Discourse-marking of concession and contrast in asynchronous online discussion

From AcaWiki
Jump to: navigation, search

Citation: Diane M. Lewis Discourse-marking of concession and contrast in asynchronous online discussion.

Download: http://sws.bu.edu/bfraser/CDM%20Papers/Lewis%20-%20DM%20of%20Concession....doc

Tagged: pragmatics (RSS), linguistics (RSS), discourse-markers (RSS), concession (RSS), contrast (RSS), discourse analysis (RSS), CMC (RSS), online communication (RSS), online journalism (RSS), discussions (RSS), online argumentation (RSS), political discourse (RSS), online news (RSS)


This linguistics conference paper discusses the concession markers "of course" and "bien sûr", contrasting their use in online forums and political speeches. Specifically, the online corpus comes from two UK (The Financial Times and The Guardian) and two French periodicals (Le Monde and Le Nouvel Observateur). (See Table 1). (Further usage data from UK conversations and periodicals is also presented from Lewis 2003, Rhetorical motivations for the emergence of discourse particles, with special reference to English of course, but equivalent French data is not available.)

There are six regular uses of "of course" and "bien sûr":

  1. emphatic affirmation or denial
  2. hedging
  3. backgrounding
  4. introducing a sub-topic
  5. concession marking
  6. irony/disapproval (to distance the speaker from the proposition).

Irony is absent from political speeches but accounts for 15% of the use online, where irony, sarcasm, and mockery are acceptable argumentation strategies, and the discourse markers "of course" and "bien sûr" are used to discount the position.

Concession is the most common use (37%) in UK political speeches yet this use "of course" is comparatively rare (21%) in English online. In French, the percentages (35% in political speeches, 34% online) are similar, but the concessions use more straightforward constructions. In political speeches, concession is structural, setting up a contrast which discounts a satellite idea while emphasizing (e.g. following 'but' or 'mais') a nuclear idea. (See Table 6 for a typical concession construction.)

The author notes that "of course" is used more frequently in online forums than in political speeches and that it is "bleached: 39% of the time it is a relevance hedge providing little extra information. "Bien sûr", on the other hand, is most common (34%) in online forums as a concession device, which the author likens to an information-structuring role because of the tightly-structured concessions; further, French has alternatives in "certes" and "il est/c'est vrai (que)".

Discussion Forums

The corpus used is online discussion forums, enabling the author to make several interesting observations about this conversational, written genre.

First, she observes that messages are like both "conversational turns" (see e.g. Ferrara, K., Brunner, H., & Whittemore, G. (1991). [Interactive written discourse as an emergent register]. Written communication, 8(1), 8.) and "scripted monological speeches".

She presents two examples. First, a forum discussion from Le Monde is analyzed from the perspective of coherence, turntaking, agreement/disagreement. She considers the first 19 messages and first indexes these messages by participant (see Table 3). Table 4 indicates which message is answered, the subject line (abbreviated A-K, with e.g. Re A, J disagreement, etc), and the opening idea. Table 5 indicates the Message openings which react--either through quoting, stating disagreement or partial agreement.

Second, she presents a Guardian-Unlimited discussion illustrating topic decay as well as how one-to-many conversations on a topic "can turn into a set of overlapping dyadic 'conversations' (which apparently happens offline as well, per [Zimmerman, D.H. and Boden, D. (1991). 'Structure in action: an introduction'. In D. Boden and D.H. Zimmerman (eds) Talk and Social Structure: Studies in Ethnomethodology and Conversation Analysis. Cambridge: Polity Press, 3-21.) Such dyadic conversations use more dialogic language (with more interpersonal references, modal and attitudinal markers (see Yates, S.J. (1996) 'Oral and written linguistic aspects of computer conferencing: a corpus-based study'. In Herring, S. (ed.) (1996) Computer-mediated communication. Linguistic, social and cross-cultural perspectives. Amsterdam: John Benjamins, 29-46.

"The dislocation of turn-taking and the jumbling of several 'conversations', far from deterring participants, can speed up the discussion." She finds evidence of "multiple intertwined dyadic conversations", indicated through the use of names (see Werry, C.C. (1996) 'Linguistic and interactional features of Internet Relay Chat'. In Herring, S. (ed.) (1996) Computer-mediated communication. Linguistic, social and cross-cultural perspectives. Amsterdam: John Benjamins, 47-63.), exact quotes (e.g. see Mondada, L. (1999) 'Formes de séquentialité dans les courriels et les forums de discussion. Une approche conversationnelle de l'interaction sur Internet'. Apprentissage des Langues et Systèmes d'Information et de Communication vol. 2, no. 1, 3-25.), and the tendency to send "several consecutive messages each directed at a different participant".

In contrast to the Herring's 'basic electronic message schema' (link to an earlier message, express views, appeal to other participants) (from page 91 of Herring, S.C. (1996) 'Two variants of an electronic message schema'. In Herring, S.C. (ed.) Computer-mediated communication. Linguistic, social and cross-cultural perspectives. Amsterdam: John Benjamins, 81-106), the author finds no uniformity in message length, language style, structure, or topic, and few appeals to other participants.

The main message structure is [reaction] + position + evidence. Reactions are often coded in discourse mesages such as oh, yes, exactly, so, well, but, ok, yep, and, actually, of course, erm, anyway. Within political forums, the general position statement is often backed up by one or more supporting statements and argumentation. Here is an example supplied from an FT.com Globalization discussion:

"In globalisation, ideas of democracy are a major cause of division. For example, a five-page essay last year on Taiwan told us ... "

Coherent sequences found in the argumentation sections include "question + answer", "concession and counterargument", "claim + evidence", "denial + correction or account". These tend towards either dialogic (e.g. Q&A) or monologic argument, though they can appear in both contexts. Examples of each are presented.

The author believes that material influences such as screen size and transmission speed are less important than "social and economic factors that new technologies are given rise to".

Future Work

Longtitudinal studies of CMC could start to address other linguistics-oriented questions like:

  • "How quickly do new online discourse communities establish their own styles and conventions?"
  • "What kind of dialect levelling takes place and how?"
  • "How do discourse practices circulate online among different communities?"

This paper is likely among the first to address online argumentation, yet does not present suggestions for future work in that area.

Context of this Conference Paper

This was presented at a panel on 'Different approaches to spoken interaction' organized by Anna-Brita Stenström and Karin Aijmer, citation from author

An updated journal article is based on this paper; see Lewis, D. M. (2005). Arguing in English and French asynchronous online discussion. Journal of Pragmatics, 37(11), 1801-1818. doi:10.1016/j.pragma.2005.02.014

Other related work on genre analysis/linguistics in new media, connectives, and discourse pragmatics

The author's dissertation was about discourse connectives: Diana LEWIS. [Some emergent discourse connectives in English: Grammaticalization via rhetorical patterns]. (U. Oxford, 2000).

(Following selected from citations on her publication page):

Theoretical and practical relevance:

Among the first linguistics studies of online argumentation. Important for considering coherence and rhetorical markers. The variant uses (and divergent semantics) of even simple terms like "of course" provides a caution against simple keyword-based indexing of online argumentation.

Further details on methodology would have been helpful for future researchers.