The structure of online discussions are partially determined by its platform. Whether it is through blogs, BBS, chat, email and other online platforms, the depth, dynamicity, communicability, accountability, communability and the behavior of the discussions vary. As an easy example, the presence of anonymity limits the responsibility and accountability of the communicator — diluting the constructiveness of the thought into a more emotional one. Huffington Post utilizes a thumbs up/down system to establish credibility, while DailyKos lets users post there own diaries. The ability of the latter to recommend diaries make it possible to motivate users to post entire essays on topics of interest — establishing a much higher signal to noise ratio than HuffPo (even if you fix the user base; and yes, there are obvious differences between an online press and a community blog).
The structure of 2channel is characterized by its anonymity, the omnicompetence of its discussion topics, and its speed (both its interface and the ability to post news stories). Its anonymity is well known, as most people don’t even use pseudonyms to opine behind the walls as nanashi (名無し), or nameless. The anonymity helps broaden its influence in Japan for two reasons: the liking of privacy (or a dislike of publicity, depending on POV) and the lack of constructive online platforms independent of 2ch. The former is self-explanatory, while the latter is as such because there is a mismatch between the society’s communicativeness and its available infrastructure. 2channel can be partially looked at as a compromise between a serious online discussion and no online discussion; where 2ch anonymity is a compromise between an online presence and no online presence. Combine this with a constant forum to discuss news, a light website design (more dynamic and exciting) and the breadth of topics covered, 2ch has become a nationwide media presence.
Understanding the structure, and hence, the dynamics of 2ch interactions can potentially elucidate significant factors that has an effect on user and community behavior. To understand certain characteristics of 2ch interactions, Matsumura et al. (2005) collected and analyzed data on 2ch threads. This is no easy task because the forum is so goddamn large, but they managed to grab data on 5,748 threads (including 1,738,418 messages in total) classified into 30 categories. They explicitly defined eight measurable indices (variables) to characterize threads:
contents ~ Measures the average message size (jargons, emoticons, and ASCII arts are unaccounted) in a thread to know the depth of discussion
activity ~ Measures the number of messages in a thread to know the activity of discussions
interaction ~ Measures the proportion of messages replying to previous messages with ‘‘>’’ mark to know the tendency of interaction between people
speed ~ Measures the average number of messages posted per day to know the intensity of discussions
jargon ~ Measures the proportion of jargons specific to 2channel used in a thread to know the consciousness of people
ASCII art ~ Measures the proportion of ASCII arts specific to 2channel to know the consciousness of people
nameless ~ Measures the proportion of messages posted nameless in a thread to know anonymity of discussion
ABON ~ A coined word in 2channel with which administrators replace wrong messages such as disclosure of personal information, measures the proportion of messages including ‘‘ABON’’ in a thread to know the disturbance of 2channel
The data is standardized so that it is comparable across variables. I plot the data below so that the whiter areas are higher values and the redder areas are lower values of each metric with respect to each of the 30 categories.
Unfortunately, the categories weren’t described in the article so I can only be semi-sure of what each of the mean. For example, one of the categories is “Commentary,” which I assume to be the broadcast (実況) threads where users comment on ongoing (say) TV programs at real time. From this interpretation, it goes with intuition that the “speed” and “activity” of the threads are high. The “ABON” rates are high for the “Adult” category for reasons I won’t say 
. The “ABON” rates are supposedly the biggest indicator of flaming, and it is high in “Society,” where they might discuss China and South Korea. Another interesting tidbit is that, “interaction” rates are high for the “Academic-science” category, whereas it is about average for “Academic-humanity.” The “activity” is also low for “Academic-science,” so it might hint that there exists a few people discussing issues in a constructive manner (humanities folks are incapable of such ).
Anyways, they did a factor analysis to study the relationships (correlations) between each variable, and validated their results by structural equation modeling (SEM). One of the things that they concluded was that anonymity (“nameless”) positively affected the activity of communication (“activity”) and negatively affected the depth of discussion (“contents”). They measure “nameless” by the proportion of pseudonyms used (vs. no pseudonym) and it stimulated postings, but degraded the discussion.
I plotted the “nameless” variable to the above variables for illustration.
Although this graph gives different quantitative results from the SEM, qualitatively, it is the same. We also see that the strength of the anonymity status in 2ch is very weak (not significant). Anonymity does help a little bit in the activity and lowers the depth of the discussion, but considering 2ch culture, it really is not a significant player.
Though this research did not give out any novel results, it does help to quantitatively lay out the heterogeneity in the discussion within 2ch. Some people tend to instinctively categorize 2ch as a flaming arena (part of its reputation deserved), but there is some decent, informative threads out there. Fairly recently, there was a pseudo-political debate show (太田総理) that proposed to limit internet usage to one hour (!) since the “internet” has debilitating effects on imagination, creativity, etc. It’s only an opinion of an archaic TV dude, but when a decent part of the population makes synonymous the “internet” and gloomy parts of 2ch, it makes it harder for a society to embrace a more mature, cultured online platform.
Refs:
- Matsumura, N., Miura, A., Shibanai, Y., Ohsawa, Y., & Nishida, T. (2004). The dynamism of 2channel AI & SOCIETY, 19 (1), 84-92 DOI: 10.1007/s00146-004-0302-5
