Template talk:Infobox book

New field request: word count

As fields go, "Pages" is not very useful to readers. It fails to provide any real point of comparison because pages are not of standardised length across the literary industry. Word count, meanwhile, is precise. This is a commonly requested addition for this template. As an editor working actively in this space, I would like to present some analysis on why pages are not especially useful to readers:

• A Game of Thrones (1ed) has 694 pages and 298,000 words; 429 words per page.
• Dracula (1ed) has 418 pages and 161,000 words; avg. 385 words per page.
• Gone Girl (1ed) has 432 pages and 148,000 words; 342 words per page.
• Frankenstein (1ed) has 280 pages and 78,000 words; avg. 273 words per page.
• It Ends With Us (1ed) has 376 pages and 105,000 words; avg. 280 words per page.

The variation is significant: It Ends With Us has 54% of A Game of Thrones's pages, but 34% of its word count.

Previous objections include verifiability (it is trivially verifiable for public-domain texts) and variation between editions (a problem that already exists with page count for the 1818 and 1831 editions of Frankenstein).

I have recently nominated Dracula for FAC and, moving to Frankenstein, believe there is value here. This would be especially valuable for classic texts, to provide a better indication of a work's size. This is not going to make the field mandatory for every book, but provides an editor with the ability to include it. — ImaginesTigers (talk∙contribs) 11:42, 9 February 2025 (UTC)

Is it easily verifiable though? Even with regard to public-domain texts. As one user wrote in 2010, "Pages [are] a regular feature of bibliographic details provided by publishers etc. and [are] very easily verifiable. Word count is less easily verified." Nikkimaria opposed the proposal back in 2016 on the same grounds. Word counts would mostly be the result of WP:ORIGINAL, and I believe result in endless nitpicking revisions. You yourself rounded the figures of your examples. When I try to find word count figures for classic texts, I find differing figures for seemingly the same editions. The figures could be presented in prose when supported by reliable secondary sources, but such a parameter seems problematic to me. The calculation method of word counts is also not exactly obvious, which would fail WP:CALC.

Pinging other users who participated in the past three discussions and are still active on Wiki: Cybercobra, MZMcBride, Odie5533, TAnthony. Οἶδα (talk) 09:46, 21 February 2025 (UTC)

It's odd to me that I am prevented from adding it. If the information is verifiable, I should be able to provide it. Framing the issue around "people will argue" is fine: if there is consensus against inclusion (as with any content dispute), it shouldn't be included. — ImaginesTigers (talk∙contribs) 10:32, 21 February 2025 (UTC)

There is no parameter for cover colour; for protagonist name; for number of chapters. All of these things are easily verifiable in at least some cases. From that perspective you're "prevented" from adding many things, although you really could add any of them elsewhere if you felt it warranted. Nikkimaria (talk) 01:27, 22 February 2025 (UTC)

"Prevented" implies you wish to add infobox parameters without consensus. But I am not sure that consensus will be built for a parameter that is not easily verifiable nor widely applicable. As mentioned above, you are free to it to the article's prose instead. I would be curious to see secondary sources that actually mention word count calculations because I've had a hard time finding any. An easy one should be Harry Potter and the Philosopher's Stone, which many unusable sources claim to have a word count of 76,944. But when I search for that number in reliable sources, I cannot find the figure. The closest I could find was an Oprah Daily article merely mentioning that Rowling's books each include more than 76,000 words. Even for Pride and Prejudice, I can find varying literary sources claiming over 122,000 or 124,000 words, but no exact figures. Οἶδα (talk) 07:38, 22 February 2025 (UTC)

I am not requesting any of those parameters, and would oppose any attempt to add them.

IMO, the fetishism over verifiability is a bit silly in this instance. Anyone can check word count. If someone disagrees with it, that's fine. They can discuss on the Talk if the number is wrong, and someone can go check. This looks like dislike of change for the sake of it to me. 🤷‍♂️ That's the nature of this site, I guess... status quo is best. — ImaginesTigers (talk∙contribs) 13:00, 24 February 2025 (UTC)

Okay, but I feel as if you ignored most of what I wrote and did not entertain my appeal for such sources. You even cited the same guideline as proof of the opposite. "Anyone checking word count" is original research and I believe involve more than a routine calculation. It is brave to assume that there woud be "consensus among editors that the results of the calculations are correct, and a meaningful reflection of the sources." I am not finding an obvious method of calculating a book's exact word count. And as I alluded to, such sources are severely lacking and even when they exist they are imprecise and contradictory. But more importantly, you seem to be convinced that there is a significant need for this parameter due to the fact that there was one request 17 years ago, another 15 years ago, and a third 8 years ago. I would not characterise that as reflecting a significant demand from the community. In fact, those discussion all resulted in either confusion or disapproval. I would hope other participants could chime in on this discussion. Οἶδα (talk) 21:16, 26 February 2025 (UTC)

I agree this is useful. Recently I had some discussion on classification of works into novels, novellas, short stories and like, based on some definitions of word count. It would be good to have this number research and displayed more prominently. _{Piotr Konieczny aka Prokonsul Piotrus| reply here} 00:20, 11 April 2025 (UTC)

I get the idea but I don't think checking it yourself is as easy as you indicate. PARAKANYAA (talk) 00:31, 11 April 2025 (UTC)

followed - title by publication date, or reading order? For later prequels

See Talk:Rozdroże_kruków#Followed_by_The_Last_Wish???. A 7th book in a series, but a prequel. Is it the first or the last in the infobox order? PS. I would think the former (publication date). Should we have an option to list works by reading order (in-universe chronology)? _{Piotr Konieczny aka Prokonsul Piotrus| reply here} 16:05, 10 April 2025 (UTC)

Code read out when using a screen reader

When viewing an article in Firefox with the template inserted, the following code is visible when using a screen reader (truncated for brevity):

ctx_ver=Z39.88-2004&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Abook&rft.genre=book&rft.btitle=%27%27name%27%27&rft.author=%27%27author%27%27&rft.date=%27%27pub_date%27%27&rft.pub=%27%27publisher%27%27&rft.place=%27...

The above code is in the template.

As you can imagine, it's distracting to read this and looks like something is broken.

Tagging in @Graham87 for further input. KaraLG84 (talk) 19:36, 26 April 2025 (UTC)

@KaraLG84: These are Wikipedia:COinS, which are a type of microformat. They've been interesting with screen readers for a while now; I know there accessibility was mentioned in this discussion in 2014. Unfortunately, I don't think there's much that can be done about it. CC'ing Pigsonthewing, as the main architect of this system, who has also participated in many accessibility discussions here over the years. Graham87 (talk) 03:56, 27 April 2025 (UTC)

COinS predates microformats; I had no part in implementing the former.

Is there not a CSS rule that can be used to prevent text from being read out? Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 08:50, 27 April 2025 (UTC)

Notified: Wikipedia_talk:WikiProject_Accessibility. : KaraLG84 (talk) 11:19, 27 April 2025 (UTC)

I'm surprised actually, a screen reader is reading out a title attribute of an empty element without a distinct request? Izno (talk) 15:42, 27 April 2025 (UTC)

Yeah it doesn't happen any more in Chrome (which is my primary browser), but it does in Firefox. Graham87 (talk) 03:36, 28 April 2025 (UTC)

Weird. Well, I've moved the couple of custom styles over to TemplateStyles so I've also added a display: none for the span. Does Firefox still read it out? Izno (talk) 04:26, 28 April 2025 (UTC)

@Izno: Not any more .. thanks very much ... that seems to have done the trick! (I tested it with the latest stable versions of both JAWS and NVDA and they both don't read out the spurious text, as expected). Slightly off-topic for this page, but could you also look in to the CS1 citation templates? They don't read out the COinS metadata in NVDA under either Chrome or Firefox now, but still do so with JAWS and Firefox (but that could well be a JAWS problem, as it's Firefox support is ... interesting sometimes). Should something be done for the more general case? Surely {{infobox book}} isn't the only such template that emits metadata in that way. Unfortunately it's harder for me to test this sort of thing thoroughly now because, as I said above, it no longer happens in Chrome and Firefox isn't my primary browser. Graham87 (talk) 05:26, 28 April 2025 (UTC)

It would be easy to do, but I'm a little leery of jumping to do so in CS1 just because that's a chunk more affected pages and etc. I think probably what should be done is to file (email? IDK) a task upstream. Again, I'm generally surprised that any titles that aren't associated with the modified use of the attribute (as with <abbr>) are being read out in any agent. Izno (talk) 05:30, 28 April 2025 (UTC)

I should've known ... in JAWS' default settings, it reads the COinS out for citation templates in *both* Chrome and Firefox. That's because, in JAWS 2025, they recently added a so-called "feature" to announce "descriptions of elements" (i.e. the title tag) everywhere (including link titles in Wikipedia where they differ from the page title!) I bitterly complained about this by email but my message got nowhere (though I was informed about how to turn the new title reading off, which I did immediately of course). Search for the words "short description to an element on a page" in their what's new page. Your addition of "display: none " stopped the COinS from reading at {{infobox book}} even with the new feature on, which is what we want. But that can't (and shouldn't) be applied everywhere. (Here it's probably worth noting that JAWS can be customised to behave differently for each website ... but choose your default settings wisely people!). Graham87 (talk) 08:41, 28 April 2025 (UTC)

I can also confirm it no longer does this with NVDA. Thanks for fixing it. KaraLG84 (talk) 10:37, 28 April 2025 (UTC)

@Izno @KaraLG84 @Graham87

Sorry to ping you 3 months later, but display:none; will also now not show the text to the reader (which I'm unsure if it would be necessary but if it is), rather give the element that contains it the role=none or wrap the output in a div that you give that role as it should give it to all it's children. waddie96 ★ (talk) 09:05, 23 August 2025 (UTC)

Correction: poorly worded, rather let me expand:

display:none; or visibility:hidden; will completely remove the content from the accessibility tree, so assistive tech (e.g. screen readers) won't see it. That's useful for decorative elements but problematic if the content is meant to be perceivable in any way like text that must be read (sample code in a code block, or even errors like the annoying pre-save transform UNIQ might actually need to be shown so the reader knows something isn't right).

role="none" / role="presentation" is the better approach when you want the element (or container) to be ignored semantically, but still have its children (like text nodes) announced. Wrapping in a <div role="none"> (or applying it directly to the container element) makes sense if you want to strip away the "extra" semantics without actually hiding the text.

waddie96 ★ (talk) 23:34, 23 August 2025 (UTC)

Posthumous

Can we add a field called "posthumous" such that if it yes, then a tag (posthumous) appears somewhere in the template, maybe next to the author? ReyHahn (talk) 08:33, 2 September 2025 (UTC)

No, that wouldn't be useful. PARAKANYAA (talk) 18:47, 2 September 2025 (UTC)

It can just be added in the article prose. – Jonesey95 (talk) 04:09, 4 September 2025 (UTC)

Status?

Can we add status, or something like that? There are number of books where status is important, like Lost or Posthumously published. Primarily i created some lost books and i wanted to present that in status. --Dr.Bookman (📖) 08:10, 13 October 2025 (UTC)

I don't think this is useful for 99.999% of books. I would oppose its addition. PARAKANYAA (talk) 08:14, 13 October 2025 (UTC)

I beg to differ, its far bigger number of books that can be affected in many ways with status. Why should be forbid something that will make Wikipedia better? --Dr.Bookman (📖) 21:13, 13 October 2025 (UTC)

I'm seeing about 700 pages in Category:Books published posthumously, or a little over 1% of the transclusions of this template. Category:Lost works is much more sparse. Infoboxes are supposed to summarize key information in the article. In the rare case that a book has a "status", which is a quite ambiguous term that would be misused by editors, that status can be mentioned in the prose, the book can be tagged with a category, or both. – Jonesey95 (talk) 21:30, 13 October 2025 (UTC)