Emotion, Archives, Interactive Fiction, and Linked Data

[Edit (Feb 24, 2013): Thanks to the fantastic work of Tara Robertson, the video of my lightning talk is now available!]

I gave a lightning talk entitled “Wielding the Whip: Affect, Archives, and Ontological Fusion” at the 2013 Code4lib conference in Chicago, Illinois. This lightning talk was one of the most difficult presentations I’ve ever given for a number of reasons, including the emotional aspect of the content itself, as well as the fact that several of the ideas I was trying to articulate weren’t fully baked. I’ve been thinking about this for the four to six months in various capacities and with different focuses, especially as I read more interactive fiction and learn more about it (as well as about hypertext in general). This post serves as an expansion of some of the ideas in my lightning talk and as a way to further the discussion around the following question:

Can we write interactive fiction and (semi-/para-)fictional hypertext that leverages linked data to create an emotional connection to the “real world”?

Emotion in the archives

Starting from the beginning of my talk, I invoked a presentation from last November by Tim Sherratt that explored the role of emotion in archives. Tim’s presentation is very compelling, and I won’t go over it in detail here. One of the most critical points that Tim made was asking why we refuse to acknowledge the place of emotion in archives (arguably, both in materials themselves and in archival practice):

Why are we so reluctant to acknowledge that archives are repositories of feeling? Is emotion meaningless because it can’t be quantified, dangerous because it can’t be controlled, or does it simply not fit with the professional discourse of evidence, authority and reliability.

Tim’s presentation also referenced two other things that influenced my thinking. The first was another talk he’d given recently at the National Digital Forum (NDF) in New Zealand that urged us to think about linked data as more than just an infrastructure project, and as a means to contextualize stories. The other was Courtney Johnston’s NDF presentation, which in part ends with her articulation of a “daydream” about a Museum of Emotions:

For a long time I have been frustrated that I can’t search our collections by emotion. Cataloguers record a factual description (‘Boy, aged approximately five, wearing woollen pullover and crying’) but rarely emotional tone or content (‘Boy, aged approximately five, wearing woollen pullover and crying with frustration’). But I want to type happy or sad or loving or bored or awed into a search box and get a stream of results.

The presentations by Tim and Courtney resonated with me because it seems that archives (in the sense of being cultural heritage institutions) can very rarely acknowledge their emotional aspects. Language used in our descriptive apparatuses is often flat in affect and feels clinical to me even when the records don’t have emotional content. As Courtney suggests, emotion could be essential in the discovery process for certain types of collections or subjects. But by far, emotion is how we develop our bond with the past represented through records.

Archives of emotion

My lightning talk also engaged with some of the ideas put forth by Ann Cvetkovich. I’ve recently started reading her book An Archive of Feelings. The book focuses on a slightly different meaning for the word “archive” when she defines “an archive of feelings” as “cultural texts as repositories of feelings and emotions … encoded not only in the content of the texts themselves but in the practices that surround their production and reception.” Cvetkovich establishes the necessity of this shift through the focus of her research on trauma:

Because trauma can be unspeakable and unrepresentable and because it is marked by forgetting and dissociation, it often seems to leave behind no records at all. … It thus demands an unusual archive, whose materials, in pointing to trauma’s ephemerality, are themselves frequently ephemeral. Trauma’s archive incorporates personal memories, which can be recorded in oral and video testimonies, memoirs, letters, and journals.

This has made me realize that we don’t have much to gain about quibbling about what “archives” are or are in many cases. This is not just an issue of “absence of evidence is not evidence of absence.” Cvetkovich’s definition includes that which is not “normally” archival because of the need to represent that which is problematic to document (i.e., in her words “to make not just texts but whole cultures visible”). In the broadest sense, claiming something may not be or is not “archives” constitutes a form of erasure of that which refuses to be assimilated (cf. pages 11-12). In a discussion with Kathryn Garcia, Catherine Lord, and Martabel Wasserman, Cvetkovich also articulates the importance of establishing context and recognizing the social relations inherent in archives:

When you do research in queer archives you discover that behind every name or better-known name, there are always so many other names. Archives allow you to start tracking those friendships and connections. Those relationships are the foundation for the production of art and for a life well lived (which might be one of the best forms of art) – and, indeed, the ability to do anything in the world. … I always want people to take their own connections seriously. There is a kind of political depression that can come from the sense of belatedness we discussed earlier: “they had this special set of relations and we don’t.” … There is still a tremendous amount of scholarship to be done just to fully excavate those queer social worlds. But I want that work not to mystify those worlds but instead to provide enabling models for you and me having this conversation or for the conversations I have with my friends.

The archive, then, ultimately also has the possibility as serving as an empowering tool - not just in capturing culture, emotion and lives, but also as a resource in terms of allowing us to build a bond to past worlds and lives. It is ultimately also a site of creation and construction of cultural works and emotion. While Cvetkovich says she means something different than the “traditional” meaning for “archives,” I think that it’s perfectly reasonable to assume that, given the right situation, they could have also serve a similar dual purpose.

Small data, big data, and platform as a social monopoly

Tim Sherratt’s NDF presentation also specifically acknowledged the need to “fight back against the homogenisation of data” in the big data world, and casting the creation of companion narratives that linked to cultural heritage resources as “small stories.” This reminded me of some of the recent work by my colleagues Amelia Abreu and Amelia Acker on developing a research agenda for small data. At the same time, I’ve had a growing concern about the use of platforms like Facebook as a means to collect and analyze data created through the behavior of those users. While this is troublesome on its own, I’ve found it moreso with the addition of Facebook Graph Search available as an end-user tool. There’s a bit of discomfort when browsing through the Actual Facebook Graph Searches blog, and even more so when you consider the tangible ramifications of the parody video “When Facebook Resurrected the Dead”, created by the blog’s maintainer, Tom Scott.

At a larger scale, my primary issue is how Facebook generates value through free labor, or in the words of Böhm, Land, and Beverungen, “work that is both free from the obligations, constraints and managerial imperatives of formal employment, and work that is done for free and without financial recompense.” Böhm, Land, and Beverungen’s analysis suggests that autonomous behavior/autonomous labor of Facebook users occurs while their social relations and lives are being subsumed under capital. They identify the advantage that Facebook has in this sector by recognizing that the company is positioning itself to monopolize network-based social interaction:

Bill Gates’ business model is based on creating a form of monopolistic control over the basic infrastructure of the knowledge economy and doing everything he can to ensure that Microsoft maintain effective control over the networks and protocols of communication that are used in the communicative circuits of production in the social factory. Facebook is the new Microsoft. It has established itself as the de-facto gatekeeper and standard of the social networking world.

We see this monopolization more and more places, most notably through frictionless sharing and the increased use of Facebook accounts as an authentication mechanism on the open web. Through this, Facebook-as-platform has reduced some degrees of expression down to the mechanics of consumption, through the “Like” button and traces left by frictionless sharing (that I uploaded a file to Scribd, bought tickets through Ticketmaster, or even just read an article). Reduction of behavior to “likes” and a few select other actions through the platform’s Open Graph Protocol is a form of protological control which, according to Gerlitz and Helmond, “enables particular reactions and activities rather than others.” Accordingly, certain types of affect and emotion are by necessity filtered out within the Facebook Graph because they are fundamentally at odds with the underlying values represented by the platform. Lin and Qiu note that Facebook users who have used the platform longer tend to express emotion less over time. One has to wonder if this has anything to do with the recognition of what and how things are actually shared and expressed through Facebook.

Public emotions: a personal digression

At the very beginning of my talk I indicated that I’d had a difficult year but glossed over many of the details. Providing this information is just as valuable from a contextual standpoint as everything else I’ve covered so far and will mention in this post, so please stay patient.

This year I’ve been dealing with an unusual bout of depression. It is certainly not unusual to be depressed (and I am no stranger to it), but the way one particular bout manifested itself was unusual - it came on suddenly and was oddly crippling to an extent I’d never experienced. My tendency, when depressed, is not to reach out to people. This time, however, I was compelled to acknowledge it - not entirely publicly, but to select friends through an email message. This in itself was a big step, and my friends who I wrote to were all great and responsive. At the same time it made me want to pull away from social media platforms, most notably Facebook. Even on my best days I find Facebook more and more alienating as it becomes allegedly “easier” to maintain friendships and relationships through a few clicks.

Last October, I also lost my friend Laura Tatum to ovarian cancer. Laura, a fellow archivist, University of Michigan alum, and former colleague from Manuscripts and Archives at Yale, was the first person for whom I have ever truly grieved, and she is still on my mind almost constantly. My circle of friends and colleagues, especially in the cultural heritage sector, has last many others as well: Michael Cohen, Lee Dirks, Michael Nash, and of course, Aaron Swartz.

Through depression and loss I have learned that keeping my emotions private was deleterous to my well-being. Making them public was a necessity, even to just a selected public. It also dawned on me that acknowledging emotion publicly could be a political act or bound with political expression, which I surprisingly discovered as also being present in some of Ann Cvetkovich’s more recent work. Expressing emotion itself could also, in some cases, become an expression or assertion of power. The hardest part of this, at least, was finding my voice.

Interactive fiction and hypertext

Some time last fall I started getting curious about hypertext as literature. It probably started as a result of reading about Michael Joyce’s “afternoon, a story” in Matthew Kirschenbaum’s book Mechanisms: New Media and the Forensic Imagination. This was all well and good, and it piqued my curiosity further. What I noticed, however, was that these “serious” works felt somehow fundamentally inaccessible. It wasn’t because I was by intimidated by their authors or their content. The level of commitment required to find them, purchase them, ensure that they’d work, however - that’s what pushed me away. I didn’t feel like I could take a risk with a work of hypertext that saw itself as “serious.” It also felt in some ways alien to the hypertext that I use everyday and in which I feel fluent: the World Wide Web.

I’d never considered myself much of a gamer but I was also curious about interactive fiction. (Where the boundaries lie between hypertext and interactive fiction is not something about which I’m certain.) I’d played a few Infocom games growing up, and I found a few others. There are a couple of things that really drew me in, however. The first was reading about First Draft of the Revolution, a collaborative work between Liza Daly and Emily Short. I was impressed with the elegance in design and content, and it felt compelling that this all could be experienced in a browser. Some time shortly after that I discovered Twine, which seemed like a great system: simple enough to get started, sophisticated enough not to get in the way, particularly if you were trying to build something along the lines of a choose-your-own-adventure piece of interactive fiction.

It all changed - radically - when I discovered Porpentine. Someone I follow on Twitter posted or retweeted a link to her post Creation Under Capitalism and the Twine Revolution. This was heavy stuff. It distilled some of the discomfort I felt with what I thought was “hypertext literature” while demonstrating that an alternative was not only possible, it was already here in the form of something like HyperCard. There is so much exquisite brilliance in her post that any pull quote won’t do it justice. So read it, and read it again. It’s taken me at least five passes and it hasn’t totally sunken in yet. The two quotes that have stuck with the most are the following:

The only sympathetic depiction we can trust is the one we make ourselves, in love and tenderness and the desire for justice. We don’t have to wait. We have the power to create those depictions right now because those depictions live in our bones. They couldn’t give us shit with a budget of $25 million dollars. We can do better for the price of food and shelter.

…and…

Twine is the closest we’ve come to a blank page. It binds itself and it can bind itself along an infinite number of spines extending in any direction. It’s hard to visualize our problems and emotions when they get interrupted by code, but we know the feel of words. They’ve dwelt in us our whole life. They are alive and they want to come out. Twine is the invitation to be personal.

Twine had to be a solution. I started reading works built in Twine. I laughed and was moved in a way I hadn’t been by fiction in quite some time - works like Porpentine’s howling dogs, Adam Dickinson’s weird tape in the mail, and Anna Anthropy’s HUNT FOR THE GAY PLANET. I was utterly devastated by Kim’s Story. I’ve started writing in Twine - I don’t quite feel like I’m on my feet yet, but I’m getting there.

Bringing linked data to Twine and hypertext fiction

Looping back around to where I began this post, I started to wonder if there was a way where I could leverage the “real world” through linked data. We very well could have something simple that would allow people to tell stories - even explicitly fictional ones - that allowed us to pull in details about the past and the present in a more meaningful way. We already can harness emotion through narrative, and if we wanted, we could tie that emotion and the narrative to recorded information - to substantiate it, to reference it, or to refute it.

Tim Sherratt’s presentations have shown what was really possible using simple tools. If it was totally reasonable for Tim to augment a previous presentation using “simple tools, no platforms, and no markup madness,” yet having linked data and have some nice UI tidbits provided by JavaScript and CSS, what would happen when we did the same with hypertext fiction, especially in something as simple as Twine? This could just be just as Tim suggested: stories + data = fighting back.

I had visions of untamed hypertext - interactive fiction built on the native technologies of the web that actively defied the some of the platform’s nature. Inadvertently, I stumbled on a few publications by Jill Walker while I was working on my lightning talk. The two that really got the gears turning were “Feral Hypertext: When Hypertext Literature Escapes Control”, and “Performing Fictions: Interaction and Depiction”. Both works gave me great ideas, and Walker’s notion of feral hypertext spoke to what I felt might be possible with Twine:

Feral hypertext has a tendency to move beneath the radar. It is easy to not identify feral hypertext as hypertext at all. Feral hypertexts are not as clearly delimited and disciplined as domesticated hypertexts are, and our language and culture aren’t designed to speak about things that lack boundaries. What feral hypertexts have in common is that they have reverted to the wild, in one respect or another. They are no longer tame. They won’t do what we expect and they refuse to stay put within boundaries we’ve defined. They don’t follow standards – indeed, they appear to revel in the non-standard, while perhaps building new kinds of standard that we don’t yet understand.

Meanwhile, “Performing Fictions” provided me with an important notion of “ontological fusion,” wherein one world (say, a fictional one) is overlaid on top of another (say, the “actual” world). This is precisely what I wanted to do. But we still have an open question of how we can actually implement these ideas into something that may be useful.

What are we building?

What I want to build is something that’s simple, that stays in the spirit of Twine. I want there be an easy way of annotating text or adding another kind of link to something outside of a Twine story. I want these outside links to be some times obvious - maybe sometimes not. I don’t know how I want to surface this information, or whether (or when) I want a Twine story to pollute the global graph. So roughly, I see us needing to come up with the following:

• An extension to Twee (Twine’s markup language) and the JavaScript code that turns it into reader-actionable HTML.
• Some thought to the UI containing the links as represented in a reader-facing Twine story.
• How to represent these outside resources in Twine node maps.
• A dead-simple way to add these things using the Twine GUI.

It doesn’t even have to be Twine, but it makes sense because it’s already out there.

I don’t want to worry about the problems of the semantic web. Provenance is an issue near and dear to my heart but focusing on that it this point stands to complicate the entire task of building this. Worrying about the “proper way” to build this stands only to get in the way and threaten the necessity of this being a grassroots project. I want this to be more than an academic exercise. If the semantic web community cares about what we’re doing, maybe what we’re doing matters more than we think.

By taking back or taking issue with linked data through our stories built on a free/libre tool, we are committing a political act. We are fighting back and we’re taking the whole world of knowledge either with us or standing in opposition to it.

Are you on our side?

What do you think we should build?