Seeing all the threads around lately, I think standardizing the useful ones would be a good idea. So here's a single thread for them. Mostly just a few new ones and some restructuring atm. I'm thinking new ones can be added and discussed here.
If we want to keep them consistent:

• Keeping them alone will cause problems with actual numbers and letters like 3, 9, a, etc.
• Suffixing them with _sign could create trouble with tags like stop_sign.
• Prefixing with hand_ or asl_ could be a solution.
• As is of course only pre-/suffixing collisions, if consistency is not so required. This may require changes later.

Format is "proposed_tag: example post (potential aliases or alternative names) (implications)"
I am using a priority system where sufficiently universal/very common names are used first, followed by ASL standard signs, followed by other names, followed by custom danbooru names.
I'd say the focus is on hand poses, not necessarily whether that exact sign was meant or not, however it should be used only for obvious signs or hand poses that look sufficiently off in the context of what they are doing (post #949683, post #891070), not for casual poses (post #1029367, post #1023939).

ASL

• asl_v: post #1051092 (asl_2, v)
• asl_w: post #1003227 (asl_6)
• asl_u: post #962868 - girl in back
• asl_0: post #1028975 - 2 right-most girls (asl_o)
• asl_1: post #1016495, post #1028975 - silver-haired girl right (asl_d) (index_finger_raised)
• Often associated: pointing
• index_finger_raised could just be the main tag, but it sounds like it could apply in multiple scenarios, not only as asl_1, and seems to be used that way too, so I stuck with an implication for now.
• Same with asl_i and pinky_out. I'm even less sure about this one's usefulness.
• asl_3: post #1050010
• asl_8: post #998211
• asl_ily: post #1042210 (\m/, ily)

Named Signs

• corna: post #1042312, post #299661 (\n/)
• finger_gun: post #934761 (asl_l)
• Often associated: pointing
• (i'm_sure_this_has_a_name): post #1000220
• inward_v: post #1040340 (backhand_v) (v)
• wanker: post #1048207

Stances/Techniques/Ritual

• I'd like to tag things like post #908063, post #1049460 and similar combat-related or ritualistic signs.

Danbooru Signs

• open_ily: post #841285

Derivatives

• double_hand_sign: Added when both hands make the same sign. post #998125
• hand_sign_over_eye: Added when a sign is made over the eye. post #1042444, post #1029822
• hand_sign/hand_gesture: Umbrella tag and default for when it doesn't fit under any of the specific ones.

Related:

• Alias posing <-> pose
• beckon
• come hither
• hand_spelling: post #924746
• hand_letter: It differs in that letters are recreated, instead of signs representing them.
• tag group:gestures

Looks great.

+1

I honestly don't feel the necessity to try and standardize these tags. Tying the naming to a sign language honestly seems to give the depictions a meaning that isn't normally going to be there. In general, the current naming scheme has not really assigned a meaning but tried to define the shape. Some of these current text icons do a better job covering the concept without assigning meaning in my opinion. It's much like how Danielx21 proposed offensive_v for the palm inward v hand gesture, even if the image isn't using it in that manner, the tag name alone is assigning meaning that isn't there.

I think it is good enough that the tags can be placed under the tag group:gestures to list out what tags are available. Having to standardize the name to indicate that they're a gesture seems a lot less important when you can simply bring up a list that says they're a gesture to begin with.

I've actually suggested migrating the current \m/ to ily before, and then converting \m/ into a more general tag the would cover all similar images that involve a hand sign with the pinky and thumb extended. Given that there are several similar hand signs that share this appearance but slight alterations I think using \m/ to house them would fine with tags like ily implicating \m/. A tag like corna would be fine, but the current images under \n/ would first, under my scheme, need to be migrated to the more general \m/, because not all images under \n/ are necessarily the horns signs hand gesture. That tag really exists more because they're images that have the extended pinky and thumb but don't fall under the current ily \m/ tag.

For hand_sign_over_eye, I think you'd need to provide examples to justify its existence, because if the images are all just v signs or similar over the eye then the current v_over_eye tag is already more than enough.

double_hand_sign seems like something that would be useful, so I would support a tag like this.

My reasoning with ASL was that it's a standard we can compare gestures to, ignoring their meaning. Same with common western namings.
I thought at least trying to standardize part of it would be better than the current way of wrapping half-random descriptions around it (like middle_w, \m/), and having tags like a bring up instances of the letter 'a' while v brings up hand signs. It seems somewhat inconsistent. Especially if any of these grow, and it seems to me a lot of hand signs aren't tagged.
See also the confusion some users see with v vs w, and \n/ vs \m/.

Having \m/ or something be a "grouping" tag could be nice for wider searches, but I thought users may want to look for more specific gestures too. As you can see, I've already tried to give the tags some leeway, grouping very similar ones.
Related; I thought about context-adaptive tags too (potentially with aliases or implications), but I'm not sure how I feel about them (leaning towards "against"). Speaking of which though, asl_l could really join asl_1 and asl_i in having the finger_gun tag imply it, rather than being an alias of it.
Overall, I think it's a bit of a balancing act regarding number of tags, subsets, variations, etc.

I agree about hand_sign_over_eye, but I assumed there may be more signs used than v, just they're not tagged. Having a quick look now, I can find a few posts: post #385583, post #287141, post #1042210.

Anyway, I don't entirely like how it's laid out yet either, but that's why it's here; just getting feedback on the idea and hopefully improving it. (or indeed getting rid of it)