A study of online discussion in an Open-Source community: reconstructing thematic coherence and argumentation from quotation practices

From AcaWiki
Jump to: navigation, search


Citation: Flore Barcellini, Françoise Détienne, Jean-Marie Burkhardt, Warren Sack (2005) A study of online discussion in an Open-Source community: reconstructing thematic coherence and argumentation from quotation practices. Communities and Technologies 2005: Proceedings of the Second Communities and Technologies Conference, Milano 2005 (RSS)




Tagged: Computer Science (RSS) argumentation (RSS), open source (RSS), CMC (RSS), quotation (RSS), open source software community (RSS), cognitive coherence relations (RSS), rhetorical structure (RSS), argumentative moves (RSS), online argumentation (RSS)


Summary:

Closely related to Thematic coherence and quotation practices in OSS design-oriented online discussions, published the same year about related issues. The main differences:

  1. This paper discusses rhetorical structure of listserv messages
  2. the other discusses 2 PEPs whereas here only 1 is considered.

Both papers are part of Barcellini's ongoing analysis of the Python community.

This paper provides a finer-grained distinction between messages, in particular distinguishing number of sources quoted (one or multiple) in addition to number of quotes (one, several, or none) (see page 9). However this distinction is not further discussed.

It further points out that "average depth of quotation is rather small".

The bulk of the interest of the paper, as distinct from the SIGGROUP paper referenced above, is in the rhetorical structure. Items are also categorized by rhetorical force, whether they are:

  1. Syntheses
  2. Disagreements
  3. Proposals
  4. Agreements

These differ in likelihood of quotation: "Analyzing the content of the quotations, we found that the most prevalent design activities quoted are Syntheses (N=49; i.e., 28%) and Disagreements (N=48; i.e., 28%) followed by Proposals (N=31; i.e., 18%) and Agreements (N=20; i.e., 12%)."

Source of quotations is also discussed: about 20% of all quotations come from each of the champion of the Python Enhancement Proposal, highly active developers and administrators, and the project leader. The relationship between the status of the person quoting and what they quoted is particularly interesting: "* The PEP champion mostly quoted Decisions; other activity; and, to a slightly lesser degree, Agreements. Conversely, he did not quote Syntheses. This is easily explained since he was the author of most of the synthesizing messages and he does not quote himself.

  • The project leader quotes Syntheses and Proposals. Conversely, he tends not to quote Agreements or Disagreements.
  • HP-As mostly quoted Disagreements, Agreements, and Coordination messages. Conversely, they tend not to quote Proposals.
  • LP-As quoted Proposals and tended not to quote Disagreements.
  • HP-Devs mostly quoted Clarifications and Disagreements. They tended not

to quote other activities and Agreements."

Further, 57% of messages were either an agreement or disagreement.

There is also a structure to the relationship between the message quoted and the comment added. Agreements can be followed by Proposals or Syntheses. Disagreements are usually followed by Disagreements, rather than Agreements. Quotes of a Coordination comment usually signal another Coordination comment. "Clarifications are usually followed by Syntheses or a previous clarification quotation".

The authors discuss how these sequences can be seen as implicit agreement and disagreement, and point out that quotations also relate to argumentative moves as discussed by D’Astous, P., Détienne, F., Visser, W., and Robillard, P. N. (2004) Changing our view on design evaluation meetings methodology: a study of software technical evaluation meetings. Design Studies, 25, 625-655.

Theoretical and practical relevance:

Rhetorical structure could be very useful for analyzing online discussions.