The bulk update request #4705 is pending approval.

remove alias cat (0) -> domestic_cat (293395)
remove alias kitty (0) -> domestic_cat (293395)
remove alias cats (0) -> domestic_cat (293395)
remove alias black_cat (0) -> domestic_cat (293395)
remove alias dog (0) -> domestic_dog (356835)
remove alias doggies (0) -> domestic_dog (356835)
remove alias feral_dog (0) -> domestic_dog (356835)
remove alias anthro_dog (0) -> domestic_dog (356835)
remove alias pupper (0) -> domestic_dog (356835)
remove alias pup (0) -> domestic_dog (356835)
remove alias doggo (0) -> domestic_dog (356835)
remove alias dogs (0) -> domestic_dog (356835)
remove alias brown_dog (0) -> domestic_dog (356835)
remove alias small_dog (0) -> domestic_dog (356835)
remove alias doggie (0) -> domestic_dog (356835)

Reason: This has been bothering me for a while. Colloquially, "cat" (and kitty) refers to any felid or feline, and "dog" (doggies, doggo, pup, etc) refers to any canid or canine. Having these generic tags aliased to domestic variants specifically has caused unknown amounts of mistags, including lions and tigers tagged as domestic cats and wolves tagged as domestic dogs. This has undoubtedly also contributed to more mistags of fictional cat-like species, like meowscarada, being tagged domestic_cat, and fictional dog-like species, like lucario, being tagged domestic_dog.

This first BUR unaliases them from their domestic tags, and a followup will alias them to felid and canid respectively (or feline and canine, if people think that's better?).

If lions and such are getting mistagged then felid would be ok since big cats aren't in the feline genus, they're in panthera. But for dogs, you might want to go with canine rather than canid. All living canid species are in the caninae family, the only canids that aren't canines are extinct and have very few images on here.

Hmm, I wonder what the dissenters say. To be honest, if someone's tagging a big cat a 'kitty', that's insane, but what would I know? ;)

alphamule said:
Hmm, I wonder what the dissenters say. To be honest, if someone's tagging a big cat a 'kitty', that's insane, but what would I know? ;)

I'm on the fence. If this went through, there would be a huge amount of domestic cats and domestic dogs that just get tagged felid or canid with their actual species tag missing and that would have to be cleaned up. So it depends on whether it would be harder to constantly clean up that situation or to constantly clean up the current situation with lions and tigers getting tagged domestic_cat.

I have no actual evidence either way but I'm guessing this proposal would produce a much larger mess than the current one.

cloudpie said:
If this went through, there would be a huge amount of domestic cats and domestic dogs that just get tagged felid or canid with their actual species tag missing and that would have to be cleaned up.

This happens anyway. domestic_cat and domestic_dog aren't actual species, they're umbrella groups for domesticated cat and dog species. They would still be missing their actual species tag that someone needs to add.

cloudpie said:
So it depends on whether it would be harder to constantly clean up that situation or to constantly clean up the current situation with lions and tigers getting tagged domestic_cat.

It's better to have posts tagged correctly and vaguely, rather than incorrectly and more specific. Especially when that incorrect specific tag may implicate other incorrect tags. This has been the policy as far as I know, that aliases should point to the most correct tag that will result in the fewest mistags, not something more specific that's correct some times and mistags other times. If mistags were acceptable from an alias as long as other posts got more exact tags, there'd be no need to invalidate many tags as they could be aliased to something that's often, but not always, correct; furry, for example, could be aliased to anthro since that's what it often means, but sometimes it doesn't, so even though it could leave some posts without a form tagged, it's better invalidated than to cause posts to be mistagged anthro when it meant something else.

watsit said:
This happens anyway. domestic_cat and domestic_dog aren't actual species, they're umbrella groups for domesticated cat and dog species. They would still be missing their actual species tag that someone needs to add.

They are species: canis familiaris and felis catus. You're thinking of breeds, which are selectively-bred lineages of domesticated animal species. Species and breed are very different things.

watsit said:
It's better to have posts tagged correctly and vaguely, rather than incorrectly and more specific. Especially when that incorrect specific tag may implicate other incorrect tags. This has been the policy as far as I know, that aliases should point to the most correct tag that will result in the fewest mistags, not something more specific that's correct some times and mistags other times. If mistags were acceptable from an alias as long as other posts got more exact tags, there'd be no need to invalidate many tags as they could be aliased to something that's often, but not always, correct; furry, for example, could be aliased to anthro since that's what it often means, but sometimes it doesn't, so even though it could leave some posts without a form tagged, it's better invalidated than to cause posts to be mistagged anthro when it meant something else.

Fair enough I suppose

watsit said:
This happens anyway. domestic_cat and domestic_dog aren't actual species, they're umbrella groups for domesticated cat and dog species. They would still be missing their actual species tag that someone needs to add.

They're actual species. The domestic cat is Felis domesticus, derived from the African wildcat (Felis lybica), and the domestic dog (including a feral breed like the dingo and the New Guinea singing dog) is Canis familiaris, or the subspecies Canis lupus familiaris if going by Wozencraft's phylogeny, derived from the Pleistocene wolf (Canis cf. lupus)* and sister species/subspecies to the modern gray wolf, although there has been some genetic leakage from closely related cat or dog species.
_______________________

• The cf. in the species name is for the Latin word confer, meaning uncertain. Basically, it says that scientists aren't quite sure how this wolf fits in relation to all the other subspecies of the wolf. All they can be sure of is that it's a wolf.

Imo, i think it will probably be useful for most. But i'm too used to just typing cat for the, well, domestic cat alias. I already know what canids and felids are and the distinction. But Gonna be bummed when i need to type domestic over and over if this goes through honestly.

benjiboyo said:
Imo, i think it will probably be useful for most. But i'm too used to just typing cat for the, well, domestic cat alias. I already know what canids and felids are and the distinction. But Gonna be bummed when i need to type domestic over and over if this goes through honestly.

That's the main reason I'm opposing this. Changing the alias like this means we'll have a lot fewer posts with any kind of base-level species, only the higher-level families, and that just sounds so much worse for searching.

We are not a science encyclopedia. If Wikipedia can get away with this, so can we.

strikerman said:
Changing the alias like this means we'll have a lot fewer posts with any kind of base-level species

But also a lot less mistags. It makes no sense to have aliases that knowingly and actively contributes heavily to mistags that require manual cleanup. Tagging convenience from an alias shouldn't come at the cost of this much improper tagging. Any other alias that leads to this many mistags would be considered bad.

watsit said:
But also a lot less mistags. It makes no sense to have aliases that knowingly and actively contributes heavily to mistags that require manual cleanup. Tagging convenience from an alias shouldn't come at the cost of this much improper tagging. Any other alias that leads to this many mistags would be considered bad.

Is it a lot of mistags? From the search you linked, about 300 posts containing lions and/or tigers had an improper domestic_cat tag. The cat -> domestic_cat alias was implemented in 2019. In that time, over 40,000 posts with just a tiger were posted.

Less than 0.01% of posts uploaded are mistagged. I'd rather not throw the baby out with the bathwater.

strikerman said:
Is it a lot of mistags? From the search you linked, about 300 posts containing lions and/or tigers had an improper domestic_cat tag.

That was also a very restrictive search, of only solo images of a lion or tiger. There would be more if we included duo or group images, but they'd take more time to actually single out cases where the lion or tiger was mistagged instead of a domestic cat being in the picture with them. It also doesn't include various panthers, cat-like pokemon, nondescript felines, etc. But even still, just because it may have a low error count now doesn't mean it's not a problem. It doesn't account for all the times it's been removed already.

And the wolf search, which was also limited to solo wolves (so not including multi-character images, or jackals, coyotes, dog-like pokemon, nondescript canines, and other "dogs"), has 40 pages of results, about 10 times more.

watsit said:
That was also a very restrictive search, of only solo images of a lion or tiger. There would be more if we included duo or group images, but they'd take more time to actually single out cases where the lion or tiger was mistagged instead of a domestic cat being in the picture with them. It also doesn't include various panthers, cat-like pokemon, nondescript felines, etc. But even still, just because it may have a low error count now doesn't mean it's not a problem. It doesn't account for all the times it's been removed already.

And the wolf search, which was also limited to solo wolves (so not including multi-character images, or jackals, coyotes, dog-like pokemon, nondescript canines, and other "dogs"), has 40 pages of results, about 10 times more.

26 pages if you exclude hybrid, because I suspect Balto (wolf/dog hybrid) is giving you false positives there. I'm not denying there isn't an issue, I'd just rather have the tagging issue than the searching issue. If people see the occasional Sprigatito when they search for "cat", so be it (and that can be quickly remedied). Better that than any kind of felid.

benjiboyo said:
Imo, i think it will probably be useful for most. But i'm too used to just typing cat for the, well, domestic cat alias. I already know what canids and felids are and the distinction. But Gonna be bummed when i need to type domestic over and over if this goes through honestly.

e6 should suggest the terms when I type them in the search field, I don't how feasible that is, but itwould fix the problem.

strikerman said:
I'm not denying there isn't an issue, I'd just rather have the tagging issue than the searching issue.

You still get searching issues; either you have fewer critters under more specific tags because of taggers not being specific in the first place (you can tag the actual breed of domestic cat which will imply domestic_cat, instead of just tagging cat for a shortcut to domestic_cat and leaving the breed untagged), or you end up with critters under more specific tags that they shouldn't be under. In the former case, someone still needs to go around to fill in the particular breeds anyway, while with the latter you have a bunch of mistags on top of missing breed tags that to be cleaned. I consider it lucky to go a few days without having to remove domestic_dog from posts of lucario, for example (I wouldn't even be surprised if Lucario is mistagged as domestic_dog more than jackal, the thing it's actually based on, because people use dog for any dog-like creature and not domestic varieties specifically).

In either case, the site policy has always been to have aliases target something that will always be correct and avoid mistags. Aliasing a vague and generic tag to something more specific never makes sense. After all, the tags are domestic_cat and domestic_dog, implying the existing of non-domestic varieties of cats and dogs, and indeed people do use the terms cat and dog in everyday speech for non-domestic types of felids and canids, like lions, panthers, wolves, coyotes, etc. If this wasn't the case, many of the Invalid tags should be re-validated and aliased to something that's close enough much of the time, and many current aliases can be made more specific, at the cost of some extra mistags that people can "easily" clean.

And if you really want such conveniences, doesn't re621 have a feature for custom tag aliases? That way you can set up your own three-letter alias for domestic_cat, without the site having to make one that invariably causes mistags because peoples' use of these words doesn't align with how the site has them aliased.