a g o n i s t i c s:   a   l a n g u a g e   g a m e

w a r r e n   s a c k
wsack @ ucsc.edu

Abstract: The images and actions used as metaphors by Chantal Mouffe and other theorists of "agonistic democracy" can be instantiated as interactive, graphical objects and dynamics. This "literal" instantiation will then be a computer game that can played by posting messages to a public, online discussion forum.

\Ag`o*nis"tics\, n. The science of athletic combats, or contests in public games. Webster's 1913 Dictionary



Argument is war. In their book Metaphors We Live By, George Lakoff and Mark Johnson explain how this is metaphorically true. The language we use to talk about arguments is a language of war. We "attack" our opponents positions and "defend" our own. We "shoot down" opposing arguments. We say that claims are "defensible" or "indefensible." We talk of "winning" and "losing" arguments. In arguing we have "tactics" and "strategies." We are "on target" or "off target" in our criticisms. We "gain ground" or "lose ground." In fact, it is not simply that we talk about arguments like this, this is what we do. Lakoff and Johnson ask us to consider a culture in which arguments are not conceptualized as verbal warfare, but as collaborative dances: participants are not opponents but partners and each counter-move is a balanced, graceful response. That would be a very different world.

Of course the latter is not an alien idea. Philosophers have long distinguished the constructive, cooperative art of conversation (dialectics) from verbal combat (rhetoric). However, the problem has often been that -- when the cool reason of conversation comes in contact with the heated emotion of argumentation -- rhetoric melts dialectic and we get a shouting match rather than a reasoned debate. What can be done?

There is an argument about arguments and it has at least two sides. On one side, the advice given is of a moral quality: To allow reason to prevail over rage, calm everyone down. Make everyone follow the rules of calm and reasonable conversation and disallow the shouts and unruly outbursts of the arguing parties. The other side is neither moral nor immoral but opportunistic. This side is usually the one politicians listen to when they are running for office or ruling a state. The other side starts with the assumption that any verbal interaction will eventually become a shouting match so the best preparation is voice training and acting lessons, so that -- when the transition to shouting is at hand -- one can shout loud enough to make one's emotional appeal. The former is the utopian, Enlightenment ideal of reasoned debate, rational politics, democracy and verbal diplomacy; the latter is our world, the world of image, charisma, negative advertising, power politics, and war.

But, if we want deliberative debate, democracy and diplomacy, how do we get from here to there? Political philosophers have been arguing about arguing for a long time. Even though the most of this territory is occupied by the two sides described above, a third "camp" is emerging. (Hmm. There's that metaphor again!) The third camp tries to break up the fight between the moral conversationalists and the political rhetoricians by attempting to get everyone off the battlefield and to reconsider the shape and forms of the field of engagement. Lakoff and Johnson do this by making us examine the language we use to describe what we are doing when we argue. Political theorists like Chantal Mouffe provide us with alternatives by pointing out that -- even if argument is war -- war is just one form (although a deadly form) of contest between adversaries. Mouffe's alternative to a utopic, moral, deliberative democracy is -- what she calls -- an "agonistic pluralism" where agon is understood as the ancient Greek term denoting "A public celebration of games; a contest for the prize at those games; or, a verbal contest or dispute between two characters in a Greek play" (OED).

Political theorists, like Mouffe, interested in the democratic potential of agonistic contests, oftentimes recast deliberative discussion as a language game -- in the sense invented by Ludwig Wittgenstein. Moreover, this reimagining of politics leans heavily on Friedrich Nietzsche's understanding of agonistics and ancient Greek philosophy. A close look at the writings of this set of political theorists (which must also include Jean-Francois Lyotard, Michel Foucault, Gilles Deleuze, and Bruno Latour) rewards one with the following insight: just as Lakoff and Johnson show how everyday thinking about arguments draws on a set of metaphorical images and actions, so do these theorists assume a different set of metaphorical images and actions to describe verbal contests -- specifically, game like images and actions. Neither are these images and actions the moral frameworks of, for example, Jurgen Habermas and other moralists hoping for perfect conditions for communicative interaction. Nor, are these images and actions the violent ones implied by the commonsense metaphor "argument is war."

What then are these images and actions? Two sorts of evidence can be gathered from a close reading of these theorists. One sort of evidence is articulated in the form of broad outlines or "sketches" for envisioning such a game. Chantal Mouffe provides an example of such a "sketch" in her article entitled "Deliberative Democracy or Agonistic Pluralism?": "pluralist politics should be envisaged as a 'mixed-game', i.e., in part collaborative and in part conflictual and not as a wholly co-operative game as most liberal pluralists would have it." More specific, detailed, "diagrammatic" evidence comes from theorists who provide us with, what Gilles Deleuze calls, "thought images." One such influential thought images is that coined by Deleuze and Guattari to describe non-hierarchical forms of knowledge and power; i.e., the rhizome. As demonstrated by online forums, like rhizome.org, such a thought image can influence an extensive information architecture. However, even more substantial than these verbal descriptions are the graphically rendered diagrams that are sometimes ventured by theorists like Bruno Latour in his book Science in Action, a Nietzschean look at the agonistic dynamics of presumably democratic, scientific debate and controversy. Mouffe, Deleuze, Latour and others have provided us with a reimagining of democratic debate as a contest to link, unlink, build and dissolve assemblages of people and things.



Game Play: Using any email program, players will be able to post to one or more online, public discussions (e.g., Usenet newsgroups, weblogs and/or Yahoo groups). The proposed system will translate players' posts into a graphical display. Depending upon the content of the message written by the player, a player will be assigned a position on several two dimensional "fields." One field will be accessible for each theme discussed by the player in the posted message. Given a theme of discussion, the player's position will be displayed in relation to the other player-posters' positions regarding the topic. Each player's position is computed given the words used by the player to describe a theme. Thus, players who describe or discuss a topic using similar terms will be grouped together in the graphically displayed "field." A clustering algorithm will be employed to recompute everyone's position after each new message is posted to the discussion. To move oneself nearer to or farther away from certain other players, one will need to write and post a message to the group articulating a specific opinion about a theme of discussion. A spectator or player will be able to see the "game" develop by watching a webpage displaying the constantly updated positions of the players. Players will have the option of representing themselves with a small photo or graphic. These graphics or photos will move when the players' positions move.

Technical Details: Many of the technical details for this proposed system have been implemented in Java or Perl code for the purposes of my previous work, specifically for the Conversation Map. See this URL for details regarding the analysis of electronic messages, the computation of themes of discussion, and their automatic translation into graphical, network diagrams (social and semantic networks): www.sims.berkeley.edu/~sack/cm. In some ways the current proposed work is a simplification of the Conversation Map: rather than an analysis of social and semantic networks, the current system will compute a combined sociolinguistic result in the form of clusters of players arranged according to their respective opinions about the themes of discussions. In other ways, this work proposes an extension to the Conversation Map: while the Conversation Map is a "batch" system that computes one map at a time, the Agonistics system will interactively recompute the "fields" of players after each new post to the discussion.



June Define a simple points-on-a-field-of-color graphical interface.
July Identify and experiment with several different ways of clustering players and measuring distance between players' positions. Standard algorithms for clustering and well-known distance measures from the literature of information retrieval will be employed.
August-September Beta-test and tune the system by analyzing a variety of existing, online public newsgroups and/or weblogs. Implement more sophisticated graphical interface that allows players more choice in how they are graphically represented. Gather and respond to poster-player feedback.
October Release final system. Make source code available for download.



This project does not require the purchase of new equipment and entails only tasks that I can complete by myself. Consequently, the only budget item I have is my fee: $2000.