Neil Gaiman is speaking on campus tomorrow. After pulling a number of strings and a couple hair brain plots, I’ll be going to see him. More on that tomorrow.

For the moment, I’m desperately trying to open a vein and write this paper for Michicago. Unfortunately, all that is coming is air. Less than five days until it’s due. This is not good. Plus I also have a midterm due in the same time frame. So, for the moment, I draw at least some solace from this mini comic from Gaiman call An Honest Answer.

… Divine, I could you a little help here (and no cat suggestions)… read the comic, I swear it makes the last bit make far more sense.

Google Scholar has been an indispensable tool in gathering research for my thesis. For those not familiar, its a google search engine that is indexing the content of academic journals (including PDF content) and websites.

Interestingly, the most powerful aspect of it, in my opinion, is that it separately categorizes citations. For example, Sherry Turkle of MIT, has published a great deal on the subject of interactions with artificial intelligence online. Her book Life on the Screen is a seminal work. Google scholar allows you to specifically search for all the papers in it’s database that cite Turkle’s work. Since that returns an ungodly numbers, you can also restrict it to the papers that cite Turkle’s work AND contain “bot” in the body of the paper.

Anyone who is doing any form of serious research and not embracing Google Scholar is, quite frankly, doing themselves a significant disservice.

Joseph Westfall’s article What is cyberwoman?: The Second Sex in cyberspace (Ethics and Information Technology, 2000) presents an interesting point that is often glossed over in discussions of online space:

Cyberspace is not a place. Although interactions occur in cyberspace, there is not ‘there’ at which the interactors can find themselves together. Cyberspace only exists for anyone communicating in cyberspace. (Westfall, 2000: p161)

The first reaction to this is probably a resounding “Duh! Of course cyberspace doesn’t actually exist.” But I think there is a lot to unpack here.

Human beings are taught to be “completers” from birth. Scott McCloud insightfully points out in Understanding Comics that early games such as “peek-a-boo” help teach infants that while a person might disappear from view, they still exist. He goes on to point out that while we can’t see the world behind us, or lets say Tokyo for that matter, we know that both are there. We complete the unseen world. And that completion is an act of projection of our imagined ideas of how that world should be.

We do the same thing in most of our technology mediated communications. When I call someone on the phone, the simultaneity of conversation invokes images of the individual on the other end of the line and some rough approximation of their surroundings. And these images impact the overall phone interaction. For example, should the person I am speaking with become mad, typically my image of them displays the anger on its face. Typically sections of the conversation are geared to helping create these images. I often ask, during the course of a conversation questions like “Where are you?” or “What are you doing?” Each of these betters refines the image I form of my partner in conversation.

So why would it be any different in the online space? While the chat room is only a mediator, its interface and the related language of chat causes it to be conceptualized as a quasi-physical local. Tackling the latter first, we need look no further than the common parlance of referring to the chat environment as a “room.” The room metaphor immediate conjures the notion of a physical space, invariably four walls, a door, possibly a window or two. Further the use of phrases like “switching rooms” or “I’m in X room” further reinforces this notion.

The software interface also serves as a constant reminder of the simultaneity of interaction. Most chat programs display the usernames of all other visible chatters who are also in the room with you at the same time. So, before any text appears within the actual chat window you are already aware of the presence of other chatters. Obviously, the flow of conversation within the window further instantiates the copresence of other chatters, as do public notifications of chatters entering or exiting the room.

All of this creates the image of a shared space that we then fill with imagined images of the people that we are chatting with. And it is this desire to create these imagined images that bots, and other chatters for that matter, rely on. The mechanics of idenity, however, is a subject for another, soon to come, post.

One of the reasons for trying this format of fieldnotes is the chance to get immediate feedback. This is going to become increasingly important as I approach both Michicago and completing my thesis. What I’m saying is that questions are good. Please feel free to ask them, because they help me to construct my ideas. For example, this was a comment that was submitted about the previous post:

Have you considered using the typical “Subject A”, “Subject B” progression in place of the usernames?

Also, analyzing a user’s username for its meaning would be rather invasive to the individual. It’s sort of like publicly analyzing the names people chose for their sons and daughters and then publishing it, thusly divulging their identity.

The question of how to handle Usernames is a tricky one. The IRB doesn’t want Usernames used verbatim because in many cases they can be linked to an e-mail address. It is my responsibility to protect my subjects anonymity (within limits) to ensure that they do not come to harm as a result of their participation in my research. In more traditional forms of ethnography that would be done by simply assigning them a pseudomyn or referring to them as a subject as suggested in the comment. However there are subtle differences between a username and a given name that problemitize that strategy.

First and foremost, usernames (screennames, handles, etc.), unlike given names (here in refered to simply as name(s) for the sake of simplicity), are almost always self selected. Therefore there is data to unpack in what someone calls themselves online. For example, on forums that I’m active on, my username is “Matt Bernius”, because I feel that if I’m going to express an opinion then I should be confident enough to connect my name with every post that I make. That’s just one, albeit simple, reading of name selection.

It’s also something that a typical name doesn’t lend itself to. While it is the case that you can analyze why my parents chose to name me Matt, the fact that I’m named Matt tells you very little about me. I’m sure that some Freudians would suggest that being given the same name as my father has profound effects on me. But that’s the type of unsubstantiatable claptrap that gives the social sciences a bad name.

The analysis of usename becomes more important when you consider the interaction genre the usename is used in. In chat rooms, every time you speak your username is displayed. This means that the username is a fundamental part of the interaction, marking each of your comments. Thus, that name is constantly reinstantiating something in the mind of the viewer. Here’s an example that has popped up in my research (mom, you may want to skip the next section). I was in a room once with a guy (I’m assuming) whose username was something like: “8InchesForYou¹.” So 8, as I’ll refer to him from this point on, was hitting on everything that moved, including the bots. His username was clearly chosen to get an idea across. And that idea both reinforces and is reinforced by every post that he makes.

This means that simply neutering the username to “subject a” for example is an unsatisfactory solution. Doing that removes contextual information from the interaction. So instead what I will do is create a close enough name that the subtext isn’t lost.

^{1. In keeping with IRB regs I’m protecting the names of the innocent (or guilty in this particular case).}