I have something I'vebeen wanting to write for a bit. I'm not sure how far I will get with it, but I want to at least start before going to bed for the night.

     Maybe, as was the case with me in high school, you join a Discord server run by people from the grade below you and, over time, they start to "ironically" post hentai all the time... Well, anyways, I left that Discord server. This anecdote is meant to illustrate that people share nsfw digital art through many channels. While there may be exceptions, it seems like these communication channels incentivize the erosion of context. By context I mean things such as, where the image came from, who or what it depicts, why it was made, when it was made (i.e., commission or monthly poll?), etc.

     Staying with my high school Discord anecdote (but speaking in parallel via a rhyming hypothetical): say, for example, someone posts a picture of a busty and dubiously "of age" girl getting dicked down by Ken from SSB. In general, my experience is that the picture will be posted as just an image (i.e., no link). On the server end, I believe this looks like attaching or copy-pasting an image to Discord, which the site then logs in its database and generates a cdn.discordapp.com link (Discord's "content delivery network"). This allows other people to view the image even if they don't have a copy previously saved on their system. Posted in this way, the image's context is overwritten (in a way). The image did not originate from Discord, but the viewable version of it is hosted on Discord's server in a way that doesn't natively point elsewhere.

     What if instead of just posting the image, the person in this hypothetical also posts where they found the image. This adds context to the image but likely is still not the full picture. That is because, more often than not, an nsfw image that appears on Discord made it there because the poster found it on a separate, also truncated-of-context image sharing channel. Popular channels for finding images include rule34.xxx, Twitter.com, and even just a separate Discord (or similar SNS) server, DM, or group chat. As I mentioned, this is still circling the drain on the issue. In all likelihood, the posted image wasn't born on these platforms either. So what should one post, to be "truly" correct? Well, if the relevant context is connecting the image to where the artist first posted it, then the answer would be some combination of hentai-foundary.com, furaffinity.net, Pixiv .net, fanbox.cc, subscribestar.com(?) (idk the correct generic top level domain), patreon.com, or another similar platform (do people still use Deviantart? was that ever actually a thing?).

     The issue with the example of linking directly to the page where an artist first posted something is that, for nsfw art, the internet is an ultimately cruel or chaotic domain. Put another way, linking to the original post is not a permanent nor always a possible solution. (Cases where an artist posts a work simultaneously to multiple platforms don't necessarily change this). While a given work may be viewable on, for example, Patreon at a given point in time, that is not to say that it won't become unviewable in the future. Patreon may change its policies, Patreon may get bought out, Patreon may be losing money or otherwise become unable to maintain their servers. Historically, capital such banks are hostile to nsfw archives, I think. This could be incorrect, though, but I would point to examples such as Tumblr getting bought out or the once threat of OnlyFans prohibiting the posting of adult content. Things change, people change -- by force or by choice ; by the behest of power or by the election of free will. Depending on how extreme the change is, a link which once pointed to the first utterance of an image now points to an "error 404, page not found" or an "this page is under review." In addition to site-imposed changes, it may also be the case that an artist changes and that that change affects the viewability of the original post. Examples include if the artist changes their name, their profile picture, or even their line of work. I've seen it happen a number of times where an artist makes nsfw digital art purely to pay the bills, and when eventually they secure a better job, they erase the record of their former line of work. That said, I believe it's more common for an artist who changes professions to keep their profile up and just move on with their life by themself. ... Things change. And I think that's okay.

     I have one final point about the complications of context in the example of someone posting an nsfw image on Discord and then I’ll move on. Chances are they don't know where the image originally came from. I often don't know where the nsfw digital images I see come from (or what characters they depict). Sometimes I do. But sometimes I don't. While reverse image search technologies can help to track down where an image originally came from, it is not a guaranteed solution. Depending on factors such as site policy changes, a reverse image search may not be able to find the original source. In some of these cases, where the original has been taken down, reverse image searches will reveal other places the image has been posted, albeit in an ouroboric, non-sourced fashion. Sometimes time intercedes. Sometimes technologies fail… Although crowdsourcing the context of an image is possible (such as posting an image on reddit and asking for its "sauce," as was at one point the colloquial way for asking for an image's source), there are limits to that too. People might not remember. People who remember might not see the post. Sometimes the community resuscitates an image. Sometimes the brain degenerates (the matter dies out; this is to speak nothing of the morality of porn).

     To evidence these claims, I will use a case study that starts with the hentai image site "hitomi.la" , a hopefully representative example. The information behind the first chunk of what I will talk about comes from this community-sourced wiki page on Hitomi: https://ja.namu.wiki/w/Hitomi.la [is in Japanese]. While I believe there may be some errors and outdated statements on the page, the breadth of the information seems generally reliable. Additionally, I don't currently possess other means of acquiring this type of background information on the server side functioning of these sites, so I will go with this for the time being. That is to say, take some of the statements with a grain of salt or understand that there may be some errors.

(maybe it's ironic that one of the fiew times I cite sources it's in a post about not citing sources and the lack of context,, idk)

So, what is hitomi.la? (henceforth Hitomi)

     Aside from meaning pupil in Japanese, Hitomi is a site that scrapes content from E-hentai.org (another hentai image site) in order to mirror certain categories of content. In the past, it has been one of the most trafficked hentai image sites in Korea. As for E-hentai, it is a site that allows users to upload image galleries, generally these are nsfw (an example of a gallery would be the scanned pages of a tankoubon manga [a publishing format for standalone books; to be contrasted with magazines, among others]). Users can tag the work (gallery) by artist, content type, or content form (other tags exist but these are the main ones). Hitomi scrapes E-hentai's site data (i.e., browse the site and take certain information) whereupon it then hosts the scraped content on its own servers. This duplication of content is called "mirroring." There are various reasons why someone would mirror files, one example would be to increase the files' resilience. That is, if something is mirrored across multiple servers, taking it down requires addressing each server individually. As for the type of content Hitomi scrapes, the site focuses on mirroring doujinshi, manga, artist cg, game cg, and image sets. Other E-hentai content, such as pictures from a cosplay photoshoot, is not mirrored by Hitomi.

I will now describe what constitutes each of these content types:

• Doujinshi are independently published shorter form comics which usually (but not always) incorporate characters from franchise media. Like fan fiction, doujinshi are not necessarily sexual, however, most of the doujinshi posted on Hitomi are. Doujinshi culture is complex, but the other thing I will mention is that, on both E-hentai and Hitomi, uploaded scans of a given doujinshi will often use tags to note when the piece was originally published or was designed to be sold. This is done in the form of "c" + a number (e.g., "c103"). This denotes at which Comiket venue the piece was sold (Comiket is a large, semiannual convention for selling doujinshi).

I assume most people know what manga is, so I'll skip that for now.

• While there may be minutia I am missing, artist cg refers to a content style that's sort of a flip book mixed with a comic. An artist draws several renders for various scenes. Each render will have a few variants (e.g., characters in the same poses with different facial expressions). Generally, artist cg will have a text and no-text version. In the text version, dialogue bubbles or captions will accompany each image. As for the no-text version, it is just the renders and variants. There's a strange temporality to artist cg that I can't quite describe.

•      Game cg are simpler to describe, but are also elusive in their own right. Most commonly, the tag refers to the image rips from a given eroge compiled into a galley. (Eroge means nsfw game / hentai game). Which eroge assets a game cg gallery contains depends on several factors, though it can also randomly vary depending on an arcane logic. In general, game cg galleries will contain things like all the character sprites, all the background textures, all the effect textures (if relevant), and all the game cgs (think static cutscenes). Background images are more common for image rips of visual novels while sprite dumps are more common for rpgmaker games. While visual novel game cg galleries generally port over all the game's image files, my experience is that game cg for rpgmaker games will be more selective. One issue with both of these is that the decontextualized sprites may make little sense by themselves. While some games have renders of a character in multiple expressions, others will overlay multiple sprites to display a character (e.g., there may be a file texture with an array of different eyes; ranging from happy to angry, which the game will call in response to the corresponding game trigger).

•      And finally, image set, which is the most elusive category and the thing that was the prime motivator for my writing this. This label most commonly refers to a dump of all the images from a given artist's Pixiv or Twitter profile. Other examples of image set galleries are curated collections of images which share a given tag or fetish element. The first one of these I saw when I checked just now was titled "ass to mouth galleries."

     If not already clear, context, for image sets, is razor thin. While it's reasonable for someone to assume the point of origin for a gallery called "[x artist] Pixiv ", it's harder to do so with galleries that only list the artist's name. For better or worse, the latter kind is the most common. There's also sometimes the case of a gallery that lists only the platform the images came from and thus leave out the artist's name. Curated image sets are even harder to trace back to their points of origin. While artist may be listed, such lists are rarely exhaustive, nor do they indicate which image comes from which artist (just that, nebulously, these images *could* be linked to this artist).

     This non-context is further heightened when image sets are mirrored to sites beyond Hitomi and E-hentai . While Hitomi has a direct connection with E-hentai , the majority of other hentai image gallery sites are essentially the wild west; almost anyone can upload almost anything. And so you get cases where people reupload galleries from hitomi/e-hentai. Generally these reuploads are numerically lossless, however, I believe they can also be subject to intermediary pruning or curation. Perhaps the person uploading to, say, hentaiera.com-- perhaps they download a gallery from Hitomi, but as they're looking through it, they end up deleting 10% of the images since they don't really like them. Then, at some point in the future, the person starts to run out of room on their phone or computer. However, since they don't want to delete their hentai stache outright, they opt to upload the folder they have on their device to hentaiera.com -- maybe the name changes this time, maybe it stays the same. Either way, some kind of context is lost (or at least the gallery changed).

     I perceive Hentaiera and similar hentai image gallery sites (e.g., luscious.net) as on an outer orbit of the hentai image media ecology. While some people upload their original work to these sites (when I checked most recently this was a handful of independent furry webcomics), the bulk seems to be galleries stitched together from the crumbs of other galleries. Because, yes, the thing is the images in a curated image set (even on E-hentai or wherever) are, nine times out of ten, images that come from any combination of game cg, artist cg, doujinshi, manga, or even just the straight up solo digital images as posted on Pixiv or Twitter. The permutational nature of the outer orbit curated galleries collides most of these boundaries into illegibility. While it's technically possible to reverse image search each image, the sheer quantity of images disincentivizes that (in conjunction with the other issues with reverse image search I outlined above). These galleries will generally have somewhere in the range of 1000-2000 images. Sure every five to ten images will come from the same source, but that's still just like so much stuff from so many places. Well, at least when I tried to do the detective work in the past, it wore me out and I gave up.

     If it's not clear by this point, what I disguise as a conversation about "context" is probably better understood as one about authorship. Although artists can try to prohibit the sharing of their work, there's often little they can do to prevent sharing from happening. Image sharing is the vast norm, after all. It's been ingrained into the media ecology as default for multiple decades by this point.

     Anyways, let me tell you about the booru structure of imageboards. My mind is flaking at the moment, but, in essence, booru are imageboards that usually have some specific kind of focus or specialty. There are nsfw booru and all ages booru. I don't remember exactly how they proliferate or are maintained, but I recall that users can upload content (for the most part) to some whatever booru and that it's generally possible to create a new booru for a specific niche (though I imagine there's some cost barrier to entry for that). In the US, the most popular booru/booru-like is rule34.xxx ; elsewhere danbooru.donmai.us and gelbooru.com are very popular as well. An example of a niche booru would be hypnohub.net, a site where people post hypnosis/mind control-focused nsfw art. While there may be artists who upload their work to these sites, it's more often the case that non-artists upload an artist's work to the imageboard. Most sites allow artists to request their work be removed, but it's unlikely that an artist will be aware of every site that their works are or will ever be posted to. Regardless, when someone uploads images to a booru, they can optionally include the link to where the image came from. It seems to be about 50/50 as to whether people do this. Usually, if you rotate through enough works tagged as belonging to a given artist, you can find where they came from, but that's not always the case. Sometimes an artist only has a handful of pictures posted on a booru and they all lack a source link.

     However, booru (speaking more so about niche ones) are more than just repositories of reposted images. While they are that, there is also an extant culture of image editing. With the example of Hypnohub, someone may edit an image to make it about hypnosis or take a hypnosis fetish image and add a caption to it. Although some people may choose to edit and post images anonymously, I believe there's also a culture of especially popular editors becoming known entities. Like, people will follow editors because they like the types of images they edit or the types of edits they do to images. (Artistry emerges once again!)

     Many booru have internal features which allow users to create image galleries. Obviously, these come to resemble the contents of E-hentai after a certain point. Although there are definite nuances in the types of images that make it into galleries on a booru versus a site like E-hentai (e.g., the nascent structure of booru prioritizes single image nsfw art; what this means for the comparison is that booru will have far fewer images from doujinshi or manga).

     Another platform for hentai image galleries is on multimedia hentai p1racy sites, such as f95zone.to and anime-sharing.com (there's also ryuugames.com, but it's less image focused). While I think the primary function of these sites is for sharing eroge, they also have image repositories. The repositories on F95zone seem to be more focused around centralizing the works of a given artist, however, there are also a number of individual doujinshi, webcomics, and maga. I'm not as familiar with the image sharing culture on Anime-sharing, however it seems like it's more doujinshi / manga focused than artist-focused. While these sites have image sharing functionalities, the reason I mention them is instead because they link back to the artist cg galleries I talked about before. If someone tracks down where an image came from, these platforms would provide a secondary means to download an image and place it into a context. At the same time, these sites equally demonstrate the volatility of hentai images online. To illustrate this, I will use the example of anime-sharing, as they seem to have less internal server infrastructure. While anime-sharing offers links to download images and games, the site provides a level of native visuality to indicate a sense of what you're downloading. For eroge, that generally means that they'll mirror the set of five or so preview images viewable on the game's DLsite.com page. These images are not hosted on anime-sharing, though, and because of their pornographic nature, they are eventually removed from the third party image hosting sites they get posted to. The images are shattered, fractured across space and time. A puzzle of sharded blood and cumstains.

      addition to shattered images, the example of eroge p1racy also illuminates the issue of shattered files. Depending on factors such as the limitations of a hosting site and the size of a game, the zip file for a game may need to be split into several parts. While it's generally simple to download all the parts, it can also be the case that one part is removed from a file sharing site but the others remain. In these cases, the resulting partial file is essentially unusable, though depending on how it was split up, it may contain legible images (and just lack a legible game).

     The last phenomena I want to talk about is nsfw art (such as fanfic) incorporating hentai images as generic visual aids to depict what the scene describes. While I think archiveofourown.org probably has a relatively small image culture, there are some other smaller fanfic sites which place a more central role on images, such as CHYOA.com. Again, the vast norm in these cases is for reference images to be used without saying where they came from. (It's worth noting that some artists make art specifically for their fics). Other times, artists use AI to give flesh to what they're talking about. The benefit of using AI is that it gives the author more artistic freedom, theoretically. With AI images, the image can come to the text instead of the text coming to the image. (By the latter I mean such things as when a fic or chapter is written in service of giving a story to an image). Fics in service of an image are a wordier iteration of the doujinshi-in-translation practice of rewriting a comic (probably it happens with things like manga and webcomics too; I just most frequently see it done with doujinshi). The images are born anew. And, honestly, you don't even need to go that far. Even just translation by itself is a decontextualization of the image.

(I meant to mention it earlier; but Hitomi has an imageboard mirror of its site's contents called nozomi.la ; like Hitomi, users cannot upload content to it. nevertheless, it shares similar tag search functionalities as the generic imageboard/booru model)

     The maelstroms of time... so anyways, I think discourse around AI images as theft is disingenuous. Or, rather, it is not cognizant of the histories governing the image sharing cultures of the internet.

     With this said, it's worth making the provision that the topic of this post is nsfw digital images. Perhaps other digital images have different histories or are governed by different image sharing cultures. I think that's probably partly the case in some respect, but I don't believe that these histories would be *so* different to the point that my points about image sharing eroding context would become illegitimate.

     I've talked a lot about how various apparatuses of digital image sharing (culture + technology) incentivize or actualize the erosion of an piece's context – i.e., the who, what, when, where, and why of how the original image came to be – but I want to clarify here, at the end, that I don't think there's anything wrong with that. I talk about anti-context and removing context, and, to be clear, this is bad faith. It's more apt to describe what's happening as REcontextualizing. Sure, the "original" context may transform or fade away, but the original context is not the only context. It can be helpful information to have insofar as that it may inform what one makes of an image, but it is not the be all and end all. To this end, I reject the notion of dubbing as an erosion of quality. I reject the notion of art theft. Because of course I would, right? I grew up in this media ecology. I first found out about nsfw (digital) images (at some age, idk) via Google images. The first rabbit hole I remember going down was with my discovery of the furry booru/imageboard, e621.net. I was very fascinated with it all. I remember sitting in school and speculating about how many types of animals and media entities might be represented on the site. I remember, for example, wondering if (furry) crab-centered porn was a thing that existed -- "dang, what would that even look like?" But, that's an essay for another day.

     Just rip their face off. Life's full of bleeding limbs. Bite your tongue and roll in your grave. The day bad objects arrive is the day erotica dies. Maybe..?

     At the end of the day, though, I play fast and loose with words and fallacies, so who knows if this holds. But it feels right to me. Knowing what I know. It's what my biased research says, anyway.

A few postscript remarks:

• I think it's worth giving lip service to some other possible explanations for why context may be left out when someone cross posts a nsfw image. The most obvious reason is probably the crushing weight of shame and the societal taboo around smut.

• I couldn't find a good spot to fit it in, but I feel it's worth giving a shout to the ahegao meme template that became popular in the 2016-2019 period (to the point that people would buy hoodies or shirts with the design printed on it; to flex their irony, or something). While I'm sure someone has documented where those images came from, the default state that most people would have encountered them was in a purely isolated format. Other iterations of this template substitute will substitute in other suggestively nsfw manga panels (in place of the base format which is a cropped panel of a girl making an ahegao face).

• I understand that artists need to make money, however, I don't think that image sharing boards are ever genuinely threatening people's livelihoods. Or, rather, their threat is never in any way proportional to that of the overarching system of capitalism, which holds the right to life as hostage against a person's ability to work. The problem is not with the individual consumer, anyways.