Art Papers  

more from the
July/August 2014 issue:

Rob Horning

by Dan Weiskopf

It Was the Best of Times, It Was the Worst of Times:
Art Education
Reaches an Apex in the American South

by Stephanie Bailey

Mechanical Turk

by Will Becker

Picturing the Self
in the Age Of Data

Text / Dan Weiskopf

The visual arts provide fertile ground for charting the evolution of our changing notions of selfhood and identity. The self has been variously portrayed as stable or protean, autonomous or dependent on its sociocultural relations to others, an intangible abstraction or something embodied in flesh and bone—or simply as an absence, papered over with delusion. In the digital age, the truest portraits are drawn in data.

Zach Blas, Facial Weaponization Suite: Militancy, Vulnerability, Obfuscation, tableau vivant, June 7, 2013, San Diego, CA [courtesy of the artist, photo: Tanner Cook]

Traditionally, portraits were guided by the ideal of likeness to their subject—by the notion that "the human body," in Wittgenstein's words, "is the best picture of the human soul."1 An apt portrait would capture the appearance of a person in a way that produces recognition; having seen the picture, you would know its subject, and vice versa. Portraiture's representational tool kit is an expansive one. Good likenesses include more than bodily appearances, and good portraits differ from mere pictures of persons: they attempt to visibly capture an individual's distinctive and essential character. Invisible mental and moral qualities shine forth in natural signs manifest in the person's bodily traits. Posture, intensity of gaze, and other expressive details convey aspects of character, just as occupation, marital status, and economic class can identify a person's place in a matrix of relations through visual codes, icons, and symbols.

The stereotype of the portrait may be grounded in the figure, but throughout the last century, the genre has drifted deeper into abstraction. Francis Picabia's series of machine portraits dispensed with human forms entirely in favor of symbolic, mechanical proxies. In Here, This is Stieglitz Here (1915), the photographer's apparatus—that is, Alfred Stieglitz's camera—stands in as the best representative for the subject himself. Later, bioscientific works such as Gary Schneider's Genetic Self-Portraits (19971998) or Marc Quinn's A Genomic Portrait: Sir John Sulston (2001)—in which a sample of the sitter's DNA in agar jelly is mounted in stainless steel—took the bearers of identity to be images of chromosomes, enlargements of microscopic hair samples, retinal images, and even mounted DNA itself.

The notion of a "portrait" is thus sufficiently labile to admit potentially any abstract substitute that can convey identity—including those grounded in scientific theorizing about the nature of identity. Changes in our self-conceptions are driven not just by social, political, physical, or economic factors, but also by technology. As Picabia's portrait of Stieglitz suggests, the machines we operate shape our conception of who we are. Contemporary selfhood has inevitably been shaped by the emergence of that most ubiquitous technology: the networked computer.

LEFT TO RIGHT: Zach Blas, Facial Weaponization Suite: Mask - May 31, 2013, San Diego, CA, 2013; Zach Blas, Facial Weaponization Suite: Fag Face Mask - October 20, 2012, Los Angeles, CA, 2013 [courtesy of the artist, photos: Christopher O'Leary]

Ever more of our lives are lived out on networked screens, which are always ready-at-hand and receptive. These handheld rectangles of light come securely framed on all sides, imposing a visual boundary between the digital world within and the more solid world without. This divide fosters the illusion that our digital and corporeal lives are ontologically separate domains. In truth, they shade seamlessly into each other. In all our networked interactions we cast a "data shadow"—an outline filled in by accumulated transactions with social, government, and commercial websites, emails, texts, and swipes at ATMs and purchase points. By recording and analyzing this cloud of individually insignificant items, algorithms can predict our preferences in food and movies, spending habits, political orientation and subversive potential, sexual matters, and mental health. Our dense loops of interaction with these networked systems give rise to a novel form of selfhood that social media theorist Rob Horning has called the "data self"—an identity posited "by the synthesis of data captured in social media," information we project onto the network that can ultimately be traced back to us.2 Crucially, this data is not an inert excrescence; it derives from our actions, feeds back, and reshapes them through our affective tendencies to seek affirmation and membership. We are at once creators, custodians, and products of networked systems.

The data self is a Janus-like construct. Its public face includes the content of our Twitter feeds, Spotify playlists, and Facebook timelines, all of which are explicit attempts to craft a picture of ourselves for our collective audience. Yet behind this is a hidden face, which consists of the profiles secretly assembled on us by sites that offer us services, such as Amazon or Netflix. Even individuals who do not explicitly post about their tastes and interests are subject to consumer analysis by a host of predictive algorithms.

As networked databases and software become more sophisticated, so do the tools to sort and visualize their contents. These platforms, which originate in scientific computing, graphic design, and engineering, have been co-opted by artists exploring the aesthetic possibilities of data visualization. In the age of big data, visual practices that draw on these technologies give us the most characteristic modes of representing the self. We are partially constituted by our data, and these graphics may provide some of the "truest" portraits of the contemporary self.

Zach Blas, Facial Weaponization Suite: Face Off, tableau vivant, June 7, 2013, San Diego, CA [courtesy of the artist, photo: Tanner Cook]

For several years, American infographic designer Nicholas Felton produced a series of dossiers on himself under the title The Feltron Annual Report (20052012). Designed and printed in a glossy, corporate style thick with charts and statistics, the works summarize endless facts relating to the artist's life that year. They were collected through use of a specially conceived iPhone app, Reporter, that periodically buzzes to demand the user fill out a survey on his location, companions, activities, food consumption, and so on. In 2012 Felton entered a daunting 4,739 reports containing 47 MB of data.

Such a dossier recalls the work of unscrupulous marketers, and what is simulated here is the result of practices we all unthinkingly engage in online. In Horning's terms, "[d]ata is the authorized way to pursue self-knowledge in the networked society; the other means are [suspect], deluded or outmoded."3 The Feltron report contains the complete statistical profile of a single person in sufficient detail to uniquely distinguish him from all others: no one else could have precisely those clothes, companions, purchases, travels, and sleep patterns. If a portrait is a visual representation that lets us identify someone as a unique individual, then the report is a canonical portrait, painted in data. Of course, looking at it won't enable us to "visually identify" its subject, but other legitimate forms of portraiture have already left behind the question of likeness. The identification that data portraits enable requires seeing the subject through the "eyes" of software.

Other works mine public aspects of people's digital self-presentation. Golan Levin, Kamal Nigam, and Jonathan Feinberg's The Dumpster (2006) presents a visualization of 20,000 blog posts in which people describe "being dumped" or "dumping" their partner. These posts are classified by the poster's age and sex, and they are coded in a way that allows linguistically and emotionally similar breakups to be compared. Aside from its voyeuristic interest, a work such as The Dumpster captures the normalizing role that social media play. By accessing similar experiences we reassure ourselves that we fit in, that our experiences are statistically unremarkable. The clear visual display of failed romantic profiles enables us to quickly determine how weird our heartbreak should make us feel. Visualization enables data to more effectively play this regulative role.

Of course, the comforting knowledge of our statistical normality can equally well obliterate the sense of our uniqueness. German photographer Peter Piller's Frau Baum (2012)—in English, "Woman Tree"—plays with this tension. The work consists of a collection of digitally blurred photographs from online dating profiles in which women have uploaded pictures of themselves standing next to trees. The discovery that what is experienced as an individual mode of self-expression actually exhibits a type—a "dryad"—is disconcerting. But as more of our lives become logged in databases, algorithms for automatically searching text and images will discover even more similarities among us, and more recondite classifications. The more we share, the better these algorithms can organize us, and the more they will reduce the space within which we can live out our individual subjectivity.

Zach Blas, Facial Weaponization Suite: Feminist Imperceptibilities, 2014, from a conversation performance, The New Romantics at Eyebeam, New York, NY, April 25, 2014 [courtesy of the artist, photo: Christine Butler]

The existence of data-based art depends on the networks, software, and practices of sharing and collecting information that are often the very target of its critical barbs. It is no surprise that individuals at the forefront of aesthetics and data visualization—including Felton, as well as Brazilian scientist and designer Fernanda Vigas and scientist and artist Martin Wattenberg—have worked in the search and social media industries. This background gives rise to a form of technological complicity in which the artworks themselves may easily transform into low-grade forms of voyeurism and surveillance. Artist Liz Sterry's Kay's Blog (2012) exploits this creepiness: it is an installation that re-creates an unknown blogger's bedroom flawlessly from her posted photos and descriptions.

Some works, however, aim to mount a substantial critique of the role of networked databases in self-formation. Software designer and media theorist Warren Sack makes this interpretive proposal: "When you look at artistic projects that map out and visualize information, do not worry so much about whether they are pretty, beautiful, friendly, or easy to use. Instead interrogate them by asking what sorts of governance they support or reflect. Are they democratic or bureaucratic?"4

Many of these projects involve directly intervening in databases, feeding them misleading information in order to uncover their hidden workings. For Data Mining the Amazon (2003), Angie Waller created fake user profiles on and seeded them with texts popular among readers of different political orientations. She then mapped out the other recommended books and music for customers who shared that political stance. The resulting clusters of works give insight not only into the preferences of actual people with those preferences, but more importantly into the mechanisms the site uses to classify them.

These experimental interventions aim to reveal how the systems we interact with are structured to model us. The algorithms these sites use to offer recommendations, purchases, links, and friends are not passive. By suggesting that we are a certain kind of person—the kind who would like this product or that experience—they subtly nudge us into embodying it. Ian Hacking, a philosopher of science, refers to this phenomenon as a "looping effect," in which classifying and labeling people changes them to better fit precisely those labels. It may be tempting to outsource to these algorithms the "choice" of what sort of person one is to be. What is troubling, however, is the possibility of their inextricability from the development of the self, as ever more experience is quietly prefiltered to better suit the interests of content providers.

Zach Blas, Facial Weaponization Suite: Fag Face Scanning Station, 2013, from reclaim: pride with the ONE Archives and RECAPS Magazine at the Christopher Street West Pride Festival in West Hollywood, CA, June 8, 2013 [courtesy of the artist, photo: David Evans Frantz]

Other works blur a distinction between representation and creation already fuzzy in the context of the data self. To exist as a person in the digital age is, ostensibly, to be linked to entries in official databases, both government and corporate. British artist Heath Bunting's Status Project (2012) exploits this thesis. For around 500, Bunting offers a "synthetic off-the-shelf British natural person," complete with a name, official residence, mobile phone number, and enough supporting documentation to eventually open a bank account. Individuals' corporeal being, if any, is distinctly secondary to their existence as collections of data.

Such works fufill the mimetic ideal of a portrait or sculpture come to life. Outside the fantasies of Zeuxis, no merely pictorial representation could pass for real so successfully in our eyes. A self made of data, however, can pass perfectly when viewed by the software of the appropriate agencies.

The wholesale automated creation of hidden profiles on individuals by states and corporations is unsavory: data is collected without users' knowledge or consent and without mechanisms for opting out; it is used for purposes the subject may not necessarily endorse; and profilers often target vulnerable populations. Oppositional data artworks aim to upend these relations and apply the underlying technology to participatory and liberatory ends. In this vein, Sharon Daniel has created several community databases, including Need_X_Change (2003), Subtract the Sky (2004), and Palabras_ (2006-2007), designed to allow members of marginalized groups (such as IV-drug users in an Oakland, CA, clinic) to control their own self-representation by contributing to a database of photographic, audio, and video histories. In contrast with the categories produced by most algorithms, these databases are folksonomically organized, meaning that participants personalize their semantic structures by creating their own tags, links, and categories. Given the inevitability of our being classified, controlling the terms on which it occurs is crucial.

Portraiture began with the body, and much of our thought about the self remains anchored there. However, sociologist David Lyon notes: "More and more, bodies are, in an ugly but apt word, 'informatized'. In numerous surveillance situations, bodies are reduced to data, perhaps most obviously through the use of biometrics at borders."5 One strategy to counter biometrics is to render oneself physically unclassifiable. Zach Blas' Facial Weaponization Suite (2011-ongoing) achieves this by creating a set of masks derived from 3-D scans of dozens of people's faces. The scans are not averaged but are combined in more complex ways, resulting in an uncanny rippling, blobby appearance. The pink, inhuman surface of Blas' Fag Face Mask serves to biometrically erase the self by submerging it in a new type of collective facial image—a form of self-presentation grounded in solidarity. The body, too, returns to us transformed by data.

Zach Blas, Facial Weaponization Suite: Mask - November 20, 2013, New York, NY, 2013 [courtesy of the artist, photo: Christopher O'Leary]

Seeing data-based artworks as continuous with the representational practices and concerns of portraiture gives a provisional, imperfect handle on how artists working with databases have coped with the historically unique problems of identity that have arisen in the age of data. Media theorist Lev Manovich argues that databases are fundamentally a non-narrative, or antinarrative, way of organizing information: "[a]s a cultural form, database represents the world as a list of items, and it refuses to order this list. In contrast, a narrative creates a cause-and-effect trajectory of seemingly unordered items (events). Therefore, database and narrative are natural enemies."6

Data portraits, then, can no longer rely on a narrative understanding of their subjects' lives and character. Sometimes objects retrieved by querying a database can be assembled into narratives, but it is equally possible to retrieve a random collection of items with no such connection. Databases and search engines can encode similarities and categories that we cannot easily name or grasp, and these algorithmically generated classifications resist our everyday practices of interpretation. With no convenient storyline to grasp and organize their experience, the challenge for artists interested in questions of identity then becomes how to use these same tools of data manipulation to portray the strangeness of postnarrative selfhood.

Dan Weiskopf is Associate Professor of Philosophy and an associate member of the Neuroscience Institute at Georgia State University. His research focuses on the nature of representation in cognition, science, and art.

Interview: Rob Horning
by Dan Weiskopf

1. Ludwig Wittgenstein, Philosophical Investigations, trans. G.E.M. Anscombe, P.M.S. Hacker, and Joachim Schulte (Malden, MA: Blackwell, 2009), 366.
2. Rob Horning, "Google Alert for the Soul," last modified April 12, 2013,
3. Rob Horning, "The Rise of the Data Self," last modified January 25, 2012,
4. Warren Sack, "Aesthetics of Information Visualization," in Context Providers: Conditions of Meaning in Media Arts, eds. Margot Lovejoy, Christiane Paul, and Victoria Vesna (Bristol, UK: Intellect, 2011), 132.
5. Zygmunt Bauman and David Lyon, Liquid Surveillance (Malden, MA: Polity, 2013), 134.
6. Lev Manovich, "Database as Symbolic Form," Convergence 5 (1999), 85.

Retrospective | Special Events | Donate | Subscribe
Editorial | Contact | Advertise | About ART PAPERS

© 2014 ART PAPERS, Inc.