Monthly Archives: November 2016

Out of Sync: Digital Humanities and the Cloud

Matthew K. Gold

Out of Sync: Digital Humanities and the Cloud

This is the text of a keynote lecture I gave at DH Congress at the University of Sheffield on September 10, 2016. I’m grateful to Michael Pidd, the University of Sheffield Humanities Research Institute, and the conference organizing committee for inviting me to speak.

* * *

In February 1884, John Ruskin delivered an address to the London Institution titled “The Storm-Cloud of the Nineteenth Century.” Ruskin began with a reference to his somewhat ominous title:

Let me first assure my audience that I have no arrière pensée in the title chosen for this lecture. I might, indeed, have meant, and it would have been only too like me to mean, any number of things by such a title;—but, tonight, I mean simply what I have said, and propose to bring to your notice a series of cloud phenomena, which, so far as I can weigh existing evidence, are peculiar to our own times; yet which have not hitherto received any special notice or description from meteorologists.

Ruskin went on, in his lecture, to do just that — to convey his thoughts on clouds based on his sketches and observations of the sky, though at least some audience members and later critics have seen embedded in his remarks a critique of encroaching industrialization. And so, in a talk focused on clouds — which Ruskin elsewhere described with much beauty and care — we see the seeds of a larger political critique.

Were I to “propose to bring to your notice a series of cloud phenomena” that are, in Ruskin’s words, “peculiar to our own times,” we would begin, most likely, not by looking up to the sky, but rather down at our cell phones. We wouldn’t discuss “storm clouds,” or even “clouds,” but rather “THE cloud,” by which we would refer to the distributed set of services, platforms, and networked infrastructures that string trellised connections between our phones, laptops, desktops, and tablets. We would point to the media systems that have caused us to consign our CDs to closets and to sign up for subscription-based music services such as Spotify and Tidal. We would point to Google drives and docs, Twitter hashtags and Facebook feeds, wifi signals, Bluetooth connections, Github repositories, files synced across DropBox, Box, and SpiderOak. Indeed — the sync — the action of connecting to and syncing with the network, of comparing our local files to those on a remote server and updating them to match — might be the signal action of the cloud-based life. We become dependent on, and interdependent with, the network — always incomplete, awaiting sync, ready to be updated. The cloud produces both security and instability, offering back-up services but keeping us always in need of updates. We look to the cloud not to see an alien sky but rather to recover parts of ourselves and to connect or reconnect to our own work.

My aim in this talk is to spend some time thinking with you about the cloud and about what it portends for the digital humanities. But it’s difficult to talk about the cloud without also talking about infrastructure, in part because of the clear ways in which cloud-based services and models are dependent upon physical conduits, things in the world, that belie the cloud’s supposedly abstract, virtual, and ineffable nature. I’d like to draw the DH community’s attention to a set of conversations that are occurring both in DH and also outside of it–in the realm of media studies, and in particular, the growing area of critical infrastructure studies. In connecting these conversations, I want to encourage us to think about how DH work relates to or should relate to issues of infrastructure — particularly as these issues involve larger concerns that have been raised in the humanities about DH work around issues of instrumentalism and neoliberalism. My premise is that thinking about DH work within an infrastructural context may allow us to both focus on the work that DHers do so well — reflexive and theoretically informed building, making, and critique — and to build or rebuild that human, social, and scholarly communications infrastructure upon sturdier grounds of social justice.

It’s easy to see that an “infrastructural turn” has been growing over the past year in DH and allied fields. We can see it in evidence in the July 2016 King’s College symposium “Interrogating Infrastructure“; in the recent DH2016 panel on “Creating Feminist Infrastructure in the Digital Humanities”; in DH2016 presentations such as James Smithies “Full Stack DH,” which described his project to build a Virtual Research Environment on a Raspberry Pi; and in experiments such as my own team’s DH Box project. In media studies, we see this infrastructural turn in recent publications such as Tung-Hui Hu’s A Prehistory of the Cloud and Nicole Starosielski’s The Undersea Network; in a renewed focus, generally, on the material nature of computers; in the growth of the field of media archaeology; in calls across the academy to pay more attention to the social and political contexts of digital work; and in efforts to recover the diverse histories of early computing. This work encourages us to take account of computational work as enmeshed in the different levels of the “stack,” Benjamin H. Bratton’s term for the set of infrastructures – media infrastructures, data infrastructures, political infrastructures, physical infrastructures and legal infrastructures — that have accreted over time into an accidental whole through what Bratton calls “planetary scale computing.” As more and more DH work moves to the cloud and becomes dependent on networked infrastructure, thinking about the protocols, dependencies, and inter-dependencies of the Stack can help us fruitfully shape our work as it relates to both allied scholarly disciplines and to larger publics.

What role should the digital humanities play in conversations about infrastructure? What particular insights does work in the field have to contribute to them? And to what extent are DHers already doing the work of infrastructure in the academy broadly, and the humanities more specifically? I will argue in this talk that DHers should engage the Cloud and its associated infrastructures critically, thinking about how the emergence of the Cloud, even as it makes possible new forms of networked connection, also foregrounds multiple risks. It’s my belief, as I’ll detail later on in the talk, that DH should step back and re-consider its use of proprietary social networks, and that it should focus on building alternate forms of scholarly publishing and communication infrastructure that help move us away from proprietary networks where every interaction is always already commodified and where the network effect all too often puts marginalized populations at risk.

In his opening remarks at the “Interrogating Infrastructure” event — an event I did not attend, but which I have at least some sense of thanks to his online notes — ­­Alan Liu positioned the topic of infrastructure as a key future direction for the digital humanities. He argued not just that the topic was well-suited to the field, but that it was one which DH was well-positioned to address. Infrastructure, as the set of social and technological systems undergirding many aspects of networked modern life, for Liu, has the “potential to give us the same general purchase on social complexity that Stuart Hall, Raymond Williams, and others sought when they reached for their all-purpose interpretive and critical word, ‘culture.’”

I think Liu is right that DHers can and should pay increased attention to issues of infrastructure and the effects of that infrastructure on the larger communicative and meaning-making networks of contemporary society. And clearly, much work on infrastructure is already blending scholarship in new media studies, science and technology studies, and the digital humanities. I think here of Matthew Kirschenbaum’s work on forensic materiality and software platforms, Lori Emerson’s work on interfaces, Jentery Sayer’s work on prototyping the past, Jussi Parrikka’s work on media archaeology, Simone Browne’s work on surveillance networks and race, and Kari Kraus’s work on speculative design. All of these scholars are already exploring the intersections of infrastructure, platform/material studies, design and new media.

The past year has been notable within the emerging field of infrastructure studies, as scholars in the fields of new media studies and science and technology studies have published a range of books that put the infrastructure of the Cloud into theoretical and infrastructural contexts. Across four of those books — Nicole Starosielski’s The Undersea Network; Tung-Hui Hu’s A Prehistory of the Cloud; John Durham Peters’s The Marvelous Clouds; and Benjamin H. Bratton’s The Stack — we see a range of approaches:

  • An examination of the physical infrastructure underlying virtual networks. Starosielski examines the undersea cables that continue to carry much internet traffic; while Hu looks at how fiber-optic network infrastructure has been “grafted” onto America’s aging railroad track system. In both cases, we see attention paid to the physical infrastructures of the internet that are often overlooked, if not purposefully hidden.
  • An exploration of how power plays across networked interfaces, infrastructures, and protocols — and particularly how the traditional laws of the nation-state become confused and overwritten across the liminal space of the web. Starosielski looks at cable stations across the Pacific, exploring past and present effects of colonized states; Hu examines what he calls the “sovereignty of data,” exploring how we “invest the cloud’s technology with cultural fantasies about security and participation.” And Bratton delineates, as I’ve noted earlier, the various layers of what he calls “the Stack.” In Bratton’s view, computation itself has become a “global infrastructure [that] contributes to an ungluing and delamination of land, governance, and territory, one from the other” (14).
  • A connection, drawn by Peters through what he calls “infrastructuralism,” of computing technologies and the environment. This involves partly a consideration of the effect of computing technology on the environment — what Bratton calls “the ponderous heaviness of Cloud computing” — and partly, through Peters’s book, a consideration of media as environment, as space and place through which we move.

Across all of these works, we see concerns over issues of power, capital and surveillance; the physical and commercial structures through which the phenomenon we refer to as “the network” is built; and the growing sense in which media and networked infrastructures have become constitutive of much of our experience in the world.

The cloud is blurring lines and connecting us in ways that have reshaped conventional boundaries. For instance, As Bratton considers issues of sovereignty, citizenship, the polis, and the network, he ponders the dividing lines between “citizen” and “user,” between subject and state, wondering whether the network itself provides for new understandings of citizenship. He asks:

What if effective citizenship in a polity were granted not according to categoriocal juridical identity but as a shifting status derived from any user’s generic relationship to the machine systems that bind that polity to itself?” In other words, if the interfaces of the city address everyone as a “user,” then perhaps one’s status as a user is what really counts. The right to address and be addressed by the polity would be understood as some shared and portable relationship to common infrastructure. Properly scaled and codified, this by itself would be a significant (if also accidental) accomplishment of ubiquitous computing. From this perhaps we see less the articulation of citizenship for any one city, enclosed behind its walls, but of a “citizen” (Is that even still the right word?) of the global aggregate urban condition, a “citizen-user” of the vast, discontiguous city that striates Earth, built not only of buildings and roads but also perplexing grids and dense, fast data archipelagos. Could this aggregate “city” wrapping the planet serve as the condition, the grounded legitimate referent, from which another, more plasmic, universal suffrage can be derived and designed? Could this composite city-machine, based on the terms of mobility and immobility, a public ethics of energy and electrons, and unforeseeable manifestations of data sovereignty . . . provide for some kind of ambient homeland? If so, for whom and for what? (10, emphasis added)

The questions seething through Bratton’s book — especially those around citizenship, subjectivity, and participation in the techno-sphere – embody the kinds of questions DHers might ask of infrastructure. As enormous forces of capital and computation engender new networked publics around us, to what extent are those publics built on the grounds of equity and social justice? As DHers participate in these new cloud polities, to what extent are we asking Bratton’s question, “for whom and for what,” as we do our work?

DH has always been wildly various and multivalent, and its practices and methods range widely (some see this as a feature; others, as a bug. Count me on the side of those who appreciate DH’s capacious frame). The increasing prevalence of the cloud in our lives and works offers us a chance to intervene in the systems of media and communication developing around us. We can and should ask where and how DH insight might best contribute to scholarly conversations around infrastructure.

One possibility is the work on large-scale text, sound, and image corpora that many DHers — Franco Moretti, Ted Underwood, Tim Hitchcock, Andrew Piper, Richard Jean-So, Mark Algee-Hewitt, Matthew Jockers, Tanya Clement, Lev Manovich, and many others — have been working on, often through larger infrastructural platforms such as the Hathi Trust. Surely, this work involves issues of the cloud, infrastructure, and culture, and surely it builds on methods central to, and perhaps in some ways unique to, DH work. DHers excel at contributing to and taking advantage of this kind of networked infrastructure for scholarly work – look at how a set of national and international scholarly infrastructure projects — such as DARIAH-EU, Compute Canada, Hathi Trust, and Europeana Research — are helping DH researchers do their work at scale and also to participate in larger public conversations.

These platforms, and this type of infrastructural work is important. And they may be the answer for DH as it thinks about cloud-based infrastructure. But – aside from the fact that, as I have argued elsewhere, large-scale text mining too often stands in the public mind as a synecdoche of what DH is and should be — this kind of infrastructural work is sometimes hampered by the complex rights issues that attend cultural heritage materials, and these platforms often have a somewhat problematic relationship to access, offering member institutions one set or quality of resources, and offering the public another. Such platforms often embrace a stance of political neutrality that may be inadequate to the increasing complexity of the cloud. And so — perhaps for those reasons, and perhaps because of the direction of my own work – I’d like to consider other possibilities for DH in the cloud, as well.

Earlier, I noted that the action of the “sync” – the point where the user connects to a cloud-based service such as DropBox, Gmail, iTunes, or Google Docs to upload and download files – as the quintessential act of the cloud. As DHers, we know and recognize that these systems do much more than update files – they check us in with that vast global network of users, update terms of agreement, provide companies with a chance to flag illegal downloads. The sync is as much an act of corporate surveillance as it is an act of routine file maintenance.

As DHers, we know this and to some degree accept this in the same way that we know and accept our participation in proprietary networks like Twitter. It seems at times a cost of living in a cloud-based world.

But when we think about what DH is and what it can be, and of how it might relate to the cloud, we might consider that DHers are, among academics, perhaps best suited to reshape the nature of academic research itself. This work — often described as scholarly communication — has focused on the creation of new publishing interfaces and platforms; on the extension of humanities work to include alternatives to text-based argument; on the use of social media and blogging platforms to share in-process scholarship in public ways; on the consideration of collaborative work in the humanities; and on a reconsideration of what scholarship itself is and should be.

Perhaps the great work of DH is to envision alternate infrastructures for technical and scholarly work that help divorce us from systems of entrenched capital, that help move us away from our shared dependence on the set of proprietary service platforms — Twitter, Facebook, Github, Slack, — that have dominated scholarly communication in the humanities (and digital humanities) to date, and to recognize this shared dependence as such. Perhaps a central mission of DH is to build alternate infrastructures that are premised upon social and political understandings of the cloud, as articulated at least in part through scholarship in new media and science and technology studies.

This kind of work could help us address one of the most ironic gestures we see in current critiques of DH: harsh, outraged attacks on the supposed “neoliberalism” of DH, delivered by scholars through commercial proprietary platforms like Facebook and Twitter, or through online publication venues that use clickbait-y headlines to capture page views in the attention economy. It’s hard to see how the platform of delivery of those critiques does not detract at least a bit from their bite.

And yet if we think we are immune from this problem ourselves, we are wrong — this is an issue that affects not just new-media scholars or conservative humanists; it is undeniably present in the digital humanities community, as well. Yes, we have Humanist-L, DH Q&A, personal academic blogs, and multiple scholarly journals that we use to share work in the field. Yes, we are building new venues for open-access publishing such as The Open Library of the Humanities. Yes, we are building out institutional and inter-institutional methods of conversation and connection such as MLA Commons and MLA CORE, not to mention institution-specific repositories.

But DHers also participate actively and enthusiastically in Facebook, Twitter, Instagram, and Slack, among others. Twitter has, for many, become the de facto meeting ground of the field. And there is an undeniable good here: a strong DH presence on these platforms has enabled DHers to share their work with larger publics. But they also suggest a missed opportunity for scholarly communication and a regrettable participation in the larger systems of capital accumulation that DH could potentially resist.

In “The Scandal of Digital Humanities,” Brian Greenspan’s response to “Neoliberal Tools (and Archives): A Political History of Digital Humanities,” published by Daniel Allington, Sarah Brouillette, and David Golumbia in the L.A. Review of Books, Greenspan argues that digital humanities work is fundamentally aligned against the “strictly economic logic’ of neoliberalism; he notes that much DH work resists the “pressure to commercialize” and in fact involves “either detourning commercial tools and products for scholarly purposes, or building Open Access archives, databases and platforms.” Greenspan remarks sardonically that that is why so many DH projects are “so often broken, unworking or unfinished, and far from anything ‘immediately usable by industry.'”

DH work, as Alan Liu and others have argued, presents to the academy a mode of engagement between the humanities and computational methods and tools that is self-reflexive and empowering. Building on that notion, Greenspan argues that:

DH involves close scrutiny of the affordances and constraints that govern most scholarly work today, whether they’re technical (relating to media, networks, platforms, interfaces, codes and databases), social (involving collaboration, authorial capital, copyright and IP, censorship and firewalls, viral memes, the idea of “the book,” audiences, literacies and competencies), or labour-related (emphasizing the often hidden work of students, librarians and archivists, programmers, techies, RAs, TAs and alt-ac workers).

“If anything,” Greenspan notes, “DH is guilty of making all too visible the dirty gears that drive the scholarly machine, along with the mechanic’s maintenance bill. . . .And that’s precisely its unique and urgent potential: by providing the possibility of apprehending these mechanisms fully, DH takes the first steps toward a genuinely materialist and radical critique of scholarship in the 21st century.”

To the extent that this critique can be baked into the building that digital humanists do — and I do think that that is one of the key aims of DH as a field and practice, particularly in the age of the cloud — Greenspan helps us see DH’s potential contribution to questions of infrastructure. Digital humanities work can indeed help us reposition the infrastructure of scholarship away from the formations we now have in place and towards a more purposeful, and more resistant digital humanism that is grounded not just in non-commercial practices, but in anti-commercial practices. In this way, the strength of DH, its ability to peek into the black box of technological platforms, can be strengthened and can help the academy as it faces the onslaught of techno-capital from all sides. The need for this kind of work is urgent, as the drumbeat of constant reductions in state funding, certainly felt here in England but also in the U.S., force institutions to adopt austerity measures of various kinds.

To some degree, DH is already doing this kind of work—and I don’t want to erase the important contributions of these projects by failing to mention them. For example, we can look at a range of representative projects — Zotero, which has provided an effective alternative to costly bibliographic software; Omeka, which has created an easy way to present cultural heritage objects; Mukurtu, which is designed specifically to take account of diverse cultural attitudes towards the sharing of heritage materials; Scalar, which encourages multi-modal and non-linear argument; Domain of One’s Own, which helps students familiarize themselves with hosting infrastructure and take some measure of control over their online persona; and a few of the projects I have been involved in — Manifold Scholarship, which is creating a new open-source platform for interactive scholarly monographs; Commons In A Box, which provides a free platform for academic social networks; DH Box, which opens DH computing software to communities without technical DH infrastructure; and Social Paper, which is planting seeds that may one day help us move away from Google docs. Across all of these projects are the beginnings of an infrastructure for shared scholarly work that offer alternatives to commercial environments and platforms.

And yet, as Miriam Posner has noted, “the digital humanities [still] actually borrows a lot of its infrastructure . . . from models developed for business applications.” For many, the mere fact that DH involves the kind of technical training that may be very much in line with marketplace demands is evidence of its complicity with the forces of neoliberalism in the academy. How can we ensure that the infrastructure DH builds is self-reflexive infrastructure for scholarly practice and communication; that its builders ask themselves Bratton’s question — for whom and for what — at every turn; that it foreground humanistic research questions and resist the persistent encroachment of capital into higher education?

I don’t have answers, but I can suggest starting points:

  • We need a re-articulation of DH technical practice as essentially reflexive endeavor. DHers tend to approach technological systems by seeking to understand them, to historicize them, to unpack the computational and ethical logics that structure them. This gives DHers a good starting place for building out more ethical tools for scholarly communication. But we need to make this case more powerfully to the public.
  • We need open and robust conversations about inclusive practices. As recent years have shown, DHers need to pay careful attention to the make-up of their own projects and conferences, seeking to counter the forces of structural racism and gender bias. We might move this conversation forward by consciously seeking to expand our project teams and ensure that our projects engage issues diversity and difference.
  • We should expand our notions of what we mean by infrastructure, Jacqueline Wernimont’s argument at DH2016 in her talk “Feminist Infrastructure as Metamorphic Infrastructure.” There, Jacque described a concept of feminist infrastructure that commits to people, that is built upon relational accountability, that embeds ideals of collaboration, collectivity, and care, and that, as Jacque notes about FemTechNet’s charter, foregrounds pedagogies that are anti-racist, queer, decolonizing, trans-feminist, and focused on civil rights.
  • We should continue to build infrastructures and infrastructural conversations that encourage the growth of global DH; Alex Gil’s minimal computing is a wonderful example of this, in that it is an infrastructural philosophy and set of technological platforms — such as Ed, his Jekyll theme designed to produce minimal textual editions. The Gacaca Community Justice archive that we heard about from Marilyn Deegan on Thursday is another wonderful example of this.
  • That we speak more about, and continue to think through, the kind of education and training that many of us provide for our colleagues and students at our universities, and to situate that work within the context of critical pedagogy, ensuring that when we teach our students, we do so by emphasizing humanities values. Our students need to use DH methods to explore and explicate ambiguity rather than to flatten it. I think we do this already, but our academic colleagues sometimes miss this point.
  • That we take seriously the proposition put forward by Geoffrey Rockwell and Stephen Ramsay that for digital projects to be taken seriously, they have to make arguments. And to the extent that they can make arguments in their conception and function, they will help explain what DH is and can be.

And that, I think, is the challenge for DH infrastructure: it needs to make an argument, and it needs to make an argument through its projects, as Tara McPherson argued in 2010. Many of the projects I mentioned above do just that — think of Alex Gil’s minimal computing, of that way it embeds an argument about access and infrastructure into its codebase. Think about James Smithies attempts to build a virtual research environment on a cheap and affordable Raspberry Pi. Consider the DH Box’s team to make DH tools available to institutions that don’t have reliable networked infrastructure. Consider how Commons In A Box offers academics an alternative to Facebook, and how it has been used by scholarly associations such as Modern Language Association to build out alternates to corporate sites like

There are limits, of course. DH Box, though it is available free software, currently runs through Amazon Web Hosting. Domains of One’s Own similarly is a project that is ultimately based on commercial web hosting space. As Tung-Hui Hu reminds us, network infrastructure is often literally laid on top of older commercial infrastructure. It’s hard to live completely off the commercial grid, to live on the bare wires of the network – especially if we want to be involved in larger public conversations. The cloud calls to us to sync with it, and that call is hard to resist.

And there are other challenges. Free software communities, at least those in the US with which I am most familiar, are dominated by white males and are not always welcoming to women and minorities (something that I think and hope is changing through organizations such as PyLadies, Black Girls Code, and similar organizations).

And the work is painful. We are using Twitter, and not Diaspora or, for a reason. The slick, seductive surfaces, the smooth user interfaces of commercial social media platforms are not just hard to resist — they are where other conversations are happening. Removing ourselves from those platforms would cost DHers exposure — and, were more academics to follow — would risk moving academic discourse even farther from the public sphere than it already is.

But as Eben Moglen pointed out in his talk “Freedom in the Cloud” — the talk that inspired a group of NYU undergrads to create the twitter alternative Diaspora — when we use Gmail, it comes with the “free” service of semantic email analysis by Google for Google; that when we get free email and document storage we get a “free” service which is “spying all the time.” That location-based tweets may be used to squash protest. We know this – everyone knows this – but we could do more to combat the force of the commercial cloud.

DH can and will be useful to the humanities and to the academy. But it has the opportunity to consider what the next generation of scholarly communication platforms is and can be. It has the opportunity, and perhaps the responsibility, to approach questions of infrastructure with political and social contexts in mind — to consider, for instance, how its infrastructure can be modeled, to use language from Elizabeth Losh, Jacqueline Wernimont, Laura Wexler, Hong-An Wu upon feminist values, embracing “genuinely messy, heterogeneous, and contentious pluralism” in its design. Or, to return to the cloud, to offer us new visions of what it means to sync with the cloud and with the world. DH can and perhaps should be a primary force for resisting the entrance of capital into the ecosystem of educational institutions, by insisting upon critical engagements with commercial technologies. We can and must interrupt the sync.

Resisting the smooth services of the corporate web — building tools, platforms, and communities that embrace core humanities values of discourse, dialogue, inclusivity, and intellectual exchange — perhaps represent another side of what Miriam Posner has called the “radical, unrealized potential of the digital humanities.”

Were we to engage in that work — and I think we are already doing it, just not as purposefully and mindfully as we might — we would in fact have made a significant contribution to the world and would perhaps help dissipate the storm clouds of our times.

* * *

I’m grateful to Lauren F. Klein, Kari Kraus, and Brian Croxall for their comments on an earlier draft of this paper.