Information Technology Reference
In-Depth Information
A rather blurry line separates the Internet-based practices of relating and
retelling widely-circulated stories authored by mass-media producers (e.g.,
Hollywood, CNN, etc.) from the practices of independently producing stories
for Internet distribution. It is the former sort of practice that is the concern
of this paper. Quotation, citation, and fragmentary repetition of stories are
the life-blood of audience discussions and analysis of mass-produced stories
Henry Jenkins discusses these audience practices as tactics of “poaching”; see
Jenkins (1992). Audience members recirculate famous lines from movies (e.g.,
“Frankly my dear I don't give a damn,” “I'll be back,” “Make my day,” etc.),
comment on the plots and characters of known stories, summarize and retell
pieces of stories for one another. The technology presented here is a first step
towards a better understanding of story quotations, citations, and repetitions
as the “threads” that weave people together into online, social networks.
A social network-based approach to story understanding differs from the
standard approaches to “story understanding” that have been pursued by re-
searchers in symbolic AI. Rather than examining stories as cognitive structures
internal to individuals, the social network perspective is to see stories as shared
ties that gather people into communities or social networks. An analogous dif-
ference in approaches to narrative theory was described by Mikhail Bakhtin in
his critique of Russian Formalist approaches to literature and in his advocacy
for a sociolinguistic method. See, for example, (Medvedev 1978). Bakhtin's
“dialogical” approach to language and literature has been widely employed in
contemporary literary theory, sociology, and media studies.
Moreover, unlike various media studies content analyses and structuralist
analyses of narrative and film, it assumes the existence of an active, creative au-
dience and uses audience activity (e.g., their discussion about a story) as the fo-
cus for gaining an understanding of stories. This distinction between research
approaches in media studies (i.e., “content analysis” versus ethnographic ap-
proaches to the “active audience”) has been recently summarized in topics
such as (Nightingale 1996). This alternative perspective shares some affinities
with AI collaborative filtering techniques. Outside of AI, in the field of sociol-
ogy, social network-based approaches to story understanding are not unusual,
but the techniques of sociology can be improved through the use and devel-
opment of an array of tools from natural language processing/computational
linguistics. The research described here folds together insights from computa-
tional linguistics and the sociology of social networks to support the design of
a new kind of story understanding technology; a technology predicated on the
existence of verbally active story audiences.
