Writing web histories with an eye on the analog past

Introduction

Despite the ‘perfect’ copies digital media make possible, one of the biggest challenges of web historiography is the dearth of websites preserved and available to historians. Projects like the Internet Archive provide limited glimpses of the past, but websites are notoriously unstable objects. It is far easier to find an example of a film from 1924 than a website from 1994. After the dot-com crash in 2000, paper wealth and new media jobs were not the only things that vanished; countless commercial websites that were once enthusiastic representations of New Economy hype became embarrassing reminders of irrational exuberance and were quickly removed when venture capital ran out. Domain names were resold, repurposed, or abandoned, only to re-emerge a short time later as one of those generic directories of sponsored links that thrive on the expired traffic of an outdated web.

Even ‘saved’ websites bear their own burdens: broken links, missing images, 404 errors, and code written for outdated browsers. Yet another reason why new media is ‘new’ is because they create artifacts so unlike prior media forms. The unique characteristics of the web and the peculiarities of the digital object will certainly require new ways of doing media history. At the same time, familiarity with the efforts and insights of media historians studying the emergence of earlier media and communications technologies can provide useful points of comparison for understanding the historical development of the web as a new cultural industry and cultural form. Recent critical attention to web history and historiography invites important reflections on the task at hand: What does ‘doing web history’ mean? What sources might we rely on and what stories might we tell? Which stories are privileged and which are harder to reveal? How can we historicize a web that is constantly revised, deleted, and ‘saved over?’

While acknowledging that the task of writing new media histories introduces new problems and possibilities, this article urges web historians to consider the work of broadcast historians trying to recover the cultural roots of radio and early television who understand well the difficulties in piecing together the past when so much of what was broadcast was sent out live and unrecorded. Some of the methods and concerns that broadcast historians have grappled with can serve as a useful guide in forging approaches for new media historiography.

I begin by surveying recent work in web history, highlighting methodological concerns in defining the object of study and addressing the process and implications of web archiving. Next, I evaluate how broadcast historians cope with the challenges of ephemeral media by piecing together correspondence and production documents, emphasizing the role of discourse in creating a context through which people make sense of a new medium, and calling attention to the power relations that undergird archival efforts and historical narratives. Finally, I address the limitations of historical approaches that compare digital networked cultural forms with old media and point to the emerging field of software studies as a way to return a focus on digital culture and digital artifacts. I offer a brief case study of Flash, a multimedia web platform that rose to prominence during the dot-com boom, in order to demonstrate how attention to software along with approaches informed by cultural histories of broadcast media can provide a new perspective for exploring the ephemeral nature of web objects and the discursive negotiations surrounding their production.

Approaching web history

One of the central problems of both contemporaneous and historical web research involves defining the boundaries of study. All research, of course, requires drawing boundaries, but the complex and multifarious forms the web enables make this task particularly troublesome. Noting the difficulty in writing about the web as a definable whole, Burnett and Marshall use the term ‘loose web’ to acknowledge the array of interrelated activities possible online. Because it is interpersonal and mass-mediated, public and private, a site of consumption and production, the user’s relationship to the web is substantially different from prior media forms (2003: 59). Yet, to study the web means to define it in some way: What do we mean when we talk about ‘the web?’ What should we look at? How should we select, collect, and save web artifacts for analysis?

The growing recognition that websites are significant components of digital cultural heritage has spawned a number of national and international web archiving initiatives. These projects use several strategies, including broad harvests that take snapshots of a large number of sites, selective methods that archive a smaller number of sites more completely, and thematic or event-based collections that focus on material surrounding a topic of interest (Brown, 2006; Brügger, 2011; Masanès, 2006). One of the largest initiatives, the Internet Archive, relies on a broad-based ‘snapshot’ approach to website preservation in which web crawler software roams the entire publicly accessible web and harvests files along the way. These are stored in a large-scale data repository made accessible to the public through the ‘Wayback Machine,’ a search engine that allows users to enter a URL and view copies the crawler harvested.

The biggest advantage of the Wayback Machine is the breadth of sites crawled: by 2010, 150 billion web pages were archived, amounting to over three petabytes of data (Kahle, 2010). But this great breadth comes with the price of depth, consistency, and accuracy. Rarely are the entire contents of a site preserved and the ‘snapshots’ are often incomplete with missing images and broken links. National web archival efforts, like Australia’s Pandora archive, Sweden’s Kulturarw3, and Denmark’s Netarchive.dk, use snapshot approaches that define boundaries along national domains as well as selective methods that use traditional appraisal techniques to identify a smaller number of ‘significant’ sites that meet particular institutional missions or goals (Brügger, 2011: 28).

Finally, thematic or event-based archives, like the Library of Congress’s MINERVA archive, collect sites around a particular topic (i.e., an election cycle or the September 11th attacks). Web sphere analysis is a multi-method approach to defining a thematic archive that emphasizes the hyper-linked, co-produced and evolving characteristics of the web (Schneider and Foot, 2005: 157). A web sphere is conceptualized as a ‘set of dynamically-defined digital resources spanning multiple websites deemed relevant or related to a central event, concept or theme, and often connected by hyperlinks’ (2005: 158). Using this method, researchers identify a number of ‘seed URLs,’ a sphere of websites related to a given event or theme and use web crawler software to capture these sites and those joined to them through links. Since the methodology of web sphere analysis is intimately bound to the archival process, it is ideal for researchers who identify a contemporaneous event for retrospective analysis. However, it is of limited use to web historians hoping to uncover earlier web practices that have already disappeared.

Because the particular nature of the web archive has significant implications for web historians, analyses of archival methods have been central to the first scholarly discussions of web historiography. Schneider and Foot, for example, explore the characteristics that make archiving web content a uniquely challenging process. They point out that the web’s distinct mixture of the permanent and the ephemeral create conditions unlike prior media forms (2005: 166–167). On one hand, files destined for the web must be saved before being transmitted, giving web artifacts a sense of permanence unlike live broadcast or theater. But web content is also transient and not expected to last. Unlike paper or DVDs, which can be saved in their original form of presentation, saving web content demands action – websites must be intentionally preserved in order to be viewed again (Schneider and Foot, 2004: 115). Furthermore, web artifacts are regularly overwritten as new versions replace what came before. ‘Unlike any other permanent media, a website may destroy its predecessor regularly and procedurally each time it is updated by its producer,’ they note (2004: 115).

These conditions raise methodological concerns for internet researchers who must designate the parameters of what to include, how many links to follow, how deep to probe, and how often to save. For these reasons, Brügger (2008) suggests the web archive may be ‘a new type of historical document’ because the process of archiving is reflected in the artifact. That is, archived versions of websites are not mere copies but unique objects created by the act of archiving (2008: 156). In addition to technical problems (missing images, broken links, unsupported file types) there are often temporal inaccuracies. As Brügger notes:

The archived web document is not only incomplete, but is also ‘too complete’; something that was not on the live web at the same time, the content of two webpages or website sections, is now combined in the archive and it is difficult to determine what the website was actually like at a given point in time. (2009: 127)

For example, the method the Internet Archive uses to fill the gaps of incomplete harvests involves grabbing a file with the same name that has the closest available date to the file a user is viewing and using it as a substitute (‘Internet Archive,’ n.d.). This means that links between pages may connect two files that never existed on the live web at the same time. Since images are separate files linked to web pages, the Wayback Machine may display a page that merges images and content from different dates, creating a representation that was never actually viewed by any user at the time the site was harvested.

What is a historian to make of the quandaries of web harvesting, the unique mixture of ephemeral and permanent, and the temporal and technical inconsistencies of archived objects? These are a distinct set of problems not incurred by other media historians. Yet, in the process of delineating these distinctions, we risk overlooking some of the more enduring consistencies that define the work of media historians invested in different objects of study. Therefore, I turn to the work of broadcast historians to lend some perspective to the practice of web historiography while further engaging the role of power in the production of history.

Tuning in to broadcast historiography

By looking at the ways cultural historians have reflected on broadcast historiography, I want to highlight two inter-related themes that emerge, which may help new media historians explore potential directions for historical web research. First, broadcast historians have actively explored the practical and methodological challenges in coping with ephemeral sources. Familiarity with this literature can help new media historians find alternative strategies for piecing together past web cultures that have escaped the archive. Second, and closely linked to the first point, broadcast historians have focused significant attention on the role of power in saving and writing about the past, particularly when exploring gender, marginal groups, and the everyday experiences of ordinary users. These concerns have opened new theoretical avenues in order to better account for the gaps – and also the surpluses – in historical records.

An underlying point in broadcast history is the contradictory observation that there is both too much and too little to look at (Anderson and Curtin, 2002; Corner, 2003; Thumim, 1998). Defining the boundaries of an object of study like television is no clear-cut task, as Williams (1974) famously observed with his concept of ‘flow’ to describe the uninterrupted ceaseless sequence of images that constitute the American broadcast experience. As Anderson and Curtin note, ‘How can one make sense of the range of programs running simultaneously on competing channels? How does one account for … the endless, irreducible, ever-multiplying flow of images and sounds that are somehow integrated into the everyday lives of viewers?’ (2002: 23). At the same time, notes Corner, ‘the questions of how much to look at and in what detail is overshadowed by the question of what there is to look at’ (2003: 277). Historians of early radio and television face the daunting task of writing histories of programming practices without audiovisual material to rely on, a practice that many scholars have likened to an archaeological dig (Branston, 1998: 51–52; Caughie, 1991: 25; Medhurst, 2008: 127).

Because early (pre-1955) television was aesthetically valued for its liveness, program recording was not adopted as early as was technologically possible. Additionally, television’s low status as a pastime produced the attitude that television’s output was not worth preserving (Bryant, 1989: 19). In fact, even when live television was recorded for the training purposes of production crew, the filmed telerecordings (and later videotapes) were often destroyed or taped over – the tapes and storage space were valued more than the preservation of programs (Jacobs, 2000: 11). In a sense, the analogy offered by Schneider and Foot comparing the procedural destruction of past websites with a daily newspaper printed each day on the same piece of paper ‘obliterating yesterday’s news to produce today’s’ (2004: 115), has some resonance with television as well.

After the British Broadcasting Corporation (BBC) acknowledged the commercial viability of television preservation in the 1950s, news and documentary programs were usually the ones deemed important enough to archive. Consequently, histories of 1950s British television have privileged public service broadcasting (Thumim, 1998: 99). The programs most likely to ‘disappear’ were entertainment programs and daytime programming geared towards women; this was due to the historically powerful assumptions about their lack of seriousness and worth and the significance accorded to male and middle-class audience preferences (Branston, 1998: 54). In the face of such severe absences, what kind of histories have broadcast historians pieced together and what kind of alternative sources have they relied on?

Two examples of historical research that may prove instructive include the study of televisual style in the era of live programming, and the study of television’s early audiences as the TV set was first installed in the home. Jacobs (2000) analyzes the aesthetics and style of early British television drama from 1936–1955 without access to the actual programs aired. Still, by reconstructing televisual design through studio plans, camera scripts, policy memos, committee minutes, and so on, he challenges the conventional wisdom that early television merely borrowed from repertory theater and radio drama without developing its own aesthetic. Jacobs argues that the absence of audiovisual archival materials created false assumptions about the reality of early television, which infuse historical accounts. While the sources he relies on may not show precisely what appeared on television prior to 1955 – after all, actors veer from scripts, performers’ inflections shift a line’s meaning, and last minute changes are routine – the work of reconstructing approximations of these past programs helps historians piece together the intentions, discourses, and formal concerns at the forefront of television producers’ experiments with a nascent medium.

In the second example, Spigel (1992) investigates television’s introduction into the postwar home during the medium’s first decade in order to make sense of the ‘often invisible history of everyday life … the ephemeral qualities of daily experiences’ (p. 2). Writing against institutional histories that tend to privilege the sphere of men, Spigel must contend with the obstacles that arise when trying to recover marginal histories that escape the archive. To explore the experiences of early female television audiences, she looks primarily to popular women’s magazines and advertisements, but she also turns to television scripts, news articles, and audience research studies, in order to understand the discursive rules formed for thinking about early television. While acknowledging that there is no sure indication of how these texts were received by individual women readers or audience members at the time, Spigel points out that these partial, mediated representations still offer something about the way women might have experienced the arrival of television in their homes (p. 4). By recognizing these sources as discursive sites, she is able to better account for the different and often contradictory meanings about family life, gender roles, divisions of labor and notions of leisure, and distinctions between public and private spheres in a socio-historical context. Her study demonstrates how the challenge of looking for traces of women’s experiences in the realm of the invisible, ephemeral everyday life can open fruitful theoretical avenues to pursue, shedding light not only on untold stories, but also strategies for understanding the ways popular sources can reveal how a mass medium develops in relation to its social context and how responses to this shape its cultural form.

These reflections on the role of the ephemeral in the construction of media history continually remind us that power figures largely in the preservation of the past and the stories generated to account for it. Decisions concerning what to archive are also decisions concerning what not to archive – these choices create both narrative constructions and, as Trouillot has argued, ‘a particular bundle of silences’ (1995: 27). While these insights are not new ones, they have yet to receive the critical attention they deserve by internet researchers and web historians. In defining websites and web spheres for analysis and preservation, we should be acutely aware that our choices of research topics, our decisions to focus on particular sites, and of course, our archival decisions will be as much a part of the historical record as the web materials we preserve.

When the head of the British Library recently made a case for the cultural preservation of Britain’s digital heritage, she cited examples of ‘lost’ artifacts including websites surrounding the 2000 Olympic Games and the White House web pages from the Bush administration (Brindley, 2009). ‘We are in danger of creating a black hole for future historians and writers,’ she warns. At the same time, as British historian Tristram Hunt has pointed out, the archival process requires selection:

It’s essential that mainstream institutions such as the National Gallery or the White House or the Ministry of Defense keep email correspondence … On the other hand, we’re producing much more information these days than we used to, and not all of it is necessary. Do we want to keep the Twitter account of Stephen Fry or some of the marginalia around the edges of the Sydney Olympics? I don’t think we necessarily do. (Hunt, quoted in Smith, 2009)

While saving too much information can be just as problematic as saving too little, we must be attentive to the political implications that color ‘the marginalia around the edges.’ Will web historians of the future confront a surplus of government or civic-oriented traces of the web just as British television historians encounter a BBC archive that privileges broadcasting’s public service record over light entertainment (i.e., women’s programming)?

So far, the selective web archives created by web preservation initiatives like the MINERVA archive have concentrated on topics of public and civic interest, such as the websites surrounding the 9/11 tragedy, national elections, the Papal transition, and the crisis in Darfur – efforts that are obviously important and commendable. Not typically archived are those websites that represent commercial culture and the endless array of minutiae that constitute the popular leisure and viral experiences of the web: Flash animations, interactive advertising, videologs, LOLcats, webcams, Rickrolling. Are these commercial forms, digital cultures, modes of production, and ordinary experiences worth preserving? Even if we decide they are, how will we navigate the particular archival challenges – technical, legal, ethical, practical – that these ephemeral digital practices present? Realizing that much has already been lost, how might we construct web histories that account for cultural artifacts and practices that have escaped the archive?

Making sense of the comparisons: Problems and possibilities

A lot of history is born digital. This should not be like early television where there is not a record.

While the work of broadcast historians may raise some crucial issues highly relevant to historical projects in new media, it is also important to acknowledge some of the very real differences that complicate the task of mapping broadcast historiographical issues onto the new media landscape. Kahle’s above remark is not entirely accurate; it depends on what one considers the ‘record.’ One of the most significant distinctions between broadcast and web historical work is that broadcast historians can rely on centralized corporate and institutional archives. Jacobs’ account of the visual style of early British television drama is only possible because of the mammoth holdings of the BBC’s Written Archives Centre, which houses the BBC’s working papers from 1922 to the 1970s and includes thousands of internal memoranda, letters, financial and program records, press cuttings, reports, and much more (Kavanagh, 1999). Although US broadcasting does not have a unified central archive that compares to the BBC, broadcast networks, advertisers, producers, and writers have donated countless files to a number of federal repositories such as the Library of Congress’s Motion Picture, Broadcasting and Recorded Sound Division. State and municipal repositories are also extensive – for example, the Wisconsin Historical Society’s Mass Communications Collection includes over 600 boxes donated by NBC consisting of network reports, correspondence, publicity materials, scripts and recordings dating from the 1920s through the 1960s (Hilmes and VanCour, 2007). In short, broadcast historians have a slew of contextual data to help make sense of broadcast institutions, production practices, technological constraints, and formal and aesthetic concerns.

Web archives that include only preserved (often partially preserved) digital files without proper contextualization will leave future web historians with the inverse problem to that faced by broadcast historians. While broadcast historians have only had access to the extensive internal and external communications surrounding the audiovisual program, web historians will find themselves with partial access to digital files and scant evidence of the behind-the-scenes communications detailing the how-and-why of cultural production. Recognizing the importance of contextual knowledge production in the collection, preservation, and presentation of digital artifacts, some scholars emphasize the role of the web interface to access public collections, calling for systems that include the capacity for annotation to better account for collective memory (Dougherty, 2007; Van den Heuvel, 2010).

Initiatives like the Preserving Virtual Worlds (PVW) project, which addresses challenges and develops solutions for preserving digital games and interactive fiction, have identified similar needs. ¹ In their final report, the PVW research team concludes that the inclusion of contextualizing information created outside of virtual worlds is a vital component of digital preservation, and collaborations with designers, developers, and user communities is key to enhancing these efforts (McDonough et al., 2010). One useful source for contextualizing the early web industry is the Business Plan Archive, started by David Kirsch in 2002 to preserve the legal records and business plans of failed dot-coms. ² Though these materials focus on the ‘failures’ and are not linked to the preservation of actual websites, this significant repository of entrepreneurial records – including emails, spreadsheets, and PowerPoint presentations – provides valuable details that can help historians understand an important moment in American business culture.

Yet, however useful these comparisons between broadcast and web historiographies, they do not fully account for the unique characteristics of the web that digital preservationists encounter. The hyperlinked infrastructure, the unique distinctions between the live and archived web object, the ‘deep Web’ pages that escape and confound the archival process – these issues must be acknowledged. As Manovich points out, new media may have numerous historical precedents in earlier media, but comparing computable cultural forms with old media ‘cannot address the fundamental quality of new media that has no historical precedent – programmability’ (2001: 47). He was an early advocate of ‘software studies,’ which would examine ‘the new terms, categories, and operations’ that characterize media that have become programmable (p. 48). Since software is what is unique to new media, Manovich argues, comparing new media to prior forms like television will never tell the whole story. Therefore, I conclude by offering a short case study that applies some ‘lessons’ from broadcast historiography to an analysis situated within the context of software studies. Specifically, I suggest that the materiality and ‘system’ of software can be approached as not only a characteristic that makes new media new, but also as a rich site for evidence of web discourse that can help piece together some of the negotiations that marked past web production practices in the absence of preserved web artifacts.

Re-visioning software: The case of Flash

Although software has sometimes been understood in terms of ‘immateriality’ because of its status as computer code that operates inside the machine, the emerging field of software studies challenges this assumption (Fuller, 2008; Kirschenbaum, 2008; Mackenzie, 2006). As the interface between hardware and users, software’s thorough integration into patterns of work, play, and communication should be recognized in terms of its real, lived experience. As Fuller points out, software is often seen as a neutral tool, ‘something to do something with’ (2008: 3). But it is also the product of social relations and cultural values. Software helps structure social processes, and in return, socio-cultural beliefs and assumptions get hard coded back into software.

As Gillespie notes in his analysis of the Macromedia Dreamweaver and Director software applications, the interface metaphors, help files, marketing materials, and guided tours that accompany these tools ‘tell stories’ to users by symbolically announcing the kinds of websites the software is designed to produce and what the web is for (2003: 113). These examples remind us that software is the product of material circumstances that leave material traces (Kirschenbaum, 2008: 15). Because the discourses surrounding software leave evidence of deleted websites (in production manuals, listserv archives, trade press, conference reports, tutorials, design books, etc.), they provide a potential entryway to the web cultures, production practices, and symbolic systems informing lost cultural artifacts. This view of software as a site that leaves material evidence of cultural discourse reinforces Volti’s observation that technologies are not merely material artifacts but ‘technological systems’ requiring the support of other elements – devices, skills, and organizational structures – which are systemically interconnected (1995: 5; cited by Gillespie, 2003). Even the simplest technologies, Volti argues, require a complex network of material suppliers, toolmakers, producers, marketing agents, and consumers capable of putting the technology to good use (p. 5). To understand the web as a technological system, we must look not only at content or code, but at a network of technical and creative laborers, the hierarchies of value that inform production practices, the rules and expectations that shape skill development, and the new media industries that support the commercial production of web culture. To explore the issues discussed in this article, I examine the web animation software Flash as a technological system historically situated within the socio-economic context of the dot-com bubble and its collapse.

Rise and fall of a new visuality

Now an industry-standard web development platform used to deliver a wide range of media including television and video content through Hulu, YouTube, and Tivo, as well as games, motion graphics, and interactive ads, Flash was first launched in 1996 by FutureWave Software as a web animation program called FutureSplash. Soon after its release, FutureSplash generated enough attention that the software company Macromedia acquired FutureWave and relaunched the application as Macromedia Flash 1.0. For the next three years, Flash gained popularity with web designers for a number of reasons. In the heat of the Browser Wars, the competition between Netscape and Explorer meant it was not unusual for designers to create up to four versions of the same website in order to accommodate the different ways browsers supported (or failed to support) various HTML tags (Zeldman, 2000). Flash offered developers a solution to ‘cross-browser compatibility’ issues by creating a single website that displayed uniformly across browsers. Not only did this substantially reduce the cost of development, it also provided designers with a new skill-set they could use to negotiate higher salaries.

But perhaps more than anything, Flash appealed to a new vision of the web, one vastly different from the static, silent, textual form that imitated the aesthetics of print. Flash’s vector format made full-screen high-resolution graphics possible for a fraction of the filesize as a bitmapped image. Its popularity rising with the dot-com bubble, Flash represented a transformative moment in web production discourse where presumptions concerning ‘quality’ web design (how the web should look, feel, sound, behave) were in flux. The production, circulation, and widespread mimicry of Flash-based visual styles stirred passionate responses that voiced particular fears and hopes for the web’s future.

On one hand, as a proprietary software product, Flash called attention to some of the most contentious debates about commercialism and the web. Usability promoters and open source advocates criticized the software, fearing it would violate the web’s fundamental principles of accessibility and openness. But Flash also captured the imaginations of professional designers navigating the new media industrial climate of heavy consolidation and restructuring that was underway in the late 1990s in anticipation of an e-commerce enabled web. Many embraced Flash as a way to intervene in an overly commercial web by shifting the attention from shopping carts and secure servers to storytelling, creative expression, and sensory experience. For example, The Remedi Project, an experimental web gallery launched in 1998, was ‘created in the belief that by suspending judgment about the web, and by abandoning our preconceptions about how to use it to communicate, we may find a better way to express our discordant voices.’ Co-creator Josh Ulm explained, ‘I’d like to see people start telling stories with the internet. With interactivity. [sic] I don’t think much about video on demand and online shopping … I’d like to see people collaborating and building ideas that we never anticipated or haven’t even dreamt of’ (Burgoyne and Faber, 1999: 51).

One example of an early Flash website that gained immediate attention is ‘Eye4U,’ which launched in 1998 to showcase the talents of a German interactive studio. Following the conventions of the time, the website opens with a ‘launch’ screen that checks a user’s plug-in status to make sure the Flash 3 player (released in May 1998) is installed. If Flash 3 is detected, viewers see the ‘enter’ button that grants access to the site. Upon entering, users are met with a visually dramatic full-screen graphical experience set to an electronic music track. Featuring bold, animated transitions with bright blocks of color that whirl and spin into eye-popping kaleidoscopes and pinwheels, Eye4U – as the name implies – is meant to highlight the web’s new hypervisual form (see Fig. 1).

OPEN IN VIEWER

Figure 1.

Screenshots of the Flash website for the Berlin-based interactive studio, Eye4u.

By 2000, the number of heavily animated Flash sites like Eye4U proliferated beyond the realm of studio sites and designers’ portfolios and into the commercial sphere for major brands like Volkswagen, Coca-Cola, Nike, MTV, Macy’s, Ford, and Disney (Kahney, 2000). As evidenced in numerous books, magazines, blogs, and design journals, the visual style of the web at this moment of dot-com collapse was the subject of many self-reflexive meditations and passionate diatribes on the nature of web aesthetics and the very meaning of the web itself as a cultural form.

But despite the software’s ubiquity, most evidence of the early Flash web is long gone. The original Eye4U site remains one of the few Flash sites from the late 1990s still accessible on the live web over a decade later, only because the site’s creators have actively preserved access to the original version. ³ Most Flash sites escape the archive because proprietary multimedia files are difficult for web crawlers to save; unlike HTML pages, one cannot use a browser to view the source code of a Flash file. If we hoped to access Eye4U through the Wayback Machine, we would see only the launch page with a broken image icon where the enter button would be. Essentially, the Wayback Machine shows only that a Flash site once existed, but it gives no indication of the site’s content, visual style, or purpose (see Fig. 2).

OPEN IN VIEWER

Figure 2.

Eye4u website as viewed through the Wayback Machine.

This is the case for thousands of Flash-based websites; an entire genre of web design has not been catalogued. As a result, some of the most visible, widely circulated, and divisive websites of the dot-com era are now lost objects. In this case, it is quite literally the loudest artifacts that have been silenced. To analyze new media forms like Flash, we need to turn to alternative sources in order to make sense of the web’s production culture. A number of different strategies for finding traces of the lost web, both online and offline, can clue us in to some of the discursive negotiations surrounding web production with Flash. Of course, oral testimony and primary sources (news, trade press, blogs, design manuals) would be the first lines of inquiry to register the pulse of Flash enthusiasm, critique, and disparagement. But additionally, much can be gleaned from the ancillary products of the Flash industry, as well as closely examining the software product itself.

By 2000, Flash was a multimillion-dollar business of not just software, but of numerous Flash-related goods: workshops, classes, books for inspiration, books for learning code, and even books by design studios that analyze and illustrate their creative process of conceiving, building, and coding Flash websites. Popular events like the semiannual ‘Flash Forward’ conference and Film Festival generated massive interest in the software and the practice of web design, generating sample files, tutorials, and training videos. These sources reveal beta versions of websites, mockups and storyboards, complete with significant reflections on the production process – all useful for understanding creative practices, formal decisions, rejected approaches, and technical limitations that shape the production of cultural artifacts.

Within this universe of Flash-related products, a dominant discourse began to cohere around the output of a group of ‘Flash masters,’ top designers who achieved the most visibility in the field. Through ‘Site of the Day’ showcase features on the Macromedia website, speaking invitations to the Flash Forward conference, nominations for the Flash Film Festival, networks of links from design portals, and numerous book contracts and chapter solicitations, certain designers’ styles became highly circulated and copied, privileging particular expressions of ‘quality’ design and shaping the way Flash users began to think of the web. Top designers were greeted and promoted as celebrities within the Flash industry, their output eagerly anticipated by enthusiastic fans. The first edition of the best selling New Masters of Flash (Jankowsi et al., 2000) was described as both a ‘global showcase’ and ‘practical tutorial’ in which ‘nineteen of the planet’s most awe-inspiring Flash designers share their influences, ideas and objectives.’ ⁴ But within this surplus, there were notable absences. Only one of the nineteen designers featured in New Masters of Flash was a woman. Flash conferences were dominated by male speakers and the most prominent Flash design portals – edited collections of links to the ‘best’ Flash sites – were created by male designers and tended to feature the work of other men.

By late 2000, Flash discourse registered a noticeable chill. Critics like usability expert Jakob Nielsen (2000) lambasted the software as ‘99% Bad,’ for ‘encouraging design abuse’ and ‘breaking with web fundamentals,’ charges that expressed a biting exasperation with the bells, whistles, and lengthy animated intro sequences now regarded as part of the ‘look what I can do’ Flash aesthetic. After the collapse of the dot-com economy, critiques of internet speculation were often translated into critiques of the web’s visual style, and these arguments repeatedly took on gendered undertones. As commentators reflected on the state of the web industry after the bust, some pointed to the links between web design and the indulgences and excesses of dot-com IPOs, a ‘boys club’ atmosphere of ‘tech fetishism, aggression, and adolescent image-consciousness that quickly became the dominant culture of web media agencies’ (Mahoney, 2001). Flash was just one expression of this image, but it came to stand in for a host of critiques surrounding the hype and hubris of the e-culture industry. In the wake of the stock market crash, Flash flamboyancy seemed superficial, out of touch, and self-indulgent – a visual manifestation of irrational exuberance.

The bursting of the bubble and the shattering of Flash’s hegemony, which started at the level of the interface, also broadened the terrain in which reform was addressed at other levels as well. The climate of a severe ‘Flash backlash’ after the crash and the self-reflexivity it spurred within the Flash community provided a new space for women designers to collectively challenge their lack of visibility in the design industry. In 2001, a number of projects were launched including Rina Cheung’s ‘Women in Design,’ Kylie Gusset’s ‘Don’t You Stop,’ and Carole Guevin’s ‘Powagirrrls,’ which aimed to catalog, publicize, and circulate women’s portfolios through link directories and featured spotlights. ⁵ The ‘Flash Goddess’ community resource site (www.flashgoddess.com) was developed by Ann-Marie Cheung in 2002 to showcase, interview, and promote women who worked with Flash. ‘You hear about the great Flash Gods all the time doing amazing work, my project started out as a personal experiment for my own information to discover where all the Flash Goddesses were hiding,’ she explains (Cheung, 2002). This final irony – that while Flash was being chastised for being too visual, women designers were struggling to be seen – reveals a gendered subtext that frames the discursive construction of web design. These efforts by women to counter their marginality introduce additional entryways for making sense of the technological system of Flash while revealing struggles over visibility, identity, and power during the industry’s transition towards what would become known as ‘Web 2.0.’

Finally, by observing how Macromedia ‘revised’ Flash software following the collapse of the bubble, we begin to see how dominant discourses and values are written into code. To combat the growing concerns over usability, Macromedia addressed Flash design critiques in the software application and its accompanying support documents. The 2002 version of Flash MX, for example, added features designed to help developers ‘be more productive while ensuring their work is both usable and accessible’ (Macromedia, 2002). The new interface added built-in standardized components to the programming library to encourage ‘consistent interface building’ and made a host of other substantial changes addressing Nielsen’s critiques (Arah, 2002). Macromedia also unveiled an elaborate usability awareness campaign on their website including a ‘Design for Usability contest’ and a usability resource center that offered tips, best practice guidelines, and whitepapers on developing user-friendly Flash content. Prioritizing usability aided Flash’s transformation from animation software to a platform for building rich media applications like Flickr, YouTube, and Hulu in the ‘Web 2.0’ era.

Conclusion

If one function of writing history is to ‘counter the inevitability of the consequences of technological and aesthetic change, as well as to chart its course,’ as television historian Stephen Lacey suggests (2006: 6), then the rise, fall, and transformation of Flash says something about the web of the present and the past. Indeed, by 2010 Flash was mired in another cantankerous debate between Adobe and Apple over its support on the iPad and iPhone. These industrial politics make preservation even more complicated as ‘the web’ connects different commercial players across mobile devices, televisions, gaming platforms, and computer screens. It highlights the power relations that structure and shape cultural production, archive selection and preservation, and ultimately impacts how histories are written – a point too often overlooked or insufficiently acknowledged in web studies.

By bringing together literature from broadcast historiography and software studies, this article explored alternative methods for analyzing the ephemeral nature of web objects and the discursive negotiations surrounding their production. The case of Flash is just one example that illustrates the complexities of digital preservation and the work of analyzing digital cultures that have not been archived. But it also calls attention to the critical need for web histories that account for the diverse range of creative, communicative, and symbolic practices and modes of production that have historically constituted the web but are not easily saved – from chat rooms and instant messages to webcams and massively multi-player virtual worlds. While much of this work may require new strategies for uncovering the digital past, these concerns also magnify the problems of preserving today’s web. In a sense, all web research is part of a historical project that speaks of tomorrow’s gaps and erasures.

Like Flash, social media and user-generated content is notoriously difficult to archive. Despite popular anecdotes reminding young people that unsavory content will forever tarnish their online reputations, preservationists at the Library of Congress are struggling to capture and display the media-rich social networking content of Web 2.0 election sites (Grotke, 2011). These efforts, including the recent decision to acquire the entire public Twitter archive, readily acknowledge social media’s historical significance. At the same time, collecting institutions understand that selection is key to sustaining web archives in the future. The question becomes: How well will these institutions predict the needs of future researchers?

Well-publicized accounts of mass deletions – Yahoo’s decision to pull the plug on GeoCities; Friendster’s report that profile content would be removed; Google’s (quickly retracted) announcement that Google Videos would be deleted – strike a public nerve and inspire grassroots campaigns to preserve old web content. But it was not the Library of Congress that scurried to save the animated GIFs populating GeoCities’ neighborhoods; instead, the self-described ‘rogue archivists’ of the Archive Center, a loose collection of independent preservationists led by Jason Scott, took up the task and released the GeoCities torrent, a massive 643 gigabyte file distributed across peer-to-peer networks. News of these potential losses and heroic recoveries spark moments of nostalgia and critical reflection, creating much-needed public discussion about the preservation of our digital cultural heritage.

Indeed, many in the preservation community understand that collaboration and shared stewardship must emerge across institutions and interdisciplinary lines to best address the needs and voices of diverse groups. Humanities and social science researchers who study historical or contemporaneous web practices have much to contribute to these debates, as the very future of internet studies is tied to digital preservation efforts. To address these challenges and to help secure a digital record that exemplifies a broader range of the web’s social and symbolic activity, we must build awareness of the political and cultural importance of digital preservation. New media scholars need to address these concerns in their research and in the classroom by integrating web historiography and digital cultural heritage into new media studies curricula. In addition, as Uricchio (2009) argues, we need to embrace the collective logics and bottom-up dynamics of participatory culture and put these social practices to work in the service of a ‘decentralized memory institute’ (p. 138). By creating incentives that harness collective intelligence, crowdsourcing, and the social sharing of user-preserved digital content (like the GeoCities torrent), we can extend the lifecycle of digital cultures that evade the preservation frameworks of cultural heritage institutions. Despite selection biases, we must also acknowledge that institutional repositories serve as trusted stewards of historical records. In order to convince those working in new media industries of the value in donating their creative output, documentation, and records, which provide much-needed context to commercially produced digital culture, the reputation of public institutions will be foundational. While there are many challenges ahead, particularly in relation to copyright law and privacy concerns, new media scholars must build bridges with the digital preservation community and work to preserve the web cultures we study if we hope to have a say in the web’s historical record. Taking a cue from broadcast historians, we should be acutely aware that absences shape histories as much as artifacts.