Issuepedia:Dispute Resolution Technology

Navigation
Issuepedia:About: Technology: Dispute Resolution

The Problem
Discussion of complicated issues has long been constrained by the linear nature of human language. Discussions often spout multiple threads which must each be followed in order for the main body of the discussion to make sense, but because of the innate difficulty of representing multiple threads in writing – much less in speech – most people will rely solely on what they can remember, or (more likely) what they think most of the audience will remember, following only the most critical few threads. When the discussion is combative rather than truth-seeking, selective thread-abandonment can be used strategically, which causes each party to avoid bringing up new threads – in effect, oversimplifying the issue as much as possible in their own favor.

Even when a discussion is non-combative, it can quickly become very difficult to follow when more than a few threads emerge. Participants must carefully label each thread to which they are responding; if even one person forgets to do this, the threads can become tangled to the point where untangling them becomes a new issue of its own, giving the dialogue a prohibitively high maintenance overhead; most discussions are usually abandoned at that point, or shift in focus so that the original topic is essentially lost.

Indeed, the idea of thread completion does not seem to be part of our culture. Back-and-forth discussions in newspaper columns, with each columnist answering only the points they want to answer (and which will fit in the available column space) seem to be an acceptable form of debate – as are so-called formal "debates", where each side is allowed a limited number of statements, and a limited amount of time for each one. There are many other examples of arenas in which we attempt to resolve major issues, and in very few of them is there any sense that threads should be completed before the main issue can be decided.

Author David Brin raised much the same point in his essay "Disputation Arenas: Harnessing Conflict and Competitiveness for Society's Benefit": Many wonderful and eloquent arguments are raised, only to float away like ghosts, seldom to join any coalescing model. Rabid statements that are decisively refuted simply bounce off the ground, springing back like the undead. Reputations only glancingly correlate with proof or ability. Imagine anything good coming out of science, law, or markets if the old arenas ran that way!

Existing Solutions
If the root of the problem is our linear, single-threaded language, then the solution must lie in somehow de-linearizing it and multi-threading it.

Picture a venue where adversaries can no longer get away with just screaming past each other, but must actively answer each others' accusations, criticisms and complaints. A place where one group's vision -- or model of the world -- can be tested, dented, appraised... and possibly improved under the watchful gaze of an interested public. A site where the disprovable can be disproved, the ambiguous can be pinned down a bit more, and good ideas may get deserved attention just a bit sooner. -- David Brin again

Hypertext seems to offer at least part of a solution. Sub-issues (threads) can be spun off onto separate pages, where they can expand as needed without crowding out or confusing the main issue. References can be provided to prevent muddying of the facts. Errors can be corrected.

Wiki seems to be another powerful tool, one that further leverages the power of hypertext. Anyone can edit, so anyone can suggest overlooked lines of inquiry, or point out faulty reasoning or incorrect facts. Because of wiki's internal auditing, anyone can also determine who said something, or who made changes to what someone else said. Wiki enforces accountability while drastically lowering the barriers to participation.

Future Solutions
While hypertext and wiki are good tools, they lack a certain ability to handle real-time discussion. "Chat" formats, such as Internet Relay Chat and Instant Messaging, are excellent for realtime discussion and have very low barriers to participation (type something, hit enter; done), but again lack any innate thread-handling ability. It seems to me that some kind of hybrid between chat and hypertext, possibly with miniature windows instead of separate pages, should be developed. Chat also does not archive very well; many IRC clients can automatically log conversations, but there are very few tools for making such conversations available for further/future reference.

some projects

 * wikis are generally useful for documenting or discussing any complex area of understanding
 * The MIT Collaboratorium attempts to solve this exact problem, but it's not clear if they have a working demo or not
 * Intel research has two Firefox plugins which present some useful ideas implemented in a rather limited way:
 * Dispute Finder
 * Think Link
 * Debategraph
 * 2009-01-14 Another way to look at the crisis "A new web tool is being used to help foster sensible debate about the conflict in Gaza."
 * An earlier report mentions Ushahidi, a downloadable crowdsourcing tool being used to report on the Congo war
 * Honest Argument allows arguments to be broken down in a hierarchical (tree) view showing how points and sub-points, supporting and contradicting arguments, all relate to each other. (discovered 2006-11-13)
 * Google Moderator: not so much a debate tool as a way of prioritizing questions from a large audience
 * Jyte is very similar to Issuegroups, except simpler and purpose-written (rather than being an adaptation of software written for another purpose). The idea is you make a claim, and other people vote for it. There is non-threaded discussion, and some statistical compilation.
 * TakeOnIt "Compare opinions of world leading experts and influencers." Looks like they do some tracking of whose opinions (external authorities) are having the most influence, i.e. most cited
 * TruthMapping.com allows deliberations in a way that naturally filters out the noise leaving the true content.

in development

 * Convince Me (alt): open-source application and Java applet for mapping out arguments. Appears to have some serious science behind it: "The Theory of Explanatory Coherence (TEC) and its associated connectionist model, ECHO, offers an account of how people decide the plausibility of beliefs asserted in an explanation or argument (Ranney & Thagard, 1988; Thagard, 1989)"
 * Written in Java, but source code is available from SourceForge -- possibly could be modified or rewritten to work within a larger platform
 * Holocene Chat, a concept by author-futurist David Brin
 * HyperChat: on hold for the time being; a Woozle project
 * Open Debate Engine: no site (or released code) yet, but the author appears to have arrived at many of the same conclusions about what is needed
 * Elephant: similar concept to TruthMapping, but more mail-list oriented
 * The Issue Based Information System (IBIS) was developed by Hoerst Rittel, a professor of Architecture at UC Berkeley to record design decisions for later use. More recently the book "Dialogue Mapping" by Jeff Conklin discusses the same idea in depth. The innovation was to insist that the base issues (questions) must precede any answers which in turn precede arguments pro and con. This then gives a convenient index into an extended discussion as well as making it obvious that multiple answers and arguments are expected.

Related Pages

 * accountability arenas
 * structured debate