An argumentation analysis of weblog conversations
Citation: Aldo de Moor, Lilia Efimova (2004) An argumentation analysis of weblog conversations. The 9th International Working Conference on the Language-Action Perspective on Communication Modelling (LAP 2004) (RSS)
Tagged: Computer Science (RSS) argumentation (RSS), blogs (RSS), language-action (RSS), communication (RSS), CSCW (RSS), socio-technical systems (RSS), knowledge management (RSS), web2.0 (RSS), socio-technical systems (RSS), pragmatics (RSS)
This paper focuses on blog conversations, looking at the dialogue-based argumentation that happens among the distributed individual blogs that make up a blogosphere. Their goal is to show the complexity of existing conversations and to pave the way for socio-technical systems which use personal blogs to support productive, organizational conversations.
They take a wide view of argumentation, focusing on conversation as an exchange of ideas taking place in a dialogue, which may (but need not) be monolithic. As they note: "Blog-mediated conversations are prime examples of conversations that do not produce traditional monolithic documents, but true dialogic texts, which reflect the involvement of multiple authorial voices" (On-line disciplines: Computer-mediated scholarship in the humanities and social sciences).
This paper views blogs as their own publishing genre (citing among others Bridging the gap: A Genre analysis of weblogs), with prominent features such as comments, trackback, and RSS, and an ecosystem which includes news aggregators.
This work is based in part on the authors' framework for socio-technical argumentation analysis Argumentation support: From technologies to tools, which they say distinguishes between three aspects:
- functionality design - the possible technological functions
- argumentation design - how the functionality is expressed in practice to shape the argumentation behavior
- argumentation routines - expected practices about who can speak and listen, what types of arguments are admissible, how to resolve arguments, etc
Further, they emphasize minimizing the gap between the way the technology is used (argumentation routines) and what is functionally possible (argumentation design).
Blog Conversation Analysis
They traced a single conversation from one week in November 2003. The conversation was traced by starting backwards from a final post from the author whose post had kicked off the thread. They followed links, selecting only posts that made the initial topic and quoted part of the posts they linked back to. Figure 2 of the paper shows a selection from this conversation, from 5 blogs.
Blogs versus Mailing LIsts
They note various features of the blog conversation, which they contrast with mailing lists
- It is distributed across many blogs (mailing lists use a single tool)
- The blogger has to actively search for replies (mailing lists distribute messages automatically to all members)
- Four days pass before the first response, but then 5 posts follow in the subsequent 2 days. (mailing lists usually have a short, intense exchange of messages)
- Posts have their own title (mailing lists usually maintain an existing subject line)
- Posts contain many comments and links, beyond the direct replies. "The form of the conversation is thus much more a web rather than the usual focused discussion tree found in dedicated forums such as mailing lists" (mailing lists are visualized as a tree)
- The final post quotes and interprets responses made to that blogger's first post. (in mailing lists, such summaries are rare, and generally only the preceding message is quoted.)
Implications & Conclusions
Summaries are common, they say because "weblogs provide their authors with a personal space next to a community space", meaning that there is both a conversation with oneself and a conversation with others taking place on each blog (citing Discovering the iceberg of knowledge work). Further, each community of bloggers has loose boundaries, making for multiple overlapping communities.
For knowledge creation, they see the relevant issues as fragmentation and slow response time, as well as finding the boundaries and topic of the conversation in the moment. They point to a number of issues regarding fragmentation: cross-channel conversations (e.g. use of email, IRC, VOIP, etc in addition to blogs), the use of both comments and new posts to respond, and the complexity of link semantics (since links without quoting can be used to give credit or support an argument, and since posts can be mentioned without formal quoting).
Suggested Future Work
They suggest several avenues of possible future work:
- using graph theory and visualizations to make networks visible (e.g. Social translucence: Designing social infrastructures that make collective activity visible and Visual unrolling of network evolution and the analysis of dynamic discourse)
- finding ways to represent the semantics and pragmatics of the blogosphere (perhaps using IBIS and knowledge maps; but they also describe Depluralising the Web, repluralising public debate - the case of the GM food debate on the Web in which Marres & Rogers also attend to the organizational dimension of issue networks)
- finding ways to use blogs as part of complex socio-technical systems for virtual communities; these systems will need to be iteratively developed. They suggest, for example, that if a certain issue is discussed by a critical mass of individual blogs of employees in a department, then the department blog should be automatically updated and a notification email sent to the bloggers and department manager regarding a summarizing conversation.
They also mention various elements they did not explore here:
- adding classical argumentation from rhetoric and legal theory (e.g. Toulmin)
- refining their theoretical framework of argumentation
- studying limitations on discourse rationality (e.g. time, power)
- Brandes, U. and Corman, S. R. (2002), Visual unrolling of network evolution and the analysis of dynamic discourse. In Proc. of the IEEE Symposium on Information Visualization (InfoVis'02), IEEE Computer Society, pp. pp. 141-151.
- de Moor, A. and Aakhus, M. (2003), Argumentation support: From technologies to tools. In Proc. of the 8th International Working Conference on the Language-Action Perspective on Communication Modelling (LAP 2003), July 1-2, Tilburg (Ed, de Moor, A.), pp. 135-141.
- de Moor, A., Keeler, M. and Richmond, G. (2002), Towards a pragmatic Web. In Proc. of the 10th International Conference On Conceptual Structures (ICCS 2002), Borovets, Bulgaria, 15 - 19 July, (Ed, Corbett, D.).
- Efimova, L. (2004), Discovering the iceberg of knowledge work. In Proc. of the Fifth European Conference on Organisational Knowledge, Learning, and Capabilities (OKLC 2004), April 2-3, Innsbruck, Available at: https://doc.telin.nl/dscgi/ds.py/Get/File-34786.
- Erickson, T., Halverson, C., Kellogg, W. A., Laff, M. and Wolf, T. (2002), Social translucence: Designing social infrastructures that make collective activity visible, Communications of the ACM, 45(4): 40-44.
- Harrison, T. M. and Stephen, T. (1992), On-line disciplines: Computer-mediated scholarship in the humanities and social sciences, Computers and the Humanities, 26: 181-193.
- Herring, S. C., Scheidt, L. A., Bonus, S. and Wright, E. (2004), Bridging the gap: A Genre analysis of weblogs. In Proc. of the 37th Hawaii International Conference on System Sciences (HICSS'04), IEEE Press.
- Jenkins, E. (2003), Dynamics of a Blogosphere Story. In Microdoc News: Online Magazine about Exercising Personal Power in the Information Age, May 20, (now via http://web.archive.org/web/20040203105253/http://microdoc-news.info/home/2003/05/20.html )
- Marres, N. and Rogers, R. (2000), Depluralising the Web, repluralising public debate - the case of the GM food debate on the Web. In Preferred Placement: Knowledge Politics on the Web (Ed, Rogers, R.), Jan van Eyck Akademie, De Balie, Maastricht, Amsterdam, pp. 113-136.
- Voiskounsky, A. (1997), Telelogue conversations, Journal of Computer-Mediated Communication, 2(4).
Theoretical and practical relevance:
Understanding how blogs support other work practices, and how a conversation is distributed among multiple blogs and blog posts. Does not focus on the blog comments. Contrasts with listserv conversations.
Reviving Jenkins is useful. Jenkins' graph of blogosphere story (Figure 1 of the paper and below) views each post as one of four types: opinion, vote, reaction, or summation. Opinion posts define a topic; vote posts indicate agreement or disagreement with another post; reaction posts respond to a single story on another site; and summation posts summarize various articles. According to Jenkins, opinion posts kickstart most discussions, then generate votes and reactions. Some, but not all, blogosphere stories receive a summation.