My Reaction
On Twitter @HardHead, @VincentR and myself have been discussing the Usenet and its use in communication research. Micheal Beam, or HardHead, posted his reaction paper from a political communication class we took together, so I figured I would follow his lead and put my reaction up. I take it in a completely different direction than him, but still interesting. To this I would also like to add there are also serious sampling issues in the Usenet pieces that I don’t directly address in the paper.
Anonymity and the Internet
The readings this week generally discussed how computer-mediated communication fits into deliberative theory. Specifically, a number of the authors examined the Usenet for its deliberative potential (Davis, 2005; Papacharissi, 2004; Wilhelm, 2000). Others more broadly examined the potential of CMC in general to spawn political communication (Ho & McLeod, 2008; Stromer-Galley, 2002). Lastly, Wright and Street (2007) looked at the design of CMC systems to understand the deliberative potential of those systems. Using these readings as a basis, this essay will argue for a greater understanding of the role of anonymity in online deliberation.
As of late, online anonymity and Internet privacy have been fairly hot topics in both the legal scholarship (see Lessig, 2006) and the policy discussions (e.g. the creation of the Internet Safety Technical Task Force)(1). Yet we, as media scholars, don’t seem to have a firm grasp on the consequences and outcomes of anonymity on online discussion. The articles this week serve as example of how anonymity is often assumed to have an effect, but is rarely directly tested.
Like the numerous examples Wright and Street (2007) presented (e.g. the presence of a moderator), the relative level of anonymity(2) present in a CMC system is a function of design. For example, a CMC environment could allow complete anonymity. In this type of environment, an individual could post anything they want without any tracking or identifying information to connect the post to them, either their online or offline identity. At the other extreme, a CMC system could be designed to allow for completely clarity of the poster’s identity. For example, the wiki we use in this class is designed to allow every comment or alteration to be directly tracked to the author. The course wiki also goes beyond just connecting posts to an online identity. It connects every post to both an online and offline identity. Between these two extremes would be a number of alternative designs, which would allow for varying levels of anonymity(3). These varying levels of anonymity could lead to varying effects.
Ho and McLeod (2008) and Stromer-Galley (2002) both looked at the differences between face-to-face political discussion and computer-mediated political discussion. In both studies, they found that there are some people who don’t talk about politics in face-to-face settings, but might talk about politics online(4). Also in both studies, anonymity is assumed but never empirically examined. For example, in the Ho and McLeod piece it does not seem like the students were ever told whether they would be talking to people they knew or not or if they would be identifying themselves in anyway during the discussion. Maybe the difference Ho and McLeod found was not a function of the varying media, but a function of the varying perceived anonymity. The students in the face-to-face condition might have assumed they would be discussing gay marriage with other communication students. These discussion partners in a face-to-face environment would be able to easily identify the speaker. Whereas the students in the CMC condition might have assumed they would be anonymous and therefore would be more willing to express their opinions. Following this train of thought, the manipulation of the medium would make no difference it would be the participant’s perception of anonymity, which cause the findings. Ho and McLoed could have partially controlled for this by telling students who they would be talking to and how they would be identified in both conditions. A better way to test this would be to allow individual’s to discuss in different CMC and face-to-face anonymity conditions. This would be a simple 2 (CMC vs. F2F) by 2 (low vs. high anonymity) design(5). They could then tell if it is the medium causing the change or if it is really just one attribute of the communication system which is causing the increased willingness to express opinion.
The studies of the Usenet also suffer from not looking closer at anonymity and how anonymity could be involved in their findings(6). It is important to note that although all three of the Usenet studies discuss anonymity, none of the pieces clearly defined or directly tested anonymity. Davis (2005) is the only one that attempted to look at anonymity. He did this by looking at the names people used when posting. He found three groups of people: those that used clearly identifiable names, those who used aliases, and those who used no name at all. These three groups could be seen as theoretically matching the levels of anonymity described above. The first group, if they use clearly identifiable names that connect to their offline identity, could been seen as closely related to the low anonymity group (e.g. the course wiki). The third group could be seen as a high anonymity group. Then the second group could vary a lot. Some people who use their alias a lot and comment on the Usenet newsgroup a lot might feel they have no anonymity. Whereas someone who just signed up for this user name and never uses this specific Usenet forum might feel a lot of anonymity. Unfortunately, Davis did not look at this in very much depth.
Neither Wilhelm (2000) nor Papacharissi (2004) looked at anonymity at all in their studies, although both could have used it. In both studies, the authors performed content analyses of Usenet newsgroups. Their choice of method obviously constrained their ability to look too critically at anonymity as a possible explanation, but they still use anonymity to explain part of their findings. Both of them seem to be theoretically arguing that the Internet is anonymous by its nature or that anonymity is a characteristic of the underlying infrastructure. This is untrue. Anonymity is a function of design and use. Anonymity through design has already been discussed. Anonymity is also a function of how individuals use a CMC system. The Usenet does allow for relatively high anonymity, but that does not mean all users are anonymous. Some will remain anonymous, but others will become completely identifiable. The perceived anonymity of the user could easily have effects on the levels of civility and politeness the user shows when engaging in discussions.
Anonymity is becoming important in the discussion about the future of the Internet, with some arguing for an identification layer built into the code of the Internet (Lessig, 2006). Therefore it is important that we, as communication scholars, understand the role anonymity plays in online discussions. I am not making any normative judgments about Internet anonymity here. It could clearly have both possible and negative aspects. High anonymity setting can be very positive by allowing people to meet in a safe-haven where they can discuss problems without having to worry about repercussion (e.g. anonymous AIDS chatrooms). Unfortunately, they can also be very negative, as we have seen in the reporting on Internet predators who target underage children. Low anonymity settings also have positive and negative outcomes associated them. I am instead arguing that we must have a greater understanding of online anonymity, before its possibility is legislated away.
Notes
(1) In January 2008, MySpace announced a deal it had reached with 49 state attorneys general to create the Internet Safety Technical Task Force to examine issues of Internet privacy and safety. Primarily, it was created to protect children from Internet predators.
(2) For this essay, I am using a simple dictionary definition of anonymity. Merriam-Webster dictionary defines anonymity as a quality or state of being which lacks individuality, distinction, or recognizability.
(3) This view of anonymity could fit into Eveland’s (2003) “mix of attributes” approach to understanding media effects.
(4) I really have issues with the Ho and McLeod (2008) piece. I know they say that this happens a lot in spiral of silence literature, but I don’t like that they did a study of deliberation without any deliberation.
(5) The CMC/Hi could have no identifying information and have the participants talking to strangers. The CMC/Lo could have identifying information and classmates. The F2F/Hi could no introduction and stranger discussing. The F2F/Lo could have participants do introductions including name, profession, etc. and classmates. Not perfect, but a thought.
(6) I would also like to note that I had sampling issues with all three of the articles. My main problem is that I think when attempting to make claims about the whole USENET the unit of analysis should probably be newsgroups. These studies then have samples sizes of 4, 10, and 10. I have problems generalizing from these samples to the whole Usenet, let alone all of CMC.
References
Stromer-Galley, J. (2002). New voices in the public sphere: A comparative analysis of interpersonal and online political talk. Javnost: The Public, 9(2), 23-42.
Davis, R. (2005). Politics online: Blogs, chatrooms, and discussion groups in American democracy (pp. 35-67). New York: Routledge.
Wilhelm, A. G. (2000). Democracy in the digital age (pp. 86-104). New York: Routledge.
Wright, S., & Street, J. (2007). Democracy, deliberation and design: The case of online discussion forums. New Media & Society, 9, 849-869.
Papacharissi, Z. (2004). Democracy online: Civility, politeness, and the democratic potential of online political discussion groups. New Media & Society, 6, 259-283.
Ho, S. S., & McLeod, D. M. (2008). Social-psychological influences on opinion expression in face-to-face and computer-mediated communication. Communication Research, 35, 190-207.
Eveland, W. P. (2003). A “mix of attributes” approach to the study of media effects and new communication technologies. Journal of Communication, 53(3), 395-410.
Lessig, L. (2006). Code version 2.0. New York: Basic Books.