It is widely understood that negotiation and debate among editors is key to the Wikipedia’s system of governance. However, the role of rhetoric, the construction and application of argument, in this process is less well understood. In this paper, I attempt to move our understanding of the role of rhetoric on Wikipedia forward by proposing a taxonomy of rhetoric derived from Kenneth Burke’s dramatism. The proposed taxonomy will classify arguments first by the dominant ratio of elements from Burke’s pentad present in the argument (scene–act, act–agent, etc.), then by the vocabulary employed to make the argument (what Burke called the terministic screen), then by the scope of terms used. I then demonstrate how this taxonomy might be developed to describe the rhetoric of the Wikipedia community by employing it to classify a sample of arguments drawn from Wikipedia’s Articles for Deletion (AfD) process. Further development and application of the taxonomy proposed here could help us refine our understanding of how and why Wikipedia editors develop and deploy certain arguments, and which sorts of arguments have the most success in swaying Wikipedian debates.
Introduction — Wikipedia and rhetoric
Existing research: The hospitable text of Wikipedia
Building a taxonomy from dramatism: Pentad, screen, scope
Using a Burke–based taxonomy to describe the AfD process
Conclusions and next steps
In the decade since Wikipedia’s inception in 2001 the site, which dubs itself “the free encyclopedia anyone can edit,” has become both a heavily used information resource (Alexa ranks it as the world’s sixth most visted Web site) and a popular subject of scholarly inquiry. Scholars, such as Yochai Benkler (2006), have used Wikipedia as the premier model of innovative, volunteer–;based methods for producing information. In a similar vein, Jonathan Zittrain (2008) calls Wikipedia an example of the “generative” qualities of the Internet medium. Zittrain warns that these generative qualities, which he argues are key to the Internet’s potential to foster expressive freedom, are under threat as the open architecture of the Web is subsumed under walled–off platforms like Facebook and iTunes, and that Wikipedia plays a vital role in preserving the open architecture of the Web. In addition to these engagements with Wikipedia as an example of a larger and more sweeping trend in digital communications, a host of researchers have devoted considerable energy to the project of small–scale investigations of the actual workings of the Wikipedia community. Examples of these studies include investigations of: how Wikipedia is cited in a variety of contexts (Peoples, 2009; Shaw, 2008; Lih, 2004; Head, 2007; Schweitzer, 2008; Chandler–Olcott, 2009); the range of subject matter covered by Wikipedia (Schweitzer, 2008; Halavais and Lackaff, 2008); the accuracy of Wikipedia articles (Giles, 2005; Bragues, 2007; Chesney, 2006; Frančula, 2009; Nielsen, 2007; Mercer, 2007); Wikipedia’s error correction methods (Ciffolilli, 2003; Viegas, et al., 2004; Kittur, et al., 2007); the growth (or decline) of the Wikipedian editing community (Kittur, et al., 2007; Ortega, 2009); and, the motivations of editors (Okoli and Oh, 2007; Forte and Bruckman, 2008a; Masum and Zhang, 2004).
This essay will seek to add to this impressive body of work by contributing to our understanding of the role of rhetoric in the Wikipedia community. It does this by describing a method for classifying arguments made by Wikipedia editors based on the theory of “dramatism,” developed by the literary theorist Kenneth Burke, and demonstrating how this method can be applied to a small sample of arguments drawn from Wikipedia’s “Article for Deletion” (AfD) process. I believe that a rhetorical taxonomy build on Burke’s work can provide the basis for a meaningful content analysis of Wikipedia’s rhetoric. For example, as I describe in the next section, it might help us to better understand what sorts of arguments are more or less “welcome” within the Wikipedia community, and in turn to unravel how some topics and editors might be excluded from the project. The work done here is not sufficient to provide a complete or through understanding of the actual role of rhetorical techniques in the Wikipedia community, but should provide an important theoretical underpinning for understanding the rhetorical moves made by Wikipedia participants. In addition, I close with some practical next steps for performing research that will build on these theoretical ideas to establish a stronger empirical understanding of rhetoric in the Wikipedia community.
The existing research establishes that debate, argument, discussion and other methods of using language as a tool of negotiation and persuasion play an important role in Wikipedia. Joseph Reagle, in his 2010 ethnography of the site, Good faith collaboration: The culture of Wikipedia, writes that the actual encyclopedic content provided by Wikipedia, “at any moment in time, is simply a snapshot of the community’s continuing conversation.” In an examination of Wikipedia’s evolving governance methods, Andrea Forte and Amy Bruckman (2008b) write that Wikipedia editors’ actions are not dictated by a static set of policies, but rather, “the creation and refinement of policy is a complex social negotiation that often takes place across many communication channels and in which power, authority and reputation play decisive roles.”
In addition, Wikipedia itself acknowledges the important role of discourse in its process of producing and editing encyclopedic knowledge. One of Wikipedia’s “five pillars,” a collection of five “fundamental principles by which Wikipedia operates,” is the Civility Policy, which gives editors guidance in engaging with one another in discussion. The policy stresses the need for editors to act, “calmly and reasonably, even during heated debates,” (Wikipedia, 2012) demonstrating that Wikipedia is well aware of the propensity of editors to end up in vigorous contests of opinion with one another. Furthermore, Wikipedia provides editors with a wide variety of channels for communication, including “talk” pages associated with each Wikipedia article, a “meta–wiki” for discussing issues involving multiple Wikimedia foundation projects, and numerous mailing lists and Internet Relay Chat (IRC) channels.
However, despite the clear role of negotiation and communication within Wikipedia’s practices of production and governance, little work to date has directly examined the role of rhetoric, that is to say the techniques of persuasion, in Wikipedia. One notable exception to this is James Joseph Brown’s 2009 dissertation Hospitable texts, which attempts “to locate an emerging digital rhetoric,” in the actions of Wikipedia editors . Brown’s work centers around describing the rhetorical situation of Wikipedia, which he argues can best be described as that of a “hospitable text.” In a paragraph early in his thesis, Brown describes the structures that make Web–based texts, like Wikipedia, “hospitable”:
Many of the Web’s emerging digital infrastructures are built upon a code of hospitality. This hospitality welcomes digital writers without putting up many of the filters that have long controlled what does and does not get published. At the level of structure, texts like Wikipedia are hospitable. That is, they invite writers inside and call for participation. This structural hospitality is not necessarily about a choice made by a writer or central entity to invite others to edit the text. Rather, it is a predicament that each reader and writer is caught up in, and it is not entirely new. All texts are hospitable to a certain extent in that they invite interpretation. But Wikipedia and certain other Web texts extend an invitation to both readers and writers. 
Brown admits that “the contributors to [Wikipedia] can be very inhospitable in practice,” but maintains that, nonetheless Wikipedia remains structurally hospitable, as continues to invite participation from a broad spectrum of writers, and remains (with the exception of protected articles) open to be edited by anonymous participants with no formal ties to the community.
Brown makes several observations about the effects of this structural hospitality on Wikipedia’s rhetorical situation that I will build from in this essay. First, Brown emphasizes that hospitality, as an ideal, does not mean that Wikipedia is a space where anything and everything is welcomed in practice. Instead, he draws from Derrida’s On hospitality to explain that the transcendent ideal of perfect hospitality, in Derrida’s terms, “the Law” of hospitality, is never fully met by actual everyday practices, “the laws” of hospitality. Brown writes, “As open as this text is structurally, in practice it still falls short of absolute hospitality. Thus, the distinction between the openness of Wikipedia at the technological and textual level and the practices of Wikipedians is an important one. Such a distinction draws a line between the Law (structure, code) and the laws (practice).”  This essay will demonstrate how we can better understand Wikipedia’s laws of hospitality by developing a method for classifying arguments used on the site. A taxonomy of Wikipedia rhetoric will allow us to better understand what arguments are “welcome” or “not welcome” on Wikipedia.
Brown himself demonstrates these laws of hospitality in his examination of ethical arguments on Wikipedia. Brown’s discussion of how editors use (and misuse) ethical arguments centers around a close examination of the scandal surrounding “Essjay,” a high–profile Wikipedia editor who was discovered to have faked real world credentials (he fraudulently claimed to have earned a Ph.D.). Brown argues that Wikipedia’s rhetorical practices, while permitting Essjay to misuse his fraudulent identity, actually minimized the damage he could do by posing as an expert because they did not allow him to stop conversation with an appeal to his phony credentials. Brown writes, “Rather than allowing anyone to stop the conversation by invoking a situated ethos of expertise (a move that is not well–suited for an interdisciplinary conversation that cannot be grounded in any one set of rhetorical assumptions), Wikipedia attempts to remain in the realm of invented ethos by demanding citations.”  In this way, Brown demonstrates how Wikipedia both rejects Essjay’s fraudulent arguments (following the laws of hospitality) and avoids screening participants prior to allowing them to participate within the community (motioning toward the Law of hospitality). This essay builds on Brown’s observations by demonstrating how we can better understand how different types of ethical argument are deployed on Wikipedia, and how these different types of argument, in turn, shape Wikipedia’s community and content alike.
Finally, and perhaps most importantly, Brown’s work provides a model for the dynamics of the Wikipedia community that allows us to see clear parallels between the inner workings of this community and the dramatistic theory of Kenneth Burke. Brown, for his part, argues that Wikipedia’s hospitality means that the author/editors of Wikipedia often find themselves working within the space of the same text with others who posses wildly different agendas. These “collisions” between agendas, Brown argues, often become “collusions” in the larger task of writing the encyclopedia. This process of collusion through collision means that Wikipedia’s community cannot be reduced to any single, essential, identity or agenda. Instead, Wikipedia’s community is an amalgam of collisions. Brown writes, “Understanding community in this way allows us to see that collaboration in the name of a common goal always happens in the face of competing interests and colliding identities.” 
Brown’s understanding of the Wikipedia community as heterogeneous and based on “collisions” between disparate identities (as opposed to being based on a single, shared identity) mirrors Burke’s understanding of the ambiguities of “consubstantiation” and the value of “cooperative competition” in rhetorical practice. In his Rhetoric of motives, Burke argues that rhetoric is, in part, a technique for building shared identities and that a desire for shared substance with others, what Burke calls “consubstantiality,” helps to drive this process. However, he writes, this desire for consubstantiality is ambiguous and imperfect. For example, in the case of two agents, A and B, with shared interests, the two agents may arrive at a shared sense of purpose, and yet:
Here are ambiguities of substance. In being identified with B, A is “substantially one” with a person other than himself. Yet at the same time, he remains unique, an individual locus of motives. Thus he is both joined and separate, at once a distinct substance and consubstantial with another. 
Thus, Burke’s understanding of the ambiguities of identity mirrors Brown’s understanding of the ambiguous and heterogeneous identities of Wikipedia editors. Furthermore, Burke argues that this ambiguous and imperfect consubstantiality, while frustrating, has an important advantage over understanding shared identity as something more perfectly unifying. Just as Brown argues that Wikipedia benefits from colluding collisions, so too Burke argues that cooperative competition can lead to a more robust and diverse rhetorical environment. “Put several [...] voices together, with each voicing its own special assertion, let them act upon one another in co–operative competition, and you get a dialectic that, properly developed, can lead to views transcending the limitations of each.” 
Thus, Brown provides us with a point of entry both into understanding Wikipedia’s rhetorical situation, and to understanding why Burke might be a good choice for a theoretical framework for further examining rhetorical practices on Wikipedia. In the next section, I further develop the argument for using Burke as a theoretical resource.
At first glance, Kenneth Burke, a mid–twentieth century theorist of rhetoric and human behavior, may not seem to be an obvious choice for understanding Wikipedia’s twenty–first century communication practices. However, a closer examination of Burke’s thinking will show how it can reveal important contours within the practices that make up Wikipedia’s complicated “laws of hospitality.” First, however, I will briefly explain why I am attempting the complicated process of using Burke to build a taxonomy of Wikipedia arguments, rather than employing an existing taxonomy of arguments to understand Wikipedia.
Existing taxonomies of rhetoric can be divided roughly into two groups: descriptive taxonomies, and structural taxonomies. Descriptive taxonomies, such as Albert Hirschmann’s “rhetoric of reaction” taxonomy, divide arguments into categories based on common themes. For example, Hirshmann’s scheme divides arguments made by opponents of social reforms into three broad descriptive categories each based on a common “thesis.” Arguments categorized as employing the “perversity thesis” hold that a proposed reform will in fact have the opposite effect from that intended. Building a descriptive taxonomy of arguments used within a subset of Wikipedia discussions would be both possible and constructive, but would possess important limitations. A descriptive taxonomy would allow for classification of the vast confusion of arguments made by Wikipedia editors, and would permit useful comparisons to be made between different types of arguments made in similar situations, or between similar arguments made in disparate situations. Descriptive taxonomies have the advantage of being holistic, taking the broad structure of arguments into account, and being relatively easy to understand. However, building such a descriptive taxonomy of Wikipedia arguments would not be readily or easily extensible. Without a logical structure to guide the creation of new categories, a descriptive taxonomy would be limited to describing future arguments on Wikipedia in terms of past arguments. Thus descriptive taxonomy’s limited capacity for growth and change make them unfit tools for the task of analyzing Wikipedia because of the site’s fast rate of change and vast scope.
Structural taxonomies would seem to avoid the limitations of descriptive taxonomies, but they have failings of their own. Structural taxonomies, such as Walton’s argumentative schemes, seek to classify arguments based on the internal structure of the argument itself. For example, Walton classifies all arguments with the structure given below as “argument from sign.”
A is true in this situation.
B is generally indicated as true when its sign, A, is true, in this kind of situation.
Therefore, B is true in this situation. 
Unlike descriptive taxonomies, structural taxonomies exhibit an internal logic that makes them almost infinitely extensible and transportable. However, because they focus on the linguistic structure of the argument, rather than its meaning, structural taxonomies achieve their flexibility and power at the cost of becoming abstracted away from the actual experiences of the participants in the rhetorical processes that they describe.
Burke, in contrast, focuses on the practice of rhetoric in his theoretical work. Most significantly, unlike structural taxonomies, which classify arguments with a systematic precision rarely experienced by those who actually make them, Burke’s thought emphasizes that ambiguity is an unavoidable element in argument. In A grammar of motives, Burke argues that “not terms that avoid ambiguity, but terms that clearly reveal the strategic spots at which ambiguities necessarily arise,” provide the best vocabulary for describing rhetorical moves. In using Burke’s thought to provide a logical basis for my taxonomy, I hope to gain the extensibility and power of structural taxonomies, and therefore the ability to deal with Wikipedia’s size and pace, while retaining the ability to accurately reflect the ambiguities of everyday rhetorical actions within the Wikipedia community.
Furthermore, Burke’s thought allows us to understand rhetoric not as an already–complete structure to be mapped, but as an ongoing action. Burke’s choice of the metaphor of drama as a source for terminology to describe rhetoric makes this clear enough. Burke writes, in the introduction to A grammar of motives, “the titular word for our own method is ‘dramatism,’ since it invites one to consider the matter of motives in a perspective that, being developed from the analysis of drama, treats language and thought primarily as modes of action.”  Thus, choosing Burke as a source for the theoretical basis of my rhetorical taxonomy allows me to chart the rhetoric Wikipedians use without fixing their rhetorical acts for all time in an abstract structure. Instead, I hope to build a structure that will be amenable to both close readings of individual arguments made by individual Wikipedians and distant readings of the Wikipedia corpus as a whole. As I demonstrate in the next section, Burke’s notions of the dramatistic pentad, the representative anecdote, and scope provide us with the conceptual resources we need to build a content analysis methodology that will allow for exactly this sort of hybrid between distant and close readings.
To understand Burke’s thinking on rhetoric, we first must understand his focus on motives. For Burke, all rhetoric is a form of symbolic action, and thus his method focuses on analyzing how actions are described and performed in language. As Burke puts it, in the first line of his introduction to A grammar of motives, his primary goal is to provide the means to answer the question, “what is involved, when we say what people are doing and why they are doing it?”  To achieve this, Burke develops a vocabulary for talking about how people talk about action: his dramatistic pentad. This pentad is a series of five terms that Burke believes can be used to adequately describe all statements about human motives, and thus human action. Burke gives a straightforward description of each of the five terms:
In a rounded statement about motives, you must have some word that names the act (names what took place, in thought or deed), and another that names the scene (the background of the act, the situation in which it occurred); also, you must indicate what person or kind of person (agent) performed the act, what means or instruments he used (agency), and the purpose. 
Burke believes that the simplicity of these terms is what gives them their generative power, he writes that the pentad, “should provide us with a kind of simplicity that can be developed into a considerable complexity, and yet can be discovered beneath its elaborations.” 
The first method that Burke employs to develop complex relationships from his pentad is that of the ratio. Each of the pentad’s five terms can be combined with one other to create a total of ten relationships between the terms, which Burke dubs “ratios.” The ratios express what Burke calls “principles of determination,” that is to say they demonstrate how one element of the pentad influences another, and how these two elements together exercise influence over the remaining terms. Burke believes that all statements about motives, and thus all arguments, can be classified according to one of the ten ratios. For example, arguments that employ the “scene–act” ratio hold that acts have been determined by their appropriateness (or perhaps inappropriateness) to the scene that contains them. Burke provides Horatio’s warning to Hamlet against following his father’s ghost as a specific instance of this ratio. When Horatio tells his prince that he fears his pursuit may drive him to insanity because:
The very place puts toys of desperation,
Without more motive, into every brain
That looks so many fathoms to the sea
And hears it roar beneath
Burke writes that Hortatio is, in fact employing the scene act ratio, and “saying that the sheer natural surroundings might be enough to provide a man with a motive for an act as desperate and absolute as suicide.” 
My taxonomy uses the ratios Burke develops from his pentad as the basis for classifying arguments made within the Wikipedia community. Following Burke, I sort arguments into categories based on the elements of the pentad which feature most prominently in a given argument’s reasoning. This determination requires human judgment and close reading, but still results in a useful means for sorting arguments into meaningful categories. I do not expect to find arguments to fit all ten of the ratios. Indeed, Burke himself focused on only a handful of the ratios, and believed that some were much more commonly deployed than others. My model taxonomy, developed in the next section, focuses on two of the ratios Burke himself developed in detail in A grammar of motives (the scene–act and act–agent ratios) and another discussed by David Blakesley (2002) in his primer The elements of dramatism (the act–agency ratio).
The ratios, however, do not provide enough detail to fully classify the many arguments we find in the Wikipedia community (or indeed anywhere else). To further place arguments within my taxonomy, I turn to Burke’s notion of the terministic screen. For Burke, the vocabulary used to discuss a particular action, and in particular the terms chosen to describe each of the five elements of the pentad, deeply effect how a particular action is understood by those describing it. He writes, “men seek for vocabularies that will be faithful reflections of reality. To this end the must develop vocabularies that are selections of reality.”  He dubs the selection of a particular vocabularies, and thus of a particular slice of a large and complex reality, a “terministic screen.”
In following Burke, I use vocabulary as a further method for sorting arguments into more specific categories than those permitted by the ratios derived from the pentad. In effect, I utilize the terministic screen built by Wikipedia editors  as they select vocabulary to further classify the arguments made by these editors. For example, in the sample taxonomy I develop in the next section, I divide arguments belonging to the “scene–act” ratio category into two subcategories. I dub these two subcategories “cruft arguments” and “legality arguments,” based on the language used by Wikipedia editors in each respective case.In the case of “cruft arguments,” editors argue that a particular act of article creation does not meet the needs of the scene of Wikipedia because the content of the article is extraneous or unnecessary. They often use the programmer’s term “cruft” (for needlessly complicated code) to describe such arguments, and it is this vocabulary choice that I have chosen to name this subcategory after. In the case of the other subcategory, “legality arguments,” Wikipedia editors use language that indicates they are concerned that a particular act of article creation could provoke a legal challenge to Wikipedia within the larger, global scene in which the project itself is embedded.
As the example above suggests, the taxonomic classification of arguments can be improved by making one important further distinction between them. Namely, we can distinguish between arguments in terms of the scope of the terms they employ. Burke writes, “one may place the object of one’s definition in contexts of varying scope. And out remarks on the scene–act ratio, for instance, suggest that the choice of circumference for the scene in terms of which a given act is to be located will have a corresponding effect upon the interpretation of the act itself.”  I follow Burke’s thinking here and distinguish between arguments on the basis of the scope of terms employed, and especially the scope of the terms employed to describe the scene, where appropriate. For example, the “cruft” classification and be further subdivided into articles challenged on the basis of Wikipedia’s own standards of importance (notably the Reliable Sources policy), in which case the focus is on the immediate scene of Wikipedia itself, and articles challenged with the claim that their subjects are unimportant regardless of sources, in which case the focus is on a global scene.
To summarize, I propose a taxonomy of rhetoric derived from Kenneth Burke’s dramatism. The proposed taxonomy will classify arguments first by the dominant ratio of elements from Burke’s pentad present in the argument (scene–act, act–agent, etc.), then by the vocabulary employed to make the argument (what Burke called the terministic screen), then by the scope of terms used. In the next section, I demonstrate how this taxonomy might be developed to describe the rhetoric of the Wikipedia community by employing it to classify a sample of arguments drawn from Wikipedia’s Articles for Deletion (AfD) process.
Wikipedia’s AfD process is one of several mechanisms Wikipedia employs (along with Speedy Deletion, and the Proposed Deletion template) to remove unwanted articles from the encyclopedia. Of all the deletion processes that Wikipedia employs , AfD is the most deliberate. Any Wikipedia editor may open the AfD process on any article he or she believes should be deleted. Once an article has been thus “nominated” editors engage in several days of open discussion about the article’s respective merits and flaws. During this debate period, editors post comments voicing their support or opposition to deleting the article, and explaining the reasoning behind their support or opposition. Editors often respond to comments left by others with counter–arguments. At the close of the debate period, a self–selected Wikipedia editor with administrative privileges (usually referred to as an “admin”) reviews the discussion, and decides if consensus for either deleting or retaining the article has been reached. Wikipedia policy stresses that administrators should not simply be counting how many editors expressed support or opposition for the article, but should be judging consensus on the basis of the substance of the arguments made. In addition, the AfD process is often the scene of heated debate, as editors either vigorously attack articles they feel damage the reputation of Wikipedia, or support articles on topics they feel deserve more visibility with equal energy. All of this means that the AfD process is a rich source of data on arguments made by Wikipedia editors.
This sample taxonomy is based on a sample of 30 AfD pages drawn from the over 2000 such pages created during the month of April 2011. The 30 pages selected were chosen from among the most active discussions during this period, as defined by the editing activity on each discussion. The selection of highly active discussions was based on the assumption that these discussions would provided the richest information about arguments employed in AfD discussions. In addition, AfD pages were selected for study on the basis of whether the article being debated was ultimately deleted or retained by Wikipedia editors. Fifteen AfD discussions that resulted in the article under discussion being deleted were selected, and 15 AfD discussions that resulted in the article under discussion being retained were selected. In both cases, the AfD discussions selected were those with the largest number of total edits in their respective categories. The AfD discussions’ final status and editing activity levels were determined by using a python script to process an archive of Wikipedia editing data in xml format obtained via Wikipedia’s Special:Export function, which permits users to download a local copy of a given Wikipedia page, complete with that page’s editing history and other statistics.
A close reading of each of these discussions was performed to determine what sorts of arguments were employed by editors, and especially what sorts of arguments were used by editors nominating particular pages for deletion. The arguments made by the nominating editor tended to influence the direction of the rest of the discussion, as they were highly visible and other editors felt the need to respond to them, either in support or opposition. I found that these arguments tended to fit under the categories of the scene/act or act/agent rations, and could be further subdivided into several meaningful subcategories, detailed below. In addition, a set of important arguments was identified that engaged with the ethos of participants in the AfD discussion. These arguments, which I have classified under the act/agency ratio, attempted to dismiss the arguments made by some editors due to the method by which the editors’ input was solicited, or the editing activity of the challenged editor. These arguments stood out both because of their relatively frequent use, and because Wikipedia’s community has developed specific, software based tools designed to assist in making these arguments.
The outline below gives an overview of the complete sample taxonomy:
1. Source Problem (local scope)
2. Subject not important (global scope)
2. Illegal (global scope)
1. BLP (Biography of Living Persons)
2. Copyvio (Copyright violation)
1. COI (Conflict of Interest)/Vanity (global scope)
2. POV(Point of View) Fork (local scope)
III. Act/Agency (local scope)
1. SPA (Single Purpose Account)
The sections below give a brief description of each category in the sample taxonomy.
For Burke, the essential feature of Scene/Act arguments is that they focus on how “the scene contains the act.” That is to say, they focus on how a particular setting for action makes the action appropriate (or perhaps inappropriate). In the case of Wikipedia’s AfD discussions, I found that editors often made scene/act arguments, effectively arguing that the act of article creation either was, or was not, appropriate to the scene. In fact, it was rare to find an AfD discussion without at least some arguments of this kind. These could be subdivided into discussions that focused on how the article in question was extraneous or unnecessary, which Wikipedia editors often identified with the programmer’s term “cruft” (which historically referred to unnecessary or poorly designed code), and discussions that focused on how the article in question might violate the law (usually copyright or libel law). In some cases, editors focused on the local scene of Wikipedia itself, and argued that the act of article creation was inconsistent with specific Wikipedia policies. In other cases, editors focused on the larger, global scene, arguing that articles either violated national laws, or lacked global importance (regardless of their fulfillment of local Wikipedia policies).
In cases where Wikipedia editors focused on the local, Wikipedia policy–based arguments against cruft, they tended to make arguments tied to Wikipedia’s notability policy, which holds that topics must receive, “significant coverage in reliable sources that are independent of the subject” , in order to be the subject of a Wikipedia article. In some instances, editors challenged articles on the grounds that the number of sources documenting a subject was insufficient, as when an editor in favor of deleting the article on Arthur S. Herman wrote that there were, “no results from Google Books or Google News archive, only 9 from regular Google,” on the topic. It was not unusual for editors to use Google search tools to investigate the sources available on the topic (indeed, the current version of the AfD template automatically includes links to relevant Google searches) and to explicitly cite the results of their searches in arguments. In other instances, editors challenged articles on the grounds that the sources supporting the topic were in some way unreliable. For example, the editor who nominated an article describing “Lousiana Baptist Institute and Seminary” wrote that the article had “zero non–trival sources.”
However, in other cases, Wikipedia editors focused not on whether or not the article rose to the local standards of the notability policy, but to a larger, global ideal of importance. For example, an editor arguing in favor of deleting an article documenting an incident in which two aircraft had a low–speed collision on an airport runway, admitted that this incident had received coverage in reliable sources, but then wrote, “Two airplanes bumped into each other. Contrary to what the popular news media would like you to think, this is not an unusual occurence.” Examples like these demonstrate the ambiguity inherent in the notability policy, and show how Wikipedia editors can slip between the local and global scenes of acts on Wikipedia. This slippage opens up important questions. When, if ever, do Wikipedia editors find arguments that a given topic lacks importance regardless of the coverage it has received in global media convincing? If editors do find this line of argument convincing, it would seem to contradict the spirit of the NPOV, which asks Wikipedia editors not to apply their own judgment in this way. On the other hand, if Wikipedia editors keep to the NPOV and rarely, if ever, make judgments about relevance independent of coverage in media sources, why does this line of argument appear in AfD discussions? Further investigation of these sorts of questions could reveal important information how the laws of hospitality operate on Wikipedia. In particular, we might discover more about how consistently the letter and spirit of the NPOV, which asks Wikipedia to be “hospitable” to all significant points of view, is followed.
A second class of scene/act arguments concentrates on how the act of creating an article may be inappropriate to the larger, global scene of the international legal environment Wikipedia operates in. These arguments tend to emphasize the potential legal threats to the Wikipedia project posed by the article nominated for deletion. For example, an editor arguing in favor of deleting an article entitled “List of descendants of Nazi officials” calls the list, “potentially libelous” and another dubs it “a BLP [for ‘Biography of Living Persons’,” the Wikipedia policy put in place to deal with the potential legal and ethical challenges inherent to writing biographical content about living people implemented after the famous Seigenthaler incident] minefield.” Most articles in the sample taxonomy that included arguments involving potential legal problems connected to Wikipedia content focused on these sorts of BLP issues, however some also mentioned the possibility that Wikipedia content should be deleted because of violations of copyright law (copyvio). Further investigation of this class of arguments could provide better insights into how Wikipedia editors understand the legal environment in which the project operates, and how this legal environment — modulated by the understanding of editors — shapes the content of the site itself.
For Burke, an act/agent argument focuses on how an act flows from some quality of the agent who performs it. For example, he writes that ascribing the success of the Russian resistance to German invasion to the character of “the Russian people” would be an act/agent argument. I found two major categories of act/agent arguments in the AfD discussion I studied to develop the sample taxonomy. Both focused on the way in which particular editors (agents) were motivated by their character or desires to perform the act of article creation, and argued that this expression was inappropriate to the purpose of Wikipedia.
In some cases, the editors responsible for creating a particular article were accused of acting out of a vane desire for self–promotion, or of attempting to promote a subject they were connected to in violation of Wikipedia’s Conflict of Interest (COI) Policy. For example, an editor arguing for the deletion of an article describing an obscure scientific theory calling itself, “Evolutionary Relativity,” wrote that the article’s “references are mostly of one author, presumably the author of this article (or fan).” In a similar case, an academic discovered writing a an article describing himself was accused of “vanity publishing.” In other instances, vanity and cruft arguments were combined, with editors arguing that a poorly sourced article would serve to nurture an undeserved sense of importance in an obscure or insignificant person or subject. In some cases, the stories Wikipedia editors told describing these imagined subjects were quite developed, and quite telling, as when an editor arguing in favor of deleting an article on “Interactions Between Micronations” wrote:
A person is lets say a 50 year old, part–time teaching assistant and lives with his parents, then sister. Behind the computer he imagines that he is the king, grand–duke or lord–emperor of some island or sandbank, designs a flag, some coins, stamps etc. He even has a t–shirt with the name of “his kingdom”. A newspaper writes a human interest story about this person and his hobby. Next the “micronation” becomes “notable” and gets an article on Wikipedia. Now this person has a cup of coffee at McDonald’s with a fellow hobbyist.
In writing this story, the editor shows us that he or she has a well–developed mental picture of the purported “hobbyist” who might try to pervert the Wikipedia project with information about a pet project. By preventing the sort of small, sad, irrelevant person who might be a “teaching assistants” who “lives with his parents, then sister” from using Wikipedia for self–promotion, the editor seems to say, the community prevents Wikipedia itself from becoming small, sad, and irrelevant. It is interesting to note that the picture this editor paints of the sort of person who must be prevented from perverting Wikipedia in many ways parallels to the image of Wikipedia editors often promoted by critics of Wikipedia itself . Further investigation of the role of these sorts of “vanity” arguments on Wikipedia might help us to more fully understand the mechanisms by which the Wikipedia community policies itself and builds a self–image.
In other cases, the editors accused the creators of articles of acting out of a desire to pursue a particular point of view within Wikipedia, to the exclusion of opposing points of view. In contrast to the vanity argument, which focused on maintaining the prestige of Wikipedia within the global scene, these arguments focused on the need to respect Wikipedia’s local policies, including the all important Neutral Point of View (NPOV) policy, and the attendant policy prohibiting the creation of so–called POVFORK articles, so named for being devoted to a particular Point of View (POV) and a “fork” or split of an existing article. For example, editors in favor of the deletion of an article entitled “Apartheid in Saudi Arabia,” argued that this article represented an illegitimate attempt to pursue a POVFORK designed to discredit other articles documenting “Israel and the Apartheid Analogy;” a topic editors agreed was controversial but which nonetheless was supported by third–party sources. The editor responsible for nominating the “Apartheid in Saudi Arabia” article for deletion wrote, “What we have here is, again, a WP:POVFORK created by crafting disparate, off–mentions of human rights issues in Saudi Arabia and trying to frame them with the “apartheid” label with the mistaken notation that tis will birng about some sort of balancing NPOV with the Israel article.” POVFORK arguments often seem to arise in association with contentious topics that have attracted self–selected cliques of Wikipedia editors committed to one side or another of a real–world conflict involved in the editing process. In another example from the April 2011 AfD debates analyzed for this sample taxonomy, editors argued for the deletion of an article on “Race and Crime,” for these reasons. Further analysis of the use of POVFORK arguments in the AfD process, and especially their relative rates of success or failure, might provide important insights into the relative power of different groups of partisans within the Wikipedia community.
The act/agency ratio is described by Blakesley as, “a ratio that comes to the fore when we consider the meaning of phrases like ‘the ends justify the means’ or ‘by any means necessary’.”  In the case of my sample taxonomy, act/agency arguments appeared when editors accused other editors of employing the technological means of Wikipedia itself, it’s pseudonymous and “hospitable” technological structure, in ways they found unacceptable. In some cases, these same technological means were recruited by editors seeking to prevent these offenses, in the form of “bots,” or automated procedures, employed to mark potentially illegitimate editors. Unlike the classes of argument discussed so far, which focused on challenging the appropriateness of a particular article, these arguments challenged the participation of a particular editor or class of editors.
In come cases, the participation of editors was challenged on the grounds that these editors were “Single Purpose Accounts” (SPAs), that is to say, editors that had created accounts on Wikipedia for the sole purpose of participating in a deletion debating or editing a particular article or class of articles. In these cases, an automated process marks accounts with “few or no edits outside this subject area,” when then participate in debate. SPA editors are seen by other Wikipedia editors as being suspect of pursuing a particular Point of View to the detriment of the larger community. The same technological means that allows these editors to easily create pseudonymous identities to participate in the Wikipedia community can then be employed to automatically create an ethical argument challenging the legitimacy of this participation. Further investigation of this process could reveal important insights into how ethical arguments based on participation and prestige shape the diversity of article content within Wikipedia.
In a closely related, but distinct set of cases, editors expressed concern that other editors had engaged in canvassing, attempting to recruit involvement in the AfD discussion either via Wikipedia’s own discussion tools or via off–wiki channels, presumably with the intent of attracting participation by supportive editors. Canvassing was often seen by Wikipedia editors as an illegitimate means of attracting involvement in a discussion, since Wikipedia’s large editing community and varied discussion spaces mean that the outcome of a debate may be determined by “who shows up to the discussion that day.” Again, this is a function of the technological means (agency) behind Wikipedia creating opportunities for acts abuse that editors believe they must guard against. In some instances where canvassing was suspected, the “Not a Ballot” template was affixed to the discussion page. This template warns new editors that, “If you came here because someone asked you to, or you read a message on another website, please note that this is not a majority vote, but instead a discussion among Wikipedia contributors.” This message serves to both educate newcomers in the ways of Wikipedia, and to perhaps scare them off. While the template goes on to reassure new editors that “you are invited to participate and your opinion is welcome,” the use of this template, in conjunction with arguments framing canvassing as unethical, serves to cast a shadow over the legitimacy of these users contributions. While Wikipedia’s format gives editors good reasons to guard against these sorts of intrusions, further study of the use of this class of argument will help us to better understand how new users are judged on Wikipedia. Most importantly, a better understanding of when and where canvassing arguments are used, and when and where they are successful, will help to better reveal if these arguments are applied equally to different user populations, or if they may have a discriminatory effect.
This essay has established the grounds for using Kenneth Burke’s dramatistic theory as the basis for building a rhetorical taxonomy for Wikipedia. Using Burke as the theoretical basis for this taxonomy allows me to build an understanding of the arguments used on Wikipedia that focuses on how these arguments are used as tools for symbolic action, and that remains close to the meanings expressed by actual users when they make arguments. Already, the sample taxonomy developed from the April 2011 AfD discussions investigated for this essay, while incomplete in scope, demonstrates some of the strategic points of ambiguity engaged in by Wikipedia editors in these important debates. Issues of notability, editor legitimacy, and editor motivation are contentious and ambiguous issues that often attract arguments from editors that frame Wikipedia’s purpose and scope in particular ways. Further investigation of these arguments could lead to a better understanding of Wikipedia’s laws of hospitality, shedding light on who and what is welcome within the Wikipedia community.
Furthermore, my initial investigation of a small sample of Wikipedia’s AfD process suggests several specific lines of questioning that might be good starting points for larger–scale, more rigorous studies of the Wikipedia community. First, there is the question of when, and if, Wikipedia editors employ arguments set in a global scene, rather than the local scene of Wikipedia itself? When are Wikipedia editors convinced by appeals to larger concerns, and when are they strictly faithful to policies of the project? Second, when are arguments based on the character or motivation of editors themselves employed to govern editing activity, and what effects do these arguments have on the overall content of Wikipedia? Are editors working on particular topics, expressing a particular point of view, or performing a particular identity more or less vulnerable to “vanity” or “conflict of interest” arguments?
As a next step towards investigating these questions I propose that a larger scale content analysis of Wikipedia’s AfD discussions be undertaken, with the sample taxonomy developed in this essay used as a starting point. This larger content analysis would have several goals: it would discover and classify forms of argument employed in AfD discussion space but not described in the sample taxonomy, it would allow for a larger analysis of how arguments frame debates in the AfD space, and it would create a larger corpus of classified discussions for use in further research into rhetorical practices on Wikipedia. Such an expanded corpus of classified arguments would permit researchers to move between statistical analysis of Wikipedia, so called “distant reading,” and qualitative analysis of individual arguments or argument categories, in the more traditional humanist mode of “close reading.” Bridging this gap is an important, perhaps necessary, step towards building a more complete understanding of Wikipedia.
About the author
Andrew Famiglietti is a Visiting Assistant Professor in the Emerging Media and Communication program at the University of Texas at Dallas. He was previously a Brittain Fellow at the Georgia Institute of Technology.
E–mail: afamiglietti [at] gmail [dot] com
1. Brown, 2009, p. v.
2. Brown, 2009, p. 2.
3. Brown, 2009, p. 16.
4. Brown, 2009, p. 98.
5. Brown, 2009, p. 151.
6. Burke, 1969, p. 20.
7. Burke, 1951, p. 63.
8. Walton, 1995, p. 49.
9. Burke, 1945, p. xxii.
10. Burke, 1945, p. i.
12. Burke, 1945, p. xvi.
13. Burke, 1945, p. 6.
14. Burke, 1945, p. 59.
15. “Editor” is the generic term most often used by members of the Wikipedia community to describe any contributor to the project, regardless of the exact nature of the contribution.
16. Burke, 1945, p. 77.
17. Wikipedia also employs a “Speedy Deletion” process that allows obvious acts of vandalism, copyright violation, and other unambiguously inappropriate articles to be deleted quickly without extensive debate, and a “Proposed Deletion” process that allows articles to be flagged as inappropriate and deleted, if no one objects.
18. That is, the sources documenting the event, person, object, etc. described in the article must not be involved with the thing being described. For example, a company’s own press releases would not serve to establish that company’s notability.
19. For example, compare to Andrew Orlowski’s portrayal of the Wikipedia community: “a picture of the body behind the ‘Hive Mind’ of ‘collective intelligence’ begins to take shape. He’s 14, he’s got acne he’s got a lot of problems with authority ... and he’s got an encyclopedia on dar interweb [sic]” (Orlowski quoted in Brown, p. 141).
20. Blakesley, 2002, p. 118.
Yochai Benkler, 2006. The wealth of networks: How social production transforms markets and freedom. New Haven, Conn.: Yale University Press.
David Blakesley, 2002. The elements of dramatism. New York: Longman.
George Bragues, 2007. “Wiki–philosophizing in a marketplace of ideas: Evaluating Wikipedia’s entries on seven great minds,” at http://dx.doi.org/10.2139/ssrn.978177, accessed 20 August 2012.
James Joseph Brown, 2009. “Hospitable texts,” Ph.D. dissertation, English, concentration in Digital Literacies and Literatures. Austin: University of Texas at Austin.
Kenneth Burke, 1969. A rhetoric of motives. Berkeley: University of California Press.
Kenneth Burke, 1951. “Rhetoric – Old and new,” Journal of General Education, volume 5, number 3 (1 April), pp. 202–209.
Kenneth Burke, 1945. A grammar of motives. New York: Prentice–Hall.
Kelly Chandler–Olcott, 2009. “A tale of two tasks: Editing in the era of digital literacies,” Journal of Adolescent & Adult Literacy, volume 53, number 1, pp. 71–74.http://dx.doi.org/10.1598/JAAL.53.1.7
Thomas Chesney, 2006. “An empirical examination of Wikipedia’s credibility,” First Monday, volume 11, number 11, at http://firstmonday.org/article/view/1413/1331, accessed 20 August 2012.
Andrea Ciffolilli, 2003. “Phantom authority, self–selective recruitment and retention of members in virtual communities: The case of Wikipedia,” First Monday, volume 8, number 12, at http://firstmonday.org/article/view/1108/1028, accessed 20 August 2012.
Andrea Forte and Amy Bruckman, 2008a. “Why do people write for Wikipedia? Incentives to contribute to open–content publishing,” Proceedings of 41st Annual Hawaii International Conference on System Sciences (HICSS); version at http://jellis.org/work/group2005/papers/forteBruckmanIncentivesGroup.pdf, accessed 20 August 2012.
Andrea Forte and Amy Bruckman, 2008b. “Scaling consensus: Increasing decentralization in Wikipedia governance,” HICSS ’08: Proceedings of the Proceedings of the 41st Annual Hawaii International Conference on System Sciences, pp. 157–166; version at http://dlc.dlib.indiana.edu/dlc/bitstream/handle/10535/5638/ForteBruckmanScalingConsensus.pdf, accessed 20 August 2012.
Nedjeljko Frančula, 2009. “Map projections in Wikipedia,” Kartografija i geoinformacije (Cartography and Geoinformation), volume 8, number 12, pp. 120–121, at http://www.kartografija.hr/kig, accessed 20 August 2012.
Jim Giles, 2005. “Internet encyclopaedias go head to head,” Nature, volume 438, number 7070 (15 December), pp. 900–901, and at http://www.nature.com/nature/journal/v438/n7070/full/438900a.html, accessed 20 August 2012.
Alex Halavais and Derek Lackaff, 2008. “An analysis of topical coverage of Wikipedia.” Journal of Computer–Mediated Communication, volume 13, number 2, pp. 429–440.http://dx.doi.org/10.1111/j.1083-6101.2008.00403.x
Alison J. Head, 2007. “Beyond Google: How do students conduct academic research?” First Monday, volume 12, number 8, at http://firstmonday.org/article/view/1998/1873, accessed 20 August 2012.
Aniket Kittur, Bongwon Suh, Bryan A. Pendleton, and Ed H. Chi, 2007. “He says, she says: Conflict and coordination in Wikipedia,” CHI ’07: Proceedings of the SIGCHI Conference on Human Factors in Computing Systems, pp. 453–462.
Andrew Lih, 2004. “Wikipedia as participatory journalism: Reliable sources? Metrics for evaluating collaborative media as a news resource,” Proceedings of the Fifth International Symposium on Online Journalism, pp. 16–17.
Hassan Masum and Yi–Cheng Zhang, 2004. “Manifesto for the reputation society,” First Monday, volume 9, number 7, at http://firstmonday.org/article/view/1158/1078, accessed 20 August 2012.
Jean Mercer, 2007. “Wikipedia and ‘open source’ mental health information,” Scientific Review of Mental Health Practice, volume 5, number 1, pp. 88–92.
Finn Årup Nielsen, 2007. “Scientific citations in Wikipedia,” First Monday, volume 12, number 8, at http://firstmonday.org/article/view/1997/1872, accessed 20 August 2012.
Chitu Okoli and Wonseok Oh, 2007. “Investigating recognition–based performance in an open content community: A social capital perspective,” Information and Management, volume 44, number 3, pp. 240–252.http://dx.doi.org/10.1016/j.im.2006.12.007
José Felipe Ortega, 2009. “Wikipedia: A quantitative analysis,” Ph.D. thesis, Universidad Rey Juan Carlos; see http://www.brzezinski.zdnet.pl/michal-brzezinski/inspirations/jose-felipe-ortega-soto-wikipedia-a-quantitative-analysis-2009, accessed 20 August 2012.
Lee F. Peoples, 2009. “The citation of Wikipedia in judicial opinions,” Yale Journal of Law and Technology, volume 12, article 1, at http://yjolt.org/, accessed 20 August 2012.
Joseph Michael Reagle, 2010. Good faith collaboration: The culture of Wikipedia. Cambridge, Mass.: MIT Press.
N.J. Schweitzer, 2008. “Wikipedia and psychology: Coverage of concepts and its use by undergraduate students,” Teaching of Psychology, volume 35, number 2, pp. 81–85.http://dx.doi.org/10.1080/00986280802004594
Donna Shaw, 2008. “Citing Wikipedia,” American Journalism Review, volume 30, number 1, p. 43, and at http://www.ajr.org/article.asp?id=4462, accessed 20 August 2012.
Fernanda B. Viégas, Martin Wattenberg, and Kushal Dave, 2004. “Studying cooperation and conflict between authors with history flow visualizations,” CHI ’04: Proceedings of the SIGCHI Conference on Human Factors in Computing Systems, pp. 575–582.
Douglas N. Walton, 1995. Argumentation schemes for presumptive reasoning. Mahwah, N.J.: Lawrence Erlbaum Associates.
Wikipedia, 2012. “Wikipedia:Civility,” at http://en.wikipedia.org/wiki/Wikipedia:Civility, accessed 20 August 2012.
Jonathan Zittrain, 2008. The future of the Internet and how to stop it. New Haven, Conn.: Yale University Press.
Received 29 May 2012; revised 29 June 2012; accepted 17 August 2012.
This work is licensed under a Creative Commons Attribution–NonCommercial–ShareAlike 3.0 Unported License.
The pentad of cruft: A taxonomy of rhetoric used by Wikipedia editors based on the dramatism of Kenneth Burke
by Andrew Famiglietti
First Monday, Volume 17, Number 9 - 3 September 2012
A Great Cities Initiative of the University of Illinois at Chicago University Library.
© First Monday, 1995-2016. ISSN 1396-0466.