Participatory Archives and Technological Mediation: A RGS Approach to Understanding Tool Use in Digital Environments

Justin Lewis, University of Nevada-Reno

(Published December 6, 2015)

Introduction

This article tries to better understand what Kenneth Burke calls "instrumental causes" or technological tools that mediate human desire. Functioning as conduits that link humans to their lifeworld, instrumental causes are underwritten by agenic capacities imbued over long histories of cultural-historical use; further, as mediating technologies, digital tools provide creative solutions to the contradictions that develop from destabilizations in any activity system. While investigating the historical and cultural role mediating technologies play in digital spheres, I foreground rhetorical genre studies (RGS) to emphasize the rhetorical role objects play in structuring human activity. Looking beyond early articulations of genre as either form orsubstance, a RGS approach to digital tool mediation highlights the essential elements of interface and medium in understanding the social, political, and cultural traces circumscribed in objects like tagging systems and augmented browsing scripts. Those traces function as specters or haunts in much the same way as Jacques Derrida describes in Spectres of Marx: the "revenants" of object use phase in and out of human activity somewhere between prior use and future desire. The incongruity between past and future use recognizes the codification of social motives inside digital tools while also accepting the possibility of systemic destabilization and future tool reconfiguration.

Following a cue from Clay Spinuzzi's work on non-official, textual workplace genres, this article examines "mesoscopic" digital tool use to understand how digital tools function as rhetorical genres. Mesoscopic tools allow individual users to consciously manipulate information to reach a goal. In the case of Spinuzzi's work, mesoscopic tools like post-it notes mediate human activity, allowing individuals to achieve their individual goals more efficiently by utilizing informal textual genres. In this analysis, I've followed a similar tack, drawing attention to the ways that digital tools function at the mesoscopic layer. My supplement to Spinuzzi's work on individual textual genres theorizes digital tools as genres that rely on aggregated social data to assist users in achieving particular ends. Such a recognition shifts the onus of mesoscopic analysis away from a single individual's personomy to the folksonomic systems that characterize tool use in digital environments. The move away from person to folk has important implications for studying tool mediation. I argue that the implicit social motive embedded in text-based genres becomes explicit in digital objects that mediate experience; as such, a return to RGS for investigating digital mediation yields important insights into the sociotechnical cultures that characterize Internet identity.

Specifically, this study demonstrates how mediating tools at the mesoscopic layer of activity play an integral role in constructing user experiences of participatory archives. These tools confer particular agenic capacities and address recurrent rhetorical situations; namely, in the Question.cd BitTorrent community studied herein, mediating technologies provide generic responses to the recurrent social need of archival navigation and media discovery. To connect the mediating role of digital tools to work familiar to writing studies scholars, I rely on rhetorical articulations of genre, emphasizing how tools embody social motives and allow individual and collective action to meet recurrent exigencies of typified rhetorical situations. As I argue throughout, digital tools like tagging systems and collages should be considered from a rhetorical genre perspective circumscribed by an activity theoretic epistemology. From this perspective, the methodology I propose reveals how technological tools inscribed with individual intentions and social motives stabilize a shared digital ecology: the Question.cd1 participatory archive.

The Research Site

Question.cd: A Piratical Community Archive

As of November 2015, the number of registered contributors to the English version of Wikipedia numbered roughly 26 million, of which 128,000 regularly contribute to the site ("Wikipedia:Wikipedians"). Since its launch in 2001, the English Wikipedia user community contributed ~5 million unique articles and in November 2015, those articles were being created at a rate of ~800 per day, creating one of the largest ever peer produced, collaboratively curated electronic text resources ever ("Wikipedia:Statistics").

For comparison, the Internet Archive2 claims 2.4 million users ("User Statistics"). Archive.org houses a staggering 14.5 million digital objects, the majority of which are digitizations of texts whose copyright status has fallen into the public domain ("Archive.org Homepage"). Working together with libraries and government agencies at 23 scanning centers in five different countries, Archive.org volunteers digitize 1,000 books per day for inclusion in the collection ("Internet Archive"). Since its launch in 2001, the Internet Archive has collected roughly 50 petabytes (500 million megabytes) of cultural materials under the CreativeCommons license, providing a rich storehouse of moving images, static images, and texts for future research, consumption, and production ("Petabox").

While a good deal of research has investigated the users in community curated archives like Wikipedia (cf. Keen; Ayers, Matthews, and Yates; Dalby; Lih; Reagle; O'Sullivan), less research has considered the tools that mediate archival creation and curation.3 Though not overtly agenic, technologies of participation on sites like Archive.org are designed not only for archival creation but also navigation and discovery. Further, these tools encourage rhetorical negotiation among site members, providing digital spaces where archive producers hash out differences concerning artifactual description or appropriate categorization. Tools of participation like those in communally curated archives are transformational technologies that reshape the contours of distributed activity but are typically marginalized in research on digital collaboration in favor of human agents.

Outside of legal peer-produced archives like Wikipedia, piratical communities also engage in the collaborative act of archive making. Napster, the first widespread peer-to-peer sharing network dedicated to indexing a distributed archive of media, contained links to over 80 million songs at the time of its demise; however, many of these artifacts were inaccessible as Napster's collection wasn't actively tended. KaZaa's p2p network suffered similar problems as individual users were unable to curate networked data, leading to a preponderance of fraudulent media objects and the proliferation of malware. After the death of first-generation p2p file sharing networks in the early 2000s, new networks of distributed archival production addressed the shortcomings of Napster, KaZaa, and others. Adopting many principles and techniques of archival curation incorporated in legal alternatives like Wikipedia and Archive.org, private BitTorrent communities offered new sites of distributed social production whose archive was composed of material still under copyright protection. At first, these sites remained fairly small, drawing members into niche communities dedicated to preserving and sharing obscure digital ephemera; however, as time wore on, these communities exploded in size and volume, attracting users interested in curating massive generalist archives. The community of Question.cd is just such a community and will provide the data used in this article to analyze tool mediation in collaborative digital environments.

In the wake of anti-piracy legislation and prosecutions in the first decade of the 2000s one might expect that an illegal file sharing community wouldn't have the capacity to create archives on the order of Wikipedia or the Internet Archive; yet, the numbers don't lie. As of May 2012, Question.cd boasted 323,969 users. In the first five years of its existence, Question.cd members uploaded 1,457,817 torrents4 at a rate of over 1,000 torrents per day. This pace exceeds the article creation rate of Wikipedia in 2010, 2011, and 2012 and equals the current text digitization rate of Archive.org. Perhaps even more remarkably, the community of Question.cd is the archive. While sites of commons based peer production like Wikipedia and the Internet Archive house their downloadable/accessible material on centrally located dedicated servers, the contents of the archive of Question.cd are distributed, residing on the personal computers of its 300,000 plus members. This means that users aren't only users on Question.cd, they're also relied upon to maintain the stability and accessibility of the archive.5 To ease navigation through the nearly 1.5 million digital artifacts accessible through Question.cd's tracker, users have developed sophisticated systems of folksonomic tagging and associative linking, relying on structured, user-generated metadata to facilitate content exploration.

While the motives of users on communities like Question.cd may differ from the exigencies that increase user activity on the Internet Archive or Wikipedia, all three could be considered participatory archives: communally generated collections whose creation and administration is user controlled and whose design is user oriented (Huvila 35). As a knowledge object, Question.cd's archive grows and is managed through processes of decentralized curation. The archive users share responsibility for extending and cleaning the archive, relying on their own subject expertise and experience to make decisions concerning artifact inclusion/exclusion and categorization. Further, because findability of artifacts trumps preservation of materials in collaborative digital archives, usability of the archive is radically user-oriented, ensuring that as the archive grows and transforms over time its navigability reflects user choices about categorization and metadata structure.

Processes such as radical user orientation and decentralized curation are crucial for the production of participatory distributed knowledge objects like digital piratical archives or collaborative encyclopedias. Yet, these processes of networked collaboration are made possible through human engagement with various mediating technologies such as editing interfaces, folksonomic tagging systems, optimized searching algorithms, and open licensed server-end software.

Methodology: An Activity Theoretic RGS Approach to Studying Tool Mediation

An Activity Theoretic Scene

Because tool development and use in the Question.cd archive is well documented and mediates all user activity, I pay special attention to the mediating technologies that facilitate participation in user-generated archives. Though the other elements of the piratical activity system will be referenced and elaborated throughout the sections that follow, a close analysis of community, rules, and division of labor is outside the scope of the present study and will only be discussed as extensions of tool-use in the Question.cd community. Like Bakhtinian utterances, tools function as internalizations of a society's cultural and historical tool use; as such, a close investigation of the cultural-historical use of tools used for participatory archive creation and curation reveals the implicit rules that govern site activity, the contours of communal inclusion, and the lateral division of labor instantiated in radically user-centered participatory archives. Further, because multiple researchers in writing studies have employed activity theory to research texts that function as tools or mediating artifacts, the general contours of a study that takes tools as its central focus is already well established. In the sections that follow, I'll sketch the connections between activity theory research on writing technologies and rhetorical genre studies before turning to the explicit methods I employ to investigate tool mediation in the Question.cd community.

Activity Theoretic Analysis in Writing Studies

Writing studies is primarily concerned with the function of written documents and artifacts in social action. This connection is most apparent in the impressive body of scholarship loosely referred to as rhetorical genre studies (RGS). To fully understand the resonances between activity theory and RGS, I'll provide a brief review of the genre as a dynamic, yet stabilizing force that structures symbolic action; next, I'll explore how dynamic genres fit into an activity theoretic ontology, emphasizing the role that mediating technologies play in shaping action in digital spaces. This section will provide a background for understanding why rhetorical genres as mediating tools in use is essential to better conceptualize the diverse and multiple world of participatory archives.

Drawing on the notion of social typification found in Alfred Schutz and Thomas Luckmann's multivolume Structures of the Life World, Carolyn Miller's 1984 article "Genre as Social Action" connected the cultural and historical genesis of written documents to their structuring capacities as genres that influence social action and embody social motives. Noting that genres are "typified rhetorical actions based in recurrent situations," Miller employs a method of rhetorical analysis that rejects genre as diachronically static, instead positing that genres mutate and transform as the needs of human users change over time. Catherine Schryer's work in the early 1990s echoes Miller, noting that genres are routinized, "stabilized-for-now" responses that prove productive in systems of cultural-historical activity (208-9). As other researchers have shown, the elasticity of rhetorical genres is remarkable, as they provide important methods of mediating human activity that vary a great degree in scale. As Stephen Witte demonstrates, textual genres might be as simple as typified responses to the recurring situation of grocery shopping or as Charles Bazerman documents, as complex as the output of an entire constellation of scientific research communities. Other researchers extend genre theory to encompass the constitution of particular experiences, highlighting how generic texts force individuals to experience the world through the lens of generic possibility (Spinuzzi 42).

What is important to note about all of the aforementioned theorizations of rhetorical genre is the role of texts as mediating tools or alphabetic technologies that internalize social motives and direct future human and non-human activity. As structuring agents, genres enact social intentions by providing the tools necessary to make real the collective ends of collaborative activity. As David Russell points out, "Genres are, in a sense, classifications of artifacts-plus-intentions" and should be investigated to uncover the culturally bound, historically circumscribed motives of the individuals using them to carry out their work ("Uses of Activity Theory" 45). In addition to the internalization of social motives, rhetorical genres are agenic objects that make possible certain kinds of acts while making others more difficult.

In total, rhetorical genres function as mediating objects that are shot through with social motives and agenic capacities. In Burke's words, they're "instrumental causes" that allow individuals to make sense and meaning from their relations to the world by providing foundations of mutual recognition and possible action among collaborating parties. By concentrating on interlocking systems of objects that define the bounds of possible activity in fields of science, business, and technology, writing studies research accentuates the sociocultural, historical, and mediatory role of texts on human activity; however, because rhetorical genres function in sites of manifold complexity, some rhetorical genre scholarship looks to activity theory to provide bounded scenes for investigation.

The first substantive linking of activity theory and RGS occurs in Russell's 1997 Written Communication article, "Rethinking Genre in School and Society: An Activity Theory Analysis." In this landmark study, Russell synthesizes Yrgo Engestrom's systems version of Vygotskian cultural-historical activity theory with Bazerman's work on genre systems. In so doing, he draws special attention to the role of texts as mediating genres in activity systems. This is especially important as it is the first time that written objects are not only positioned as mediational means in activity theoretic systems but also function as rhetorical genres that mediate possible action (Fig. 1). Russell posits that many genres exist because they embody social desires that meet recurring situational exigencies of the discourse communities where they circulate; further, he also contends that texts mediate and transform the activity of individual actants in broader systems of culturally embedded, historically inflected human action.

(Fig. 1. Simple activity system from Russell's "Rethinking Genre in School and Society: An Activity Theory Analysis" with "writing" highlighted for emphasis.)

Drawing on Bazerman, and Mikhail Bakhtin before him, Russell's synthesis of activity theory and RGS allows him to move beyond an emphasis on the dyad, widening scope to accommodate the broader discourse communities wherein genres function as mediating tools.

From Alphabetic to Interfacial: Digital Tools, Activity Theory, and Rhetorical Genre Studies

"Rethinking Genre" and subsequent work by Witte, Cheryl Geisler and others expanded the intersections of activity theory and RGS, exploring the agenic capacities of texts as mediating tools that reveal documentary reality of organizational cultures (Berkenkotter 53). Yet, while sites of research multiplied, writing studies scholars maintained a fairly strict attention to alphabetic texts when undertaking activity theoretic-RGS research. Spinuzzi's work in Tracing Genres through Organizations, Paul Prior and Jody Shipka's "Chronotopic Lamination" and Geisler's "Textual Objects" all employ methodologies derived from Russell's article and position writing as a mediating technology from which discursive norms, collective motives and collaborative action are derived. Relying on activity theory to study text as something more than dialogic, these scholars expanded rhetorical inquiry by locating the heteroglossic interpenetration of social languages first theorized by Bakhtin in wider scenes of human activity. Yet, despite the progression of writing technologies into various forms of new media, writing studies scholars' uptake of RGS analysis circumscribed by an activity theoretic orientation have remained largely focused on the alphabetic.

The methodology employed in this article departs from this trend by investigating digital objects as rhetorical genres that mediate activity in participatory archives. To bring activity theory and RGS together to study non-alphabetic digital tools, I rely on a rearticulation of the notion of genre, expanding its bounds to accommodate the non-alphabetic while also recognizing the important role of these tools in mediating human action.

In "Genres of Organizational Communication: A Structurational Approach to Studying Communication and Media," Joanne Yates and Wanda J. Orlikowski note that genres are composed of both substance and form. Substance refers to the "social motives" or "frames for social action" that are expressed in moments of generic use whereas form refers to the "observable physical and linguistic features of the communication" (301). To get at digital objects functioning as rhetorical genres/mediating tools in activity systems, a slight modification to Yates and Orlikowski's theorization of genre is required. The modification proposed herein relies on new media scholar Andreas Gregersen's uptake of rhetorical genre theory. In "Genre, Technology, and Embodied Interaction: The Evolution of Digital Game Genres and Motion Gaming," Gregersen explains that although Yates and Orlikowski's inquiry works well for alphabetic texts, it doesn't necessarily hold for gaming genres. Expanding the notion of rhetorical genre, Gregersen's offers medium as a dimension of form and a supplement to substance. Generic mediums are one aspect of form but are at the same time a thing apart from it. In Gregersen's words,

On the one hand, genres are defined partly by similarities of form, one subdimension of which is medium—the medium is thus understood as part of the genre. On the other hand, media are to be distinguished from genres, the latter of which are understood as typified communicative actions. A reasonable reconstruction of the argument is that genre, once defined as typified actions, does not belong to specific media, but rather can move between media. Yates and Orlikowski themselves offer the following observation: "Though a genre's form may at one point include the medium, that genre may also expand into other media" (1992, p. 319). This quote hints at the kind of situation I want to substantiate here with regards to digital games. Some genres seem constituted by specific technologies, others less so. Such relations will change over time—and, importantly, technologies may also expand into other existing genres as part of such processes. (96, emphasis added)

The addition of medium to the definition of genre advances Yates and Orlikowki's articulation beyond the alphabetic by paying special attention to the (im)materialityof generic form. Gregersen includes this addition because his work draws attention to interactive technologies such as video game controllers and their role in new media game genres. I rely on Gregersen's inclusion of medium because it moves genres beyond structural and substantive components into the reconfigurable space of piratical user activity, allowing my study of genre to take into account the role of interface in digital tool use.

As Collin Brooke notes in Lingua Fracta, interfaces are the "ever-elastic middle" that "incorporates, and indeed constitutes" the outside of things (25). In this way, the interface functions as a space of elasticity and reconfiguration, shaping static objects while also providing them stabilized-for-now dwelling spaces. Brooke says as much, noting that, "[The] interface [is] itself a momentarily situated encounter among users, machines, programmers, cultures, and institutions" (42). Johndan Johnson-Eilola traces the genesis of computer interfaces in Datacloud, conveying the transformation of interface from the static medium of hardwired connection to the distributed nature of dynamic interfaces in contemporary networked information economies. Although Brooke's notion of interface remains mostly theoretical and Johnson-Eilola's historical account hews close to specific examples, both explorations of interface highlight its role as technology that mediates individual action, is responsive to change, and renders static specific iterations for particular uses. In this way, interfaces—especially the interfaces that include digital tools that mediate our experiences in electronic spaces—function much the same as rhetorical genres. Using the articulation of rhetorical genre as form + substance + medium, this article proceeds to analyze digital objects as rhetorical genres functioning as mediating tools within the interfacial spaces of piratical activity systems (Fig. 2).

(Fig. 2. Rhetorical genre as form + substance + medium/interface)

The Problem of Scope: Levels of Activity

To effectively bring RGS and activity theory together toward analyzing how tool mediation shapes user activity in participatory archives, I need to narrow the locus of investigation. In activity theory parlance this narrowing refers to "scope" and delineates the contexts of activity that are already operant in activity systems. In Tracing Genres through Organizations, Spinuzzi notes that sociocultural theories such as activity theory draw on an integrated-scope approach that tries to account for the iterative constitution of the entire activity system by exploring the role of habitual operationalized actions, goal-directed individual actions, and systemic collective activity.

Operationalized activity answers the question of "How?" in activity systems and is typically characterized by unconscious, routinized human or machine actions. Goal-directed activity answers the question of "What?" in activity systems and is typically characterized by the conscious, directed action of individual participants. Systemic activity answers the question of "Why?" and is comprised of the collective actions of all members participating in the activity system (Fig. 3).

Tool-use in participatory archives is often conscious, goal-directed and answers the question of "What?" This form of instrumental agency occurs at what Spinuzzi calls the "mesoscopic" level of activity, or the "tasks in which people are consciously engaged" (33). Mesoscopic activity is the easiest to observe as the active manipulation of technologies to facilitate action are consciously put to use by members of any given activity system. Microscopic activity is operationalized action and might be best conceived as unconscious manipulation of tools. For most researchers, relatable examples of microscopic activity might be the unconscious manipulation of a computer keyboard or a door handle. Macroscopic activity is collective action towards object creation. The entire Wikipedia or Question.cd archive is mediated by macroscopic tools like server-side software and content management systems, though a sustained investigation into macroscopic mediation is outside the scope of this study (see Lewis 2016). Interestingly, the levels of activity are fluid and contain in their constitution the capacity for transference. A mesoscopic activity can degrade into unconscious microscopic operation; further, what one actant considers a routinized operation another might not be internalized by another actant and hence functions as an action at the mesoscopic.

Level	Activity Theory Term	Example
Macroscopic	Activity (cultural-historical, unconscious)	Collective construction of the Question.cd archive
Mesoscopic	Action (goal-directed, conscious)	Conscious individual use of tools to navigate the Question.cd archive
Microscopic	Operation	Unconscious use of keyboards to input information into one of many Question.cd interfaces

(Fig. 3. Levels of Activity)

Summing Up: A Methodology for Integrating Rhetorical Genre Studies and Activity Theory for Studying Tool Mediation in Participatory Archives

To recap, the methodology employed herein circumscribes the work of RGS by locating mediating technologies in broader systems of activity. The analysis that follows sketches activity at the mesoscopic level, placing an emphasis on digital tools as rhetorical genres constituted by form, substance, and medium. Ultimately, the goal of this methodological synthesis highlights the role of mediating technologies as rhetorical genres in social web environments like that of the participatory archive. For each of the tool analyses,6 I make the following moves:

• Identification: Identify digital objects that function as rhetorical genres and structure mesoscopic activity in piratical participatory archives.
• Substantive Account: Locate the digital object diachronically by considering the substance or social frames of action each object fulfills. This provides a cultural-historical account of human-tool interaction by attending to the recurrent exigencies that motivate tool use and revision.
• Structural Account: Consider the form of each new media object to discover its structural features.
• Interfacial Account: Consider the medium of these new media objects to discover how they function in broader interfaces and ecologies of digital activity. Because tools function together in genre ecologies, interfacial exploration of medium will occur after the substantive and structural accounts are complete.

Mesoscopic Activity: Three Investigations into Generic Mediation

Navigation is perhaps the most important action used to make sense of archives. As spaces of fascinating cultural-historical significance, archives preserve communal meaning and history, carry forth unique cultural significance, provide systematic tools for interpretation, and are malleable to changing needs and desires of the communities that use them. In the broader activity systems that constitute BitTorrent communities a need exists for navigational structures that assist in archival exploration and discovery. Luckily, unlike at the Ancient Library of Alexandria or The National Archives in London, specialists aren't required to facilitate this process. Because BitTorrent archives are participatory, tools that draw on user knowledge and expertise override any top-down, systematic categorization schema. The digital objects that facilitate goal-directed, individual action in this study are tools like tagging systems, collages, and requests. In the sections that follow, I'll explore how these objects function as rhetorical genres, emerging from cultural-historical activity and inscribed by ongoing activity, stabilize the activity in which they are used, and solve particular problems for the individuals who use them (Spinuzzi 39). As mediating tools, the navigational objects are consciously selected, interpreted, and produced to mediate goal-directed actions in participatory archives.

Tagging Tools

Vast digital archives and small, personalized collections utilize descriptive metadata to provide structure and relationality among media. Metadata, or data about data, takes myriad forms and existed before the advent of digital media; however, the widespread integration of .xml,7 a markup language designed to carry data but not display data, allows users to easily append "tags" or small bits of classificatory information to digital objects. While traditional systems of digital organization include structural and preservation metadata, incorporating user-generated descriptive metadata transforms archival categorization, opening up navigation to the rhizomatic and associational links generated by users, not specialists. The bottom-up, non-hierarchical structure of tag-based "personomies" enables users to create individualized classification schemes for information retrieval. When aggregated and networked, these personomies form folksonomies, or what Rice calls "new media taxonomies" that categorize information according to desire, taste, personal interest, communal knowledge, imagination, and so on" (119). As Nicotra points out, because folksonomies are collaboratively created, frequency-based networks that harness the wisdom of crowds, they are essentially social and essentially inventional (271-3). Sifting data into searchable, navigable forms, folksonomies provide social methods of navigating content, allowing users to explore the web using the folksonomic descriptive metadata other users also find useful. Providing new methods of information management and information architecture, the proliferation of social tagging sites like Delicious and 43Things first opened up archival interface design in fascinating ways.

Launched by Joshua Schachter in 2003, Delicious was the first widely adopted social tagging service. Organized around the ability to contextualize and archive links from around the web, Delicious not only provided users with individually-generated tagging systems for categorizing web content, it also allowed users to share those categorizations. Built around the notion that all tags are theoretically "equal" to one another, Delicious allowed folks on their service to taxonomize their web habits and then share those folksonomies. In other words, Delicious allowed users to develop individual categorization schema for navigating the archive of the World Wide Web while at the same time sharing those schema with other Delicious members. The development of shared categorization—not classification—created a media ecology unto itself, prompting Delicious users to explore the web through the Delicious folksonomy rather than organic search engine navigation or direct URL visitation. Other services such as 43Things and 2dobeforeidie adopted folksonomies into their design, allowing users to connect not only around shared categorization of content but also around a variety of shared goals.

Delicious, 43Things, and other social networks organized around folksonomic tagging provide tools for meeting the demands of recurrent problems with stabilized-for-now, tool-mediated responses. In the case of Delicious, Schachter hoped to bring order to the chaos to his link-hosting website Memepool. In its infancy, Memepool was a small collection of web links that Schachter found interesting. Over time, other web users found their way to Schachter's site, enjoying his content but also sending him additional links to supplement his own. The categorization and inclusion of these reader-generated links resulted in more than 20,000 entries or "far more than any folder system [of the time] could handle" (Surowiecki). Schachter soon realized that tens of thousands of web users were looking around for "good links" or links that would help them more easily conquer the signal to noise ratio problem of Internet content. As a result, he developed Muxway and then Delicious, inviting users to collate and share their favorite links through social tagging tools. In RGS terms, the substance of Delicious as rhetorical genre is its ability to meet the recurrent problem of successful web navigation with the stabilized tool of user-generated tagging.

The cultural-historical genesis of folksonomic tagging systems like Delicious continued after its own relevance and utility faded. Numerous other companies, including the wildly popular Pinterest, exploit the same recurrent social need and devise a host of interfaces to assist individuals in categorizing and sharing their web content. In the educational sector, libraries have found multiple methods of integrating social tagging to enhance their content and facilitate user navigation. The University of Pennsylvania's PennTags8 was the first to do just that; however, libraries at hundreds of other institutions are also making use of user-generated descriptive metadata to facilitate exploration and discovery.

The folksonomic tagging system at Question.cd bears substantive similarity to the rhetorical-generic tools used on Delicious, 43Things, and other social bookmarking sites. Relying on user-generated metadata, two different tagging inputs allow users to draw associations between specific media and other media that is in some way connected. In Fig. 4 we see two user-generated tagging systems. The first (Input 1) is a familiar tagging tool that allows users to input other genres that bear some similarity with the music being considered. In this example, the media is a collaborative album by DJ Danger Mouse and Sparklehorse entitled Dark Night of the Soul. Users have tagged the album as a variety of different genres, including "experimental," "indie," "hip.hop," and "collaboration" (Output 1).

(Fig. 4. User generated tagging system on Question.cd)

When a user clicks on any of these tags they are redirected to a results page that includes all of the torrents that share the same tag. The other tagging system included on Question.cd is useful to this specific community. The "add artist" tool (Input 2) allows users to include artists that also participated in the production of this album-length release. Because Dark Night of the Soul includes performances by a host of performers besides Danger Mouse or Sparklehorse, these artists are tagged by users and linked by the tool itself (Output 2). By clicking on any of these artists, users are redirected to a results page that includes any media in the archive released by that artist and available for download. The "add artist" tool functions as a specialized tagging system, which allows users to draw connections between artists that may not be apparent from an album's release information. Substantively, the tagging tools on Question.cd meet a recurrent need for users: navigation of a vast archive of user-generated content. These tools mediate user activity and facilitate particular kinds of action, notably: exploration, association, and formalization of preexisting, though often invisible, relationships across media.

(Fig. 5. Delicious interface)

Formally, the structural features of the Question.cd tagging system bear striking resemblance to other folksonomic tagging tools. Like the Delicious interface (Fig. 5), Question.cd's tagging system requires single word tags separated by commas. Two or more words tags are welcome; however, the words must be separated by periods. The outputs of both systems are in list form and provide links for users to click and visit.

Folksonomic tools like those used by Delicious, Pinterest and Question.cd are substantively and formally similar: they, and many other folksonomic tagging tools like them, meet user needs and facilitate user action by providing stabilized tools for navigating vast digital archives. In this sense, they are rhetorical genres or frames of action that allow individuals to complete conscious, goal-oriented tasks in technologically mediated environments. As tagging continues to permeate social media and personal collections, it's plausible that the act of tagging might move away from conscious, goal-directed action and toward unconscious operation. In other words, tagging may not remain a mesoscopic action forever, degrading into unconscious microscopic operation.

Collage Tools

In Lingua Fracta, Brooke discusses how digital tagging tools like Delicious allow users to establish collections, or "the individual assembly of a large group of whatever items we might choose to collect" (109). For Brooke, tag-based collections and other forms of user-generated descriptive metadata function as tools that trace emergence and provide methods for distant reading of large-corpus data. Taken as a social aggregate, tagged links to web content in Brooke's work create coherent narratives from the Babel of digitized information that exists on the web. I turn to collections here because tools that create individualized assemblages are important mediational technologies that operate at the mesoscopic level of user action. Collages have cultural-historical roots that extend deeper than digital folksonomies and mediate goal-directed user activity in participatory archives. On Question.cd, the collage functions as a tool of navigation and discovery that directs user action by providing means of personalization while also filtering the archive to meet the needs of various niche audiences.

Historically speaking, collage was likely in use as far back as the invention of paper in China; however, the rise of Modernist art in the early 20th century is the era that most historians associate with the formalization of collage-as-method. Pablo Picasso and Georges Braque first used collage in their work to deconstruct the binary between painting and sculpture. By layering the surface of canvas with glued-on patches of paper, fabric, and found objects, future artists pushed collage in multiple directions, creating multimedia work like wood collage, decoupage, photomontage and even digital collage. Of course, collage has also been put to use by musicians and filmmakers.

Artistic collage bears marked resemblance to Brooke's "collection" in that they are both assemblages of partial forms united under an organizing principle or concept. Collage is ubiquitous on the net and, like tagging, its genesis is no longer only individual but also social. Interestingly, there's little difference between the rhetorical nature of analog and digital collage. An analog collage such as Hannah Hoch's Cut with the Dada Kitchen Knife through the Las Weimar Beer-Belly Cultural Epoch in Germany (1919) excises printed photographs of industry, leaders, and the Dada movement to create a critique of Weimar Germany. As collage, Cut with the Dada Kitchen Knife functions synecdochally, using parts of specific things (e.g., the head of General Hindenberg on the top of the dancer Sent M'ahesa) to refer to wholes (e.g., the failure of the Weimar republic under the leadership of Hindenberg in the post-WWI reconstruction period). A digital collage such as those found on Pinterest are also synecdochal, providing excerpted images from links that are organized under user-generated tags or more formalized site generated categories. The difference between analog and digital collage lies in chronos. The duration of Hoch's piece is frozen in time and speaks, importantly, to a particular historical moment. Because of their social, dynamic and aggregative nature, the pinboards of Pinterest are rooted in deixis, or what Brooke calls the "timeless present tense," constantly updating and expanding through the continually generated descriptive tags of site users.

The collage tool on Question.cd is a digital object individuals consciously manipulate, creating collections of media that deepen their media profile, facilitate new pathways of archival navigation and expand media discovery. User created collages are searchable and organized around a variety of site-sponsored categories such as "Genre Introduction," "Discography," and "Label" as well as around individual interests such as the expansive "Personal" as well as "Theme" and "Staff Picks" (Fig. 6). The uptake of collage tools in piratical participatory archives is a rather recent development, mirroring the playlist creation tools of social music sites like Last.fm, Spotify and Rdio.

(Fig. 6. Collage browsing interface)

The collage presentation interface is visually rich and bears structural resemblance to other digital collage interfaces such as Pinterest boards and the teen-centric weheartit.9 In the "Girls Who Tilt Their Head" collage (Fig. 7), users are presented with album covers related to a unifying theme, namely, "A collage for releases with Cover Art consisting of girls who are tilting their head." While the theme seems rather esoteric, the creator of the collage explained that the purpose of the collection was to help users discover new music in "a visual way, outside the usual genres of rock, pop, rap, etc."

(Fig. 7. "Girls who tilt their head" collage)

In much the same way Pinterest functions, Question.cd community members can click on any of the album covers to read more on that artist and specific release while also being given the option to download the album. With the exception of the "Personal" and "Staff Picks" collage categories, other users are invited to contribute to the collage, expanding its contents by drawing new connections inside and across music, software, and ebooks. To add to the collage, users are presented with a small box in which they enter the URL of the torrent to be included. In essence, this small interface performs the same work as the "Pin It!" button or the "Pin It!" interface used by the collage platform Pinterest (Fig. 8).

(Fig. 8. Comparison of Question.cd & Pinterest)

In generic terms, substantive qualities of the Question.cd collage tool are complex and arise from multiple social needs that exist in participatory archival spaces. While many users still rely heavily on the search function to sift the archive and find the media they seek, some users use the site as a discovery mechanism. The collage tool meets that social need, allowing users to explore the archive through personomies and folksonomies outside of textual tags. Instead of being bound to search algorithms that return a set of collocated terms, the collage tool lets users play in the archive, expanding their knowledge of the relational ecologies that unite media at varying levels of scale, niche, and desire. The result is a ludic tool that provides stabilized, mediated responses to the social need to filter, discover and explore. Beyond community desire to traverse the archive in novel ways, the collage tool fulfills a communal-individual function as well: the further development of digital identity centered around personalized media collections. Because Question.cd is a community, individual users often direct their actions toward communal inclusion. In much the same way as expanding a Facebook profile builds digital identity, the creation of collages by Question.cd members produces new modes of expression and deeper, multidimensional digital personas while also drawing affinities among site users, creating niche subgroups around particular interests.

Substantively, digital collage affords users innovative tools to navigate information glut. On Question.cd, these tools facilitate particular kinds of goal-directed action, notably archival navigation, curation, and exploration. They also help fulfill the demands of creating community in digital spaces. Formally, the digital collage user interface is much akin to the tagging tools discussed above, providing users a simple input mechanism for drawing associations among media. As stabilized, tool-mediated responses to recurrent rhetorical situations, collection utilities like the collage tool on Question.cd are remarkably simple, yet sophisticated digital rhetorical genres that facilitate user action, enhance navigation, and expand digital identity in purposeful ways.

Request Tools

The Bulletin Board System, or BBS, was the first widespread communication technology that allowed private distribution of content over digital networks. In contrast to the World Wide Web, BBSes were small communities organized around niche interests like text-based games, ASCII art, hacking information and pornographic images. Early on, BBSers utilized .NFO files and threaded, text-based discussion forums to make requests of other users for content. textfiles.com, a remarkable catalogue of BBS data primarily gathered from the 1980s, chronicles some of these early requests and shows how BBSers made requests, "filled" requests, and evaluated fills with commentary related to the quality of a shared file. The key to BBS request success rested with the communal aspects of the site, requiring users to maintain a dedicated presence and participation in bulletin board activity.

Adoption of web browsers during the mid to late 1990s resulted in widespread decline of BBS user participation. Graphical interfaces, improved user navigation, and integration of multimedia enticed users and opened up the ability to "surf" to wider, non-technical audiences. While communities did continue to exist in web-based interfaces such as AOL and Yahoo!, requesting, filling, and evaluating shared, and often pirated, information wasn't an integral part of most web users' experiences. Usenet did fill some of this void; however, the capacity for Usenet to facilitate affinity-based niche communities like those in the BBS era was limited. Peer-to-peer filesharing networks like Napster, KaZaa, and Gnutella in the late 1990s and early 2000s also lacked interfaces that encouraged community development, instead focusing on the rapid expansion of the archive at the expense of digital relationship building among network users.

After the release of the BitTorrent protocol in 2001, smaller niche-oriented file trading groups integrated social web tools like forums and user profiles to build community in digital file-sharing spaces. At first, requests for music files, applications, and ebooks centered on "Offers" forums wherein users posted text files of their digital media collections, inviting other users to request uploads. "Offers" forums were so popular that BitTorrent community site designers recognized the social desire for a requests function and tools were developed to that end. The "Requests" tool on Question.cd fills that social need and illustrates how digital objects not only fit into cultural-historical models of genre development but also integrate new functionalities to accommodate shifts in the digital world. In other words, the requests tool is flexible, adaptive to the interplay of social forces and emergent technologies that circulate in media ecologies of participatory archives by transforming to provide new stabilized-for-now responses to recurring social desires.

If Question.cd community members fail to find the media artifact they're after, they often turn to the "Requests" tool (Fig. 9). Users input information that corresponds to both the structural metadata that parse the archive as well as descriptive metadata and open-field content. These inputs are sometimes automated and sometimes manual; for example, if a user happens to know the Worldcat identification data (Input 1) for the media they're searching for, they can input that information and the requests tool will autopopulate release information, including artist title, album title and release year. The "Tags" box (Input 2) allows users to provide tags in the same manner as the aforementioned tagging tool; however, generic designations such as "funk," "hip.hop," "progressive.rock," and "world.music" allow users to classify the request in the formalized nomenclature of the site. The "Formats," "Bitrate," and "Media" tick boxes (Input 3) provide users the ability to choose among a variety of encoders (e.g., .mp3, .FLAC), data fidelity (e.g., 192 kb/s, 320kb/s, 24bit Lossless), and release formats (e.g., CD, Soundboard, Cassette). These three media characteristics are important to site users, especially the audiophilic. The open-field "Description" box (Input 4) provides a space for users to add any commentary they wish and often serves as a space to include track lists or critical reviews. Finally, the "Bounty" boxes (Input 5) allow users to gift prospective uploaders with an upload bonus that enhances their sharing ratio, rewarding them for extending the archive.

(Fig. 9. Requests tool)

It's important to note here that the substantive features of the "Requests" tool aren't relegated to individual artifact acquisition and the expansion of the archive; rather, "Requests" also fulfills another social motive of site members: community citizenship. Most private BitTorrent communities require users to maintain a healthy "ratio" of sharing; in other words, for every megabyte of content downloaded, the site requires a certain percentage of that megabyte to be reuploaded to other users. On Question.cd, users are required to share their content on a graduated scale up to 60% of the total data they've downloaded. For example, if user X downloads 20 gigabytes of data, they're required to share at least 20%, or 4 gigabytes, of that data back to other community members. The ratio system is designed to maintain the stability of the archive over time, ensuring that as older media are buried deeper in the archive they don't become inaccessible. The "Bounty" function of the "Requests" tool allows requesters to transfer a bounty or defined amount of their own upload credit to another site member for filling their request. As such, the "Requests" tool not only facilitates participatory archival expansion but also signifies site member participation and investment, rewarding those that upload with additional citizenship credit.

Users hoping to increase their participation on Question.cd regularly visit the "Requests" tool to find new material to upload. The tool provides additional means of achieving that end. After a user fills out the requisite information and creates a request, a page displaying the desired media is created to guide future user action (Fig. 10).

(Fig. 10. Requests tool output)

In this example, a user requested the ebook Defense against the Black Arts: How Hackers Do What They Do and How to Protect against It. The requested item page contains identifying characteristics such as the Worldcat identification number, a short description of the book and a picture of the cover; however, the page also contains two powerful links that connect potential request fillers with the media itself. The "Find in Library" link (Link 1) sends users to the WorldCat database, allowing users to find what libraries in their local area might have the book for checkout. The "Find in Stores" link (Link 2) redirects users to a Google Products search that is autopopulated with the book title. The "Find in Stores" link encourages site users to purchase the requested media before sharing with the wider community.

The "Find in Library" and "Find in Stores" functionalities were only recently added to the "Requests" tool and provide a good example of how new media rhetorical genres transform to meet the changing dynamics of technology and social desire in participatory archives. Recall for a moment that the tools in this section are mediating instruments that direct user action and transform user intention. When considered in activity theoretic terms, tagging systems, collages, and the requests tool assist in archival navigation and expansion as well as media discovery. As such, they mediate these desired outcomes—or "objects"—and circumscribe the possible methods of achieving those ends (Fig.11).

(Fig. 11. Mesoscopic tool-use in the Question.cd activity system)

But ends sometimes change, and even when they remain stable, the means of achieving ends transform, subject to mutations in technology and society at varying levels of scale. The changes in both means and ends create what activity theorist Yrgo Engesrom calls "contradictions" or ruptures that occur when nodes of an activity system transform, creating "dynamic tensions" within the system that must be resolved through destruction of constraints and construction of new systems. In the case of the "Requests" tool, Question.cd members recognized that the number of media being requested far exceeded the rate at which requests were filled. In fact, the backlog of requested materials grew so large so quickly that users began to question the utility of cataloging such information at all, instead opting to return to the forum-based, threaded request systems typical of early BitTorrent communities and BBS networks. The Question.cd development team floated different proposals in site forums, proposing numerous alternative tools to the ineffective "Requests" function. Eventually, after extended deliberation with the broader community and with smaller code and development teams, the updated "Requests" tool was released with WorldCat and GoogleProducts integration. The new media rhetorical genre of the "Requests" tool mutated, transforming to meet the needs of the broader community. In the process, it restabilized the activity system of archival expansion and facilitated a dramatic uptick in "Requests" fulfillments. The inner contradiction in this case was a mediating tool that failed to successfully assist individual users in achieving goal-directed action. As a result, the contradiction was resolved through new forms of tool mediation, or, to put it in RGS terms, the "Requests" tool was developed as an "improvisational strategy triggered by the interaction between socialization . . . and an organization" resulting in a restabilization of genre as a "flexible set of recurring practices (textual and nontextual)" (Schryer 450).

OiNKPlus: A Genre Ecology at the Interstice between Mesoscopic and Macroscopic Activity

In her College English article "Integrating Rhetorical and Literary Theories of Genre," Amy Devitt argues that although singular textual genres are important to consider in response to particular contexts, a richer way to conduct genre analysis might be to consider how genres call other genres, creating complex interaction among multiple genres in response to equally complex rhetorical situations. In another College English article the same year, Anis Bawarshi echoes Devitt, noting that although rhetorical situations call forth socialized activity in the form of generic use, usually more than one genre is operating in any social situation (351). In his 2003 monograph Tracing Genres through Organizations, Clay Spinuzzi also explores the ability of any given genre to mediate the activity of more than one activity system; likewise, he also explores the opposite arrangement: the ability of any activity system to be mediated and transformed by multiple interlocking genres. Spinuzzi refers to the complex coordination of multiple genres to mediate activity as a genre ecology.

While the section on "Mesoscopic Investigations" has hitherto explored singular genres extracted from their imbrications in genre ecologies, this final analysis of generic tool mediation in the mesoscopic will explore an interlocking set of digital tools that direct user action and assist in the same systemic activity: navigation of the archive and discovery of media artifacts. In this section I'll also return to medium, in addition to substance and form, to describe how new media objects function as rhetorical genres interfacially. It is my contention that interlocking digital rhetorical genres can best be observed, described and analyzed in the interface as this is the immaterial space that interweaves rhetorical genres.

Greasemonkey and OiNKPlus: Augmented Browsing and User Scripts

Although Internet Explorer and Netscape Navigator dominated early web browsing platforms, the addition of the Mozilla Foundation's Firefox browser expanded the tools available for navigating the web. Soon after the launch of the open-source Firefox browser in 2004, users began developing extensions and add-ons to augment web browsing. Using a variety of Java scripts that automatically transform and personalize web content, Firefox browser add-ons allowed users to change the default browser interface, reconfigure the design of some web pages by tweaking site CSSs, and integrate RSS, FTP, social bookmarks, and developer tools into their browsing experience. Though most contemporary web users don't think about extensions very often, add-ons like the Weather.com toolbar, AdBlock Plus, and GoogleDrive are augmentations used by millions.

Greasemonkey is an open-source Mozilla extension that allows users to install small programs called "scripts" that automate some tasks. These scripts augment the browsing activity of web users, allowing them to tweak their web experience to better suit their needs. After installing the Greasemonkey extension from the Mozilla site, web users visit userscripts.org to download and install open-source scripts to transform their browsing experience. As of November 2015, individuals on userscripts.org contributed over 137,000 scripts, augmenting everything from Google search results to the appearance of Amazon product pages ("Scripts").

One such script that I'll consider in this section is OiNKPlus. With over 116,000 downloads, the OiNKPlus script functions as a "media enhancer" and "artist discovery" tool that augments the exploration of participatory archives on BitTorrent sites like Question.cd. According to the script author, OiNKPlus was developed because,

Music discovery is an essential part of private torrent sites. Although OiNKPlus was originally developed exclusively for OiNK, after the site was taken down by the BPI and the IFPI, I decided to go all-out and support not only the most popular private torrent sites, but some of the most popular public ones too. ("Indieana")

In activity theoretic terms, the OiNKPlus script incorporates a variety of mediating tools to facilitate conscious, goal-directed action by torrent community members. As users look for newer, more sophisticated ways to explore and discover content in the archive, conventional data sifting technologies like search boxes don't deliver like OiNKPlus does. Inasmuch as OiNKPlus incorporates a variety of digital objects as mediating tools that facilitate archival navigation and discovery, we can also consider this complex assemblage as a genre ecology: a multidimensional, multi-object, multi-genre response to the recurrent action of participatory archival navigation and digital media discovery.

OiNKPlus – Components

The OiNKPlus Greasemonkey script incorporates a variety of individual tools (Fig. 12).

(Fig. 12. OiNKPlus extension)

The "Similar artists" tool pulls metadata from the aforementioned "Similar artists" input box included in every torrent description. Employing user-generated data from "Similar artists" input, OiNKPlus aggregates and collocates data, emphasizing the artists most often identified by site users as related to the band in question. In the example provided above (Fig. 12), users identified twenty-eight artists whose generic similarities are fairly congruent with the type of music composed by Talkdemonic. When users click any of the similar artists, they're redirected to a new page on Question.cd that contains all of the media uploaded as that artist. Denoted by a small tag icon, the "Tags" section draws on the aforementioned user-generated tags to guide potential downloader's expectations about the kind of music Talkdemonic produces. The "Abstract" section harvests data from social music site Last.fm to provide users with a short biographic summary of the band. The "MySpace player" section draws on the MySpace social network, embedding the player found there to preview particular tracks before they go about downloading the entire torrent. The "External Links" section allows users to automatically query popular informational sites like Last.fm and Wikipedia to learn more about the artist; additionally, the "External Links" also provide links to commercial sites like iTunes, Amazon, and Bandcamp, allowing users to quickly access and purchase material by the artist should they feel compelled to do so. Finally, the "Elsewhere" section allows users to automatically search other private bitttorrent sites such as Pancakes.fm as well as public trackers like ThePirateBay and Mininova.

Taken collectively, the OiNKPlus tool provides Question.cd community members numerous avenues for exploring artist content via preview functions while also allowing users to draw new associations among artists whose music is generically consubstantial. Drawing on both formal and user-generated metadata, the tool augments user experience, allowing community members to meet two exigencies: participatory archival navigation and archival discovery. Interestingly, OiNKPlus also provides associational visualizations, guiding members whose primary form of engagement isn't necessarily list-based. The "Similar Artists Map" (Fig. 13) draws on both Last.fm similar artists links as well as the user-generated similar artists tags to create a visual representation of media connections. Utilizing the same frequency-based visualizations as popular tools like tag clouds, artists who were tagged more frequently have thicker connective lines and their font appears larger.

(Fig. 13. Similar artists map)

The final component of OiNKPlus breaks from an emphasis on archival navigation and music exploration, instead highlighting the interconnectedness of digital and analog worlds. The "Upcoming Concerts" section (Fig. 14) allows users to see where the artist or band will be appearing or performing in the near future.

(Fig. 14. Upcoming concerts section)

When the "Go to Thread" option is clicked, users are taken to a forum on Question.cd that connects individuals hoping to attend the same performance. Herein community members coordinate meetups to attend the show and in the process build social connections in the material world. The "Upcoming Concerts" section of OiNKPlus functions in much the same way as the social ecologies that Byron Hawk describes in his article, "Curating Ecologies, Circulating Musics." According to Hawk,

[D]igital media allow musicians to curate their own ecologies of practice in order to open pathways for the production of social ecologies around and through music . . . . [these kind of immaterial connections] fold back over onto physical ecologies and gathered publics in the form of the shows and concerts attended by folks who are involved in [the band's] digital practice. (165-6)

While the "Upcoming Concerts" section doesn't necessarily help users achieve participatory archive navigation, it does enhance artist discovery by providing virtual community members the opportunity to network, and in the process, shifts the ecological impact of the OiNKPlus script away from digital worlds into analog environments.

OiNKPlus – Medium/Interface

The substantive and formal characteristics of OiNKPlus components mirror the tagging, collage, and requests tools at the mesoscopic level of activity. Substantively, the script allows users to meet the social needs of archival navigation and artist exploration and discovery. Formally, most of these tools use simple user input boxes to record user-generated metadata that is then transformed into various data-sifting outputs like the Similar Artists Map visualization. When we supplement this exploration of mediating tools as rhetorical genres with a consideration of medium or interface, we get a clearer picture of the power of digital tools at mesoscopic levels of activity. In Lingua Fracta, Brooke argues that "the mutability of new media means that we should be shifting our focus from textual objects to medial interfaces" before providing an ecological rendering of the rhetorical canons for new media (6). Brooke's project is of particular importance to articulating the interface as the ever-elastic, interstitial space wherein new media objects interlock and overlap to create new media genre ecologies. These ecologies, and their interfacial wrapping are crucial to understanding tool-mediation in digital activity systems like the participatory archive at Question.cd.

Brooke argues that as analog textual artifacts remain fairly static, interfaces change—and they change for a variety of reasons. In his study, Brooke traces the transformation of the World of Warcraft interface, highlighting how user game progression necessitates transformation and customization of the interface to more successfully meet user need and desire in addition to quest-specific demands. In much the same way, Question.cd users making use of OiNKPlus are able to reconfigure the content sections, toggling different components on or off and rearrange entire sections to better assist them in navigating the content embedded in the medium. Interestingly, the OiNKPlus interface also transforms in another way: unlike user-initiated interfacial reconfigurations, shifts in aggregated metadata culled from site users transforms the output of the OiNKPlus script, reflecting the associational links site users draw among media over time.

If rhetorical genres modify according to shifting rhetorical circumstances and embody situational expectations, the interfacial medium of OiNKPlus is responsive these processes at the level of the individual and the social. Further, as an aggregation of multiple preexisting new media objects brought together to fulfill the recurring social motive of archival navigation and media discovery, OiNKPlus coordinates the activity of what Spinuzzi calls "genre ecologies" or Orlikowski and Yates call a "genre repertoire": a set of genres/agents that given communities understand and know how to manipulate. To that end, the OiNKPlus script is a mediating tool functioning as rhetorical genre ecology, providing users with a series of outputs that enhance participatory archive navigation, facilitate artist discovery and sometimes even promote interaction with the physical ecologies of community members.

Conclusion

Analyzing digital tools as rhetorical genres that assist users in achieving conscious, goal-directed action helps writing studies scholars better understand the novel ways that tool mediation is inscribed with cultural-historical histories of use. This approach problematizes instrumentalist approaches to digital tool mediation by positioning digital objects as instrumental causes, not merely instruments. Tagging systems, collage utilities, and other digital tools that rely on social data allow researchers to trace social motives in web communities in novel ways, shining additional light on the iterative relation among discourse communities, tool use, tool revision and individual action. As participatory archives continue to spread and thrive, researchers hoping to study the social operations of folksonomic organization and textual curation will benefit by adopting and RGS-inspired approach to digital tool mediation. Implementing this methodological orientation should produce more nuanced accounts of the sociotechnical dynamics of mass participation in Web 2.0 environments.

While this article analyzes the informal organizational practices of users curating massive participatory archives, other digital objects functioning as rhetorical genres also demand attention. Going forward, writing studies researchers might consider the metacommunicative social actions of literacy practices like hashtagging and other content-curation tools. These digital objects also embody histories of cultural-historical use, perform substantive rhetorical action, are characterized by recurrent form and are tethered to the technological affordances of the medium/interface. Considering the metacommunicative as a counterpart to folksonomic organization, future research into digital tools might shine bright lights onto the individual and collective aspects of our emergent, socialized digital experience.

• 1. The participatory archives I analyze in this study are illicit sites of cultural production. Because site members are involved in breaking international copyright law, Question.cd and Pancakes.fm are pseudonyms used to protect the invite-only pirate communities from which the data for this study is drawn. As such, specific user counts and archival statistics drawn from Question.cd and Pancakes.fm will lack citation.
• 2. http://archive.org
• 3. For an extended study of archival curation, see Kennedy (forthcoming).
• 4. This number isn't indicative of the total pieces of media included in the Question.cd archive as each individual torrent often includes multiple individual media files. For example, the .pdf ebook of William Shakespeare's Hamlet contains only one file, the .pdf; however, the torrent for Neutral Milk Hotel's 2012 EP Ferris Wheel on Fire contains eight individual song files.
• 5. Of the ~1.4 million torrents on Question.cd, 75% are "seeded" or made available by multiple users. This means that if one user disconnects their computer from the shared network, access to the files will be made available by another user. The redundant and distributed nature of archives utilizing the BitTorrent protocol is the single most important technological advance in p2p sharing over first generation direct peer-to-peer platforms like KaZaa or Napster.
• 6. Unfortunately, this analysis cannot give special attention to human actions that occur at the microscopic level of activity. This omission is mainly an observational issue as I am unable to actively record the kinds of actions that have been rendered operations by piratical site users in distributed locations. What I can conclude, given my own experience in these activity systems, is that actions such as typing, web browsing, and mouse use are operations that facilitate actions at the mesoscopic level but remain unconscious to community members.
• 7. EXtensible Markup Language
• 8. http://tags.library.upenn.edu/
• 9. http://weheartit.com/