Should "mother-son" get editted to "mother_and_son", to fit with the other relative-set tags? Are the any others like this?

These tags are a bad idea in general, but if it's going to exist it should at least be consistent, so yes I mass edited this.

Why they are a bad idea? Just asking out of curiosity.

Thread is a bit meandering, but there was some discussion at forum #11092. They're only really useful if used to describe sexual pairings. Used in other contexts, they're pretty much just spoiler-bait.

It's useless even for sexual pairings, since it's not something visible, and we reject anecdotal tags for good reasons. I'd vote for removing them completely, same with incest, siblings, and whatever other fluff we have that I'm forgetting at the moment.

I'm probably not as sensitive to spoilers as others, and don't have as much as much a problem with these tags, but I understand the issue.

How should we deal with things like twins or family though where it's an observable feature (such as two characters and an "if-they-mated" child), or something like that. In that case I think the tags would be meaningful, and should probably be kept.

葉月 said:
It's useless even for sexual pairings, since it's not something visible, and we reject anecdotal tags for good reasons. I'd vote for removing them completely, same with incest, siblings, and whatever other fluff we have that I'm forgetting at the moment.

You can't remove incest because there are many images that refer to it explicitly (have a look at Nounanka's stuff, for instance).

葉月 said: It's useless even for sexual pairings, since it's not something visible, and we reject anecdotal tags for good reasons.

I more or less agree, to be honest.

I'd vote for removing them completely, same with incest, siblings, and whatever other fluff we have that I'm forgetting at the moment.

But, for example, fujibayashi_ryou and fujibayashi_kyou getting it on is unquestionably incest. It's not anecdotal. For original works, this is certainly a lot less clear though, and an adult woman having sex with a young boy is much more clearly shota than incest, lacking any sort of external or textual confirmation.

LaC said:
You can't remove incest because there are many images that refer to it explicitly (have a look at Nounanka's stuff, for instance).

I'm not convinced that writing "HERE BE INCEST" on a pic makes the tag useful in any way. The problem is that you can't draw incest. You can say it is, but it doesn't change the pic in any way. It's like that russian incest porn, it's only incest in the filename, nothing changes inside, and no-one sane believes they're actually related.

jxh2154 said:
I more or less agree, to be honest.

But, for example, fujibayashi_ryou and fujibayashi_kyou getting it on is unquestionably incest. It's not anecdotal. For original works, this is certainly a lot less clear though, and an adult woman having sex with a young boy is much more clearly shota than incest, lacking any sort of external or textual confirmation.

As I said above, you can't really draw incest. That makes it, IMHO, a useless thing to mark and not something that's useful to search for because, in the end, we'll have no pictures of incest. Textual descriptions, perhaps, but the pics themselves will be shota, yuri or just sex. Thus it is my opinion that having the tag is useless.

And in any case, I definitely want to see siblings (but not twins, as that's something easily visible) killed. It serves absolutely no purpose besides inviting spoilers.

葉月 said: As I said above, you can't really draw incest.

...

And in any case, I definitely want to see siblings killed

I'm not entirely opposed to this, but I think what distinguishes incest from siblings for me is that the former implies an action, the latter is merely a passive relationship, a state of being. I agree with deleting siblings.

(but not twins, as that's something easily visible)

We would have to explicitly restrict this to identical twins, then, or else it'd just be a subset of siblings, which we propose removing. Right now the wiki description is broader than that.

葉月 said:
I'm not convinced that writing "HERE BE INCEST" on a pic makes the tag useful in any way. The problem is that you can't draw incest. You can say it is, but it doesn't change the pic in any way.

Well, you can draw a character saying あけおめことよろ, but there's no evidence that the scene actually takes place on January 1. The picture itself might have been drawn on another date! Maybe the artist drew it as far back as November in order to have it ready to post when he went partying.

For established characters it's hardly anecdotal what their relationship is. It'd be like saying the Azumanga Daioh cast isn't necessarily in the same class, just because Azuma drew them in the same schoolgirl outfits and wrote a bunch of words around them that suggested they were talking to each other...

For original characters you have more of a point, but I still don't think it should be removed entirely. It may be more guesswork and I would err on the side of caution, but if the artist says they're siblings, then (unless the image is of a dinosaur and a tomato) I'm sort of inclined to take the author's word for it. And on the subject of the incest tag... Regardless of how I feel about incestuous relationships (I find them revolting and can't imagine anyone enjoying anything about them), I'd still want to keep it. These tags are useful for people who are looking for that sort of thing, and that's what tags are for, right?

Nuking them just seems puerile to me.

jxh2154 said:
I'm not entirely opposed to this, but I think what distinguishes incest from siblings for me is that the former implies an action, the latter is merely a passive relationship, a state of being. I agree with deleting siblings.

We would have to explicitly restrict this to identical twins, then, or else it'd just be a subset of siblings, which we propose removing. Right now the wiki description is broader than that.

Yes. Only visible twinness is tag-worthy. Otherwise it's no different from 800 years old lolis, and we already have established policies on tagging those.

スラッシュ said:
For established characters it's hardly anecdotal what their relationship is. It'd be like saying the Azumanga Daioh cast isn't necessarily in the same class, just because Azuma drew them in the same schoolgirl outfits and wrote a bunch of words around them that suggested they were talking to each other...

Anecdotal meant as "based on something said outside and not anything apparent from the pic" here. Besides, being classmates is a visible feature, especially in Japanese schools, where they have uniforms. Being siblings isn't, really (unless they try hard and make them look all but identical, but then it's probably the case of twins, like with fujibayashi_kyou and fujibayashi_ryou, or hiiragi_kagami and hiiragi_tsukasa).

For original characters you have more of a point, but I still don't think it should be removed entirely. It may be more guesswork and I would err on the side of caution, but if the artist says they're siblings, then (unless the image is of a dinosaur and a tomato) I'm sort of inclined to take the author's word for it.

I'm not saying we should disbelieve the authors. But even if we take their word for it, what useful information has it introduced? You don't look at a picture and say "damn it, that would be a totally awesome picture if only they were siblings", now do you? What I'm claiming is that, while the information might be completely true, it's also useless, and in many cases harmful, if being relatives is in any way significant to the original's plot.

And on the subject of the incest tag... Regardless of how I feel about incestuous relationships (I find them revolting and can't imagine anyone enjoying anything about them), I'd still want to keep it. These tags are useful for people who are looking for that sort of thing, and that's what tags are for, right?

Except that the tags will be the only thing to make them incestuous (that or letters on the pic itself, but I regard that as a form of tagging, kind of hard_translated for tags). The pics themselves carry none of the information. You can look at an untagged pic of a character and say "I see this is X depicted". Or at a set of untagged pics of an unknown character and conclude that they all depict the same person. But you can stare all you want at a supposed incest pic, and all you'll be able to infer is that there's a couple.

Likewise, when you see a picture of Cirno, you think: "This is a little girl with blue hair. She is saying this is her year. Why is she saying that? Maybe this year she's going to try hard to accomplish her goals. That could be it. It is strange that the digit nine is circled, though. I wonder what that might signify, if anything".

Also: seiyuu_joke.

葉月 said:
You don't look at a picture and say "damn it, that would be a totally awesome picture if only they were siblings", now do you?

Actually, I do look at things like post #338454 and think "Sisters kissing? That's hot." There were some other comments to that effect on that post, so I don't think I'm the only one here.

You also seem to be saying (correct me if I'm wrong) that dialogue doesn't really make a picture incest, since words don't visibly change anything about the picture. I disagree here. A picture's dialogue changes the context in which it is viewed. Without the text, post #356087 would just be Tewi blushing, but with the text she suddenly becomes tsundere. Without the dialogue, post #246051 is just shota, but with the dialogue it's incest.

葉月 said:
(unless they try hard and make them look all but identical, but then it's probably the case of twins, like with fujibayashi_kyou and fujibayashi_ryou, or hiiragi_kagami and hiiragi_tsukasa).

I'm going to pull a dick move here and mention that hiiragi_kagami and hiiragi_tsukasa are not identical twins. In fact, if they didn't have the same hair color, someone seeing them for the first time would probably not even think them related.

Does this show that you can have similarity enough to imply that people are siblings? I agree that without foreknowledge of the characters, it is hard to claim them to be related or not. Such as the cast of Rozen Maiden, they're all "sisters", yet Suiseiseki and Souseiseki are the only ones with any clear resemblance.

Suiseiseki said: In fact, if they didn't have the same hair color, someone seeing them for the first time would probably not even think them related.

Uh, I'm not sure about that. And you can't just dismiss hair in anime, it's like 90% of what distinguishes one character from another.

Copyright/character tags and tags like trap don't fit in a scheme of tagging only by appearance either.

I think flexible tags are the main asset of this site. I say we keep the tag system as what it is now, a best-effort method of searching and adding metadata, and stop talking about mass removal of tags every other day.

In my opinion, the only worthwile restriction is "no spoiling tags".

Besides, then you have the people here that are so dis-aligned with incest, even among fake characters, that they might have such pics blacklisted. Heck, I've seen comments that've said "would be hot if it wasn't incest", showing the non-flexibility of a mindset.

I think tags like incest, seiyuu_joke, trap do have a purpose. While on the whole I am pretty strict about the "keep to visual elements" concept, I like to think of these tags as an intermediate step between purely visual and metatags.

Example: post #350527

width:800 rating:safe user:jxh2154 - these would describe the upload itself.

school_uniform blue_eyes gradient_hair - these would describe relatively unambiguous visual elements as drawn.

But there are a whole host of tags between the two. They identify the image in ways you can see, but that aren't purely visual or purely metadata.

kushieda_minori amamiya_manami - these identify the individuals depicted. In this case it's obvious, but you can't always rely on that. I've tagged characters I could only identify as such because the artist said so. Artistic license can seriously alter a base design.

toradora gakuen_utopia_manabi_straight - these identify the source from which the individuals were taken. You cannot draw "Toradora". It's closer to rating:safe metadata than school_uniform visual elements.

ttomm - this credits the individual who created the work. This image does not depict ttomm, but it's a valid tag still. One of the most important you can add, in fact.

seiyuu_joke horie_yui - now for the tags that caused this problem. These are neither physical objects visible with the image, nor metadata describing the file, nor tags identifying the source or individuals within the image.

So what are they? They're simply tags that further classify the image, that identify concepts upon which a viewer's experience of the image might rest. seiyuu_joke is not a physical object that can be depicted, it's a concept. Why are Minorin and Manabi together in this image? Because they share a seiyuu, Horie Yui. The joke is the appeal. Why are explicit images drawn of Kagami and Tsukasa getting intimate produced? Because it's incest, and that satisfies a particular fetish.

Incest also classifies an action taking place, just as contemporary classifies a style of clothing relative to the character wearing it, despite itself being a mere concept that cannot be drawn. Deedlit from Lodoss in jeans and a tshirt is contemporary while Taiga from Toradora in the same outfit is not. Kagami and Tsukasa having sex is incest, sex, and yuri, while Tsukasa and Konata having sex is only sex and yuri.

Position and action tags are not as purely visual as they seem either. Think of jumping - are they really? Maybe they're being shot out of a canon, or playing around in zero gravity, or falling out a window. But we're not going to refrain from tagging it "jumping" because of such concerns.

This isn't an argument for giving free reign to tag all possible concepts! Manabi is someone's sister, but you can't tag this image with sister. Minorin is someone's classmate, but we don't tag that either. They're both highschoolers, but we don't tag that either.

Concepts are added after considering forum debates, spoiler concerns, tag spamminess, user tastes, and simple need, and it's impossible to have a blanket rule one way or the other for such a broad category of tagging.