Char said:
This isn't what we're really saying though. Using the same logic above, I wouldn't be able to tag an image of a real person with that person's name because the person's name doesn't appear in the image. To me, this is also like saying that I can't tag "cup" on a picture, and am instead limited to only tagging what makes up the cup (e.g. plastic, cylinder, opening). The difference is that I can SEE that it is a cup. All the evidence for a cup is right there in front of me. I can SEE that the person/character in an image is definitely "person_x", all the evidence is right there. But, I can NOT tell that "character_y" in a specific picture is a herm if I CAN'T see the traits that would make him/her a herm.
I'm confused about this - why doesn't this work in opposite direction, ex. "I see Artica Sparkle, and infer 'herm'"? There's plenty of evidence for Artica Sparkle's character being a herm, even on this site - that is, why can't we (by default) presume that the same character will have the same gender in different images? To me, it's like saying that I see enough to infer that a cup is a mug, yet I can't see the handle, and therefore can only tag "mug" but not "handle"...?
Updated by anonymous