Page 71 →Chapter 4 Content Contributors, Vandals,
and the Ontology of Curation

In the first edition preface, Chambers is explicit about his reliance on numerous previously published resources. But as he prepared the second edition over the next decade, he went himself one better by proactively working to develop the wide collective of contributors that he described in his preparatory considerations. By soliciting information from the working class and the poor, he hoped to considerably expand the craft knowledge contained in the Cyclopædia. Doing so would fulfill a longtime goal that the natural philosophy community had not yet been able to satisfy because of the care that craftspeople necessarily took to protect proprietary “mysteries,” or craft techniques that were passed from master to apprentice orally or through physical demonstration.¹

Besides being imminently practical, the practice of crowdsourcing material aligned with the ideals of the Freemasons who materially supported the project. The Masons had early on adopted a proletarian policy toward membership and contributions of knowledge. Because the society developed from the guild system and was initially composed of working masons and architects, they retained a commitment to including craftsmen. As early as 1663 their orders and constitutions decreed “that no person, of what degree whatsoever, be accepted a Freemason, unless he shall have a Lodge of five Freemasons at least, whereof one must be a Master or Warden of that Limit or Division where such Lodge shall be kept, and another to be a workman of the trade of Freemasonry.” While they actively recruited “affluent, fashionable, and influential aristocrats,” Berman has noted that the inclusion of working stonemasons remained “a practice current through to the second and third decades of the eighteenth century.”² Chambers’s call for contributions of craft knowledge and contributions from craftspeople who had first-hand expertise in “mysteries” reflects a similar dedication to including the working class and respecting their expertise as equivalent to and at times integrated with scientific knowledge.

Serialization provided another means of gathering new material and corrections from the reading public. Since this particular audience was receiving and ostensibly reading three fresh pages of the text each week as it was being produced, they were in position to offer potential additions and corrections. If a large text was being produced over multiple years, it was possible to incorporate Page 73 →this feedback, especially in topics that had not yet been typeset. Chambers very likely included this audience as he wrote the text for this pamphlet, since the working class and craftsmen whom he explicitly invited to contribute would have been just the sort of readers who were in a position to benefit most from access through serialization. These contributors would have remained heavily male and privileged as a group, though, given the fact that contributing required leisure time and literacy that were not always available to women and servants whose primary duties were limited to the private sphere.

This type of distributed work was best suited to being undertaken by a unified community, Chambers suggested in the Considerations, but none was to be found at the time owing to the lack of monastic communities and the “Royal Academies and Institutions” having focused their attentions elsewhere. Besides, in his estimation, these dedicated communities required too many meetings in order to tackle this sort of project, too much ceremony and negotiation of feelings, and too much ado. Better to establish a system of “voluntary, or occasional communications, where public-spirited persons, at their leisure and liberty, furnish materials and intelligences to the undertakers of useful designs. . . . This method seems no where more practicable to better advantage than in the work before us.”³ Bringing together contributions from such a vast array of contributors would result in “the best book in the universe,” he argued, successfully appealing to national identity and pride in this invitation: “Many of our private clubs might vie with [other countries’] academies, and much of the conversation at certain coffeehouse tables, with their conferences in form.”⁴ This invitation implied, specifically, that the English could do better than the more formal French intellectual society and thus suggested that contributors should devote their energies to his project if for no other reason than competitiveness.

The lack of surviving records obscures the extent to which the public took up this egalitarian invitation, which was a remarkable precursor to the invitation extended by Wikipedia (and demonstrates the extent to which nonhuman agents shape even our reconstruction of this story), but it is possible to track changes across a limited sample. The result seems to have been that the breadth of the entire project was extended more than the scope of individual articles.⁵ The articles sampled show little content revision or substantive editing between the first and second editions. The few content changes that occurred in the articles sampled consisted primarily of occasional edits for technical clarity and improved cross-indexing. However, there were significant typographic changes. These small, consistent alterations result in a high incidence of edits in texts while size and breadth remains relatively static (see table 1).

One of the most striking differences between composing processes in the Cyclopædia and Wikipedia is the way that articles begin their lives. Each of the Page 74 →Chambers articles sampled appears in the first edition as a recognizable, fully fledged article rather than as a briefer text that was clearly meant to be expanded. Some other articles in the 1728 edition are appreciably short, but nearly all of them consist of at least a tidy paragraph. This approach is significantly different than that in Wikipedia, which relies on many small changes over time for expansion of its articles. This difference is driven in part by the technological actors available to each project. A bound, print encyclopedia that is published in a new edition once each decade (or less) depends on a unified, substantive revision that is not extendable or correctable after publication. Correcting the text and broadening its scope in successive editions was also a philosophical and economic decision that made sense both in terms of documenting findings emerging from observational natural philosophy and selling that information to a public who required the incorporation of significant new information in order to be persuaded to purchase an expensive text.

Incorporating necessary corrections and working toward “universality” were primary concerns for Chambers, particularly in light of his desire for an encyclopedia that presented a coherent, comprehensive view of the world. In the Considerations he wrote, “Indeed, much remains to be done in order to answer the design of a Cyclopædia, or Universal Dictionary, in its just latitude. . . . Our universal works are generally so scanty that they leave room for other universals of the same kind without interfering. Thus Universal Dictionaries of Arts and Sciences, and those others called Universal Historical Dictionaries, have long reigned as collateral and independent things, as if history and geography were not one of the number of sciences, or as if any science could be given to advantage without Page 75 →its history. . . . ’Tis evident the treaty of partition is but ill concerted, where generals are torn from the particulars belonging to them and reason and rules are thrown into one book, and facts and instances on which they are built, into another. This makes them both lame of necessity.”⁶ By continuing to broaden the scope of the text in each edition, he and his publishers hoped to present a more unified collection of knowledge and course of study for their readers, facilitating a comprehensive view of a logically ordered, Newtonian world. Wikipedia, with its very different material aspects, which afford tremendous expansion and open vandalism, has a more far dynamic set of performances to shape and control.

The ontology of Wikipedia is found in central policies that are applied across the community and text. These policies, known as the Five Pillars, are open to interpretations that are in turn shaped by conflicting primary philosophies about what sort of information an encyclopedia with the potential for nearly infinite expansion should contain. They are also shaped by disagreements about the public nature of articles’ lives as they are incrementally drafted and polished. Performances of authorial agency in these negotiated process are contingent and variable, moving the community and its project forward but also backward. Positive constructions of affiliation and production occur, but just as often the destructive potential of a text and community that are too big to control come to the fore in instances of alienation, discord, and vandalism.

Wikipedia is frequently rhetorically positioned as “an encyclopedia that anyone can edit,” and in its early days this was unequivocally true: while Wales and Sanger were overseers of the site, all contributors were considered more or less equal, and contributions to all pages were immediately visible on the web.⁷ These days editors are encouraged to log in with an account, and most, but not all, pages remain open and available for editing by anyone. Access to articles on contentious topics or topics that are at the top of the news cycle may be restricted, as are articles that are currently designated as Featured and spotlighted on the front page of language-specific versions.⁸ (As I write, the Featured Article is on the majungasaurus, a dinosaur that lived in Madagascar sixty-six to seventy million years ago.) Access to the front page of the site is heavily restricted, along with some policy pages. Still the majority of the encyclopedia remains open for editing by the public, and the project depends entirely on volunteer contributions.

This underscores one of the ways that technological affordances influence the labor of composition. Because of the centralization of performances necessary to compose a print text, primary authorial agency is ultimately localized to the project editors. As a result public contributions are not readily apparent Page 76 →at the interface level, and unless the curator takes explicit measures in offering acknowledgment to contributors, he or she will almost always be pointed to as the central composer of the text. If Chambers did indeed perform the bulk of this work in composing full articles for the initial edition, as he says he did, then the considerable labor and skill involved rightfully earn him that recognition. But this does not render him the sole author of that text, which was curated through careful integration of previously published and freshly contributed information. In Wikipedia the community—especially the central committees—conducts the distributed work of curation through incremental, contingent performances.

Despite the hundreds of thousands of registered editors, only a relative handful contribute significantly to the project. Over the years the estimation has fluctuated: in 2005 Wales estimated that “2% of the users do 75% of the work,”⁹ while a 2007 empirical study suggested even more startlingly that one-tenth of 1 percent of editors—approximately forty-two hundred people—contribute nearly half of the site’s content.¹⁰ The dedication demonstrated by relatively few editors is one of the elements that render the distributed work of composing Wikipedia curation rather than random composition by individuals dropping in out of the digital skies, making changes, and then never returning. This cohesive, central group of editors has both understanding of and commitment to the pillars of Wikipedia. They also closely monitor contributions in order to curate the project as an encyclopedia rather than a loose collection of links and opinions. Over time, this group has accrued ethos through long service, forming a cabal of long-timers that controls central policy and arbitration decisions. Their performances are exponentially dominant in the system by virtue of setting the rules for other actors and shaping decisions about content and interface that influence both the curated text and the actions of both human and nonhuman curators.

As a group Wikipedia editors comprise a heavily male demographic, with only approximately 10 percent of the group estimated to be female.¹¹ In an effort to encourage women to become involved as editors and to build articles on notable women, the community launched a Gender Gap Task Force to address systemic issues. A number of reasons for this imbalance have been suggested, with the most prominent reason being the uninviting ethos of the central community, which has been characterized on a spectrum from simply “unwelcoming” to a “crushing bureaucracy with an often abrasive atmosphere”¹² to “a rancorous, sexist, elitist, stupidly bureaucratic mess.”¹³ New users with few edits associated with their names are automatically flagged and frequently subject to having their contributions reverted or marked for deletion either because they do not understand the guidelines for contributions or because their newly-registered identities are not recognized and hailed by established Wikipedians. Challenges to content are semiautomated and abrupt, dispensed in boilerplate language that many Page 77 →newcomers find alienating. Women editors who prefer more personal communication styles may feel particularly unwelcome. In theory anyone can edit Wikipedia; in practice all actors are not equal within the system, and neither are their performances or contributions. Identifying as male within the system and having a very thick skin are considerable advantages.

As a result of this imbalance within the community demographic, Wikipedia’s content tends to be more developed around topics that are understood as sterotypically male, resulting in a systemic bias in content coverage. One often-cited example concerns the many extensive articles devoted to each episode of The Sopranos as well as each of its recurring characters, as opposed to the single article that covers Sex and the City. A number of universities and organizations have also worked to address this systemic problem by holding edit-a-thons on or near the date of Ada Lovelace Day, honoring the first woman computer programmer. The Royal Society has been one such sponsor, working to build articles on every female society fellow.¹⁴ Other edit-a-thons have helped build participation from other underrepresented demographics. While these efforts have improved the scope of the text to some extent, they have not fundamentally changed the demographic of the community. Another factor, which Adam Klein¹⁵ has devoted attention to, is the inevitable inclusion of racist propaganda in Wikipedia and other online resources in a coordinated effort at “information laundering” by racist communities. These contributions work to legitimize hate and also create a hostile environment for potential editors who identify as members of the targeted ethnic groups even as other Wikipedians work to revert racist edits. Systemic bias in various forms extends throughout the text and, combined with a prominent community culture that understands Wikipedia as a “battlefield of ideas” that benefit from being fought out, powerfully limits curatorial contributions by women and other underrepresented demographics.

These cultural mores have shaped the size of the contributing community and, by extension, the rate of production. Wikipedia saw robust growth of its community in its first half decade, with exponential expansion of both registered users and text each year. By mid-2007 journalism professor and Wikipedia historian Andrew Lih was among many who pointed out that casual contributors were leaving the project, feeling that they and their work on minor topics (that is, those deemed not notable) had been rejected.¹⁶ The total number of active editors has declined by half since then, hitting a low of thirty-one hundred in 2013.¹⁷ As Priedhorsky and coauthors have indicated, the contributors who do stay tend to be invested and committed to topics they are passionate about and have subject-area expertise in. Unsurprisingly communities form around common curatorial projects and interests. These smaller communities provide space for editors to have a better chance of forming closer social ties that in turn help bind them to Page 78 →the distributed project of improving information on their mutual interests. Thousands of WikiProjects curate topics in diverse umbrella categories such as “human rights” “games and toys,” “arts and culture,” or “cephalopods,” and geographically specific umbrellas such as “Africa” or “the Americas.” Subcategories further direct contributors to more specific areas of interest, so that a Lego enthusiast drilling down from “games and toys” may choose to work specifically on the Lego project, which focuses on all Lego-related articles on Wikipedia, but not on the Bionicles Taskforce, which is categorized as a related but entirely different concern. These fine distinctions in taxonomy guide content findability, of course, but also drive community formation and expansion by providing editors with a means of finding the communities they can best contribute to.

The curatorial work undertaken throughout the encyclopedia is defined by a long-standing, collaboratively developed philosophy about what Wikipedia should be, and also splintered through varying viewpoints on just what such an encyclopedia should include. Wikipedia’s Five Pillars delineate its core principles:

• •Wikipedia is an encyclopedia.
• •Wikipedia is written from a neutral point of view.
• •Wikipedia is free content that anyone can edit, use, modify, and distribute.
• •Editors should treat each other with respect and civility.
• •Wikipedia does not have firm rules.¹⁸

Here we see a concrete conceptualization of this crowdsourced encyclopedia as a curated project with centralized, specific guidelines for acceptable material that results in a filtered and ordered reference text. This and myriad other policy pages also constitute a rhetorical means of explicating its expert methodology, as Hartelius has noted.¹⁹ These pages explicate and encourage a centrally defined commitment to responsible curation and ethical community relations, shaping the performances of both human and nonhuman agents by defining tasks and behaviors. This foundational philosophy anchors the community, providing a basis from which they actively reject other descriptions that are often applied by outsiders, such as accusations of providing a soapbox, “being an experiment in anarchy or democracy,” or, most importantly, being “an indiscriminate collection of information.”²⁰ An earlier version of the first pillar was even more explicit about its emphasis on nonoriginal contributions that could be sourced to carefully selected, previously published material: “All articles must strive for verifiable accuracy: Page 79 →unreferenced material may be removed, so please provide references. Wikipedia is not the place to insert personal opinions, experiences, or arguments. Original ideas, interpretations, or research cannot be verified, and are thus inappropriate.”²¹

Remarkably the Five Pillars contain no references to truth, fact, or rigor. Instead they delineate a commitment to providing unbiased information curated by an open, civil, flexible community. In theory (although not necessarily in practice), truth and fact are not negotiated within the purview of Wikipedia. Rather the encyclopedia consolidates and reports what the larger culture it functions within has deemed to be true or factual by virtue of citable publications, and it works to do so without bias. Neutrality has become a stand-in for factuality, on the assumption that an unbiased report that avoids opinion also by nature avoids censorship, promotion of vested interests, and other rhetorical means of skewing reality. This is a substantial difference from the Cyclopædia, which did not pretend to anything approaching neutrality in its advancement of a Newtonian worldview.

The neutral point of view (NPOV) pillar shapes other aspects of contributions that might influence neutrality. Originality is explicitly banned, to the point of specifying which possible peripheral forms of originality are not allowed: not only original ideas, but also interpretations, research, opinions, experiences, or arguments. All information must be traceable to previously published external sources and cited as such, and in fact such dense citation is part of the basic criteria for Good or Featured Article status. Editors who devote their energies to building rigor within the project contribute by tagging entries that require verification and/or citation. The Fact tag, which inserts a [citation needed] hyperlink after sentences that need citation, is one such tag. In one instance an editor inserted this tag liberally in sections of the Minerals article, noting that this was meant as a helpful gesture: “large-scale fact tagging, typo-fixing, etc.” Six minutes later, the same editor added, “another fact tag-will remove the material if not fixed in the next day or so, but don’t want to disrupt authors ‘work in progress.’” In doing so this editor was demonstrating one aspect of the social, participatory nature of encyclopedic authorship, particularly within Wikipedia: constantly working with collaborative writing and thus frequently negotiating changes rather than just making them autonomously.

Privileging external sources as “facts” over the original contributions of editors who may be credentialed or subject-matter experts is sometimes interpreted as antielitism and has remained a point of contention among contributors to the text, as Sanger, Lih, Hartelius, and many others have noted.²² One of the more famous incidents resulted from the author Philip Roth’s attempt to correct the Wikipedia article on his novel The Human Stain.²³ Roth was astonished to find that his request to correct misinformation about the basis of the novel was rejected: “‘I understand your point that the author is the greatest authority on their own Page 80 →work,’ [wrote] the Wikipedia administrator—‘but we require secondary sources.’”²⁴ Roth rather correctly felt that he had no recourse but to write an open letter to Wikipedia, which the New Yorker promptly published. The published letter was then considered a valid secondary resource by the Wikipedian community and the correction was made.

Many researchers, scientists, and subject-matter experts have found themselves in the same position over the years, leading to popular charges of antielitism within the project by Larry Sanger as well as media commentators.²⁵ The community, buttressed by the pillar tenet that “anyone can edit, use, or modify” Wikipedia, rejects any formal qualifications or earned expertise in favor of sourced citations and neutral point of view. Any valuation of expertise lies in shaping and becoming expert in Wikipedian policy as well as contributing to the text through rigorous policy applications. As Jemielniak observed in his ethnographic study of the Wikipedian community, “In the Wikimedia movement, perfecting the procedural rationality can be observed in the concentration of debates on egalitarianism, inclusivity, freedom of information, and respect for significant minorities (all related to how collaboration is conducted) and not reaching the goal of a moment when every human being can freely share in the sum of all knowledge (which is the movement’s purpose, but the amount of time spent discussing how to reach it is significantly less.)”²⁶ Elucidating the ways that curatorial performances are ideally constrained is a high-priority task for the community—higher priority than composing text, it often seems. Any editor who hopes to make a persuasive contribution that will stand over time or to argue for any substantive change must account for myriad policies in the process.

However, the Fifth Pillar—Wikipedia has does not have firm rules—complicates these matters. Policies drive a clear consolidation of power by central communities, but the instruction to “ignore all rules” performs a more complex, multilayered rhetorical function. Paradoxically it is itself a rule. The original, public intent behind this pillar and its relevant injunction, “Be bold!” was to create a digital space and community that welcomed users without requiring them to learn a host of rules and to foster innovation by welcoming a variety of editorial performances. Officially this pillar always and ever keeps the community’s focus on its central purpose: creation of a free encyclopedia.²⁷ It renders all policies contingent, which can indeed drive open growth by relaxing constraints. In practice its application often subverts its rhetorical intent and accrues additional power to members of core communities, who use it as a last resort during protracted arguments. If threading the needle of project policies fails to persuade an actor who makes equally persuasive policy-based arguments, then it is always possible to fall back on the argument that there are no rules and that core consensus has been reached, thus vehemently discouraging further dialogue.

Page 81 →One example of these policy-based negotiations is the question of whether or not to include an image of the prophet Muhammad in the article devoted to Muhammad, which has concerned the community for years, resulting in extended discussion about how neutrality might be defined in this case. Insistence on maintaining a neutral point of view (NPOV) in article texts has remained an abiding concern over the life of the project. Avoiding bias is a seemingly clear criteria, but its ethical application is not always at all clear. Islam traditionally forbids depiction of the prophet in any context, including reference texts. Wikipedia conventions require the inclusion of relevant images on each article in order to enhance educational value and account for multiple modalities of learning. Consequently a number of revert wars that removed and reinserted depictions of the prophet have led to heated discussions. At the moment the lead image is a calligraphic representation of Muhammad’s name, but images of the prophet’s face appear in article subsections. Offended readers are offered instructions on how to run a script that hides the images from their view, but the depictions remain on the page. Muslims requested removal of the images based on religious respect, while editors argued that to do so would violate the NPOV rule and constitute censorship. The core English Wikipedian community’s vigilance against censorship stems from the community commitment to openness and from a particularly American ideal of free speech and open circulation of information. It also exemplifies the contemporary, Western expectation that reference texts will remain uncensored—an expectation that is not always reflective of the actual state of reference texts.

Other requesters have argued that the images should be removed because they were inaccurate based on scriptural descriptions of Muhammad’s appearance. Editor Hasim221990 wrote an impassioned argument from this perspective, referencing specific quotes in the Islamic Hadiths. The inaccurate images held no historical value, he added.²⁸ This argument based on received religious truth was met with policy-based rebuttals by editor Amatulić: “The fact remains that you are engaging in original research. No one has ever claimed that the images are representative of how Muhammad really looked. Your implication of this is a straw man argument, as is ‘reliable’ versus ‘holy,’ which is a point no one brought up but yourself (and false, too, since not all Hadith are considered reliable by mainstream Muslims). In addition, there is no editorial policy that states a representation of an historical figure must be completely accurate. It is a well-understood and uncontroversial fact that artistic license is employed by artists, and it is obvious for these images considering when they were painted. That does not invalidate their use to illustrate the article.” Truth or factuality is explicitly not a criteria in Amatulić’s interpretation of NPOV, and neither is accuracy. Neither is historical significance, he goes on to say, since that is also a matter of opinion. Rather what matters most is citation of valid external resources and reaching Page 82 →consensus: “What matters most is community consensus and reliable sources. You have failed to present a single reliable source that disputes anything about any specific image presented in the article, we have only your WP:SYNTHESIS²⁹ (a policy violation, read about it) saying that sources accepted by Muslims describe Muhammad a certain way, therefore the images must be removed. All we have are your opinions about how we should apply the sources you espouse, which are religious sources, in spite of your protestations that they are not.”

A number of appeals to pathos have been made in the discussion, asking Wikipedians to consider the ways in which this decision makes Muslims feel unwelcome as readers or contributors, and hinting that blatant disrespect for religion foments conflict that can only lead to further attacks like the Charlie Hebdo shooting. “It sounds more logical that a website which focuses on making articles from a neutral point should know that having a successful policy requires a certain amount of respect towards religious, social and similar traditions,” wrote editor Emin Čamo. “Also considering that these depictions have caused a lot of riots within the Muslim World, do you not think that this is violent propaganda? Since it, consciously, is hurting the adherents of Islam.” Altering the content based on these perceived slights would be not only a violation of policy, editor Alanscottwalker replied, but also a negation of the educational mission of an encyclopedia: “Beyond ‘offense,’ a fundamental issue is ‘educational value,’ and that’s not an NPOV issue, it is an issue of educational mission, and Wikipedia does editorially agree that there are multiple modalities of learning, and that therefore Wikipedia aims to present ‘multimedia learning’ including verbal, and visual (and even sometimes audio). (See eg., WP:IMAGE RELEVANCE. [sic]) Couple that with the fact that Wikipedia is relating biography of a man in this article, and that the target audience for Wikipedia is more or less, the words of the boards ‘offensive images’ consultant, modern secular societies.”

The informative mission of the encyclopedic genre is paramount in this counterargument. Textual forms have agency, as Campbell reminds us, and the educational performance inherent in this type of text shapes not only content but also the performances available to contributors. The educational mission of encyclopedias as has been interpreted for centuries demands the use of images, and the WP:IMAGE RELEVANCE policy that Alanscottwalker cites institutionalizes this generic convention within the purview of Wikipedia. “Images must be relevant to the article that they appear in and be significantly and directly related to the article’s topic. Because the Wikipedia project is in a position to offer multimedia learning to its audience, images are an important part of any article’s presentation. . . . Images are primarily meant to inform readers by providing visual information. Consequently, images should look like what they are meant to illustrate, even if they are not provably authentic images. For example, a photograph of a Page 83 →trompe-l’œil painting of a cupcake may be an acceptable image for [the article titled] Cupcake, but a real cupcake that has been decorated to look like something else entirely is less appropriate.”³⁰ Verisimilitude counts, but authentic truth is unnecessary in this definition. More important is the educational mission and making that education inviting and available to readers with a spectrum of learning styles.

Less important is making the project welcome to a variety of contributors who may be offended by content or accommodating a non-Western viewpoint in this Enlightenment based encyclopedia. In this exchange I have located only one participant who has suggested that including depictions of Muhammad constitutes a lack of neutrality in the form of anti-Muslim bias. A Western, presumably non-Muslim perspective is the perspective from which bias is institutionally detected and subjected to policy. This dialogue remains unresolved at this writing, having intensified in the wake of the January 2015 shootings at the offices of Charlie Hebdo, a Parisian magazine that had repeatedly published cartoon depictions of Muhammad. Nearly sixty-five thousand Muslims and their supporters signed a petition asking the editors of Wikipedia to remove all images of Muhammad on the grounds of religious respect.³¹

Conflicting philosophies about most optimal and ethical scope of Wikipedia loom large in the curatorial consciousness of the community. These conflicts again reveal the potential for both constructive and destructive authorship. The contingent nature of Wikipedian policy is also evident in WikiPhilosophies, which are positions that vested community members adopt regarding how the content of the encyclopedia should be ethically managed at the macro level. A number of conflicting stances are available to community members who are invested in the broad curation of encyclopedic content in terms of scope, collaborative practice, structure, and standards.³² The two most prominent debates that have reverberated through the community over many years and reached the mainstream media are deletionism/inclusionism and eventualism/immediatism.

With sufficient curatorial contributions, platform maintenance, and server space, Wikipedia can potential expand to include articles on almost any topic that an editor is willing to write up. In many ways doing so would be the logical conclusion of the eighteenth-century quest for a universal compendium. But would an infinite resource still be a rigorously curated one? Wikipedians wrestle with this question daily, and as befits a broad community of practice, they do not always agree on what constitutes fulfillment of the pillars or how the central curatorial tenets inform the goals of the project. These differences result in the Page 84 →articulation of differing “WikiPhilosophies,” or beliefs about what is best for the project. For some editors public affiliation with such beliefs are central to their Wikipedian identities and thus warrant visual codification through badges that are prominently displayed on the Userboxes on the sidebar of their User page.³³ Among the most prominent of these differences are the organizing philosophies known as deletionism and inclusionism.

This fundamental conflict over what constitutes appropriately curated content has come to be known as the deletionism/inclusionism debates, which have been covered by prominent media outlets such as the New York Times. Inclusionism is “the philosophy that as much of the material submitted to the site should be kept as possible.”³⁴ Inclusionists’ rallying cry, “wiki is not paper,” refers to the nearly infinitely expandable capabilities of wikis. Because limitless growth is possible in Wikipedia (provided there is sufficient server space and power), the inclusionist philosophy dictates that the editors’ moral responsibility is to foster all potential expansion and thus provide as much information as possible across a potentially universal array of topics. Minor topics and article stubs are acceptable contributions that should remain published, in their view.

Conversely deletionists “feel that an article should be in reasonable shape and about a clearly notable topic before being included; questionable material should be deleted more rigorously.”³⁵ Their reply to “wiki is not paper” is “wiki is [indeed] not paper, but neither is it an attic.”³⁶ The somewhat ambiguous curatorial criteria of notability is generally understood as “notable to an outside observer”³⁷ and can be “determined by coverage of the topic in upstanding, independent sources.”³⁸ From the viewpoint of deletionists, a more tightly curated encyclopedia results in higher quality and consequently more authority, since labor is necessarily concentrated over a smaller number of articles. Inclusionists assert in turn that “the process of deletion carries a ‘systemic bias’ or ‘process bias’ by which the process itself becomes a haven for those who wish to use it—emphasizing deletion rather than offering reasonable alternatives for keeping or redirection.”³⁹ Deletionists’ commitments are to a more tightly curated encyclopedia that offers the best information on topics that they understand as prominent or noteworthy within the context of their worldview. They face accusations of excluding minority voices and topics on a number of levels and of destroying swaths of encyclopedic text.

While these opposing curatorial viewpoints are easily stated, they are not easily enacted. “Notability” is a primary requirement for a topic to warrant its own article and remain included in Wikipedia. The guidelines define notable topics as “those that have gained sufficiently significant attention by the world at large and over a period of time, and are not outside the scope of Wikipedia. We consider evidence from reliable independent sources to gauge this attention.”⁴⁰ While seemingly straightforward enough, notability is a concept open to broad Page 85 →interpretation by users according to their own worldviews and experiences of what is important. Consequently the concept and its possible interpretations receive heavy discussion on the daily lists of articles flagged for deletion.

A number of researchers have argued that deletionism has become more prevalent throughout the project for years, and occasional categorical purges have continued. Among these are “notability purges,” such as the mass deletion of all web comic artists in the fall of 2006.⁴¹ The conflict has been sufficiently intense to spur the formation of the Association for Inclusionist Wikipedians and the Association for Deletionist Wikipedians, each with a formal WikiProject that enrolls members to patrol the site as a team for articles flagged for deletion or moved to arbitration.⁴² Not every Wikipedian has adopted a side; there also exists the Association of Wikipedians Who Dislike Making Broad Judgments about the Worthiness of a General Category of Article, and Who Are in Favor of the Deletion of Some Particularly Bad Articles, but That Doesn’t Mean They Are Deletionists. Active editors from all sides of this debate roam the text, using watch-lists for automatic notification of new edits and advocating for their side in the daily Articles for Deletion discussions. Effective rhetorical choices are vital in this forum; the deletion process guidelines note that “justification and evidence for a response carries far more weight than the response itself.”⁴³ Many pages flagged involve topics related to public entertainment: authors, forthcoming movies or video games, band reunions, less-known cultural figures, and the like. But others involve civic issues or questions concerning public memory. The outcome of these discussions is necessarily dependent to some extent on what the majority of participants find notable based on their own perspective or community identifications. A number of researchers have argued that given the heavily male and white makeup of the Wikipedian community, systemic bias is an inevitable outcome of these negotiations.

An early example of these debates is the no-longer-published page on Biloxi resident Harvey Jackson, who became the face of Hurricane Katrina for several days when, in an interview with the television network ABC, he described the hours that he hung onto his wife’s hand as they clung to a roof in the floodwaters. She was swept away from him, and, at the time of his interview, her body had not been recovered. His face and story appeared frequently in both mainstream media and citizen journalism reports on the hurricane devastation, with CNN calling him “the face of the disaster.” As the Wikipedia articles on Katrina were built in real time during the event, an article covering Jackson’s tale was created as a stub and then built out. However no such page is found in Wikipedia today; it was marked for deletion not long after its creation on the grounds that Jackson was insufficiently notable.⁴⁴ The discussion that led to this decision illustrates the messy work of textual curation in encyclopedic projects, especially one that Page 86 →addresses current significant events involving living people as subjects. Following are excerpted votes from the more than seventy-four cast on the deletion review for this article.

Weak delete or transwiki. Wrong project. There’s nothing wrong with the content except that it’s not encyclopedic, because nobody is going to remember Mr. Jackson in a year or two. The fact that people are voting “keep for the moment” is a perfect indication that this is transitory content: interesting now, but not for long. That makes it news, so it belongs on Wikinews. Isomorphic 04:31, 31 August 2005 (UTC)

Delete. Perhaps when the Harvey Jackson Trust opens, but at the moment he’s no more notable than any other named or photographed interviewee in a disaster report, i.e. he isn’t notable. Besides, how do we know he isn’t lying?-Ashley Pomeroy 04:57, 31 August 2005 (UTC).

Merge into Effect of Hurricane Katrina on Mississippi or Effect of Hurricane Katrina on Biloxi, after Effect of Hurricane Katrina on New Orleans. Samaritan 08:00, 31 August 2005 (UTC)

Delete per Nandesuka. Getting your 15 minutes of fame does not make for encyclopedic notability. Fernando Rizo T/C 16:25, 31 August 2005 (UTC)

Merge into effects/media coverage. I’ve seen the clip featuring Jackson on CNN, Vesti, and ZDF, so it seems to have been one most widespread clips of early coverage, and is therefore noteworthy, but probably not noteworthy enough to deserve its own article. 66.177.132.73 08:57, 1 September 2005 (UTC)

Delete. Millions of people lose loved ones in tragic circumstances every day. The fact that a TV crew happened to stumble upon Mr. Jackson does not make him notable. Similarly, every day the news media reports on non-notable people whose lives are momentarily thrust into the spotlight. If Jackson gains notability, then we can add an article.-Cn2b 06:03, 31 August 2005 (UTC)

Delete. Would Wikipedia keep an article about a Bangladeshi Harvey Jackson? We need to, constantly, be aware of the problems with Ethnocentrism and Systemic bias. -Tsaddik Dervish 13:40, 1 September 2005 (UTC)

Keep, in some way, shape, or form. When people remember events, they remember emotions. This is the most notable interview, and will likely be the best remembered, in 10, 20 years from now. -user:zanimum [no time stamp]

Page 87 →Eventually the article was deleted, and its removal has stood over the years since Katrina. This deletion occurred as part of the daily curatorial business of Wikipedia, which is dependent not just on gathering and filtering information before composing texts, but also negotiating cultural tenets and standards.

Another component of these discussions is article quality. The pillars make no mention of truth or facts as foundational aspects of Wikipedia. Rather they imply rigor by stating that Wikipedia is an encyclopedia, and the community supports expectations of rigor associated with the genre through adherence to the Neutral Point of View pillar, by insisting on notability, and by associating dense citation with quality through fact-tagging patrols. Through consistent negotiations over the parameters of content and consistent fact tagging, the community works to curate a rigorously contained and referenced encyclopedia. Given the distributed and open nature of the building process, community members are constantly asked to confront practical complications of the basic composition process. As in classroom writing, article composition on Wikipedia proceeds from drafting to revision. When, then, should an article be made public within an open encyclopedia that hopes for rigor? The affordances of the system dictate that it is immediately and always available for both editing and reference. These complications led to another ontological conflict within among Wikipedians: eventualism versus immediatism.

The Cyclopædia articles were published fully formed. While the codex writer must submit a finished text to the publisher with confidence that it is sufficiently complete and accurate enough to stand for several years, the wiki writer plainly does not. Rather they can choose to contribute or delete as much or as little as they like to a wiki page, demonstrating an ambiguous and at times unambitious writerly agency. The low interface barrier and ethos of openness make it easy to surf into an article on a topic of interest, read through it, and think, “Oh, but I know this other fact!” Depending on available leisure time, access, abilities, and energies, contributors can click over to the page editor, add their small fact, save the page, and never come back again. Or they can spend an evening tightening the text and adding citations. They can go further and develop a more formal attachment to this page, adding it to their user account’s watch list, monitoring every change made to it and immediately swooping in to revert vandalism. They can also join a task force dedicated to strengthening the content of the broader topic area, such as the Military History WikiProject that has included the Fortification article in its purview. Any of these performances is entirely feasible, and involvement is left to individual discretion.⁴⁵ That discretion, though, is always shaped by other human and nonhuman actors.

Page 88 →At each level individual agency is constrained by intrinsic connections to broader project parameters and community constraints. Together individual writers are linked to the other writers, texts, social conventions, project philosophies, and technological affordances that make up the distributed model of Wikipedia. This organic network changes the composition process in significant ways from the print production model. The wide and intense collaborative performances by human and nonhuman agents mean that it is truly impossible to point to a central curator like Chambers, even if one editor did more work than some of the others. It also means that there is no unified “edition” and rarely even an obvious “draft” of an article, but rather time stamped iterations that can be changed slightly at any moment—or in the seconds that make up a moment. While individual editors perform these changes, the article is effectively written by an asynchronous swarm of agents. Wikipedia is, as both Shirky and boyd have proposed separately, a process rather than a static product. The byline that attributes each article to “Wikipedia, the free encyclopedia” rather than any individual editor or WikiProject is appropriate because the articles are in fact generated by the compositional process of Wikipedia. This process includes negotiation, reversion, and vandalism—a range of performances that remind us that distributed agency is not always linear, progressive, or even generative. Because of the multilevel, structural freedom that is built into the system, the development community must constantly negotiate not just content and editing issues, but issues concerning architecture, taxonomies, metadata, and the undergirding philosophies of the project. In other words the central tasks of curation are always being socially negotiated, even though the Five Pillars provide a philosophical basis for this work.⁴⁶

While each of the Chambers articles sampled appears in the first edition as a recognizable, fully fledged article, many Wikipedia articles begin life as “stubs” that consist of only a sentence or three. A stub is a very brief summary, posted as a way of getting an article started and thus creating a page others can add to. “Stubs are incomplete—by definition, they lack something vital—but they are often useful and well written. Wikipedia disrupts accepted compositional practice by publishing what are essentially rough drafts that can consist of limited, rough text that writing teachers would consider to be freewriting or outlining. It also publishes more polished work, of course, but this incremental process of drafting and polishing very frequently happens in public and over the course of a number of years. Approximately 70 percent of Wikipedia articles are still classified as stubs.”⁴⁷ Unlike print the iterative potential of wikis allows multiple editors to make small, incremental changes to a text that add up over the life of the article. Since these incomplete pages provide spaces for continued expansion, they are officially viewed by the community as open invitations for contribution, especially by casual contributors who may happen on the page via a Google search and find Page 89 →that they have something to add. This sedimentary accretion of words and facts illustrates a central philosophy of Wikipedia: eventualism, “the idea that things will eventually improve if you leave them around long enough.”⁴⁸ When early detractors argued that article stubs were worthless and that errors immediately rendered the project suspect, Wikipedians argued in turn that the project was designed for long-term growth, and that slowly articles would grow and errors be corrected, leading to eventual value. The conflicting philosophy of immediatism argues that immediate value is a primary concern of any reputable encyclopedia. In this view any poorly written or insignificant articles should be immediately removed in order to protect the overall quality of the project. Immediatism is not necessarily incompatible with inclusionism; rather it holds that any contributions should be complete and properly formatted before being posted to the site.⁴⁹ Adherents argue that edits do not just affect texts but drive education and policy: “Good edits inform the public and clarify policy, leading to good editors and therefore good articles. Bad edits misinform the public and confuse policy, leading to bad editors and therefore bad articles.”⁵⁰ Texts demonstrate agency, as Campbell has argued. In the view of immediatism, each edit has suasive influence beyond the immediate argument of the article, shaping community constraints and individual performances.

Page 90 →Despite this insistence stubs have been and remain an extraordinarily common way of starting articles. The Wikipedia articles in my sample are just such an example. Each began life as a stub, and three began with typically minuscule word counts. At this writing the articles are nearly fifteen years old—quite old in terms of Internet time—and have all grown exponentially. The first creation of the Garden article began with a single word on a page: Garden. That brief article was quickly deleted by another editor who thought it was a possible duplicate of the Gardening page, but the creator swiftly clarified the intention to create a distinct page. Both the Minerals and Trigonometry articles began with more substantial word counts, although the Trigonometry stub was immediately flagged for plagiarism, and the offending second paragraph was removed, bringing the stub start length to sixty-nine words at the second edit.

Only rarely has text growth been due to additions of large chunks of text. Rather text growth is the result of thousands of small edits: analysis of the Trigonometry article included more than 3,552 edits,⁵¹ and other edit histories are of similar length. Not all are generative, since some involve performances of vandalism that were then reverted by either human or nonhuman editors, but many contributed to the length and quality of the articles. Championing eventualism by allowing stubs to remain part of the larger text denotes an implicit trust in the power of distributed collaboration to produce textual products that improve rather than degrade over time. The “many eyes” theory developed by open-source software communities argues that open code that is made available for scrutiny by as many pairs of eyes as possible will inevitably be debugged and extended. The same is true in Wikipedia, whose open text and accessible interface makes it easy for the many eyes passing over each page to improve texts over time. Potential for these distributed performances are built into the system on both the textual and technological levels, forming the foundation of eventualism as a philosophy. This trust in the veracity of distributed compositional work and distributed agency challenges the centralized authority that has traditionally been accrued to the single, unified authorial signature that Foucault describes as an essential element of the Author function. Instead it expects textual contributions to be distributed, somewhat anonymous, and largely valid and validated. Largely but not entirely, that is, since not all performances push the project forward. Performances of rhetorical agency are not linear or necessarily generative, and the problem of vandalism has long been an area of focus for the community.

Vandalism is to be expected in digital texts that are open to editing without any login requirements. These days it is a simple fact of life on the web, along with Page 91 →the inevitability of spam. But textual vandalism is not limited to easily changed digital artifacts. It also occurred in print encyclopedias, often as intentional performances by publishers and editors who found that vandalism served their own ends. One might suppose that curators who are committed to maintaining quality and material contributors who invested in expensive projects would never have good reason to vandalize these texts, but one would be wrong. Stakeholders in print encyclopedias have not only performed strategic vandalism but also included it as a matter of project policy.

While the material production of the Cyclopædia appears to have moved forward without undue drama, it is important to note that not all possible contributions are what we would recognize as generative or necessarily even within the mission or scope of an encyclopedia that consolidates factual information. While vandalism is exponentially more prevalent in Wikipedia, it was also a possibility in the print production process. Somewhat limited opportunities for vandalism are available to both the curator and printer, should they choose to exercise them, but few professionals besmirch texts they have invested considerable effort in creating. Still some contributions are strategically mischievous, and some are strategically malign. Some are economically motivated, others produced through political or religious influence.

One common bit of mischief perpetrated by producers of encyclopedias is known as the Mountweazel, which describes a long tradition of including a false article in an encyclopedia in order to catch other encyclopedists who might pilfer the text wholesale.⁵² I have thus far found no evidence of this type of article in the Chambers, but it would not have been unlike him or his publishers to include such an item. This mischievous act is similar to common vandalism in Wikipedia involving the insertion of “misinformation” but differs significantly in its intent, which is driven by economic interests and meant to entrap unscrupulous competitors rather than purposefully mislead audiences.

A more insidious form of vandalism is censorship, which can be conveniently enacted during the production process, especially when the encyclopedist’s first view of a finished text that incorporates last-minute additions from public contributors is at the proof stage. The Encyclopédie’s fraught production process, undertaken during the French Revolution and its culture of religious and political censorship, saw one of the most famous instances of encyclopedic vandalism when the publisher, André le Breton, censored the text while the writers and editors were still working. Collison explained the ease of such censorship, given the production process:

When the galleys were returned to the printer the chief proof-reader, under secret instructions from [primary investor] le Breton, read through Page 92 →the text and marked “doubtful” passages for le Breton’s attention. These passages were any which in the proof-reader’s opinion laid the publisher open to action on the score of libel, blasphemy, treason, etc., . . . As Diderot never saw his manuscript again, nor the final text until it was printed, this process had been going on for some time before he noticed it, although in fact there were very few articles apart from his own and Jaucort’s which were subjected to this unofficial but no less efficient censorship.⁵³

Diderot was so outraged by this betrayal that he deserted the project, supposing that a high level of censorship had been applied throughout the texts when in fact the real impact of this unwelcome collaboration was not as extensive as he thought.⁵⁴ He was eventually persuaded to return and completed twenty-eight volumes. The Cyclopædia, being published well after the concluding years of the Restoration, enjoyed the benefits of being published in a significantly more stable era. It suffered no recorded instances of production-level vandalism. It also enjoys a material stability that is impossible for Wikipedia: the content of a fixed text can be altered only during production, in terms of adding or deleting figures, inserting misinformation or nonsense, and other text-level alterations. The text can certainly be censored and the material form broken or burned, but there is little potential to alter the words themselves after the printing and distribution processes.

A significant amount of work by both human and nonhuman Wikipedians is devoted to policing vandalism, which is one of the most deleterious differences between our two primary artifacts. The low interface barrier of wikis can be extraordinarily productive, but it also invites nongenerative and even destructive performances. Agency is not always productive or positive, Campbell reminds us: it is also “perverse, protean, and ambiguous” and “can be malign, divisive, and destructive.”⁵⁵ Vandalism is a commonplace in Wikipedia, an expected but not invited form of contribution that alters the text and demands a range of rhetorical response from editors. The result is a struggle between competing cultural ideals and discourses, between the dominant Wikipedian philosophy of constructive openness and rather different philosophies of what can be done with an open site that anyone can alter.

The Wikipedian community has from the beginning been adept at catching and reverting vandalism. Because of the vast amount of text that needs to be policed and the high incidence of vandalism, a number of nonhuman actors participate in this work. Bots automatically revert common offensive words, usually within the same minute the vandalism occurs. Editors also monitor a central, algorithmically generated watch list of suspected vandalism instances, and those who have set up individual watch lists for pages they work closely on are also notified. One of the earliest studies on vandalism in Wikipedia found that the Page 93 →majority of vandalism instances were repaired in an average of 2.8 minutes.⁵⁶ The 2007 study of Priedhorsky et al., which was conducted after several years of increased vandalism rates and which analyzed one of the largest corpuses of any Wikipedia study at that time, roughly confirmed the findings of Viégas et al. about repair, finding that 42 percent of incidents were repaired “essentially immediately (i.e. within one estimated view).” However, some vandalism remained unrepaired for longer periods: 11 percent of instances persisted beyond one hundred views, 0.75 percent beyond one thousand views, and 0.06 percent beyond ten thousand views. Many eyes correct many issues, but not necessarily all of them.

Vandalism is often discussed in the media as though it were a unified action, but studies demonstrate that it actually encompasses a broad range of performances.⁵⁷ Different types of vandalism shape the text in different ways, however temporarily: for example nonsense edits like “asdflkj asdf;lj” simply add clutter, while misinformation such as altering census figures by one digit deliberately subverts the goals of the encyclopedia. Full deletions of the text, while also subversive, are easily reverted with one click of the mouse, but offensive edits that are woven into the text can require closer editing to remove and potentially alienate readers who view the page version while it is posted.

Boredom is a common impetus for vandalism, along with frustration with page topics, as this instance demonstrates: Here, the original text describing the historical development of trigonometry was deleted and replaced with the declaration “Trigonometry is the work of terrorists and is frowned on by our society.”

This page was not randomly targeted for vandalism. Instead this performance is likely the work of a disgruntled student facing down a trig assignment during an all-nighter, given the edit time of 3:56 A.M. This vandalism pointedly replaces helpful (but not vital) information on the page with an expression of dislike of Page 94 →the topic while simultaneously referencing the recent international concern with terrorism. This sort of vandalism to pages on common curricular topics is remarkably routine during school hours, frequently leading Wikipedia administrators to temporarily block school IP addresses.

These typical forms of vandalism have little rhetorical appeal. The actors performing them often have little investment in or acknowledgement of the Wikipedian community, its policies, or its social norms. The appeal for these performances lies in the captivating potential of violating expectations for the encyclopedic genre, leaving gobbledygook or misinformation for later readers who may expect to find a rigorously curated text. In the project’s earlier years, vandalizing pages simply in order to demonstrate the high potential for vandalism or see how long facetious edits would stand was also a fairly popular impetus, particularly among researchers and pundits who were publicly commenting on the project through blogs or even more prominent media outlets. In 2004 Internet Studies researcher Alex Halavais made a number of experimental alterations and posted a typical announcement of them on his blog. “No matter which side of the debate you find yourself on, this sounds like an interesting experiment. So, I have made not one, but 13 changes to the Wikipedia site. I will leave them there for a bit (probably two weeks) to see how quickly they get cleaned up. I’ll report the results here, and repair any damage I’ve done after the period is complete. My hypothesis is that most of the errors will remain intact. Does that invalidate Wikipedia? Certainly not! If anything, the general correctness and extent of Wikipedia is a tribute to humankind. It suggests the Kropotkin may be right: that the “survival of the fittest” requires that the fittest cooperate. It means that there are very few Vandals like me who are interfering with its mission.”⁵⁸ This sort of vandalism served a far more public purpose: that of public argument for (in Halavais’s case) or against the openness or reliability of what was then a new and contested digital resource.

Other types of vandalism serve far more rhetorical purposes, particularly during kairotic moments that receive intense media attention. This attention frequently drives real-time composition of articles that simultaneously receive high readership traffic as the event unfolds. Members of the public have taken advantage of these instances to distribute political commentary to what is perhaps the widest audience they might have access to in their lifetime. During Texas senator Wendy Davis’s eleven-hour filibuster for abortion rights in the summer of 2013, supporters around the world offered their support through social media. More than 175,000 people watched the live stream of the proceedings, and “Wendy Davis” and “#standwithwendy” trended globally on Twitter that night. One supporter edited the information box on the Wikipedia article on Davis to list her occupation as “The LeBron James of filibustering.” After the Senate changed the time stamp on the vote to backdate it prior to midnight and therefore claim that the Page 95 →bill had passed, Republican senator Robert Lloyd Duncan’s Wikipedia page was edited to include “Warping Space-Time” as one of his professions. Both of these humorous edits were quickly removed, but not before they were screen-capped and widely distributed in a number of time lines of the evening’s events created by prominent digital media outlets such as Slate and BuzzFeed.⁵⁹

These edits both functioned as public support of Davis: the first labels her by comparison as someone who is both a champion and a most valuable player; the second publicizes Duncan’s complicity in fraudulently time-stamping a vote. Given Wikipedia’s well-established reputation for reversion of noncompliant edits or vandalism, these contributors likely knew that these changes would not stand for long and therefore would not be part of the longer-standing record of Wikipedia except as part of the page histories. However, owing to their good-natured humor, they were captured and amplified through more permanent public records of this newsworthy event. These instances of vandalism are meant as public commentary and encouragement for larger movements and events outside the realm of Wikipedia, and these well-intentioned vandals counted on kairotic high traffic and attention for effective distribution. They may not have hoped for wide circulation, but they took advantage of the interface affordances, their understanding of Wikipedian page conventions, and the moment in order to express affiliation and support. Regardless these contributions pushed against social expectations for the encyclopedic genre, Wikipedian policies, and the seriousness of the Wikipedian community itself, and they were quickly reverted.

The ontology of these encyclopedias is complex, contingent, and riddled with competing interests and destructive performances. Collectives are productive more often than not; otherwise actors would not exert the considerable energy required to negotiate collaboration. These collectives have been extraordinarily productive, creating some of the most extensive textual products, affiliations, and policy systems in this genre. Just as often, though, performances unfold in negative ways that sabotage curatorial work and snap the loose ties of a network, driving potential actors from the project or temporarily destroying sections of the text itself. It takes not only considerable human energy to maintain these texts and communities, but also a deep reliance on the capacities of nonhuman actors.