1Memory Machine Myth: The Memex, Media Archaeology, and Repertoires of Archiving
The Internet Is Not a Memory Machine
The Internet, the World Wide Web, and desktop and mobile digital telecommunications devices comprise a system of networked computing that is often framed as a giant memory machine, a comprehensive and infinitely expansive archive, which automatically saves users’ posts and emails; the sites they have visited; and the text, image, and video content they have uploaded, downloaded, emailed, or blogged/reblogged/tweeted/pinned/tagged. From this supposed total archive, users will presumably be able to retrieve elements and traces of (what will be) their digital histories far into the future.
However, the system of networked computing fails as an archive much of the time. The Internet, far from autosaving all that we do and share there, is what Wendy Chun (2008) calls an “enduring ephemeral.” “The internet may be available 24/7, but specific content may not” (167), Chun writes. In other words, the near-constant availability and functionality of the network itself may suggest that everything that traverses the network is permanent and durable, but this association between the persistence of the Internet and the persistence of online content is a delusion, a false equivalence. In this chapter, I will argue that when the Internet does work well as an archive, it is because of the initiative and interventions of what I call “techno-volunteers”: self-appointed, mostly nonprofessional individuals and collectives who regard some digital cultural productions and events as worth preserving, and who choose to devote their skills, time, effort, and often their own finances to constructing and maintaining online archives. Networked digital culture is “saved” for future generations, not primarily through any automated operations, but through the labors of human actors.
Origins of the Memory Machine Myth
The myth that the Internet, combined with networked computers, constitutes a massive digital memory machine is as old as the concept of the Internet itself. Vannevar Bush, who served in World War II as Director of the Office of Scientific Research and Development (OSRD) (in which role he coordinated the military research activities of scientists at government and private institutions throughout the United States), first proposed the research project that would become first the ARPANET, than the Internet, in an essay called “As We May Think” that appeared in the Atlantic in July 1945. In the essay, Bush outlines his idea for a device called the “memex,” which would be a desk-sized machine linked to a large network of information. The memex, writes Bush, will be a “mechanized private file and library,” “a device in which an individual stores all his books, records, and communications, and which is mechanized so that it may be consulted with exceeding speed and flexibility,” and will serve as “an enlarged intimate supplement to his [a human user’s] memory” (Bush 1945, sec. 6). Bush describes the memex as a large desk housing scrolls of microfilm, on which are recorded documents, notes, and “associative trails” (sec. 8), that is, links between specific documents, or mental pathways through sets of material, forged by the user (see figure 1.1).
Figure 1.1 A drawing of Vannevar Bush’s proposed “memex” machine. From LIFE, September 10, 1945, 123.
Bush’s memex concept influenced many of the scientists and engineers who architected the Internet and other networked computing technologies, such as J. C. R. Licklider, who wrote seminal papers in the early 1960s proposing interactive networked computers that would be operated through graphical user interfaces (GUIs), and whose 1965 Libraries of the Future quotes Bush’s “As We May Think” on its first page; Douglas Engelbart, whose Augmented Research Center team invented the computer mouse and did pioneering development on hyperlinking, networking, and GUIs; and Ted Nelson, who coined the terms “hypertext” and “hypermedia.” Nelson makes Bush’s impact on his thinking clear when he begins a 1972 paper on hypertext with the sentence “Bush was right” (Nelson [1972] 1991, 245) (this paper of Nelson’s is entitled “As We Will Think,” a direct homage to Bush’s “As We May Think”).1 Licklider, Engelbart, and Nelson’s work in the 1960s and early 1970s, answering the call issued by Bush in his 1945 essay, gave rise to the Internet and to the system of networked, graphical, hyperlinking, personal computing that forms the basis of contemporary digital culture.
Bush can thus be credited as one of the “fathers” of the Internet. As head of the OSRD, which oversaw the Manhattan Project and the manufacture of the first generation of nuclear weapons, he was also one of the “fathers” of “the bomb.”2 It is not coincidental, I think, that Bush’s “As We May Think” appeared in print just one month prior to the US military’s August 1945 release of atomic bombs over the Japanese cities Hiroshima and Nagasaki. At the time that he wrote “As We May Think,” Bush almost certainly had full advance knowledge of the plan to use atom bombs on Japan, and in the essay, he alludes to the horrors of nuclear warfare and its world-destroying potential—a potential that he knew the entire world would soon learn about—and contrasts it with the world-saving potential of the memex. Bush writes that US scientists, having banded together so effectively during wartime for “the making of strange destructive gadgets,” should now turn their talents toward “objectives worthy of their best.” In other words, having become creators of killing machines for use in a global armed conflict, one that “appears to be approaching an end,” scientists should, in the postwar period, turn their talents to the creation of machines that will have “lasting benefit” to humanity (Bush 1945, sec. 1). His proposal for such a machine is the memex. Bush states: “The applications of science … have enabled him [the human] to throw masses of people against one another with cruel weapons. They may yet allow him truly to encompass the great record and to grow in the wisdom of race experience. He may perish in conflict before he learns to wield that record for his true good. Yet, in the application of science to the needs and desires of man, it would seem to be a singularly unfortunate stage at which to terminate the process, or to lose hope as to the outcome” (sec. 8).
What may take place next in the course of human history, Bush seems to indicate, is the end of human history, that is, the end of the human. The termination of the species may be brought about by the invention of weapons of mass destruction, made possible by scientific research. But Bush proposes that, just as his network of American science and technology researchers has produced weapons that can usher in the final phase of human history—global nuclear war—so can the same network turn itself, in peacetime, to the purpose of augmenting human intelligence—with the memex—to the point that humans will be able to “wield” “the great record” for “his [the human’s] true good” (Bush 1945, sec. 8). Bush pits the memex against the atom bomb in a timed race: either human research will bring about the destruction of humanity, or it will increase humans’ capacity for learning and thought by such a degree that humans will, at last, be wise enough to serve their own “true good” (a world without war, perhaps?). In Bush’s worldview, both the atom bomb and the memex are all-encompassing technologies. Bush seems to perceive postwar humanity as confronting a binary choice: total war, or complete archive. One or the other will be humanity’s future, in Bush’s immediately pre-postwar prognostication.
The Memory Machine Myth after Bush
The Internet thus began as a fantasy of the perfect archive, a technology that would preserve the vast record of human knowledge in its entirety. That fantasy has accompanied the network from 1945 to the present, becoming pervasive in both scholarly and popular discourses.
Noah Wardrip-Fruin, commenting on Ted Nelson’s “dream of hypermedia” as it was laid out in Nelson’s groundbreaking Literary Machines (first published in 1980 and revised and published nine additional times between 1981 and 1993), points out that Nelson’s dream was in large part a vision of “the ultimate archive.” Writes Wardrip-Fruin (2000),
Consider the dream of hypermedia, put forth by Ted Nelson and others over the last three decades: That, in a not-so-distant future, we read and write (view and draw, hear and compose) most everything from and to a world-spanning network. That everyone have the ability to produce their own documents, and connect them with any other public documents. That the author may constantly create new versions of his or her own document, and individuals may create their own versions of any public document. … That historical backtrack and degradation-proof storage allows us to visit any version, any moment in the network’s history. To have the ultimate archive, and yet have each element of the archive constantly in process. Dynamism without loss. Impermanence enfolded within permanence.
Nelson’s hope continued Bush’s: they shared the dream that the system of networked computing would prove to be a more robust, capacious, complete, and durable archive than had ever existed before.
Today, it is not uncommon to come across academic texts that subscribe to the Bush–Nelson line of thinking about the Internet as an archive. For example, the authors of Digital Humanities write, “Ubiquitous networks have led and will continue to lead to evolutions in pedagogy precisely because they involve the outsourcing of memory. … We would be ignoring precedent completely if we assumed that the allatonceness of a vast and increasing digital archive accessible anywhere at any time will not affect the way that we learn” (Burdick et al. 2013, 25). The authors do not explain or defend their equation of the Internet with memory, but treat it as self-evident.
In popular discourse, the perception of the Internet as an automated archive circulates widely via truisms such as “The Internet remembers everything.” Adults warn youth that they must be circumspect in their speech and actions, since anything they say or do can be recorded and distributed online, reaching larger audiences over longer periods of time than any young person probably intends. “The Internet is written in pen” has become an oft-repeated saying. The notion that the Internet archives children’s and teens’ foolish antics is reminiscent of a much older type of threat directed at the young, that all their follies and missteps would be documented in their “permanent records.” A 2012 Huffington Post article by Hemanshu Nigam, former Chief Security Officer of News Corp and MySpace, employs this exact phrasing: “By our own conduct, we build a permanent record of everything we do online. Whether we want them to or not, family, friends, recruiters, employers, enemies and criminals may easily access our lives with a single click of a button. What might seem like a good idea at the time often leads to embarrassment and long-term personal and professional devastation. The Internet remembers, and that is a fact that we must remember too” (Nigam 2012). A host of similar blog posts and op-eds, warning of the impossibility of erasing documentation of one’s immature indiscretions from social networks, populate news, parenting, and career advice websites.
Warnings about the Internet’s auto-archival nature are sometimes couched within scenarios of tragedy, as in the blockbuster action film The Dark Knight Rises (TDKR), the final installment of director Christopher Nolan’s trilogy of Batman films. In TDKR, one criminal character, cat burglar Selina Kyle, voices her dismay that she cannot “start fresh,” as crimefighter Batman urges her to do. Selina laments, “There’s no fresh start in today’s world. Any twelve-year-old with a cell phone could find out what you did. Everything we do is collated and quantified. Everything sticks.” In this framing, the Internet-as-memory-machine dramatically undermines the ability of individuals to continually reinvent themselves, an ability that is presented as simultaneously a freedom and an imperative in Western postmodern, postindustrial, neoliberal society. Given that the constant flux of twenty-first-century global capital demands endless flexibility and unfixity from workers at every economic stratum—capital mines the willingness of workers to “start fresh” multiple times over the course of their lives—it is understandable that parents and educators are deeply concerned about what they assume to be the Internet’s built-in archival functions, as a young person’s adaptability, on which their academic and career success depends, may be thwarted at any moment by the system’s recollection of a wrong move from their past.
Warnings of the Digital Dark Age
However widely the myth of the automatically archival Internet has spread over the past seventy years, the fact is that the system of networked computing utterly fails as a memory machine. Professionals in the field of library and information sciences (LIS) have issued warnings about digital data’s tendency to degrade and disappear since the mid-1990s. They have collectively proclaimed that the Internet and computers do not constitute the greatest archive in human history, but rather the reverse. The current historical moment, they argue, may be a “digital dark age,” a time of which future generations will have scant records, owing to the short lifespans of our current digital platforms, devices, and applications (as compared to the lifespans of older technologies, such as paper).
As far as I can determine, the first note of caution about digital decay was sounded in 1995, when Michael Lesk, a member of the US Task Force on Archiving of Digital Information,3 gave a talk called “Preserving Digital Objects: Recurring Needs and Challenges,” at Australia’s Second National Preservation Office (NPO) conference on Multimedia Preservation. Lesk told a tale of archivists putting their faith into new media that appeared to be betraying them. The abstract of his presentation reads, “Acid process wood pulp paper, used in most books since about 1850, … threatened cultural memory loss. But digital technology seemed to come to the rescue, allowing indefinite storage without loss. Now we find that digital information too, has its dark side” (Lesk 1995). In other words, Lesk argued, digital storage media at first appeared to be archivists’ ideal solution to the degradation of paper, but then turned out itself to be highly degradable. After Lesk’s talk came a report by Paul Conway (1996), then head of the Preservation Department of Yale University Library, which pointed out, “Information in digital form—the evidence of the world we live in—is more fragile than the fragments of papyrus found buried with the Pharaohs,” because “the permanence, durability, and stamina of newer recording media” have declined steadily over the course of the twentieth century, making the digital age one in which we have “information density” but few options for permanently preserving that information and keeping it accessible. One year later, in 1997, Terry Kuny appears to have coined the phrase “digital dark age(s)” in a paper presented at the Sixty-third IFLA (International Federation of Library Associations and Institutions) Council and General Conference. Kuny (1997) argued that “we are moving into an era where much of what we know today, much of what is coded and written electronically, will be lost forever. We are, to my mind, living in the midst of digital Dark Ages” (1).
Then, in the late 1990s, voices outside of the LIS professions—from the mainstream press, the business world, and humanities research—joined in the chorus cautioning that digital technologies are often antiarchival. In 1998, science and technology journalist James Gleick wrote in a New York Times Magazine story, “Many of the world’s librarians, archivists, and Internet experts are warning that the record of our blooming digital culture is heading for oblivion, and fast. They note that we have already begun losing crucial scientific data and essential business records. … In the electronic era, we are stockpiling our heritage on millions of floppy disks and hard drives and CD-ROMs. These flaky objects go obsolete dismayingly fast, with new technologies rolling in on product cycles as short as two to five years” (Gleick [1998] 2002, 197). In 1999, Stewart Brand, an early new media entrepreneur who founded the influential Whole Earth magazine, published an essay in Library Journal called “Escaping the Digital Dark Age.” Brand called for “a long-term strategy for storage,” a remedy to the fact that “there is still nothing in the digital world like acid-free paper.” Brand quoted an admonition from Peter Lyman, a librarian at the University of California, Berkeley: “We know there is a 500-year life to microfilm properly cared for. But what do we do with digital documents? … We need a digital equivalent to microfilm, a 500-year solution” (Brand 1999, 46).
One of the most widely cited critics of the myth of the archival Internet is Wendy Chun, who dissects Bush’s “As We May Think,” criticizing Bush’s prognostication that a future technology of networked memory will “make possible the overarching archive of human knowledge in which there is no gap, no absence—a summation of human knowledge” (Chun 2008, 159). Far from fulfilling Bush’s goal of a perfect archival technology, writes Chun, “Digital media is not always there. We suffer daily frustration with digital sources that just disappear. Digital media is degenerative, forgetful, erasable” (160).
Chun characterizes networked computer memory as not wholly forgetful, but as unreliable and unpredictable in what it remembers. Reconciling the “Internet remembers everything” narrative with the digital dark age narrative, Chun states, “If things constantly disappear, they also reappear, often to the chagrin of those trying to erase data” (Chun 2008, 167). The system of networked computing sometimes resurrects un-looked-for, unwanted information, and often deletes valuable and longed-for information. Vannevar Bush’s perfect memory machine has never existed.
The Wayback Machine: The Real Memex?
And yet, even in digital dark age discourse, one finds echoes of Bush’s belief that a combination of the Internet and programmable computers will produce an automated archive. Chun ascribes a kind of ultimate saving power to a different “machine” than Bush’s memex: the Internet Archive’s Wayback Machine. In 1996, the Internet Archive, an organization I will describe in more detail in chapter 2, began developing software to crawl and download all publicly viewable websites and take “snapshots” of them, in order to record the websites at different intervals over their life spans. This software, and the service that stores, preserves, and makes accessible what it “harvests,” is collectively called the Wayback Machine. Chun writes, “Like search engines, the Internet Wayback Machine (IWM)4 comprises a slew of robots and servers that automatically and diligently, and in human terms obsessively, back up most webpages. … However, unlike search engines, the IWM does not use this data to render the internet into a library but rather uses the backups to create what it calls a ‘library of the Internet’” (Chun 2008, 168). Chun, quoting from the Internet Archive’s website, compares the Wayback Machine to libraries, and emphasizes the Wayback Machine’s mission of maintaining digital cultural memory:
The need for cultural memory drives the IWM and libraries more generally. Noting the loss of early film archives due to the recycling of early film stock, the archivists state that they are building an “internet library” because “without cultural artifacts, civilization has no memory and no mechanism to learn from its successes and failures.” … The IWM is necessary because the Internet, which is in so many ways about memory, has, as [Wolfgang] Ernst argues, no memory—at least not without the intervention of something like the IWM. (Chun 2008, 168–169)
It is curious that, despite her opposition to Bush’s idealistic goal of designing a comprehensive memory technology, Chun describes the Wayback Machine as a mechanism that accomplishes Bush’s hopes. The line that she excerpts from the Internet Archive’s website—“without cultural artifacts, civilization has no memory and no mechanism to learn from its successes and failures”—reiterates Bush’s sentiments quite precisely. The storage of civilization’s memory in machines that will permit humans to review their entire collective history and their accumulated knowledge, in order to plan and execute the best possible future for humanity, is exactly Bush’s dream for the memex.
Chun acknowledges that the Wayback Machine does not capture every detail of every webpage that it archives, “because webpages link to, rather than embed, images, which can be located anywhere, and because link locations always change, the IWM preserves only a skeleton of a page, filled with broken—rendered—links and images” (Chun 2008, 169). But despite the Wayback Machine’s imperfect recall, Chun nevertheless attributes to the Machine the same all-encompassing archival power that Bush attributes to the memex. Where others regard the Internet as the descendent of Bush’s memex concept, Chun recognizes the Wayback Machine to be the only “real” memex, or memex made real (in other words, in Chun’s view, the Internet is not the memex, as most people think—but the Wayback Machine is a means by which the memex’s aims are accomplished). Positioning the Wayback Machine in relation to the digital dark age discourse summarized above, Chun writes, “Blind belief in digital memory threatens to spread [a] lack of memory everywhere and plunge us … into the so-called digital dark age. The IWM thus fixes the internet by offering us a ‘machine’ that lets us control our movement between past and future by regenerating the internet on a grand scale” (169).
I greatly admire the Internet Archive’s Wayback Machine and appreciate that the Machine’s snapshots of websites at different points in time makes it possible for online users to occasionally visit websites that have been taken down, or to see websites in previous incarnations, even if the “saved” versions are, as Chun points out, somewhat “lossy” (with missing images, broken links, etc.). However, I dispute Chun’s idea that the Machine successfully preserves digital cultural memory in an automated fashion. Both Bush and Chun fail to recognize that automated archival machines can be easily defeated by their own, or other machines’, operational assumptions. For example, the Wayback Machine’s policies cause it to delete its records of virtually all websites whose original domain owners have ceased to renew those domain names, which results in losses on a far greater scale than Chun describes. One online fan fiction archivist, whose fan pseudonym is Morgan Dawn (2012) and who was interviewed for my oral history project, explains this particular failing of the Wayback Machine: “Once you lose your domain ownership—so let’s say ‘Morgan.com,’ I allow it to lapse. Somebody, usually bulk resellers, will snap it up, and they’ll park a [new web] page [at that URL]. ‘This domain available for sale.’ They will always put a robot text [robots.txt] file on that parked page. The Internet Archive, Wayback Machine, honors robot text files and next time it scans my—‘my’—former website, it will see the robot text and it will retroactively remove my entire history. And this has been a problem that’s been known since 2007.” Reinforcing Morgan Dawn’s assessment, a blog post on the Economist website in January 2014 calls robots.txt “the [Internet] archive’s kryptonite” (Fleischman 2014). One of the driving questions of Chun’s (2008) essay is, What leads to “the resuscibility or the undead of information” (171)? The Wayback Machine, though it sometimes proves very useful for the recovery of “dead” websites, also adheres to policies that cause sites to be erased from its index, sites that it had previously recorded, rendering it an untrustworthy archive that will likely become more unreliable as time goes on, as more domain names expire and more gaps in the Machine’s “memory” appear.
Fears of a digital dark age are therefore not necessarily opposed to the myth of an automatically archival Internet. Chun seems to suggest that one answer to the ephemerality of online data is to automate the periodic saving of data through some memex-like technology. In other words, Bush’s goal of making a networked technology into a comprehensive archive may have failed, but other automatic archiving machines, however imperfect, may succeed in preserving the cultural memory of these early years of the digital age. A wish for a technological solution to the constant disappearance of human knowledge, culture, and experience lies at the heart of both the myth of the Internet-as-archive and the dread of a digital dark age.
What is missing from both of these discourses is the figure of the human archivist.
Techno-Volunteerism
Automated archival technologies acting alone have not, to date, yielded exceptional results. It is human archivists, working on and with digital tools and networks, that make digital ephemera endure—when it does endure, which is not often. Matthew Kirschenbaum writes,
As electronic objects begin to accumulate archival identities (by virtue of the libraries, museums, and other cultural repositories increasingly interested in or charged with collecting them), it will become essential to understand … where the most significant challenges of digital preservation finally lie. … Those challenges, while massively technical to be sure, are also ultimately—and profoundly—social. That is, … effective preservation must rest in large measure on the cultivation of new social practices to attend our new media. (Kirschenbaum 2012, 21)
Here Kirschenbaum argues that many of the answers to complex questions about digital preservation lie not in the technical end of archiving, but in the social aspect—in people. People must work together to preserve “electronic objects”; technology alone will not accomplish it. Nonprofessional Internet archivists have dedicated thought, intent, and time to developing the “new social practices” of archiving for which Kirschenbaum calls. At some point in each of their lives, these people decided that what they wanted to do with their spare time was to construct and maintain Internet archives of cultural content. I call these self-designated digital archivists “techno-volunteers.”
I will use the terms “techno-volunteers” and “techno-volunteerism” to signal a distinct break from the technological determinist (or “techno-determinist”) thinking that has suffused theories of digital archives from 1945 to the present. Technological determinism is a school of thought that emphasizes the agency of technologies in moments of sociocultural change; a simplistic summary of a techno-determinist outlook is, “Shifts in technology cause shifts in society and culture.” The methodological school most often opposed to techno-determinism is social constructivism (or social construction of technology [SCOT]), which focuses on how humans’ agency and social structures decide what kinds of sociocultural impact specific technologies have; social constructivism’s preferred object of analysis is “The social shaping of technology,” a framework initially proposed by Donald MacKenzie and Judy Wajcman ([1985] 1999). According to this perspective, “Our technology of production is in many ways the result of our social relations,” and “technical choices are simultaneously social through and through” (143).
New media studies usually frames the beginnings of the techno-determinist vs. social constructivism opposition as a battle of ideas between Marshall McLuhan (cited as the preeminent technological determinist) and Raymond Williams (as the primary advocate of the social construction perspective) that took place in the 1960s and 1970s. Narratives about this debate then point to the 1980s development and launch of actor-network theory (ANT) by Bruno Latour, Michael Callon, and John Law as a kind of resolution of the dichotomy, as ANT awards the same possibility for agency to all actors, or actants—both humans and nonhumans—that combine to form a sociotechnical system. Nonhuman actants may not have intentions, as humans do, but they nevertheless “make others do things” (Latour 2005, 107); for example, speed bumps cause human drivers to slow their vehicles, and so must be taken into account as significant players in social action. In his study of cell phone usage, Gerard Goggin articulates the viewpoint shared by many new media theorists, that ANT offers a middle ground between techno-determinism and social constructivism: Goggin writes that ANT revises “formulaic oppositions between technology and society,” as it refuses to subscribe wholly either to the deterministic position or to the “countervailing reaction that society determines technology” (Goggin 2006, 11). Over the past thirty years, ANT has usefully drawn attention to “the necessity of a composition of forces to explain [an] action” and has productively asserted that humans and nonhumans are always co-actants, with frequent variance as to which actant, and which type of actant, is the “prime mover” in an action (Latour 1994, 35).
However, in studies and theories of archival technologies, determinism has always been a popular stance. Above, I described how Bush, Bush’s adherent Nelson, Bush’s critic Chun, and numerous popular writers and texts, have all promoted an idea of the Internet as an automatic archive. Wolfgang Ernst, outlining his proposal for a method of “media archaeology,” conceives of recording technologies, beginning with the photographic camera, as “registering the past coldly, in contrast to painterly animation and historical imagination” (Ernst 2013, 47)—painting and history writing being two modes of recording the past that preceded photography. Here is another of Ernst’s descriptions of photography as a memory machine: “With the emergence of photography, the idea of the theatrical gaze literally staging the past is displaced by the cold mechanical eye, a technologically neutral code rather than a subjective discourse” (Ernst 2005, 592). While photography is a vastly different technology from the Internet, they are both rendered into devices that record the past in a “mechanical” and “neutral” way in the discourse of Bush and his successors and in the discourse of Ernst and media archeologists.
While human decision and action are presumably necessary to activate the memex’s and the photographic camera’s documenting and storage functions, Ernst downplays the participation of the human user to the point of claiming that photographs are a non-“subjective discourse,” and Chun frames the Internet “snapshot”-taking system of the Wayback Machine as non-reliant on human choice or intent, as do Nigam and others who perceive the Internet as a “permanent record.” Even when Ernst writes directly about Internet archives, rather than memory technologies such as photographic cameras, he uses a language of automaticity; for example, discussing “born-digital media art,” Ernst asks, “How does dynamic art archive itself?” (2013, 82).
In this book, I will relate the stories of many digital archivists and archive users to reveal how the old category of “archives” is being renewed in and through new media, generating new cultural forms and conflicts. By drawing heavily on my research team’s interviews with human actants in digital archival systems, I will be employing a “theatrical gaze,” which Ernst is eager to shunt aside in favor of the “cold gaze” of media archeology (see Parikka 2013, 8), in “staging the past” of digital archiving for the reader. While much valuable work on media technologies has resulted from Ernst’s cold gaze, with its attention to the physical materiality and engineered operability of machines, a theatrical gaze will capture the participation of human as well as nonhuman actants in the history of Internet archiving. The efforts of what I call techno-volunteers have been so crucial to online archiving that to ignore or marginalize this labor would be to completely misperceive what it means, and what it takes, to “save” the Internet.
Repertoires of Digital Archiving
To say that I advocate techno-volunteerism over techno-determinism as a lens for understanding digital memory means that I wish to emphasize that Internet culture is best preserved by self-designated archivists who perform the labor required to create and sustain online archives, rather than by any built-in functionalities of the network or by any software system.
I argue that techno-volunteers have managed to archive online cultural production through developing archival repertoires, that is, through practices and ways of doing that are passed from person to person. “Repertoire” is a mode of memory that, as Diana Taylor (2003) points out, is often held to be the antithesis of the memory mode of “archive” (19–22). “Repertoire” transmits knowledge through processes of embodied mimesis, one person imitating what another person does, while “archive” transmits knowledge through recording technologies, such as handwriting, printing, sound records, photographs, film, and so on, one person decoding the knowledge that another person has encoded in fixed form. But Taylor resists the notion that the two modes of transmission form a dichotomy, and points out that archive and repertoire frequently cooperate or operate simultaneously.
I argue that techno-volunteers have developed ways and means of archiving the Internet that form the backbone of digital cultural memory. Without these practices, few, if any, digital archives would be in existence. Thus, if the Internet is ever usable as archive, it is made so by repertoire. Internet archiving depends on techno-volunteers’ archival repertoires. Thus, it cannot be said that print culture’s elevation of archive as a superior memory mode has persisted, unchallenged, into the twenty-first century, as digital culture has become more prevalent. Techniques of repertoire have, so far, mattered at least as much as the technics of archives in the preservation of digital networked culture.
Archiving with digital technologies requires human enactment of the archivist’s repertoire. (Also, digital archives require a great deal of non-repertoire-based performance: online archiving has been so experimental throughout its first few decades that it often demands creative improvisation, and archivists sometimes wish they had more of a repertoire to rely on.) Print culture opposed archive to repertoire and assigned very different values to them; it placed archival modes in a privileged position over repertoire, linking notions of objectivity, facticity, fixity, whiteness, and modernity to archive, while relegating repertoire to the lesser position, associating repertoire with unreliability, subjectivity, fluidity, non-whiteness, and ancientness or “tradition.” In the digital era, repertoire factors so heavily in the making and sustaining of digital archives that its inextricability from the archival mode, and its significance in processes of cultural preservation, cannot be contested.
One might say that print and analog media archives were, and still are, just as dependent on repertoire as digital archives. After all, human archivists have always been necessary for the archiving of cultural materials—a major objective of LIS education programs is to transmit the repertoires of archival labor required by traditional memory institutions to successive generations of workers. But the elaborate repertoires and rituals of librarians and archivists have tended to remain what Erving Goffman ([1956] 1959) would call “backstage” at memory institutions. Digital archives, in contrast, have foregrounded archival repertoires, for three major reasons.
First, because the first generation of Internet archivists have brought these repertoires into existence, their practices and ways-of-doing are more evident to users than the regularized, professionalized practices of LIS workers at traditional archives and libraries. Even if digital archive users do not “see” the vast majority of the work performed by archivists, users do see the differences between various archivists’ methods, the upgrades and alterations that archivists make to their sites, and the moments when archives break (for example, when archives lose massive amounts of data) or succeed (for example, when archives experience large influxes of new content or recover data that was presumed lost)—in other words, the irregularities and lack of standardization between digital archives at this time, when the entire genre of online archives is still so young, announce the fact that humans are driving these archives, and those humans are still in the process of inventing and refining the repertoires of digital archival work.
Second, archivists often do a great deal of what Goffman ([1956] 1959) would call “front stage” work, representing their archives to the public and interfacing directly with users, and in general serving as the “face” (or, at least, the name and email address) of their archives. However, oftentimes only one archivist—the lead or founding archivist—is the primary contact for the archive’s users; in many cases, a team of workers is required to support an Internet archive, and this support staff’s work often is invisible to the public. Also, even if users know who the archivists of a given repository are, and frequently communicate with them, they may know nothing of, or severely underestimate, the type and quantity of labor that the archivists put into keeping their archives up to date and their interfaces easy to use.
The third reason that archival repertoires are more prominent in digital archiving than in print and analog archiving is the tendency of digital archives to fail. As I have been emphasizing, building on Chun’s “enduring ephemeral” argument, digital data is so prone to disappearance that constant intervention is required to refresh data storage and keep it retrievable. At the least, data must be migrated to new servers when old servers cease to function optimally. Also, the rental costs of server rack space must be paid, ownership of website URLs must be renewed, sites should be mirrored (redundancy is one of the best methods for staving off accidental data disappearance), and when a lead archivist decides to quit her archival responsibilities, she should recruit her replacement(s) and oversee the smooth transition of the archive into new hands. If librarians had to not only enter books into their records and put them on the shelves of a library, but also had to move the entire library to a new building every few years, and also had to pay the rent and other fees associated with having the library building in the first place, and also had to make sure that for every single book in the library, a copy existed in another library with which they were in direct contact, and if librarians had to personally designate their replacement before resigning their post, or else risk the closure of the entire library, then both librarians and the library-going public would be far more conscious of how much repeated human labor and intervention—which I am calling archival repertoire—goes into the maintenance of a library. (Of course, the type of migration and mirroring labor that I am describing would be far more physically taxing, and would require much more time and many more financial resources, for employees at a brick-and-mortar library than it is for digital archivists, but digital archivists must still devote significant volumes of energy, time, and money to their repositories—just because the digital archive’s repertoire is physically easier, and cheaper, to enact than the built-library repertoire does not mean that it isn’t work.) Users of Internet archives know that archivists are needed to make those archives operational—if only when the archivist goes on a vacation, gets too busy with his or her “real” job or personal life to actively maintain the archive, or leaves the archive altogether because it becomes too much of a drain on his or her personal resources. The day that an online archive dies, or goes dark (that is, becomes inaccessible to the public), is the day when an archivist ceases to enact the repertoire, and this is usually the day when the user base becomes painfully aware of the repertoire’s existence, and the necessity of the repertoire’s repeated performance.
Repertoires and Scripts in New Media Studies
I am not the first to borrow the term “repertoire” from performance theory and apply it to new media phenomena. The field of digital design studies has thoroughly incorporated “repertoire” into its vocabulary, often using it to refer to the sum of practices and ways-of-doing that designers employ in their work (for example: “A rich repertoire of templates and a developed language for design qualities are two essential components of professional design ability” [Löwgren and Stolterman 1998]; “A repertoire of practice refers to the sum of available tools, techniques, strategies, tactics, ways of working, expertise and know-how from which a practitioner may draw, choose from, and/or combine to suit both known and novel situations or address a particular purpose” [Burrows and Morgan 2010]). This translation of design activity as “repertoire” has been especially useful for design pedagogy, as instructors can encourage students to build their repertoires by demonstrating their own “repertoires of practice” or “repertoires of experience.” (For instance, the abstract for a paper by design educators Philippe Saliou and Vincent Ribaud [2004] reads, “Performing good design is a difficult task. To take up this challenge, practitioners rely on their repertoire of experience. Students, however, do not have any such repertoire. We propose an approach aimed at bootstrapping the repertoire.”)
However, a richer use of “repertoire” in design studies can be found in Erling Björgvinsson, Pelle Ehn, and Per-Anders Hillgren’s 2012 essay “Agonistic Participatory Design.” Björgvinsson, Ehn, and Hillgren write about the struggles that often arise among participants in an emerging system, and stress the importance of the designer’s leadership in negotiating the various demands placed upon the system: “The design researcher role becomes one of infrastructuring agonistic public spaces mainly by facilitating the careful building of arenas consisting of heterogeneous participants, legitimizing those marginalized, maintaining network constellations, and leaving behind repertoires of how to organize socio-materially when conducting transformative innovations” (Björgvinsson, Ehn, and Hillgren 2012, 143). Substituting “digital archivist” for “design researcher” in this formulation enables me to surface a number of core components of digital archival labor:
The digital archivist engages in infrastructure building, or infrastructuring.5
The archivist infrastructures public spaces that are often “agonistic,” that is, engaged in conflict (see Mouffe’s [2000] argument in The Democratic Paradox that “agonistic pluralism” is a better model for democratic interaction and action than consensus building).
A large part of the archivist’s task in building these spaces is moderating the conflicts that arise, such that “marginalized” opinions are registered and taken into account and the “network constellation” made up of “heterogeneous participants” remains operational despite the heterogeneity.
Therefore, the archivist works from a repertoire, a repertoire of how to organize a system socially and materially (that is, technically), a repertoire of how to bring new digital archives into being, how to grow them, and how to sustain them.
The repertoire of digital archiving, from this perspective, consists at least as much of managing human relations, including arbitrating majority/minority disagreements, as it does of providing technical services. As an Internet archivist succeeds in enlarging the public space (the archive) she has constructed, and in dealing with the space in a way that honors its “agonistic pluralism” on a continuing basis, the archivist not only adds to her own, individual repertoire of how-to-do-digital-archiving, but potentially to other Internet archivists’ repertoires as well: like design students, digital archivists learn from observing one another’s repertoires of practice.
Design studies’ application of “repertoire” to infrastructuring—the notion that infrastructuring emerges from, and enhances, a designer’s (or archivist’s) repertoire, and that it involves negotiating between people as well as solving technological problems—reinforces my perspective that the growing phenomenon of digital archiving is better understood when viewed through the lens of techno-volunteerism rather than the lens of techno-determinism. Techno-volunteerism emphasizes the role of human actants in the creation of systems that preserve digital content and Internet content. Archivists are, as I have argued, incredibly important actants in the building of online archives, and their repertoires must include not only technical skills but techniques for receiving and processing users’ feedback in the form of demands, complaints, requests, and compliments. In addition, this ongoing dialogue between designers/archivists and archive users illustrates that users of online archives are themselves techno-volunteers whose interactions with online archives help to shape them.
The recognition that users contribute to the design of platforms has been fostered by the subfield of design studies called “Participatory Design” and has also been productively discussed as “collaborative media” by Jonas Löwgren and Bo Reimer in their 2013 book by that title. Löwgren and Reimer (2013) attempt to put to rest the question of technological determinism versus social constructivism by asserting, “People using collaborative media products and services—people who produce media texts with the help of the products and services and people who consume texts—continuously take part … in the design process” (148). In other words, the “uses” of a technology are never inherent in that technology, and are never completely fixed in place by the designers of that technology, but arise, over time, from interactions between the technology itself, its original designers, and the users who become its codesigners over time. Löwgren and Reimer point to another appropriation of performance language by new media studies—the concept of “scripts”—to make their point that “The designer [is] one actor in a large participatory process, not the actor” (144). Löwgren and Reimer cite ANT theorist Madeleine Akrich’s argument that, “like a film script, technical objects define a framework of action together with the actors and the space in which they are supposed to act” (208). Akrich does not mean to imply that the “script” that technical objects give to actors wholly determines the performance of the actors with those technologies; like actors in a stage or media production, there is “incessant variation” between “the designer’s projected user and the real user,” and “the user’s reactions … give body to the designer’s project” (209).
Affordances Need Performances
What Akrich describes as a “script” written into a technology’s workings and interface, most designers would call a set of “affordances.” Löwgren and Reimer credit Donald Norman for importing the term “affordance” into the field of interaction design in 1988, “as a way to understand what it is in a thing that makes it interesting or relevant for a potential user” (Löwgren and Reimer 2013, 25). Yochai Benkler (2006) lauds designers’, engineers’, and STS (science and technology) scholars’ enthusiastic adoption of “affordances,” as Benkler argues that the concept allows technology theorists to move away from “a naïve [technological] determinism.” Benkler writes, “Different technologies make different kinds of human action and interaction easier or harder to perform. … Neither deterministic nor wholly malleable, technology sets some parameters of individual and social action. It can make some actions, relationships, organizations, and institutions easier to pursue, and others harder” (17).
But it is possible to move even more definitively away from techno-determinism than Akrich and Benkler do. Rather than being concerned with what is “in” a technology that makes human action/interaction “easier or harder to perform,” we can emphasize that affordances need performances. Affordances for archiving may inhere in networked computing—Bush insisted that computing have the potential for vast information storage and retrieval in “As We May Think,” and his essay influenced subsequent hardware and interface designers to develop this potentiality—but archival affordances do not bring archives into being. They may make building archives “easier to pursue,” as Benkler would say, but as this book will detail, the building and (especially) the maintaining of digital archives is not at all easily done; little is easy about the task of online cultural preservation. Technologies’ affordances, or “scripts,” may suggest uses for those technologies, but it is the human actor—let us replace “actant” with the older and more suggestive term “actor” here—who must perform those scripts, and it is the human body that “gives body,” as Akrich says, to the potential inherent in a technological system. The human actor’s performance—the mind’s decisions and the body’s actions—actualizes what is virtual in machines (their possibilities, their theoretical capacities), and because volunteerism—human will—is involved, the actuality never maps directly onto the virtuality. That is, there is never a precise, one-to-one correspondence between a technology’s virtual aspects, encoded by its designers, and the technology’s actual functioning, as performed by users.
A script is never, and can never be, performed exactly as intended, even if the intentions of the “authors” are incredibly clear (which is hardly ever the case). Rather, a script is always brought to life by, in, and through human performance differently than the script’s writers can predict. In fact, like every script, technologies’ scripts are performed differently by every human actor every time a performance takes place. The laws of performance (different-with-every-actor, different-every-time, even if there is an underlying resemblance or repetition between iterations as every actor who engages with a specific technology works with the same “script”) trump the theory of technological determinism, in the case of digital archiving and in the case of much technological use.
Concepts and terms of performance, as employed in design studies’ and ANT’s theories of technological development, thus highlight the fundamental parts that people play in creating infrastructures of digital cultural memory. “Performance” connotes modes of transmission that are not fixed (as are text and recorded media), but are processual and evolving, that are repetitious but are also unique in each instance—and so is an apt descriptor for how digital archiving currently occurs. “Performance” implies that human actors must embody and execute scripted functions—and so is useful as a metaphor for the necessary collaborations between humans and nonhumans that produce digital archival infrastructures. The readiness with which some branches of new media studies are appropriating the language of performance indicates an opportunity for performance studies to assert its centrality in the digital age. But it also challenges performance scholars to decouple performance from liveness and physical presence, and to turn their theoretical attention to the forms of repertoire and enactment that are arising in and through new media, forms that do not depend on face-to-face, real-world contact.
Although many have dreamed that computing machines could be built to house and preserve human knowledge, or have assumed (hopefully or anxiously) that the Internet and networked computing systems automatically save what users publish online, Vannevar Bush’s ultimate memory machine is still a myth. The digital archives that techno-volunteers construct, drawing on their repertoires of practice and experience, and building on those repertoires as they learn to negotiate the divergent needs and desires of their users and to navigate constant server-side and coding challenges, are currently the only persistent, reliable forms of digital cultural memory. These archives are only as persistent and as reliable as the humans who make them. Even so, they are, so far, more dependable than machines.
