23:04, 29 June 2022 (UTC)[reply]
I love Wikipedia. On what seems to be the least contentious discussion possible, a C vs c battle got us a discussion apparently exceeding 35,000 words and 280 kilobytes of page content (so far that is). What an incredible waste of time and effort on everyone's part. And in some cases, (gross) incivility — an editor calling someone "liar" and someone constantly pinging the relister, attempting to sway them isn't something I hoped to see in those huge threads. Anyway, back to the point. I'd say Overturn [as uninvolved]: To me the close looks more a WP:SUPERVOTE. Ofcourse WP:Wikipedia is not a democracy, and closes are made by weight of arguments rather than number of !votes. But, here I cannot see the "pro-lowercase" side have strong enough arguments to override the perfectly valid opinions of 70% commentators. The number of Cs, as of today, does outweigh the number of cs in ngrams. And Wikipedia lives in the present, not in the past and not in the future. Grammar issues were raised by both sides, but I'm no language specialist to judge whose argument takes the pie. But, what I know is, with a capital C that shows valid and acceptable usage globally, I see no genuine *harm* being caused by the uppercase title. WP:If it ain't broke, don't fix it, especially when there *isn't* any real consensus in favor of change. Take with it WP:TITLECHANGES. If a discussion doesn't get consensus for a change and when something is *this* controversial so as to log 35k words of debate, it is good to leave it as it has been for 21 years! —CX Zoom[he/him](let's talk • {C•X})15:43, 30 June 2022 (UTC)[reply]
Last_Comment, MOS:CAPS says Wikipedia relies on sources to determine what is conventionally capitalized; only words and phrases that are consistently capitalized in a substantial majority of independent, reliable sources are capitalized in Wikipedia. There are exceptions for specific cases discussed below. Then it started to give exceptions. The most important exception are WP:AT and proper names. To my understanding, "consistently capitalized in a substantial majority" isn't applicable to WP:AT names like IVC which uniquely identifies one and only one civilization. Being a name (proper name), the IVC qualifies to be capitalized for article title as done by academics. Northeast heritage (talk) 16:42, 30 June 2022 (UTC)[reply]
Overturn (opposed in the RM) per Kautila, RegentsPark, F&F and others. I think the major sources on the subject do capitalize, which ngrams don't catch - there is a vast literature. Close looks like a supervote. It's clear that all the regular contributors opposed, & the supports had never edited the article previously. Johnbod (talk) 21:05, 2 July 2022 (UTC)[reply]
This kind of "I think" opinion was presented in the RM, and is hardly relevant to the MR. F&f's "I think" first referred to Dyson as one that capitalized; I showed that Dyson does not. If "all the regular contributors opposed", it's a classic "specialist" style preference of people who like to cap what's important to them. No need to repeat that here. Dicklyon (talk) 04:32, 3 July 2022 (UTC)[reply]
Overturn (uninvolved). This appears to be a supervote as there was extensive discussion in the article, and ngrams were just one aspect of that and there was very clearly no consensus among participants about their relevance or validity. The other aspects were not addressed by the closer at all, despite forming the basis of many arguments. Either "not moved" or "no consensus to move" would be valid readings of that discussion, but there was very definitely no consensus to move. Thryduulf (talk) 10:25, 3 July 2022 (UTC)[reply]
Endorse (uninvolved). I agree with the above editors who endorsed, especially those who pointed to WP:NOTAVOTE. I would add: The only argument that has been raised for reopening the RM is F&F’s point about ngrams including non-RS. Even if that’s the case, however, it does not justify overturning the RM. MOS:CAPS makes clear that the burden of demonstrating an exception lies on those seeking it; failing a showing that a substantial majority of RS capitalize the term, the default is to lowercase it. F&F has not made such a showing; at best, he has merely suggested that we don’t know what a substantial majority of sources do. Clearly many editors disagree that F&F has achieved even that much, but the point is that they have not made the showing required of them, so there is no reason to reopen the discussion.
One additional point:
Whether or not I will come back to edit the Indus Valley civilisation article remains to be seen. Very likely other content creators who didn't see the point of this silly sniping excursion will have similar ambivalence. So what then has been the net gain?
This is textbook WP:HIGHMAINT behavior and should be strongly discouraged. Ultimatums like that are inappropriate no matter the issue, and if contributors at Indus Valley civilisation wish to storm off the project because they don’t get to capitalize a term that is important to them, they are welcome to go. Wallnot (talk) 20:40, 4 July 2022 (UTC)[reply]
Comment: I've been emailed twice about this misinterpretation—that in a huff, I had somehow made a threat of sorts and went off on a vacation on false pretenses.
Well, my Wikibreak is about bigger issues related to my family as my talk page post makes clear, and other posts had made clear a week earlier, the day of the US Supreme Court announcement.
After making my Wikibreak announcement, I have not found it easy to stay away from Wikipedia. POV-warriors have emerged 13 to a dozen to edit the pages whose neutrality I have worked hard to ensure, sending me stumbling into responding, and then regretting my actions later.
But there is one thing I haven't done. And that is to think about this article.
Am I human? Yes. Can't I force myself to work equably on something that holds no attraction for me, regarding it as a duty? I can't.
Overturn to no consensus defaulting to "not moved" <uninvolved>. The closure is mostly correct, and many of the overturn !votes above aren't terribly persuasive. In particular, the closer rightly zeroed in on the MOS:CAPS issue and discounted !votes that weren't grounded in policies and guidelines. I also agree that there is no consensus on whether a substantial majority of sources that use a capital C can be found, but I part ways with the closer over what that lack of consensus means. Ordinarily, "In article title discussions, in the event of a lack of consensus the applicable policy preserves the most recent prior stable title". The closer seemed to think that MOS:CAPS inverted this rule, defaulting toward no capitalization when there's no consensus. But MOS:CAPS simply doesn't say that. Unlike guidelines that do deviate from the usual rule (WP:BURDEN, WP:ELBURDEN, WP:BLPUNDEL, etc.), it doesn't mention the "burden of proof" at all. Since MOS:CAPS doesn't say what to do when there's no consensus on whether the substantial-majority threshold is met, the ordinary rule – default to the most recent prior stable title – applies. And that means that I agree with Adumbrativus that since there was no consensus on the MOS:CAPS issue, the article shouldn't have been moved. Extraordinary Writ (talk) 06:46, 7 July 2022 (UTC)[reply]
MOS:CAPS opens: Wikipedia avoids unnecessary capitalization. This reasonably creates a burden that caps must be necessary. It continues: Wikipedia relies on sources to determine what is conventionally capitalized. This then becomes a matter of verifiability (a core policy) and a need to show that a term is conventionally capitalised. WP:VER includes WP:BURDEN and WP:ONUS. The burden and onus is then to show that a term is conventionally capitalised. MOS:CAPS then continues to explain how this can be shown. So, when the closer woud state (and you would affirm): there is no consensus on whether a substantial majority of sources that use a capital C can be found, there is no consensus that the capitalisation is necessary and that the burden to show it is has not been discharged. Regards, Cinderella157 (talk) 11:54, 7 July 2022 (UTC)[reply]
MOS:CAPS indeed opens: Wikipedia avoids unnecessary capitalization. But this does not create a burden, just a policy. A policy cannot be made to bear a burden. The burden is created by WP:VER which demands that any claim that is made is verifiable. Only a claimant can be made to bear a burden. (e.g. the prosecution bears the burden of proving that the defendant is guilty beyond all reasonable doubt) Thus the question for the closer was whether there exists a clear consensus (beyond reasonable doubt) on the claim by the requester that the capitalization is un-necessary. The closer failed here, in part. Chaipau (talk) 20:19, 7 July 2022 (UTC)[reply]
I used an analogy to demonstrate who bears the burden. But WP:BURDEN makes this even more clear: The burden to demonstrate verifiability lies with the editor who adds or restores material (bolded, as in original). In other words, the burden lies with the editor making the change, and in our case, the editor(s) seeking change. Chaipau (talk) 21:42, 7 July 2022 (UTC)[reply]
But that’s the point. The “change” is modifying the MOS rule, which in this case means carving out an exception to MOS:CAPS. Thats not the same as addition or removal of content. As we have shown, MOS makes clear that the burden lies on the editors arguing for the exception, because the default is to lowercase it. No one has persuasively argued that Fowler et al. actually showed that a substantial majority of reliable sources consistently uppercase IVc.
Under your reading of BURDEN, an editor would have to seek consensus every time they applied a straightforward MOS rule. That defeats the purpose of having guidelines at all.
Wallnot (talk) 21:58, 7 July 2022 (UTC)[reply]
No, the change requested is IVC -> IVc. WP:VER is not required to make changes to policies. Looking for exceptions in policies is not making changes to policies. I was replying to Cinderella157 on the claim that the policy creates a burden. A policy cannot create a burden. Chaipau (talk) 22:07, 7 July 2022 (UTC)[reply]
Sorry for the sloppy language. I should have said that MOS:CAPS does not create a burden. WP:VER does. And it places it on editors, not policies. Chaipau (talk) 22:16, 7 July 2022 (UTC)[reply]
The change requested is dictated by a long-standing guideline. Those seeking an exception to a straightforward application of a long-standing guideline bear the burden of demonstrating that the exception is warranted; when the guideline itself provides that exceptions will be made only in limited circumstances, those seeking the exception need to show that those circumstances exist.
Since CAPS exceptions require a specific set of circumstances, and no one has showed that those circumstances exist, it follows that no exception can be made.
Also, as you yourself point out, there’s nothing in BURDEN to suggest that it applies to changes to guidelines. By its terms it’s a rule only for content. Wallnot (talk) 23:17, 7 July 2022 (UTC)[reply]
Please go over the language of the WP:BURDEN. I don't want to repeat myself. I can only point out that you are applying it very creatively.
Second, the burden is on the change agents to seek consensus on the sections of MOS that are applicable and the exceptions that aren't, as per VER. There has been a robust discussion on many of these, and if you went through the discussion you'd notice that there has been no consensus on any of these claims. The closer just cast a supervote. Chaipau (talk) 23:28, 7 July 2022 (UTC)[reply]
Since you are claiming "long-standing guideline", I was curious when the name IVC appeared on Wikipedia and what the state of MoS was on that day. The day was 15 November 2001. The IVC page was created with a capital "C", but the MoS did not even have a CAPS section. IVC is longer-standing than MOS:CAPS! The day (24 Feb 2006) the CAPS subpage was created, the page did not even have the phrase "unnecessary capitalization" in it. Chaipau (talk) 00:09, 8 July 2022 (UTC)[reply]
Great, now that the guideline has been changed, it sounds like it’s time to bring the page title in line with it.
To your other point, I have read the guideline. By its terms, it applies to:
all quotations,
all material whose verifiability has been challenged,
all material that is likely to be challenged, and
all contentious matter about living and recently deceased persons.
The page title is not any of those. This is not a debate about content but about the application of an MOS guideline. Nowhere in BURDEN or WP:V more broadly does it say it applies to the interpretation of guidelines.
I also disagree with your claim that there is a lack of consensus as to which MOS rules apply. The starting point of the debate is that MOS:CAPS applies (because it plainly does). The question is whether this falls under one of the exceptions. Right now, no one has showed that a substantial majority of reliable secondary sources capitalizes IVc. So the answer to that question is no. Wallnot (talk) 01:15, 8 July 2022 (UTC)[reply]
The way you are interpreting rhe rule flips it on its head. If you require editors who want to apply MOS:CAPS to first show that a substantial majority of reliable sources lowercase it, youre rewriting the rule. Thats not what it says, so thats not what we should do. Wallnot (talk) 01:17, 8 July 2022 (UTC)[reply]
That’s a non sequitur since where you place the burden is outcome determinative for purposes of this MR and important to the RM as well. Wallnot (talk) 00:43, 9 July 2022 (UTC)[reply]
The evidence shows otherwise. The process has been under WP:RM and this current discussion under WP:RMCI. The WP:RM process does not admit an automatic and default position for MOS:CAPS. Note:
On June 8 a RM/TR request was made to move IVC to IVc with the claim it was uncontroversial by the eventual requestor. And it was moved.
Within the hour the move was challenged and it was moved back to IVC.
Had MOS:CAPS enjoyed an elevated position there would have been no revert to IVC and the burden would have fallen on the IVc -> IVC position to seek consensus for the move. On the other hand, the situation was for the IVC -> IVc to seek consensus.
Furthermore, WP:THREEOUTCOMES suggests that the page could have been moved only when consensus is found to rename the page. When there was no consensusOf course, as elsewhere on Wikipedia, this usually means that no action is to be taken at the present time. In other words, only when the movers are able to show consensus in there a move. In other words, the burden was on the movers, not on the others. And this is arule that applies across Wikipedia as pointed out. Chaipau (talk) 13:04, 12 July 2022 (UTC)[reply]
I think we’ve identified the source of the disagreement then. The burden in the RM is indeed on the editors seeking the move. But in the RM itself, the outcome—consensus for move, consensus for no move, or no consensus—depends on MOS:CAPS. Since this is WP:NOTAVOTE, and no one has made a showing that a substantial majority of reliable sources capitalize the term, the consensus is for the move. If you can’t present any evidence justifying why we should ignore MOS:CAPS, then it applies, and there is consensus for the move, no matter how many editors come out of the woodwork to say we should capitalize it because it’s important.
But there’s no need to relitigate the RM, because the only question in this MR is whether something new has come to light that would call into question the outcome of the RM. No one has presented any new evidence indicating that a substantial majority of reliable sources capitalize the term, so I don’t see how that could be the case. Wallnot (talk) 14:54, 12 July 2022 (UTC)[reply]
Yes. Since the closer placed the burden on those who opposed the RM instead, the closing is based on considerations outside the THREEOUTCOMES and is thus flawed.
And No. MOS:CAPS enjoys no special role in this RM, and it deserves only the weight due to it. RMNOMIN suggests "unless there is a very good reason to ignore rules". Though stated in a different context, this is in the spirit of the fundamental WP:5P5. MOS:CAPS is subject to consensus and like all other rules can be ignored.
I can only observe that this is a novel spin on WP:RMNOMIN. It would state: Remember, the participants in any given discussion represent only a tiny fraction of the Wikipedia community whose consensus is reflected in the policy, guidelines and conventions to which all titles are to adhere.MOS:CAPSis one of these guidelines - through WP:AT and WP:NCCAPS. There is no reasonable case made to argue otherwise, either here or at the RM. MOS:CAPSdoes[enjoy] a special role in this RM. It deservese the weight due it. WP:RMNOMIN continues: Thus, closers are expected to be familiar with such matters, so that they have the ability to make these assessments. By this, the closer is required to give weight due the broader community consensus. MOS:CAPS would require caps are necessarily and conventionally used. If there is no consensus that a term meets these requirements, the MOS:CAPS defaults to lowercase. The closer has closed accordingly. You would miss that point. MOS:CAPSis subject to the broader community consensus and can be changed. But, until it is, it reflects the broader community consensus. It is weighed accordingly. Yes, like all rules, it can can be ignored. But unless there is a very good reason to ignore rules [emphasis added]. But it comes with a caveat that has not been established. This tenet of your argument acknowledges the rule but, without good reason, is self-defeating. In my view, this whole argument is a dead parrot horse. It isn't going to voom no matter how hard you flog it (not even with 40,000 volts). It isn't going to fly even if you pour 50 gallons (neither US nor imperial) of Red Bull down its throat. Some humour/banter Cinderella157 (talk) 11:01, 13 July 2022 (UTC)[reply]
Thanks for the banter. I don't need long winded convoluted arguments here.
I have specifically said that WP:RMNNOMIN is not the context in this case. I have just used it to illustrate the applicability of WP:5P5 universally in Wikipedia and that MOS:CAPS is no exception. WP:5P5 is no dead horse.
No amount of distractions will get you past the fact that the close made mincemeat of WP:THREEOUTCOMES.
No distractions here, and No one has made mincemeat of THREEOUTCOMES. As Cinderella points out, there’s an existing consensus behind CAPS. The RM failed to establish a consensus that the exception criteria to CAPS had been met, so the consensus is to apply the existing guideline. You haven’t presented any "very good reasons" to IAR, and, anyways, that was a discussion for the RM, which is over now. Wallnot (talk) 19:49, 13 July 2022 (UTC)[reply]
That’s not canvassing, it’s textbook WP:APPNOTE, which allows for posting at The talk page or noticeboard of one or more WikiProjects or other Wikipedia collaborations which may have interest in the topic under discussion.Wallnot (talk) 12:36, 8 July 2022 (UTC)[reply]
Note As the initiator of the WP:MVR of 25 June 2022, I'd like to sum up the voting statistics 18 days later:
Thus far 27 people have voted, 17 to overturn and of these seven were uninvolved. In opposition, 10 people have endorsed and four were uninvolved. In RM begun by Dicklyon on 8 June 2022 that led to this MRV, 16 people had opposed the move, seven had supported it. The page was moved by mellohi on 23 June 2022.
My best argument remains that at the time of the closing the closer of the RM was not aware that Google ngrams do not reflect the "independent, reliable sources," whose support, or lack thereof, the injunctions of MOS:CAPS abstract. Indeed, until I made the point in a conversation above on June 27, 2022, no one on MOS had considered this idea. Despite my asking the closer several times they have not responded. In the rare instances in which ngrams do reflect academic usage, as in the "Indus Valley Tradition," a term invented in the mid-1980s, and employed for the culture lasting from 6500 BCE to 1300 BCE of which the Indus Valley Civilisation (lasting from 2600 BCE to 1900 BCE) is the mature urban part, it is clear what they say. I have lost interest in the article. In my view, such unannounced sniping excursions by editors who are unfamiliar with an article's content do a great disservice to Wikipedia. They have a chilling effect on the creation of new content. Someone should close this MRV whichever way they want. No one has voted in a week and the damage to the article has already been done. Fowler&fowler«Talk» 21:22, 13 July 2022 (UTC) Updated. Fowler&fowler«Talk»21:34, 13 July 2022 (UTC)[reply]
I would like to second the sentiment expressed regarding the "chilling effect" this entire exercise has had on content creators. This is an honest confession, and Wikipedia needs less of this chilling effect. Chaipau (talk) 02:06, 14 July 2022 (UTC)[reply]
Since bolded bullet points seem to the order of the day, might I also respond to this with a reminder that discussions on Wikipedia are WP:NOTAVOTE? The closer will assess the arguments made, not the numbers of those responding, and as far as I can tell the central tenet of the close that the MOS:CAPS "significant majority" bar for capitalising has not been reached, remains the only legitimate conclusion here. Perhaps the trend will change in the future and we can revisit, but for now the close was correct and should (as I said above) be endorsed. — Amakuru (talk) 11:35, 18 July 2022 (UTC)[reply]
Really Amakuru you are telling a 16-year veteran about Wikipedia not being a democracy? Pretty shameful. I want this nonsense to end. A laughing stock it is. Fowler&fowler«Talk»11:41, 18 July 2022 (UTC)[reply]
Again Chaipau and Wallnot. May I request both to buzz off? Enough of this haranguing. Chaipau I don't need your support; Wallnot I don't need your dickering. The ball game is over. Neither of you has contributed any content to the IVC article. Let someone decide whichever way they want. Fowler&fowler«Talk»08:24, 14 July 2022 (UTC)[reply]
FWIW, I am not trying to help/support you, but echoing another editor's general comment to you on their talk page. I am also personally interested in the general course of this case. I have contributed to other Indus related articles marginally and I hope to contribute here in some form in the future. Chaipau (talk) 23:37, 14 July 2022 (UTC)[reply]
Overturn to no consensus defaulting to "not moved". Fundamentally this boils down to a group of copy editors relying on Ngram, which they see as providing definitive proof that trumps the expertise of subject matter experts. Ngram are fundamentally unreliable as a means of establishing capitalisation. A) they're vulnerable to manipulation [7],[8] (I can show others which are even more dramatic.) and B) they are vulnerable to contamination by false positives. I would go so far as to say that they should never be used in this regard. Ngram are no substitute for reviewing the literature and establishing the prevailing view. IMHO an absurdly high standard is being demanded by the proposers of these moves, they're acting as a lobby and trampling over topics where the number of editors with sufficient expertise is small, ignoring SME in their crusade. Copy editing should be about assisting content creators, it has become a monster stifling content creation and become an end in itself. The original close is fundamentally flawed, it doesn't represent the strength of argument and relied too heavily on the comments of editors basing their comment on a fundamentally flawed evidence base. I also note that the proposer of many of these moves was indeffed for disruptive page moves, perhaps it is time to re-examine that decision. I would add I think the editors concerned are well meaning but they are misguided in what they're doing and they have become disruptive. I would urge them to rethink before this ends up at arbcom or ANI and I do think the level of disruption over what should be a trivial matter is getting to the point of warranting such intervention. WCMemail13:04, 20 July 2022 (UTC)[reply]
Overturn. The closer was not an administrator and this was a close call and contentious, therefore it is WP:BADNAC/2. It is also a supervote, the closer did not assess consensus of the discussion but instead went forward with their idiosyncratic viewpoint.PrisonerB (talk) 08:01, 21 July 2022 (UTC)[reply]
Endorse uninvolved (having not read any other MRV !votes yet). With apologies to the sentiments given by User:Fowler&fowler, particularly the perceived rudeness of a non-editor of the article coming in the change the most important content, the title, it is long standard practice and consensus, having withstood many challenges, that capitalisation of terms that could be proper names is only done where virtually all reliable sources capitalise the term. If there is at least a strong minority of reliable sources that do not capitalise the term, then Wikipedia falls on the side of not capitalising. As per Dicklyon and his evidence, a strong minority of reliable sources do not capitalise civilisation in “Indus Valley civilisation”. Virtually all reliable sources (though not all) do capitalise Valley. The closer succinctly and correctly closed the discussion. —SmokeyJoe (talk) 05:41, 6 August 2022 (UTC)[reply]
The process is not a vote and does not exist for relitigating an RM. It is to determine if the closer erred, and the closer did not err. — SMcCandlish☏¢ 😼 06:33, 11 August 2022 (UTC); edit: I removed "Endorse" from this; I forgot I'd already responded to this MRV shortly after it opened. — SMcCandlish☏¢ 😼 14:29, 11 August 2022 (UTC)[reply]
Closer beware This editor has already voted and was replied to. I have half a mind to change the "Endorse" to "Comment,' but am leaving it in here to highlight the many-splendored manifestations of the need to be conspicuous. But it works only when the rest of us don't also break talk page guidelines
The logic of the double vote however can't be doubled. The closer was not aware that Google n-grams do not quantify the "independent, reliable sources" of the MOS:CAPS injunction.
So one month after the last vote count by yours truly, I have to report that: Thus far 31 people have voted, 19 to overturn and of these nine were uninvolved. In opposition, 12 people have endorsed and six were uninvolved.
Um, no. Closers do not have to be admins except under specific circumstances (that don't apply here), and discussions of this sort must eventually close, whether they are contentious or not (if they weren't contentious, they wouldn't need discussion in the first place; we have a manual move process for non-controversial page moves). — SMcCandlish☏¢ 😼 14:29, 11 August 2022 (UTC)[reply]
The above is an archive of the move review of the page listed in the close of this review. Please do not modify it.
No English-language reliable source using the new article name has been presented, and all English-language sources in the article use the name "Greg Hsu". Page was moved with one support, one opposition and one critical comment. Can't see any indication of consensus here. —Kusma (talk) 12:55, 21 June 2022 (UTC)[reply]
Endorse as closer. I've never heard of this guy, no idea who he is (before reading the article), but I saw a move request with a lot of back-and-forth about what sources say his English-language name is. Ultimately, nobody actually wrote oppose , and the proposer's arguments seemed logical enough, and it's part of our rules at WP:RM that an unopposed move should go forward. I think relisting is a reasonable outcome, but this move was ooooooooooooooooooooooooooooooooooooooooooooooold when I came across it in WP:RMB. A request that is open for an entire month without anyone writing "oppose" - I mean, disagree with me if you want, but I hope you see why I moved it. RedSlash20:14, 21 June 2022 (UTC)[reply]
So your excuse for moving the article despite 100% of all known secondary sources supporting the other name is that I did not bold my oppose? —Kusma (talk) 20:26, 21 June 2022 (UTC)[reply]
Overturn (not moved) Arguments for keeping the article at "Greg Hsu" are stronger. Comments don't have to say "Support" or "Oppose", closes should be based on the arguments not the count of the bolded words. PaleAqua (talk) 20:51, 21 June 2022 (UTC)[reply]
Overturn to no consensus. WP:RMNOMIN applies only when "no one has objected". Kusma raised objections to the move (specifically that it wasn't supported by sources), and the closer should have considered those objections regardless of whether they were accompanied by a bolded !vote. Extraordinary Writ (talk) 00:42, 26 June 2022 (UTC)[reply]
Overturn (uninvolved). This shouldn't have been moved as there was no consensus. As others have said, the closure should be on the content as it's not a vote. -Kj cheetham (talk) 11:42, 26 June 2022 (UTC)[reply]
The above is an archive of the move review of the page listed in the heading. Please do not modify it.
