As many of you probably know, we have kept a rather strict standing on how to tag anatomically_correct for many years appropriately. I'm sure we've all both come across posts where our opinions say, "This should be anatomically correct" or "This species doesn't even exist," depending on the image. Currently, we updated the wiki to work a little closer towards a "TWYS" standpoint, to give a little more wiggle room for fake species, but it's possible that we could have a better solution.
Currently, I have a proposition and would love to hear your input.

Create a separate anatomical tag for fake species only. "Anatomically_Appropriate" or "Anantomically_Probable" for example.

I feel like something similar to on_model would be appropriate. If you draw a sergal with a prehensile clitoral hood, that's on-model for the species. What it would be called to differentiate itself from on_model though, I'm not quite sure.

It all seems wishy-washy to me, and open to vague and subjective interpretations. If a species doesn't exist, we can't at all say what genitals it should have. Particularly if a species doesn't have defined genitalia, like pokemon. Given that pokemon are all (except one) are egg-laying, couldn't one say cloacas are "appropriate" or "probable" for them as with most other egg-laying species? How about species that can't breed and may not have genitals? It just comes down to opinions.

The only way I could see tagging expanding in this way is if we follow what a fictional species can be tagged as. A rockruff can be tagged canine (despite not being allowed to implicate it), so one could say a canine_penis or canine_pussy would be "appropriate". While a blaziken can be tagged avian, but there's no consistent avian genitalia so there's nothing that would be considered "appropriate" or "probable".

Though this too can get confusing. What of something like vaporeon, that's a hybrid marine/mammal creature? Eevee and the eeveelutions in particular can be drawn in ways to emphasize being more feline or canine (or even lagomorph), so would "appropriate" genitalia depend on how they're drawn (more dog-like eevee + canine_penis = appropriate)? Where that would open the door to further arguments over whether a given depiction is or isn't "close enough" for the type of genitalia given. Or since eevee's designed as an amalgamation of mammals, no one thing can be appropriate/probable regardless of how they look in a given image? Which would limit growth and open the door to arguments about what the original design for a species is actually based on.

And given this is fiction, there's nothing to say these species are static. Alternate designs can be added later that makes them more like other animals than they originally were designed after, while still being the same species, calling to question existing appropriate/probable anatomy for it.

moonlit-comet said:
I feel like something similar to on_model would be appropriate. If you draw a sergal with a prehensile clitoral hood, that's on-model for the species. What it would be called to differentiate itself from on_model though, I'm not quite sure.

That's a bit of a different situation, as sergals were designed with this anatomy. What I'm referring to are situations of art imitating life, things like pokemon and palworld. Animals designed to mimic something real.

versperus said:
That's a bit of a different situation, as sergals were designed with this anatomy. What I'm referring to are situations of art imitating life, things like pokemon and palworld. Animals designed to mimic something real.

Oh, I must have misunderstood.
"Appropriate" might be good if it's on something like a fictional canine with canine-like anatomy, but I agree with Watsit in the sense that "anatomical accuracy" is subjective when it comes to fictional species based on real animals. There's not really a perfect solution to be had here because of the subjectivity.

watsit said:
It all seems wishy-washy to me, and open to vague and subjective interpretations. If a species doesn't exist, we can't at all say what genitals it should have. Particularly if a species doesn't have defined genitalia, like pokemon. Given that pokemon are all (except one) are egg-laying, couldn't one say cloacas are "appropriate" or "probable" for them as with most other egg-laying species? How about species that can't breed and may not have genitals? It just comes down to opinions.

The only way I could see tagging expanding in this way is if we follow what a fictional species can be tagged as. A rockruff can be tagged canine (despite not being allowed to implicate it), so one could say a canine_penis or canine_pussy would be "appropriate". While a blaziken can be tagged avian, but there's no consistent avian genitalia so there's nothing that would be considered "appropriate" or "probable".

Though this too can get confusing. What of something like vaporeon, that's a hybrid marine/mammal creature? Eevee and the eeveelutions in particular can be drawn in ways to emphasize being more feline or canine (or even lagomorph), so would "appropriate" genitalia depend on how they're drawn (more dog-like eevee + canine_penis = appropriate), or since it's designed as an amalgamation of mammals, no one thing can be appropriate/probable regardless of how they look in a given image?

Well, that's the thing. Having a separate tag for it would give more le-way for people to use that "how this is depicted would make sense" approach. We wouldn't have to be heavy-handed on its usage, and it would keep this interpretation bias out of the population of the real-world counterpart. It would let users find reasonable examples of fake species depicted in a way that would make sense and let those who only want what's real not worry about getting their results diluted by things that don't exist.

versperus said:
Well, that's the thing. Having a separate tag for it would give more le-way for people to use that "how this is depicted would make sense" approach. We wouldn't have to be heavy-handed on its usage, and it would keep this interpretation bias out of the population of the real-world counterpart. It would let users find reasonable examples of fake species depicted in a way that would make sense and let those who only want what's real not worry about getting their results diluted by things that don't exist.

I understand it would be a separate tag, but there's the underlying problem of what would be considered "appropriate" or "probable" anatomy for a non-existent species. It still needs some way to define what it can and can't apply to, which comes with a number of difficult questions if the tag isn't to turn into a mess of inconsistent usages and be the source of more tagging wars (and honestly, even if those questions can be answered, I'm not sure I have enough faith in uploaders not using it as meaning "non-human genitalia on a fictional animal", as anatomically_correct often is already).

watsit said:
I understand it would be a separate tag, but there's the underlying problem of what would be considered "appropriate" or "probable" anatomy for a non-existent species. It still needs some way to define what it can and can't apply to, which comes with a number of difficult questions if the tag isn't to turn into a mess of inconsistent usages and be the source of more tagging wars (and honestly, even if those questions can be answered, I'm not sure I have enough faith in uploaders not using it as meaning "non-human genitalia on a fictional animal", as anatomically_correct often is already).

It would follow the same sort of guideline we're currently using that lets fake species in. "TWYS," has to have a reasonable IRL counterpart it's depicting. It can't be something like "Well, gengar is vaguely cat-shaped, so I'll give it a cat dong and call it anatomical" Things that really don't look like anything shouldn't be applicable.
We'd have to have an in-depth wiki page built regarding it of course, and it also comes with the question of "does the community want this?"

versperus said:
It would follow the same sort of guideline we're currently using that lets fake species in. "TWYS," has to have a reasonable IRL counterpart it's depicting. It can't be something like "Well, gengar is vaguely cat-shaped, so I'll give it a cat dong and call it anatomical" Things that really don't look like anything shouldn't be applicable.

What if it's "I'll draw a gengar to look a bit more cat-like and give it a cat dong", would that be anatomical? Or the question I had regarding Eevee and the eeveelutions:

Eevee and the eeveelutions in particular can be drawn in ways to emphasize being more feline or canine (or even lagomorph), so would "appropriate" genitalia depend on how they're drawn (more dog-like eevee + canine_penis = appropriate)? Where that would open the door to further arguments over whether a given depiction is or isn't "close enough" for the type of genitalia given. Or since eevee's designed as an amalgamation of mammals, no one thing can be appropriate/probable regardless of how they look in a given image? Which would limit growth and open the door to arguments about what the original design for a species is actually based on.

Does it depend on how they're depicted in a given image, or the species' canonical design?

versperus said:
We'd have to have an in-depth wiki page built regarding it of course, and it also comes with the question of "does the community want this?"

I can't say I'm fond of the idea, it just opens up way too many questions and leaves too much up to subjective interpretation. Even the recent wiki edits for anatomically_correct have me scratching my head.

Fictional species that greatly resemble a real-life species such that they can be tagged as a member of that species, however, can be anatomically correct

What is meant by "greatly resemble"? Sure, houndoom is an easy example, but is wigglytuff enough of a lagomorph to be considered anatomically correct with lagomorph genitals? What about quaquaval? Ho-oh? Incineroar? Lucario? What about a dragon that's made to resemble a komodo_dragon and has matching genitalia? How can we determine whether a given fictional species is close enough to some real species to warrant being tagged anatomically correct? Simple examples are easy, it's the difficult examples that are important.

watsit said:
What if it's "I'll draw a gengar to look a bit more cat-like and give it a cat dong", would that be anatomical? Or the question I had regarding Eevee and the eeveelutions:
Does it depend on how they're depicted in a given image, or the species' canonical design?

I can't say I'm fond of the idea, it just opens up way too many questions and leaves too much up to subjective interpretation. Even the recent wiki edits for anatomically_correct have me scratching my head.
What is meant by "greatly resemble"? Sure, houndoom is an easy example, but is wigglytuff enough of a lagomorph to be considered anatomically correct with lagomorph genitals? What about quaquaval? Ho-oh? Incineroar? Lucario? How can we know whether a given pokemon is close enough to some real-life species to warrant being tagged anatomically correct? Simple examples are easy, it's the difficult examples that are important.

Well, yeah, if someone draws a particular image to follow the profile of something real, it's viable. Like I said, if we do go forward with something like this, it should entail a little leniency and be taken with a TWYS approach. Look at the image like you've never seen what the person drew before; things like Eevee don't *have* to be hard set to being tagged with cat or dog anatomy. If you can understand why someone might think it's appropriate for that particular image, then it's viable because, as you said, the thing doesn't exist in the first place.

versperus said:
Well, yeah, if someone draws a particular image to follow the profile of something real, it's viable. Like I said, if we do go forward with something like this, it should entail a little leniency and be taken with a TWYS approach. Look at the image like you've never seen what the person drew before; things like Eevee don't *have* to be hard set to being tagged with cat or dog anatomy. If you can understand why someone might think it's appropriate for that particular image, then it's viable because, as you said, the thing doesn't exist in the first place.

So we get to fight over whether renamon looks enough like a fox/canine, or whether a given eevee(lution) looks more canine or feline, or neither, in each individual image they appear in with genitalia. That doesn't sound like a good time.

If it's too loose, I fear such a tag will become a dumping ground for animal genitalia with very loose interpretations of what looks like a matching fictional animal, resulting in it being useless. There's just too much in the air and not enough answers (and what answers there have been make it sound too easy for the tag to become pointless, and/or a source for more tagging wars) for me to say it's a good idea.

versperus said:
Like I said, if we do go forward with something like this

The anatomically_correct wiki is already changed (and locked) to include some fictional species. It's already going forward, it seems.

Giving fictional species certain stamps of anatomical correctness when that anatomical feature has no correct way of being depicted is just asking for trouble and misunderstanding.

It's the classic fallacy of composition trap.
Just because a unicorn looks like a horse and horses have a horse/equine penis does not mean that a unicorn also has a horse/equine penis.

Besides, what about hybrid species like a horse and dog hybrid? What would be anatomically correct in their case?
Canine penis because of the dog part or equine penis because of the horse part, or are both "correct"?

Having anatomical correctness only apply to examples where we could draw from real-world knowledge would cut down the amount of confusion.

And frankly, I don't think fictional species need their version of "anatomical correct" like "anatomically_appropriate" or "anantomically_probable" as that leaves way to much stuff up to interpretation, and will again just lead to confusion.

watsit said:

The anatomically_correct wiki is already changed (and locked) to include some fictional species. It's already going forward, it seems.

Though it does seem there's a couple of disagreements between staff, with the explicit references to fictional species inclusion added by @Donovan_DMC and @OneFattyCatty, and @slyroon's disagreements given above

snpthecat said:
Though it does seem there's a couple of disagreements between staff, with the explicit references to fictional species inclusion added by @Donovan_DMC and @OneFattyCatty, and @slyroon's disagreements given above

It was changed because that's the way it's currently applied and was the majority winner in an internal vote, we were debating giving a record to someone for removing anatomically correct from pokemon. We decided no, and that the wiki needed to be clearer that right now fictional species that are comparable to real species can be anatomically correct

I'm personally mixed in my opinion, but I'd rather the wiki be explicit in how the tag is currently applied so mixups don't happen again before anything is fully set in stone

watsit said:
So we get to fight over whether renamon looks enough like a fox/canine, or whether a given eevee(lution) looks more canine or feline, or neither, in each individual image they appear in with genitalia. That doesn't sound like a good time.

If it's too loose, I fear such a tag will become a dumping ground for animal genitalia with very loose interpretations of what looks like a matching fictional animal, resulting in it being useless. There's just too much in the air and not enough answers (and what answers there have been make it sound too easy for the tag to become pointless, and/or a source for more tagging wars) for me to say it's a good idea.

I personally would apply a fairly strict definition to it so as to limit it to things that are not hybrids in any sense, and actually closely resemble the species it's based on. It would be a niche use tag, and ideally that would limit how much it gets fought over.
Though if I had it my way, I would just include these cases in the original tag without making a separate tag for it, personally. This is still an acceptable solution, but I don't really see a need to separate them. Somebody searching for an anatomically_correct equine is very likely not going to care if it's on a horse or a unicorn. Likewise, an Arbok with hemipenes showing up in the search next to real snakes hardly seems like a problem. You can just filter out pokemon or our beloved new mythological_creature tag if you don't want to see those.

it seems very odd to me to tag a pokémon with anatomically_correct for genitalia when, like, canonically no one seems to understand how pokémon breeding works, and as far as anyone knows, it dosn't even involve sexual contact.

donovan_dmc said:
It was changed because that's the way it's currently applied and was the majority winner in an internal vote, we were debating giving a record to someone for removing anatomically correct from pokemon. We decided no, and that the wiki needed to be clearer that right now fictional species that are comparable to real species can be anatomically correct

I'm personally mixed in my opinion, but I'd rather the wiki be explicit in how the tag is currently applied so mixups don't happen again before anything is fully set in stone

Huh? Anatomically_correct has never applied to fictional species, and there's a number of us that endeavor to remove the tag whenever we see it misused like that. Why is it suddenly now a problem that it needs a rushed definition change? Allowing it to be used this way, without having answers for the number of problems such a change raises or even telling people about it, is going to make things unclear and cause more mixups, especially when a lot of the misuse won't be covered by the new "greatly resemble" clause anyway. It doesn't seem appropriate to make such a significant change to a long-established tag because it needed to be cleaned up from some misuse, and will still need cleaning up anyway.

watsit said:

The anatomically_correct wiki is already changed (and locked) to include some fictional species. It's already going forward, it seems.

Just to clarify. The wiki has been locked for at least 5 years. Well before this discussion began. And the examples were added before the discussion was moved here.

So no need to treat it as a done deal yet. Though currently I appear to be the only staff member that disagree with the fictional species can be anatomically correct, even if they resemble real world animals.

donovan_dmc said:
We were debating giving a record to someone for removing anatomically correct from pokemon.

Well, I'm not gonna win any prizes for guessing who that might be.

Seeing as I'm the one who apparently sparked this whole affair I'm gonna go ahead and repeat what I said in a DM to staff concerning this topic:

Labra said:
Well, the reasoning behind my tag removal choices would be, first and foremost, because that's how I interpreted the wiki. As per the previous definition I didn't expect that it was supposed to mean "non-real species, unless they have enough resemblance to real-life species", seeing as that, for example, could theoretically also include some dragons, because they have some resemblance to monitor lizards. A clear definition of Pokémon = can't be anatomically_correct is a much easier line to draw than having to guess at which point a fictional species is close enough to a real-life one to qualify as anatomically_correct. The fact that other, more established users such as Watsit seemed to follow that same definition made it appear to me that this was the intended use.

As for my personal opinion on the subject... As it stands right now, anatomically_correct is both undertagged and overapplied, meaning that it's frequently omitted when it clearly applies, but at the same time is very often applied to more-or-less fictional species where its use is of questionable utility. Not to mention that anatomically_correct as an umbrella tag is already somewhat prone to misinterpretation, as it gets regularly applied to posts that don't even feature any genitals, because people assume that it means "this character has correct anatomy".
Especially since tags like canine genitalia, equine genitalia, etc. were established last year, anatomically_correct has lost even more of its already diminished purpose. Usually, you'll get more reliable and thorough result when, for example, searching canid + canine genitalia/canine penis/canine pussy than when searching canid + anatomically correct genitalia etc. For me, the optimal use case for anatomically_correct would be "real-life species with their corresponding genitals", which would enable searching for said cases without having to filter out huge amounts of fictional/franchise species. In my opinion, adding fictional species to a tag like anatomically_correct, especially if those species don't have canonically established "correct" genitals, like Pokémon, waters down the potential usability of the tag to a point where it comes very close to redundancy.

slyroon said:
Just because a unicorn looks like a horse and horses have a horse/equine penis does not mean that a unicorn also has a horse/equine penis.

And I would tend to agree with that, but the whole section of (mostly MLP-related) unicorns/pegasi tagged as anatomically_correct is a huge can of worms that I'm now very glad I haven't opened so far.

Eh, there's a lot of grey area, but at the same time I feel like a lot of fictional species can be portrayed in ways that fit within the bounds of modifications you can make to a real animal, through style or addition of fantasy parts, and still have it be taggable.
For someone who is looking for (or trying to blacklist) anatomically correct canines are they really going to be upset that this
post #462182
shows up alongside these?
post #899248 post #4546290

regsmutt said:
Eh, there's a lot of grey area, but at the same time I feel like a lot of fictional species can be portrayed in ways that fit within the bounds of modifications you can make to a real animal, through style or addition of fantasy parts, and still have it be taggable.
For someone who is looking for (or trying to blacklist) anatomically correct canines are they really going to be upset that this
post #462182
shows up alongside these?
post #899248 post #4546290

Perhaps not, but it's a slippery slope without clearly defined boundaries. They are, ultimately, fictional species, and as the wiki says, "It can't be anatomically_correct if you cannot define which genitalia is even supposed to be correct in the first place." Most fictional species don't have defined genitalia, and they're owned by companies that market them to kids and would rather not delve into such a topic. Pokemon in particular have the unique twist in that they're hatched from eggs, suggesting non-mammalian anatomy. It just has a lot of unknowns that requires a lot of assumptions. On top of that, for fictional species that do have defined genitalia that may not match the species they're otherwise based on, creates extra confusion (e.g. a fictional feline species is explicitly defined to have humanoid genitalia, would that mean that particular fictional feline species with feline genitalia would not be anatomically correct? or would it still apply despite not being lore-correct?).

Given also the suggestion that it would depend on how a given image depicts the species (where you could assume wolf + canine_penis = anatomically_correct always, you can't assume renamon + canine_pussy = anatomically_correct always), puts it in a deeply grey area where there's a lot room for disagreement with sensible arguments on both sides. There's no way for us to determine whether a given depiction of some species "greatly resembles" a real species, or merely somewhat resembles or kind of resembles. It just creates a lot of uncertainty without clear guidelines for how to work with it, and puts a large burden on taggers to figure out how to handle it correctly.

watsit said:
Given also the suggestion that it would depend on how a given image depicts the species (where you could assume wolf + canine_penis = anatomically_correct always, you can't assume renamon + canine_pussy = anatomically_correct always), puts it in a deeply grey area where there's a lot room for disagreement with sensible arguments on both sides. There's no way for us to determine whether a given depiction of some species "greatly resembles" a real species, or merely somewhat resembles or kind of resembles. It just creates a lot of uncertainty without clear guidelines for how to work with it, and puts a large burden on taggers to figure out how to handle it correctly.

I don't think that's a great argument. If you're going to argue that there's no way to determine if a drawing of a fantasy creature looks like a horse it kind of defeats the whole "tag what you see" deal. How do you determine if a picture depicts an equine penis at all? Maybe it's a long humanoid penis and the head was drawn weird. How do you determine if a picture really shows a character with boobs and a dick versus chest tumors and a tentacle?
At some point there's an assumption that yes, people know what things look like. You can say that it needs to retain recognizable features and body proportions if you need to.

Also I think of it in terms of utility for searching and blacklisting. If someone has blacklisted 'feral canine anatomically_correct' and does not care if they see dog dicks on anthros or feral dragons, would they want these images to slip through their blacklist?
post #462182 post #2345356 post #2531401

Also we don't take lore into account for young. If it looks young, it's tagged young. It doesn't matter that a fantasy species lives forever and is physically and mentally mature if it has childlike proportions. I don't think giving fictional species exemptions to twys is a good idea.

regsmutt said:
At some point there's an assumption that yes, people know what things look like. You can say that it needs to retain recognizable features and body proportions if you need to.

Given how loosely species can be tagged, applying this same standard to fictional species would basically remove all restrictions on tagging fictional species as "anatomically correct". What one person "knows" a thing looks like, another person can "know" it doesn't, and another can be unsure of. Everyone's going to have a different line for whether a given depiction "greatly resembles" a real species or not. People are already very loose with their application of anatomically_correct, tagging it on all sorts of fictional creatures that don't have humanoid genitalia. I've even seen it tagged on dinosaurs. Extending anatomically_correct to species that don't have defined genitalia, and having it apply on a per-image basis instead of being based on species whose anatomy is in question, is just ripe for problems.

regsmutt said:
Also I think of it in terms of utility for searching and blacklisting. If someone has blacklisted 'feral canine anatomically_correct' and does not care if they see dog dicks on anthros or feral dragons, would they want these images to slip through their blacklist?
post #462182 post #2345356 post #2531401

If they object to feral canines with canine genitalia, 'feral canine canine_genitalia' would be a better search. anatomically_correct implies anatomy that is correct for the species, which depends on knowing the correct anatomy for a given species. Would they care about images like this getting through that blacklist?
post #4190926 post #4680027
Hyenas don't have canine pussies like that. Hyenas can resemble canids, particularly in art, but they are not canids. Mightyena, as its name suggests, is based on a hyena, but can be drawn to look like a canid-like hyena or hyena-like canid.

This all requires some concessions and compromises. And I think it's fair to say: it's okay to tag fictional species as canine/feline/etc if they look like that, but anatomically_correct relies on being a species with defined anatomy (be it real life or at least a canonical in-universe definition). Its okay to apply it to anthros of an otherwise suitable species. The combination of canine/feline/cervid/etc along with canine_genitalia/feline_genitalia/cervid_genitalia/etc is sufficient to cover fictional species with genitalia that matches what they look like, as we don't otherwise know what is "correct" for them.

Incidentally, I believe 'x anthro with x genitalia' was the motivating factor for the tag. As animal genitalia became more prevalent on anthros in furry art, people would call a wolf anthro with wolf genitals "anatomically correct" to distinguish it from human genitals, for example (despite their anthropomorphism suggesting human genitals could be just as correct). It also applying to ferals seems more of a byproduct, as ferals typically have matching genitalia much more commonly. One could make an argument it shouldn't be applied to ferals at all, similar to tags like nude, or all_fours for quadrupeds. For species where anatomical correctness can be determined, it almost always applies to ferals which makes it redundant (and we have tags like humanoid_genitalia_on_feral and mismatched_genitalia to cover the opposite). FWIW, the x_genitalia tags were a relatively recent introduction (being given implications within the past two years), which would have made finding 'x-like thing with x genitalia' more difficult in the past. And some of us continue to argue we should allow fictional species to imply an appropriate taxonomic class for what they are designed after and resemble, stuff they can already be tagged as, which would help plug holes in such searches.

regsmutt said:
Also we don't take lore into account for young. If it looks young, it's tagged young. It doesn't matter that a fantasy species lives forever and is physically and mentally mature if it has childlike proportions. I don't think giving fictional species exemptions to twys is a good idea.

Species aren't wholly TWYS, as quality, detail, and general anatomy aren't always the forefront of an artist's work. Some leeway is given for 'if they say it's X species, it can be tagged as X species as long as it's not apparently wrong'. As it is, anatomically_correct is already pushing the boundaries of TWYS since it implies animal junk is correct for anthropomorphic animals while humanoid junk is not, despite them being animal/human hybrids and either being equally plausible should they exist in real life. The tag has always been for '<species> with <species> genitalia (excluding humans)', not '<species>-like animal with <species> genitalia', as that would open up a large can of worms with many sharp edges and pitfalls that don't have an easy way to be dealt with.

I think there's a case for this idea.
Let's call it on_specification, or on_spec, similar to on_model.
It will be applicable to fictional species with official guidelines (for example, wickerbeast , or avali ).
Because if someone draws them according to the guidelines, then there should be a tag for that.

watsit said:
...And given this is fiction, there's nothing to say these species are static. Alternate designs can be added later that makes them more like other animals than they originally were designed after, while still being the same species, calling to question existing appropriate/probable anatomy for it.

That's why it would be a separate tag. If you're really worried about that, it could be whitelisted.

waydence said:
I think there's a case for this idea.
Let's call it on_specification, or on_spec, similar to on_model.
It will be applicable to fictional species with official guidelines (for example, wickerbeast , or avali ).
Because if someone draws them according to the guidelines, then there should be a tag for that.
That's why it would be a separate tag. If you're really worried about that, it could be whitelisted.

The problem with calling it on specification is that it implies there is an established spec for things that by design, do not have genitals.

versperus said:
The problem with calling it on specification is that it implies there is an established spec for things that by design, do not have genitals.

Not really sure I understand you here.

If species lack something by design - let's say pupils - then artwork with pupils does not qualify for the tag.
If something is not mentioned in the design doc, then it's assumed to be up to artist's interpretation, and can be ignored when evaluating for the tag.