![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
Holmes & Related Fandoms is dead, long live Holmes & Related Fandoms
Some time recently, AO3 tag wranglers de-canonized the “Sherlock Holmes & Related Fandoms” metatag. This involved detaching the fandom subtags for almost 100 different Holmes-related adaptations, reboots, and spinoffs, and then making the “Related Fandoms” metatag into a synonym of “Sherlock (TV)“.
I don’t actually know when the switch was made (though it couldn’t have been more than 4 weeks ago), but within the past week, fandom-at-large (a) noticed, (b) was horrified, and (c) inundated AO3 Support with a flood of complaints.
The fallout is being discussed all over the place — here’s a nice roundup of some reaction links — and there’s a heck of a lot of misinformation swirling around. Gonna try to set some of it straight here.
Disclosure: I’m a wrangler. An average one, not a supervisor or a chair. I don’t wrangle any Holmes-related fandoms, and was not part of the Holmes &RF discussion. This is supposed to be a facts-and-info post; any personal opinions that do show up are my own, not speaking on behalf of any other wrangler(s).
There are some links to individual fics in here — they’re not supposed to represent “good tagging” or “bad tagging”, they’re just examples of “a kind of tagging somebody has done.” Please don’t be weird to the authors.
1) Tag wranglers do not change the tags on a work. What we change is what those tags redirect to.
Whatever fandom tags you have typed in on your fic — you always keep those! Wranglers do not edit or replace the text of your fandom tags in any way.
(Caveat: if someone is just blatantly lying, that can be a TOS violation. If someone tags all their fic with the 10 most popular fandoms on the archive, without actually writing about any of them, that would probably get flagged as spam. We’re talking about good-faith tagging here.)
A lot of what wranglers do is making “Tag 1” a synonym of “Tag 2”. This doesn’t replace “Tag 1” on any individual fics. It just makes the filtering for Tag 1 redirect to Tag 2. For example, “BBC Sherlock” has been synned to “Sherlock (TV)” — so any fic that’s only tagged “BBC Sherlock” will show up in the Sherlock (TV) search results. But you can see that nobody force-added “Sherlock (TV)” to the tags on that specific fic.
Similarly, while the Holmes &RF tag was synned to the TV tag, it didn’t rewrite the tags on any fic that only had the &RF tag. Or any fic with multiple different Holmes fandom tags but no Sherlock (TV) tag. So when the synning was reversed, it didn’t require any extra changes to the database — those fics automatically went back into the search results they were in before.
And, look — most of the time, tag synning is good and useful for everyone! Of course fans shouldn’t have to look up “BBC Sherlock” and “Sherlock (TV)” separately — along with “Sherlock (T.V.)” and “Sherlock (TV Series)” and “Sherlock (TV show)” and a hundred other different ways to say the same title — just to get all the fics about this show. Of course you want them all in the same place.
As with many things, you only hear about tag synning when it’s causing people problems to complain about. When it’s working smoothly, you have a good experience and you don’t even notice.
2) Big generic/umbrella fandom metatags can come in a lot of different formats.
Along with syns, wranglers also make metatags and subtags! If you filter on a fandom metatag, you'll get all fics with all the subtags together -- but you can also filter any of the subtags as their own separate thing.
When you have a group of fandoms with different titles, but they’re all based on the same original source, the metatag often ends with “& Related Fandoms“.
Examples:
- Dracula & Related Fandoms (with subtags like Dracula – Bram Stoker (Novel 1897), Young Dracula (TV), and Hellsing)
- A Song of Ice and Fire & Related Fandoms (with subtags like A Song of Ice and Fire – George R. R. Martin, Game of Thrones (TV), and House of the Dragon (TV))
When it’s a franchise where the different versions have the same title…or, uh, mostly the same title…the metatag often ends with “– All Media Types“.
Examples:
- Fullmetal Alchemist – All Media Types (with subtags like Fullmetal Alchemist (Anime 2003), Fullmetal Alchemist (Live Action), and Fullmetal Alchemist: Brotherhood & Manga)
- Batman – All Media Types (with subtags like Batman (Comics), Batman: The Dark Knight Returns (2012-2013), and Under the Moon: A Catwoman Tale – Lauren Myracle)
Some fandom metatags were canonized before wranglers got diligent about consistently using the “&RF” or “AMT” suffixes. So they might be just the name of the overall fandom, nothing at the end.
Examples:
- Star Trek (with subtags like Star Trek: The Original Series, Star Trek: Picard, and Star Trek: Lower Decks (Cartoon))
- Bishoujo Senshi Sailor Moon | Pretty Guardian Sailor Moon (with subtags like Bishoujo Senshi Sailor Moon | Pretty Guardian Sailor Moon (Anime & Manga), Bishoujo Senshi Sailor Moon | Pretty Guardian Sailor Moon (Musicals), and Code Name: Sailor V)
All of these are the same general “kind” of tag — a big umbrella tag for an overall franchise.
When you see an official AO3 post like this one use the term “AMTs” for “All Media Types”, they’re talking about big metatags in general, even the ones that don’t use this specific phrasing.
2) Official AO3 wrangling policy is “try to dismantle fandom tag trees.”
When fans write in to Support to request a metatag for their fandoms, they get answers like the response this user got when asking for a Monogatari – All Media Types tag.
(Support has a template answer for this, with officially-approved phrasing, which they adapt slightly depending on what fandom you asked about. That’s pretty standard for them. It would be exhausting if they had to write a fresh response from scratch every time a new user makes the same request.)
The main point: “At this time, tag wranglers are not creating All Media Types tags, and are in fact dismantling them when possible.”
There are some specific reasons given, such as the idea that fandom metatags “confuse both creators and browsing users” and “can make searching and filtering more difficult.” I’m not aware of how these things were determined. Were they based on Support tickets? On seeing lots of confused public Tumblr/Reddit posts? Did we do a poll somewhere? I don’t know.
If I ever see an official AO3 statement that elaborates on these points, I’ll come back and address it here. If you’ve seen one, feel free to comment with the link.
After the public outcry against the dismantling of the Holmes metatag, the one thing we can say for sure is — whatever metrics we’re using to figure out “what will be the least confusing/difficult/generally bad for users?”, they aren’t always right.
3) Big metatag trees don’t cause any technical problems for the AO3 servers.
They used to!
Back in the day, if a fic was written for “Fandom X Movie”, but that had the metatag “Fandom X Trilogy”, which had the metatag “Fandom X Cinematic Universe”, which had the metatag “Fandom X – All Media Types”…that used to strain the servers 4x as much as a fic in a fandom with no metatags.
The code issue is no longer a problem. (I’m not a coder, I can’t explain why in any detail — but the official response from AO3 when people ask about it is, that code has been fixed.)
(ETA: Here's a public-facing response from an AO3 news post, where the head of AO3 Systems was asked about the issue! This was in July 2020.)
So if you hear people saying “I heard it was bad for the servers”…they probably did hear that at some point. And it might have been accurate when they first heard it. But that info is now out-of-date.
Unsurprisingly, this is the origin story of AO3’s “we should look for fandom metatags we can remove, and not create new ones” policy. Of course the tag wrangling committee wanted to be nice to the servers — we didn’t want to make the site fall over.
I don’t recall any official internal discussion about “okay, now that there’s no longer a technical reason to avoid metatags, should we stick with that policy? Or is it creating a worse experience for users?” Not every wrangler thought it was a good idea, you were allowed to say so, and some people did — but the policy didn’t get revisited in any official way.
In the wake of the Holmes blowup, TW Chairs have announced we’re finally going to have that official discussion.
No idea yet what the results will be! We’ll find out.
4) Fandom tag trees only get changed with the combined approval of (a) the wranglers for that fandom, and (b) Tag Wrangling supervisors/chairs.
There’s no one policy that’s automatically mandated for every kind of fandom on AO3. They’re all decided on a case-by-case basis.
Sometimes wranglers will ask for a change, and Chairs will have to approve or deny it. Other times, Chairs will suggest a change (usually passing on a Support request), and the wrangler(s) of the fandom involved will accept or reject it.
(Possible secret third thing: there’s a claim on FFA that Chairs weren’t letting anyone actually wrangle characters/relationships/freeforms in the Holmes &RF metatag. That’s not normal! With most fandom franchises, if the wranglers of the individual fandom subtags want to keep the metatag, they just work together to sort tags from the metatag into the correct fandoms — there’s no technical or administrative interference to stop them from doing that.) (Disclaimer: this is from an anonymous comment, so take that as you will.)
ETA: Asked some questions about this in org chat, and here's my takeaway:
- Pre-Holmesaggedon, the actual policy about on-hold AMT fandoms was confusing and poorly-conveyed, with wranglers getting mixed messages...and a lot of people reluctant to ask questions, in case the result made things worse for their fandom(s)
- Post-Holmesageddon, the stated policy is "yes, wranglers are welcome to assign themselves to these AMTs, without any expectation that it'll end in the AMT being dismantled. Chairs will proactively reach out to the wranglers of relevant sub-fandoms, so anyone who's been discouraged away from this before is actively encouraged to go for it now"
As you might have guessed from section (2) above, Chairs will typically approve requests to dismantle an existing metatag, but not to create a new one.
The most well-known exception is probably Video Blogging RPF. A lot of disparate fandoms used to get synned to that tag. (It’s the equivalent of, in the Holmes fandoms, if all the individual fandom tags were synned to “Sherlock Holmes & Related Fandoms”. You could get all of them together, but you couldn’t narrow it down to just the TV series fics. Or filter out just the TV series fics.)
And then: this May 2022 announcement that series like Dream SMP, Hermitcraft SMP, and SMPLive would start getting their own separate fandom canonicals! Video Blogging RPF would become the metatag that showed you all the fics together, but each fandom could now be searched and filtered separately.
This change happened after years of both wranglers and users requesting a metatag tree, here. As far as I know, there wasn’t an overwhelming write-in campaign all at once, like with the Holmes &RF tag tree — just regular complaints over the years. I don’t know what the turning point was that got Chairs to approve the request, when they previously hadn’t.
Sorry, Monogatari fandom, I’d tell you if I knew.
5) If you write a request to AO3 Support…be cool about it, come on.
Look, I get why AO3 users are frustrated. Both with fandoms where the tags suddenly change in a way that messes up your ability to use them, and fandoms where the tag status-quo has problems that nobody seems to be fixing.
You have no way of knowing if you’re the first user to point something out, or the 500th user to make the same complaint. You don’t know if your request will get passed on to the relevant wranglers at all, or vetoed by Chairs upfront. And when you get a copy-and-paste boilerplate response, it’s easy to forget there’s a person behind that other screen at all.
But there is. Every Support request gets opened by a real human being. Who probably has no connection to the thing you’re complaining about, but who has to read through whatever you wrote so they can figure out what response to send.
Don’t lash out at them, don’t get abusive, generally don’t be a dick.
(ETA: A Support vol shares some numbers on FFA. An average month has 2,000-ish tickets total. As of this update, they've had about 1,000 AMT-related tickets over 4 days.) (This is another anon comment; I will say that it tracks with things I've heard from non-anonymous volunteers.)
Further questions? Ask away.
I’m not here to share anything confidential, or gossipy/grudgewanky. (No “can you believe Wrangler X wanted to do so-and-so with Fandom Y??” type stuff.)
But there’s a lot of normal, non-secret information about How Tagging Works that many AO3 users just…don’t know. Or aren’t sure about. Or you’ve heard conflicting things and don’t know how to confirm which one is real. Or whatever.
So, hey, ask that stuff here. I’ll do my best to answer. Anything I don’t know, I’ll go looking for someone who does.
Some links:
- Tag Wrangling FAQ (for general AO3 users)
- Wrangling Guidelines (for insight into how wranglers will sort your tags)
- OTW Board official Discord server (the OTW is the organization that runs AO3)
- Quick’n’dirty AO3 metatags survey (this is not official, I’m just curious)
no subject
I have a couple thoughts just based on my experience with both online taxonomies and posting on AO3.
Firstly, with regards to point 1: it is technically correct, of course, but I think it's lacking in understanding of what a tag actually means to a creator. For context, I'm one of the people whose women's hockey RPF fic is now categorized under "Men's Hockey RPF" (link).
A tag isn't just the text on the work; when I add a tag to my work I'm cognizant that I am categorizing my work for future readers to find. So while it's obviously true that the text remains identical, the meaning is in fact changed by synning. At the time, a bunch of us used "Hockey RPF" as a generic catch-all, and we tagged women's RPF with "Hockey RPF" on purpose. This context - a bucket of NHL and women's Olympic team RPF being comingled - was destroyed by synning "Hockey RPF" to "Men's Hockey RPF". So I'd argue that "Tag wranglers do not change the tags on a work" is false. You do change the tag. You don't change the tag's literal string text in the database, but you do change the tag, e.g., the string plus its meaning at the time when I used it, because the meaning is derived from the tag's functionality. It's not just text.
Which leads me to:
So, obviously this is true, but I can think of scenarios where it might not be true, like two fandoms for the same canon that were historically separate - different languages, different country-specific fandoms, etc. In that case I'd argue that the taxonomy should be leveraged rather than synning the terms, because those two groups may not consider themselves the same fandom. It might be technically correct to syn them, but it's not contextually, socially correct, and contextually and socially correct tags were supposed to be part of the point of wrangling.
It's worth noting that this complication isn't unique to AO3. An example from work: we would coach customers to set up synonyms for regionally different terms for the same product, but we also cautioned them to consider localization; if they were literally selling the same SKU but someone might use two distinct terms to search, okay, set up a synonym. But if the two terms reflect a meaningful distinction (region-specific parts that aren't compatible with one another), then you may not want to set them up as synonyms, because then someone searching for their region's term is going to get a bunch of junk they can't even use.
Also, though, this currently isn't how a lot of synning happens on AO3. Case in point, the wank wasn't about synning "Sherlock TV" and "BBC Sherlock", it was about synning Sherlock with the AMT tag. So I think it's important to emphasize that if you're synning two terms that are not literally, provably, objectively and contextually the same, you are performing a destructive action, "destructive" here meaning that context is being lost when the two terms are synned.
Secondly -
I appreciate your transparency about where this impression came from. But I would strongly recommend TW and the OTW as a whole look into literature on UI/UX and taxonomies (just a random Google produces this link for example). Obviously I'm not a volunteer, but based on my experience as a longtime user of the site, there are two issues with TW's current understanding of browsing:
1. The actual structure of the data is not the only way to improve user experience. If users are confused about a metatag, the fix might be outside your remit as it might be an actual interface problem (surface more of the tree, make exclusions easier, etc). With a large and diverse dataset, the "perfect" taxonomy doesn't exist, and it can be very, very beneficial to take a step back and think through how users might accomplish the task at hand using all tools available to the org (UI updates, backend updates, and wrangling) rather than one committee's. Which leads to
2. It seems that wrangling leadership would benefit from exploring a few software development & product management concepts, namely user stories and user flows. Structuring your inquiry into a problem according to user stories/user flows can really help illuminate what alternate solutions exist (assuming people come to understand the AMT synning as a destructive and undesirable action, which - I hope so!). What I have noticed about wrangling's guidelines and processes over the years, at least what's exposed to end users, is that there is a very serious deficit of understanding of the problem from a user's perspective. There's plenty of documentation about the importance of internal consistency and correctness, and I get that...but, as a user, synning of tags that are not self-evidently identical is genuinely very confusing. And it seems like a hack, to me, compensating for historic inadequacies with search or infrastructure, more or less completely detached from how people actually use the archive. This isn't a problem unique to the OTW! It's the kind of problem that UI/UX as a field seeks to address, and looking at how other people integrate UI/UX into product management could be really beneficial.
(Sidebar: it's obviously not your committee, but I really hope Support is looking into automated workflows for some of this stuff - there is just no need to respond to all those tickets by hand, or even really to prioritize them for any response other than that the change has been reverted.)
Last but not least, again, thank you for posting this, it was a thorough and interesting read and will be a great link for people I know who are confused about what happened or how tagging actually works.
no subject
A lot of the original post was inspired by specific misinformation that I've seen floating around. So if you're already up on the minutiae of tag wrangling, most of this is directed at people who are, like, five steps behind you.
Example: there are people claiming "AO3 wranglers will literally rewrite the tag text on your work." When I explain that it's not true, I'm not making a secondary argument about any other possible meaning of "change the tag"! I'm just straightforwardly trying to help catch up the AO3 users who didn't know about this one.
no subject
no subject
The experience you describe here is frustrating. It's user-unfriendly. As a writer, I don't like that happening with my fics either! If I was Grand High Dictator of AO3, I'd pass a ruling to change it.
If people are mad about Actually Happening Thing X, but then they write in to Support to complain about Misunderstanding Y, of course the boilerplate response will be "good news, we don't do Misunderstanding Y." And then the AO3 machinery doesn't even process the idea that those users are really mad about Thing X. Which doesn't help with any efforts to get Thing X changed.
no subject
Definitely seems like that's the pattern! It's complicated with AO3, I have seen the sort of mythmaking and "wtf that just straight-up isn't true" stuff that gets spread on Tumblr and Reddit (and I imagine dealing with it as a wrangler is beyond frustrating). But at the end of the day, "users think the system works like X and they're mad about it but it doesn't work that way" should always prompt a "huh, why do they think it works this way?", you know? And how to structure that inquiry or even spot when the misalignment is happening is the meat & potatoes of a lot of user research in the software space.
To put a bow on it - I do appreciate your clarifications, as I said, I just also hope that things don't stop at clarification, and that TW leadership's decisions are a little more anchored to how people use the site in the future.
no subject
I'm trying to stay pretty reserved in how I talk about this, because I don't want it to come off like I'm making predictions or guarantees about what TW leadership will do. But yeah, I share your hope that this blowup leads to some meaningful policy changes in the long run.