I'll try to clarify:
There's plenty of posts tagged with cat_humanoid but not cat, do they apply to "humanoid" vs. "anything else" only? So there should be very few posts that have both, unless a standard cat is in the same picture with a catgirl, right?
What about tagging the exact species? Lots of posts with cattle_humanoid but not cattle, but if I wanted to add holstein_friesian_cattle to a humanoid character, it'd also get tagged as cattle.

Not sure if the question is clear...

Related question: feline_humanoid implies feline and felid_humanoid implies felid. Why? cat_humanoid does not imply domestic_cat.

In my opinion, *_humanoid tags should not imply *. People who search taxonomical classifications alone want to see anthro and/or feral characters, don't they?

I don't think that *_humanoid tags should imply their respective * either.
That would make "fully-furry" content separating from "humans with animal ears" more difficult, and vice versa.

bitwolfy said:
I don't think that *_humanoid tags should imply their respective * either.
That would make "fully-furry" content separating from "humans with animal ears" more difficult, and vice versa.

Should I write a BUR removing the implications? It would be pretty lengthy

gattonero2001 said:
Should I write a BUR removing the implications? It would be pretty lengthy

It's likely done the way it is for a reason. A feline_humanoid should be tagged feline just like feline ferals and feline anthros are. If anything you could argue that the various *_humanoid tags shouldn't exist (there's no dragon_anthro or dragon_feral tags for dragon+anthro or dragon+feral, but dragon_humanoid exists for dragon+humanoid?), but it is what it is. Unless the mods/admins say otherwise, it's best to leave them as they are.

watsit said:
Unless the mods/admins say otherwise, it's best to leave them as they are.

Ok, but how should I tag posts moving forward? A catgirl would be cat_humanoid domestic_cat or just cat_humanoid?
If we decide that in the end *_humanoid shouldn't imply *, then currently it's a bit of a mess when you want to get more specific in the species: a humanoid tagged with maine_coon implicates domestic_cat

@Watsit

Perhaps there is not an actual reason. Mods and admins also commit mistakes (for example, aliasing the artist name kellervo to the generic/ambiguous character name keller). Without proper documentation, we cannot know. Besides, community agreement should be an acceptable reason for change and writing a BUR would help us measure it.

@ShadyGuy

I believe that in the situation that you are describing (animal humanoids of specific breeds) it would be better to create specific tags (for example, holstein_friesian_cattle_humanoid and maine_coon_humanoid) and request the appropriate implications (in this case, to cattle_humanoid and cat_humanoid respectively).

@Watsit @bitWolfy

gattonero2001 said:
Besides, community agreement should be an acceptable reason for change and writing a BUR would help us measure it.

That'd be a pretty huge BUR:

• Go through every species tag and remove *_humanoid -> * implications and aliases
• Go through every species tag and create its *_humanoid equivalent if it doesn't already exist
• Go through every newly created *_humanoid tag and add aliases and implications (ex. since maine_coon implies domestic_cat, maine_coon_humanoid should imply cat_humanoid)
• Also clarify on the wiki? Or make a dedicated wiki page about the difference.
• Huge amount of clean-up of posts that are tagged with both

I'm not saying I don't agree, because I do, but that's a lot of work

shadyguy said:
I'm not saying I don't agree, because I do, but that's a lot of work

I have some experience writing BURs and I am willing to be responsible for this one.

However, it will probably not be as long as you are thinking, since my original intention was to remove implication *_humanoid -> * and nothing else.

Breed-specific humanoid implications (like maine_coon_humanoid -> cat_humanoid, for example) should be requested as the tags themselves are created, since trying to alias/implicate nonexistent tags will result in an error.

Proper species humanoid tags already do not implicate species, so the problem is restricted to higher taxonomical classifications (felid and feline but not cat or lynx).

The bulk update request #646 has been rejected.

Reason: Discussion in topic #28526

EDIT: The bulk update request #646 (forum #304020) has been rejected by @Millcore.

It makes sense to find avian humanoids when searching avian, and the same for other species. This reduces searching effectiveness, and puts the onus on the tagger to, for instance, manually add feline to feline_humanoid posts despite it always applying, the kind of thing implications are designed for.

watsit said:
It makes sense to find avian humanoids when searching avian, and the same for other species. This reduces searching effectiveness, and puts the onus on the tagger to, for instance, manually add feline to feline_humanoid posts despite it always applying, the kind of thing implications are designed for.

I think, the point of this is that feline_humanoid should NOT be tagged as feline at all.
A feline_humanoid is not really a feline. It's a human with cat ears and tail.

watsit said:
It makes sense to find avian humanoids when searching avian, and the same for other species.

I disagree for reasons stated in my earlier comments on this same topic.

This reduces searching effectiveness, and puts the onus on the tagger to, for instance, manually add feline to feline_humanoid posts despite it always applying, the kind of thing implications are designed for.

In my perspective, feline only applies to feline_humanoid posts when there is at least one anthro or feral feline character. Most people who search for taxonomical classifications without the humanoid suffix do not want to see humanoids.

Would you say that the onus is on the tagger to manually add domestic_cat to cat_humanoid posts, since there is no implication?

bitwolfy said:
I think, the point of this is that feline_humanoid should NOT be tagged as feline.

Both feral and anthro felines are tagged feline, so why wouldn't humanoid felines? This would also cause people who searching or blacklisting feline to miss or still get humanoid felines, reducing the effectiveness of the tags.

gattonero2001 said:
Most people who search for taxonomical classifications without the humanoid suffix do not want to see humanoids.

Says who? If I search for feline or avian or canine, I don't care if it's feral, anthro, or humanoid, as long as it's feline, avian, or canine-based. As it is, hybrids should be tagged with their component species when it's identifiable, so if a character is part feline, it should be tagged feline.

gattonero2001 said:
Would you say that the onus is on the tagger to manually add domestic_cat to cat_humanoid posts, since there is no implication?

is kind of weird because it's not a real species. It's a collection of species we've ostensibly domesticated (although plenty of families of such species may still live undomesticated lives and be undomesticatable). A cat isn't always a domestic cat, and a cat humanoid is largely also human, so calling a humanoid cat "domestic" doesn't really work.

watsit said:
Both feral and anthro felines are tagged feline, so why wouldn't humanoid felines?

There are no feline_anthro or feline_feral tags because implementing them would require creating an entire system of *_anthro and *_feral tags, which is not worth the effort.

The *_humanoid tag system, on the other hand, has already been created.

This would also cause people who searching or blacklisting feline to miss or still get humanoid felines, reducing the effectiveness of the tags.

In that case, those people can search ~feline ~feline_humanoid or blacklist feline_humanoid.

gattonero2001 said:
There are no feline_anthro or feline_feral tags because implementing them would require creating an entire system of *_anthro and *_feral tags, which is not worth the effort.

The *_humanoid tag system, on the other hand, has already been created.

So then do you agree that it's otherwise the same, it's just that the *_humanoid tags already exist so we might as well use them? Then why shouldn't humanoids be tagged as feline/canine/etc since ferals and anthros are?

gattonero2001 said:
In that case, those people can search ~feline ~feline_humanoid or blacklist feline_humanoid.

Extra work for getting what should be default behavior. As I said in my edit, hybrids should be tagged with their component species when it's identifiable, so if a character is part feline, it should be tagged feline. Why shouldn't this apply for humanoids with identifiable feline traits too?

watsit said:
Says who? If I search for feline or avian or canine, I don't care if it's feral, anthro, or humanoid, as long as it's feline, avian, or canine-based. As it is, hybrids should be tagged with their component species when it's identifiable, so if a character is part feline, it should be tagged feline.

I admit that "most" could not be accurate and it will take time to verify that based on public reaction to the above BUR, but so far bitWolfy and ShadyGuy agree with my position.

Humanoids are not hybrids, so that argument does not apply to this situation.

domestic_cat

is kind of weird because it's not a real species. It's a collection of species we've ostensibly domesticated (although plenty of families of such species may still live undomesticated lives and be undomesticatable). A cat isn't always a domestic cat, and a cat humanoid is largely also human, so calling a humanoid cat "domestic" doesn't really work.

Domestic animals aside, would you say that the onus is on the tagger to manually add wolf to wolf_humanoid posts, since there is no implication?

I disagee. Fantasy species fall into base species groups. Humanoids are treated the same way as fictional species are.

gattonero2001 said:
I disagree for reasons stated in my earlier comments on this same topic.
In my perspective, feline only applies to feline_humanoid posts when there is at least one anthro or feral feline character. Most people who search for taxonomical classifications without the humanoid suffix do not want to see humanoids.

Would you say that the onus is on the tagger to manually add domestic_cat to cat_humanoid posts, since there is no implication?

I think you are confused. Humanoids are treated like other fictional species and races. They fall into the base species group for that animal. Cat humanoids don't get implicated to cat, because their base species group is feline. Also I think you are painting everyone with a broad brush without evidence.

Users shouldn't be manually adding the missing species tags. The implications are set the way they are for most species for a reason.

Some things about the BUR itself:

- mammal_humanoid is a mammal no matter how you define it. Humans are mammals.
- avian humanoids should be avians no matter what. This is a massive base species for all kinds of winged creatures. It should not be split up.

watsit said:
So then do you agree that it's otherwise the same, it's just that the *_humanoid tags already exist so we might as well use them? Then why shouldn't humanoids be tagged as feline/canine/etc since ferals and anthros are?

Your proposal reduces search granularity and therefore search effectiveness. If the *_humanoid system had never been created, I would be against creating it now. Since it has already been created, the benefits of its existence can be appreciated without dealing with the issues involved in its implementation.

If the *_anthro and *_feral tag systems had already been created, I would be against having them imply taxonomical classifications as well.

Extra work for getting what should be default behavior. As I said in my edit, hybrids should be tagged with their component species when it's identifiable, so if a character is part feline, it should be tagged feline. Why shouldn't this apply for humanoids with identifiable feline traits too?

You believe that it should be the default behaviour. Other users, like myself, might believe that it should not. The "extra work" that I proposed is barely an inconvenience.

If animal humanoids were considered hybrids by e621's tagging system, that specific implication would have already been created.

thevileone said:
I think you are confused. Humanoids are treated like other fictional species and races. They fall into the base species group for that animal. Cat humanoids don't get implicated to cat, because their base species group is feline.

No, the "base species group" of feline_humanoid is feline. That does not mean that the implication is necessary or even desirable.

Also I think you are painting everyone with a broad brush without evidence.

I already rectified that in the comment right above your own.

"I admit that "most" could not be accurate and it will take time to verify that based on public reaction to the above BUR (...)"

Users shouldn't be manually adding the missing species tags. The implications are set the way they are for most species for a reason.

Why does canine_humanoid imply canine but wolf_humanoid does not imply wolf? That does not make sense.

gattonero2001 said:
No, the "base species group" of feline_humanoid is feline. That does not mean that the implication is necessary or even desirable.
I already rectified that in the comment right above your own.

"I admit that "most" could not be accurate and it will take time to verify that based on public reaction to the above BUR (...)"
Why does canine_humanoid imply canine but wolf_humanoid does not imply wolf? That does not make sense.

Canine is a family of species, and a wolf is a specific type of animal. Humanoids are grouped under the base species groups. You wouldn't tag dog pokemon as dog, you would call them canines instead. The same reasoning applies to humanoids as well.

Why can we not tag wolf pokemon as wolves, or dog pokemon as dogs?

As to the BUR itself, the humanoid species could be kept separate and it would avoid some of the cross-contamination caused by large species chains. There are benefits to doing so, but I feel there are also benefits of keeping them somewhat together too.

You should check my edit to my last post though concerning this BUR.

I am curious though if you plan to help fixing the 80K images that would need species clean up if this went through. Do you plan on helping fix it? I don't think this BUR should go through unless there are people ready to put in the effort required to remove the extra species tags.

gattonero2001 said:
Humanoids are not hybrids, so that argument does not apply to this situation.

They're not hybrids as far as the hybrid tag is concerned, but it is otherwise in the name. A feline_humanoid is, by definition, part feline. The only difference between it an anthro is how much the human attributes dominate.

gattonero2001 said:
Domestic animals aside, would you say that the onus is on the tagger to manually add wolf to wolf_humanoid posts, since there is no implication?

Maybe. If it were up to me, I may implicate wolf_humanoid -> wolf, but it would depend on if the mods/admins have a good reason for why it's not already.

gattonero2001 said:
Your proposal reduces search granularity and therefore search effectiveness. If the *_humanoid system had never been created, I would be against creating it now. Since it has already been created, the benefits of its existence can be appreciated without dealing with the issues involved in its implementation.

Benefit for one person is a hindrance for others. You're requiring people to add more search terms to find feline-based characters (with the obscure ~or syntax at that) and have the foreknowledge that it's necessary here, rather than being able to search feline for felines. Certain tags are deemed to be too granular, too specific, so are aliased to or implicate broader tags.

gattonero2001 said:
If animal humanoids were considered hybrids by e621's tagging system, that specific implication would have already been created.

feline_humanoid implicates both feline and (indirectly) humanoid, so the constituent components are tagged. They don't get the hybrid tag itself because that's reserved for non-human hybridization (otherwise anthros would count as hybrid as well, which would make the tag functionally useless), but they still get tagged with the identifiable elements.

I think we should decide on either of the two options, because as it is now it's not clear at all. Let's recap:

*_humanoid should implicate *

• PROS
• Easy to do, just BUR implicate the remaining *_humanoids to * and you're done
• Searching for * also yields *_humanoids, which can be seen as a plus
• Otherwise this search would need to be done using the OR (~) operator
• CONS
• Less precise search: searching for anthros would also return humanoids
• You'd need to manually -*_humanoid to every search you do if you're not into that

*_humanoid should NOT implicate *

• PROS
• More search precision
• You get exactly what you searched for, and nothing more
• CONS
• Huge amount of work required aside from the BUR: all "catgirl" images would be mistagged, as would be all other posts with tags that already implicate the species.
• To get around this we could do "from now on tag images like this", but eeeh, doesn't seem like a solution to me

I have no idea about the feline and felid, canine and canid debate

+1 for "*_humanoid should NOT implicate *", as I have stated before.

The improvement in search precision is enough justification for the changes in my opinion. The BUR has already been written and most of the tag correction could be done with scripts, leaving a relatively small amount of work to be done as a manual project.

shadyguy said:

*_humanoid should NOT implicate *

• PROS
• More search precision
• You get exactly what you searched for, and nothing more

You don't, actually. Without the implication, when you search for something like feline or canine, you miss posts with feline and canine characters because they happen to be a humanoid form instead of anthro, feral, or taur. A user would have to be aware they're not included, and know about the ~or syntax for searching. And even then, depending on what else a user may be searching for, they may not be able to do ~x ~x_humanoid (IIRC, all ~tags are considered part of the same "or" set, so ~gynomorph ~female kemono ~feline ~feline_humanoid would essentially search for "kemono AND (gynomorph OR female OR feline OR feline_humanoid)".

gattonero2001 said:
most of the tag correction could be done with scripts

How do you figure? Given all feline_humanoid posts are currently tagged with feline, how would a script know whether feline is still valid? A person would have to actually look at the image to see if there's any anthro, feral, or taur felines in the image or not.

I'm still against deimplication. A person searching feline or canine should be reasonably assured they're getting all posts with feline or canine characters regardless of form, and not have humanoids missing because they're not anthro, feral, or taur.

I personally have mixed feelings about this whole thing. It'd be a lot easier for me to have a clear "yes *_humanoid posts should imply their species" if anthros, ferals, and taurs also had their own tags. In which case all forms could imply canid, felid, equid, or whatever else as one giant umbrella tag. But that's not how things are handled here. Which leaves me kind of wishing the humanoid tags didn't imply the species tags, both because it'd be easier to search for pictures that have both a feral/anthro/taur in it + the humanoid, and also because when I think of a cat, I don't think of a human with cat ears and a tail slapped on.

Also because on the extreme end of humanoid (which is also common especially in anime art styles), many of them barely look like the animal they're meant to be at all, and often times it's hard to tell what they're even meant to be. There's been times I've seen something that's supposed to be a wolf-girl but it just looks like any other cat-girl but with a fluffier tail and that's... it. Which is a big reason why I avoid trying to tag humanoid posts where I can, because the whole species issue is just a nightmare.

@Watsit

One could write a script to remove feline and felid from every post that is tagged cat_humanoid solo and/or cat_humanoid -anthro -feral -semi-anthro -taur, for example.

Vulkalu's post above my own contains arguments that I agree with.

TheVileOne said in Discord that wolf_humanoid used to imply wolf, but that implication (along with many others of similar format) was removed. I would like to see the documentation with the reasoning for that decision. It would probably apply to the removals that I suggested as well.

vulkalu said:
Also because on the extreme end of humanoid (which is also common especially in anime art styles), many of them barely look like the animal they're meant to be at all, and often times it's hard to tell what they're even meant to be.

That's a bit of a running theme in furry art, for ferals and anthros too. Granted it's usually possible to at least infer "generic canine" from "generic feline" from "generic furred dragon", but sometimes not.

gattonero2001 said:
@Watsit

One could write a script to remove feline and felid from every post that is tagged cat_humanoid solo and/or cat_humanoid -anthro -feral -semi-anthro -taur, for example.

Those would need checking. It's not unusual to find solo tagged on an image where there is a second character but only one has main focus, or where people just neglected adding the anthro/feral/etc tag. Yes they're mistakes on the taggers' part, but those posts shouldn't lose a valid tag because of that. That's why there's tagging projects for these kinds of things, because they need a human eye to verify what's applicable.

I tend to agree with gatto that they should be separate, but Watsit is right, the tagging work required would take an eternity to do, so I'm still on the fence.

I would like to know the opinion of users who have experience using scripts before discarding them as a potential tool. Surely this kind of situation is one of their main functions (i.e. cleanup after implication removal)?

gattonero2001 said:
I would like to know the opinion of users who have experience using scripts before discarding them as a potential tool. Surely this kind of situation is one of their main functions (i.e. cleanup after implication removal)?

It's possible, but would require a ridiculous amount of work. Tag scripts are not fully automatic - they require the user to click on every thumbnail to which the script should apply. And feline_humanoid alone has 20k+ posts.
I can automate this process further by writing a bit of code, but Watsit is correct in that the posts would have to be checked manually. Most of the tags are correct, but some (like post #2578780) are obviously not.

My main gripe here is that I'm not fond of BURs that are this huge in scope being hidden in the middle of a discussion thread. If you make it a separate thread that is marked for what it is, and people do wind up voting on it for approval, I guess I'll join the effort to unfuck everything that's going to be mistagged because of it.
Personally, I think that this is the established method, and there's no reason to change it. It's easier for a user to blacklist animal_humanoid or simply remove it from a search query if they do not want to see that contents.

watsit said:
That's a bit of a running theme in furry art, for ferals and anthros too. Granted it's usually possible to at least infer "generic canine" from "generic feline" from "generic furred dragon", but sometimes not.

Oh yeah, definitely. Anthros and ferals are far from exempt from that. But like you said, usually you can at least infer what general family they're meant to be in. I can't say the same for a lot of humanoid posts, where all you have to go off of is a pair of fuzzy triangles on the head. Obviously, not all humanoids are like that, but it's what I tend to see.

That said, I'd also like to clarify my stance here (since I realize I failed to do so earlier). Because anthros, ferals, and taurs, are not getting their own tags, I'd at least vote to de-implicate specific species from the humanoid tags that have those implications. Let them imply canine or canid but not, say, wolf. Kind of like how Pokemon is treated (obviously with the exception that no pokemon species actually implies canid, felid, etc.).

Maybe we should get some (Moderators/Admins)' thoughts on this?

The bulk update request #646 (forum #304020) has been rejected by @Millcore.