Among the major preservation strategies raised in Rinehart & Ippolito’s Re-Collection, re-use and reinterpretation are the most tantalizing, and seemingly most radical. The fear of somehow being untrue to the spirit of a work, or to an artist’s intent, make these approaches look riskier than others. A few writings on digital sound and moving image hint at what it means to de-center the artist in preservation and looking to users for cues.
In Jason Eppink’s history of and interview with The Signal about GIFs, we see work consistently distanced from its creators. This is in part because the origins of images aren’t so easy to trace on the internet and in part because, as Eppink says in the interview, “There’s still very little to gain from making GIFs.” He goes on to say, “We expect the image to have an author because of the fundamental relationship of authorship to the economics of producing cultural artifacts. But today images are as cheap and prolific as the air that we utter our words with.” GIFs, to Eppink, manifest the near erasure of authorship by use and reuse. He offers an extreme vision of looking beyond artists for the primary stakeholders in preserving digital art. It makes me wonder if the gain in GIFs might lie in distribution: like how one of the interviewees in this Off Book video about YouTube describes people sharing funny internet things as “wanting them associated with their identity.” And sometimes social capital morphs into something higher-risk, as with feuds over meme-sharing and -stealing on Instagram.
Here’s a bit from Jonathan Sterne’s chapter “Format Theory” that I took as further reason not to focus too narrowly on creators and intention: “Because these kinds of codes [underlying formats] are not publicly discussed or even apparent to end-users, they often take on a sheen of ontology when they are more precisely the product of contingency” (p. 8). In other words, things aren’t necessarily made a certain way out of values- or meaning-based reasons. What contingency kludges together, specific use can improve or infuse with meaning. I was also struck by the contrast Sterne draws between the “ubiquitous,” “banal,” and “pedestrian” presence of MP3s and the passage he quotes from Lisa Gitelman ending, “Specificity is the key.” Pervasive technology might mean shared experiences, but not identical ones. Maybe format theory is best served by comparative studies or format / use genealogies, highlighting divergence as well as trends. There are local and individual variations, and variations on those variations — GIFs on GIFS on GIFs.
The pieces mentioned here intersect, in my mind, with a talk Jarrett Drake gave this week about archival description. He argues that archival practice is due to stop privileging provenance and move towards a new organizing principle (or principles). Provenance is about creatorship; valuing it above everything else results in archival description that centers records creators and the relationships between them. Archivists might propose to “collect more broadly,” but inviting the oppressed and underrepresented to participate in oppressive systems does little to effect change. How, instead, to open participation in rebuilding and reworking archival principles? His answer, deliberately not providing an answer:
“The truly transformative principle that is needed for archival practice and archival description cannot come from one person or from one invite-only forum, but such a principle necessarily must develop organically, slowly, and anti-oppressively with a radical cross-section of academic, disciplinary, racial, ethnic, gender, cultural and class backgrounds represented. In this sense, a new foundational archival principle, should it be worth anything, must be developed beyond the bounds of the archival profession.”
In other words, it’s not about being in the room where it happens, but about opening up the room and the process.
Reading this while thinking about sound and moving image distributed via browsers and apps, the collapse of user and creator categories is a major factor that could shape new kinds of archival description. YouTube is celebrated not only as a “wild west” of user-uploaded video, but also as fertile ground for new brands and businesses. How users relate to digital objects, user-creators, YouTube production companies, and each other seem to be the most important aspects to capture. It also seems worth exploring how digital objects relate to one another with or without the intervention of people. None of these phenomena can be adequately reflected in provenance-focused archival description, making digital art curation a more valuable site than ever for experimentation and enrichment.
This quote from Jason Eppink from the interview really stuck with me: “I did track down all the original sources these reaction GIFs were derived from, but in the end decided not to include them in the exhibit because that was beside the point; it was contextualizing them incorrectly.”
The fact that Jason was able to track down the original sources but decided against including them really highlights the change you’re talking about – the shift from author to user as being more important to contextualize a lot of these digital artifacts. Eppink was not intending to preserve the GIFs, but if the GIFs were to be preserved and if including the author information would be “contextualizing them incorrectly,” should archivists explicitly exclude author information even when they have it?
Great question. I think deliberately leaving out information could make sense even if one already has it. Discarding information about digital cultural heritage could be the right thing to do for cultural sensitivity or privacy, but I don’t see too much description as a concern beyond those realms. Curious to hear other opinions, on this and everything!
I was just reading something this morning (what the heck was it?) about how for some digital objects, trying to separate metadata and content doesn’t make any sense. Tweets were a good example given, in that 99% of a tweet is its metadata. It’s possible to run a script to pull out and separate only tweet content (140 characters or less) from a body of tweet data, but even that small effort seems pointless. Instead, why not focus on what to highlight through the systems mediating access to those materials? To bring it back to the issue at hand: Even if there were too much rather than too little authorship information available about GIFs, how that information is organized contributes to how it surfaces or doesn’t. So maybe we focus our efforts there rather than on hard decisions about whether to keep or not keep certain fields at all.
Not expending the effort to sort out where a GIF came from seems like the better time saver to me, even when preservation is the objective. (Which I know is different than what you’re talking about.) It seems to me that preservation is necessarily about managing resources and description in support of an argument, and if that argument de-emphasizes a singular origin story, why not use the time and space another way? And if time was already sunk into background research — oh well. Putting in more time and effort to undo or erase that work doesn’t speed up access, so, probably best to move on.
The importance of the context of the image will come up next week, when we talk about remix videos. There the scene or context can have a larger importance to the message of the vid, but they usually also work on a literal level as well. I think some gifs do work like this, but definitely not to the same extent.
Which leads me to wonder – vids have had to argue for copyright exceptions (and have won them) in order to fight back against DMCA takedowns on sites like YouTube (some content creators are privileged over others). Have GIF creators or users come up against this issue? I can’t recall hearing about the content of the GIF causing anyone to be pursued for copyright infringement by the original-media rights-holder.
Had to read around on this (casual) and one example I’ve found is about sports GIFs on Twitter and complaints by the NFL and UFC. Elsewhere, I’m seeing some support for fair use and the four factors as criteria to use when thinking about gifs and copyright. And by “support” I mean perhaps “resignation and trepidation because there’s no official word and no guarantees.” Will post again if I find more.
So I know we’re talking right this minute about GIFs as remix or appropriation, but I wonder if people might tend to lump all GIFs together as subject to the same unspoken rules for use and reuse. Is the short clip from Hamilton above seen or treated the same way as these gorgeous original art gifs by Carl Burton? (Some of which were apparently created for the second season of Serial, further complicating ownership and authorship.) What if I saw them both on Tumblr? This is where platform specifics could become important context, not because they shape creation but because they shape consumption and subsequent reuse by influencing users’ expectations of what they can or should do with a GIF. Wild speculation, really.
Are Burton’s works originally gifs, or is that the format they were uploaded to tumblr, etc in? They look to me like vector-based animations.
Tumblr is especially problematic, I think, because it actively enables stripping away identifying source information, which is a problem with any work on Tumblr, not just GIFs.
I think if it’s a remix/recontextualizing, then it’s going to have different arguments than ‘borrowing’ original work simply for redisplay would, perforce. If it was an original GIF that was being recontextualized, I’d see that the same as the first case, like rage comics.
I like you point about how much of our theory and description of the genre is being collapsed as traditional creator/viewer roles collapse. And while to a certain extent this is true, it should be remembered that, for example, the Lizzie Bennett Diaries (which I am picking on because it won an Emmy) was a fully funded production mimicking the YouTube style. It’s talking heads were professional actors, not the content creators. The Kevin Pollack Chat Show (which won a Streamie a couple of years ago) is a youtube series that mimics old-style interview shows and has multiple cameras. My point is, how much is the traditional mandate really collapsing? And if it is, how should it affect our work? What alterations to the craft of cataloguing should we make to accommodate changes in an industry?
Yeah, I wonder: What’s a good way to use cataloging and arrangement/description to connect the documentation coming from different regimes of archival practice? I guess this is part of the potential of linked data: the ability to bring in authorities from all kinds of disciplines and interrelate them, plus a way to track down said authorities. Although a counter-argument there is that authorities are all about named entities, and the entities aren’t always what matters. Creating metadata for our projects might mean encoding different kinds of relationships than what’s possible or practiced in a finding aid. For example, say Rebecca is creating a record for the Lizzie Bennett Diaries. What if there’s a way for her to encode the project’s influences as part of the metadata for this collection; like if she, as curator, recognized borrowed conventions of YouTube video diaries or nods to specific / actual diarists, and wanted to note that in a structured way.
Another example: One thing I and so many others love about Hamilton is the sheer density of allusion to hip-hop history in every single song. Hamilton songs are on Genius, but that annotation is lyrically focused. If it was possible to unlock every sound moment that made me go, “Ah!” through metadata, what kinds of relationships would be expressed? I don’t think “sampledIn,” “borrowedBy,” or whatnot are enough. Many of the works we’re seeing in this class challenge relationship expressions in similar ways. I’d have to go back to Re-Collection to see how their metadata schema addresses relationships.
And the last piece, I think, would be to change by whom, for whom, and for what purposes cataloging happens at all.