Ethnography of ICQ

Tuesday, November 27, 2001

Analysis of log one done. Moving on to the other two over the next three days. By my count, each participant should take about six hours to finish.
Entered topic and time data into a chart and didn't really find an correlations, but think it's interesting to see how much certain people talk, when there are spikes, when it runs particularly short. Greetings and closings are also interesting, especially in light that so far, 7G opens and ends far more conversations than Lady A. Looks like an issue of power, methinks.

Powerpoint/paper outline, first draft
Introduction to topic--Limited ethnography of communication on ICQ.
Ethnography--based on the SPEAKING acronym of Dell Hymes and the examples given using that acronym by Nancy Bonvillian. (pass around SPEAKING sheet)
Limited because there are certain parts of the acronym that do not fit with my chosen medium of communication--A, I, N, and G specifically.
Thesis: ICQ conversation is normal, informal conversation that has evolved to fit the strengths and weaknesses of the medium.
ICQ is an Internet based messenger system that is primarily used for talking to other people. Conversations are done through message boxes--a box containing the message comes to the user, they read it, and then reply in a different box. This means that unlike other messengers, ICQ does not show the progressive conversation on the same screen. Rather, this conversation is logged by the program in History files, which can be accessed from the message box. They can also be converted to text files if the user wishes to read the whole of the conversation.
For my project, I studied three pairs of people for this project: one m/m (7G and StN) one f/f (Lady A and Lady H), one m/f (7G and Lady A), taking fifteen conversations from February, May, and August of 2000. I chose these months primarily because there was consistent conversation and because I had existing log files for all of them.
Starrynight(StN): 21, University of New Mexico student, Asian Studies major, male
7Ghent(7G): 20, University of New Mexico student, Linguistics major, male
Lady Ariae (Lady A): 18, IUPUI student, English major, female
Lady Houki (Lady H): 17, high school student, interested in Star Wars, female

Data review
Situation: ICQ, a real-time instant messaging system
Participants: In all of these conversations, only two.
ICQ users can choose their conversation partners in four ways:
a) From friends in real life
b) Introduced by a mutual friend
c) Know from some other setting, such as a chatroom or a classroom, and wish to converse outside of that medium
d) Find randomly by searching for similar interests or location
StN and 7G know each other in real life, 7G and Lady A were introduced by a friend, and Lady H and Lady A met on a bulletin board. This has an impact on their conversations; for instance, StN and 7G talk about events and people that they know mutually, making of some of their conversations unclear to the outside reader; 7G and Lady A tend to talk about things and people that they both have knowledge of, as well as more of their lives; Lady H and Lady A tend to, at least in the beginning, focus on one topic: the bulletin board and the stories Lady A posts there. Later on they branch out.
Ends: To communicate. An underlying goal is to become friends with the person you communicate with.
Key: In the case of these users, friendship.
Norms: Three norms are usually observed by ICQ users, with some underlying norms that pertain to usual conversation.
Greeting: Opening a conversation. The conventional greeting of "Hi, Hey, Yo!" is pretty standard, followed by the standard, "How are you?" and oftentimes, "What are you up to?" in some form. Greeters usually have the first topic--"How are you?" and so in the beginning, whoever greets controls the conversation.
Topic changes: These are usually accomplished by a putting a space between the old topic and the new, and then opening the new with some sort of question/statement of topic introduction.
Well, I didn't like her much anyways.

So, how's the paper coming along?
Ooooh...that sucks...

Guess what? I'm failing Math!

Occasionally, two users will pick up a conversation where it ended, so you might have an opening that functions also as a topic introduction. This is rare, and it's because occasionally users don't get the last message until they log onto the system again. If this is a user's opening, then they tend to tack on some sort of greeting phrase to their reply. Or they just ignore what was said all together, go through the greetings, and then return to topic presented.
I hope you have a good day! (last message)

Actually, my day sucked and I want to go to bed. But how are you? (opening message)

Another key to this is prompt reply to a topic. Because ICQ works off an Internet connection, a leeway is given between messages--about 3-5 minutes, occasionally up to ten. After that, the one user will prompt the other with another message to check and see if they're still around.

Due to ICQ's spatial medium, and because a space between sentences indicates that there's a different topic, participants can carry on discussions on up to three topics at once. This usually doesn't last very long, though.

Closing: Some sort of way to tie off the conversation. This is also standard to normal conversations between people--"Good-bye, bai, good night, sleep well." Users occasionally develop a personal closing. For instance, between Lady A and 7G, "Night and sleep well!" was a sort of ritualized closing. Between Lady A and Lady H, oftentimes they would close with, "Talk to you tomorrow!" or some variant of that.

StN pretty much violates all of these norms. He doesn't greet, switches topics without much lead in, tends to wander off and not reply to his ICQ messages, and fails to close.

Other things to find places for (help!, please)
Emoticons and Actions
Emoticon: A figure "drawn" with letter and punctuation that resembles a face in real life. EX: ^_^, >.<, :-), ;-), O_o
Action: An event, usually enclosed in ** or //, that the user is essentially miming taking place. EX: *hug*, *kiss*, *smile*
emoticons used occasionally as mood indicators, and often as defusers. For instance, if something strong is said:
I really HATE that. :) <----that will function to make the statement not as serious.
Also take the place of body language on many occasions.
Actions: used to emphasis a statement made, to show affection or solidarity, and to fill in non-verbal cues.
*grrrr*punches paper through wall* I can't write this stupid thing!
I'm so sorry, chica! *hugs*

Gender differences between m/m--f/f conversations--where do I fit those in the paper? What section do they belong to, or do I cover it after I cover everything else?
Can I have a section called "Miscellaneous?" or "Various other thingees"?

Conclusion: Though ICQ is conversation through text, it has the same essential characteristics as normal speech. Differences are made to accommodate the format and medium, and include things such as topic shifts and emoticons and actions.

posted by Lady Ariae 11/27/2001

Sunday, November 25, 2001

Okay, that was ambitious. I'm about a 1/3 of the way done with my analysis of log number one, and found out I need to cut down quite a bit. My total file of three months for each conversation is 1078 pages long--mostly me and Ki, actually, since we tend to have LONG messages. The change I've decided to make is to look at 15 "days" of conversation--three from each section of the month (1-4, 11-15, 22-26)--and then do my averages. Looking down at my previous log, the only thing I haven't charted is who initiates the topic changes, though I have been counting the rest of the things. I'll try to reproduce some of the charted data here. (this is actually horizontal on the page, but I needed to put it vertically here for it to work...)

Feb. 2000: 7Ghent and Lady A
Date 1
Greeting 7Ghent
Topics Seven/Four
Conversation Time(min) 119
Min/Topic 10.1
Emoticons n/a
Actions n/a
Closing 7Ghent

I haven't actually checked out Emoticons/Actions yet. They're going to be part of my Keys analysis. I'm not doing the full SPEAKING acronym--right now, it's just SPEKN (Situations, Participants, Ends, Keys, Norms)--because I think that's all I can establish and use from my research so far. Also, E/A play a roll in looking at gender stuff...which is why I want to separate the categories. While emoticons are almost useless in many conversations (they act as emphasis or defusers), actions are key to knowing the mood of who you're talking to. And they're such a part of the online culture that any analysis of conversation would be pointless if I didn't show how their meaning can affect communication.

More tomorrow.
posted by Lady Ariae 11/25/2001

Tuesday, November 20, 2001

My analysis of log number one will be up over break, and hopefully my ICQ logs will be fixed by then, too. Might do log number two as well. I corrected and added some things to my first three sections, but that introduction will need a revision, neh?

Is it all right to structure my paper on the SPEAKING? As in, sections S, P, E, A, K, I, N, G, or is that too much? I don't anticipate any of those sections being too long, minus the Norms one, but it might be overdoing it with how many people I'm working from. still, an overview perhaps? I thought it would make an interesting structure...or, really, the only structure I could think of. I've also thought of going just with Norms, but then that's not really an ethnography, either. Gah.

Quick things I am looking at in logs:
How topic changes are intiated, and who does it.
Time on a topic.
Time of conversation.
Some use of actions/emoticons

posted by Lady Ariae 11/20/2001

Sunday, November 18, 2001

First three sections, rooooooough draft.


I. Introduction
As the use of the Internet increases, more and more people are using it to keep in touch. There are various methods to do this, including e-mail, cellphone messaging, chat rooms, and instant messengers. Instant messengers provide an interesting new linguistic venue, for while the conversations are usually informal, there are subtle differences that have evolved in order to compensate for the format. This paper intends to show how these new rules have been worked into the communication, making for conversations that are both very similar and radically different from ones had in real life.

To do this, I have done an ethnography of communication on ICQ, one of the most popular instant messengers, using Dell Hymes' SPEAKING acronym. Ethnographies of communication are studies of conversational patterns in certain sections of society, usually following along the eight ideas suggested in Hymes' paper. They details not only the norms of communication, but also the societal mores that figure into conversation. This paper will not be a full ethnography of communication; however, its structure is based on the SPEAKING acronym. Each section will be prefaced by one of the letters, a short explanation of it, and then how ICQ conversations relate to that.

II. Review of Literature
In Dell Hymes' article "Models of the Interaction of Language and Social Life," Hymes describes what he feels to be a gap in socio-linguistic research. According to him, researchers either concentrate only on the structure or the social patterns of a language. In order to close this gap, Hymes created the SPEAKING acronym. Each of the eight letters deals with a certain concept that Hymes explains in detail. These eight concepts are: Situations, Participants, Ends (community recognized goals), Acts (substance or sequence), Key (prevailing mood for meaning), Instrumentalities (means by which acts performed), Norms, and Genres. He concludes by restating that both forms and the social supports for those forms must be studied, but like most theories, this one will need testing.
Nancy Bonvillian's chapter on ethnography of communication from her book Language, Culture, and Communication presents a summary of the eight SPEAKING principles. She gives details on each principle, and in some cases, gives examples of what could be included under these principles.

III. Methodology
For this project, I used three sets of logs: one between two males (7Ghent and Starrynight), one between two females (Lady Ariae and Houki), and the last a mixed sex pair (7Ghent and Lady Ariae.) I picked the months February, May, and August from last year because the conversation was consistent in all three, and there was enough of a time difference to track any changes in norms. All three logs were converted from the ICQ database to a text file format; from there, I copied the three relevant months for each set of people to one file, and made notes in a separate file.
The participants at the time of the recorded logs ranged in age from 17 to 21. Starrynight and 7Ghent are both New Mexico residents that attend college at UNM; Starrynight is a sophomore for most of the log, and 7Ghent is still a freshman. They know each other personally, and are best friends. Both are skilled computer users and have used ICQ for over a year; Starrynight uses the Mac version of ICQ and 7Ghent uses the PC version, which may account for some of the glitches in their logs. Starrynight is 21 at the time, and 7Ghent turns twenty during the February log. Houki is a high school junior in Idaho during most of the conversation--in August she becomes a senior--and is 17 all throughout the log. She is also a skilled computer user, started using ICQ in July of 1999, and has been talking with Lady Ariae over ICQ for four months by February. They met on a bulletin board dedicated to Star Wars fanfiction, and began using ICQ to encourage each other's ideas. Lady Ariae is 18 and from Indiana; a college sophomore when the logs begin, she becomes a junior over the course of the summer. Though not as good with computers as the other three, she has been using ICQ since early 1998, and is very familiar with the system. She met 7Ghent in late March of 1999 and the two had been conversing steadily up to the point where the project logs begin.
posted by Lady Ariae 11/18/2001

Saturday, November 17, 2001

So, it's been awhile, but I have done work. Really. There's some good news and bad news, though...
Bad news: The group I set up--the male-male, fem-fem, and then mixed sex pairings in those groups--didn't work. One of the members refused permission on a shared log, so I'm back to just the male-male, fem-fem, and one male-fem pair. This could put me behind, since I had keyed myself towards those groups. Also, my ICQ prog. is down and I can't access other histories for people. Blergh.
Good news: Things are taking shape in my head. The three other people I chose to analyze I already have text logs for, ie logs I exported to text file. Is just four (including me) people enough for this project? I think so...I'm gonna look at three separate months of conversation just for progression...and that's at least 150 pages of material per month. The only really wonky person is Starrynight, and I think it makes a perfect contrast to the rest of us. It's a nice little cross-section, and it's all people I'm familiar with.

Rather than clutter this post, I'm going to go ahead and start writing the Intro, Lit. Review, and Methodology sections (perhaps in that order...perhaps not) on the next post. And just for fun, things i want to get done tonight:
Write said sections.
Read and ANALYZE (I've just been reading so far) the 7-Starrynight log.
Post up the results on that.
Maybe sleep.

'Til later!
posted by Lady Ariae 11/17/2001

Tuesday, October 30, 2001

Research has been done, woo! This meaning that I finally finished the Dell Hymes article in which he lays out the principles of an ethnography of communication, and the whole SPEAKING deal. Starrynight has also laid out the deal on SPEAKING, and I think he made up an excellent example using it, so I'm going to copy it for my paper. With his permission, of course. I'll C/P it here for fun. :)
Ends (community recognized goals)
Acts (substance or sequence)
Key (prevailing mood for meaning)
Instrumentalities (means by which acts performed)
is the acronym. StN's ex is a flame war, and goes like this:

Scenes (physical and situational circumstances): Any non-realtime electronic textual communication; email, bulletin board system, usenet, etc. An argument (usually with no clear outcome to begin with) must have degenerated to accusations and name-calling.
Participants: The disagreeing parties. Not limited to two.
Ends (community-recognized goals): To publicly vent one's frustrations at a faceless "other." Also, flame wars generally signal the end of a given thread of discussion, as most readers become rapidly disinterested.
Acts (Substance or sequence): Starting with an unresolvable difference of opinion (Mac vs. Windows, Playstation2 vs. X-Box, Star Trek vs. Star Wars, etc.) a simple argument can't resolve the issue; the only way to feel good about one's opinion is to sufficiently insult the disagreer (assuming that's a word.) Insults of increasing severity are traded until everybody gets bored and leaves.
Keys (prevailing mood that provides meaning): The whole point of a flame war is to let everybody feel that they've won, so it's a cathartic form of communication, rather than one which has some kind of clear resolution. It's very much like slamming a door as you storm out of a room.
Instrumentalities (means by which the acts is performed): Using a computer, and the most clever insults you can think up, without resorting to speculating as to the target's sexual or political orientation (i.e., calling him a faggot Mac-lover makes you look dumb, and calling him a nazi faggot Mac-lover makes you look even dumber.)
Norms (shoulds and should-nots): Generally it's bad form to clog up a group with a prolonged flame war. Once you've insulted the guy's intelligence and upbringing, it's best to bring the thread to a close and find something else to argue about.
Genres (literary forms): Scathing prose, although if you could compose a really insulting poem, that would probably work, too.

Given that he also uses an electronic form of communication, I think it's a good example to really illustrate SPEAKING online. Or I could just be fooling myself because he's far more brillant than I. Ah vel.

Not yet so concrete paper introduction...


The growing Internet is becoming a dominant force all part of people's lives. It's present in homes, schools, offices, public buildings; there are few places that have not "wired into the Web." Along with the growing Internet is the growing use of chat programs to communicate, such as the Instant Messanger (from AOL, Yahoo!, or various other places) or a program like ICQ, which uses the letter message format. These chat programs allow for conversation across distances without large phone bills or the lag time of e-mail, which makes it feel exactly like a true person-to-person conversation. Because of this, ICQ logs read like transcripts of conversation, and so often follow the norms of conversation; but, due to the electronic medium, have differencies to compensate for distance and lack of emotional reading.

I'm gonna stop there, 'cause I DON'T LIKE IT. Minus the title. Will work on this LATER. Jya ne!
posted by Lady Ariae 10/30/2001

Sunday, October 28, 2001

Dear Lord, I've gotten myself into quite a project here. I just read 432 pages of log between Ghent and Starrynight. I have 300+ pages of log to read for me and Ghent/Ki, and that's JUST the month of August. (I cut down to just a month for users I'm familiar with; with Ghent and StN, I had to read the whole thing up to that month. Kinda voyeristic this project is, too...*g*) If my ICQ program EVER cooperates, I'm going to get the logs pasted into the file for Korana and Kaizoku. I may have to do the whole thing by hand, especially if I want to highlight or play with colors. BUt I have some observations down:

  • Between Starrynight and Ghent, Starrynight is the main controller of topic. Occasionally, if the conversation is something Ghent knows more about, he'll change the topic, but Starrynight is the initiator of most topic changes by words. Ghent tends to initate topic changes or topics by presenting a link to something and inviting comment on it. Starrynight tends to make "out of the blue" topic changes and then invite comment on it; call it an abstract (thought based) topic control, rather than concrete (object based) topic control. Either way, Starrynight still dominates.
  • In the beginning, these two rarely exchanged a greeting, and almost never a good-bye. Towards the middle and end, they had a set "heya" or "how goes?" to greet, and then "jya ne" for a good-bye. Still, a good portion of the conversations had no good-bye phrase. They just stopped. I attribute this to Starrynight, since if a good-bye was said, it was usually initiated by Ghent.
  • Conversation topics center around social life, computers, and Everything2, an online database of knowledge.
  • Starrynight also uses emoticons more, but he uses the Eastern set. When Ghent uses emoticons, he uses the Western set, letter form.
    (EX: The Eastern set is based on Japanese Internet users--their "smilies" are all rightside up and display no teeth. For example, a simple smile is :) Western and ^_^ Eastern. Letter form of the Western set means that they uses *g* or *grin* or *smile* or whatever for emoticons and actions. Most likely, I'm the influence for that, at least on Ghent.)
  • For a contrast on topic control, between Ki and I, Ki controls the topic...but in a very segue-ish manner, in the sense that she almost ALWAYS asks, "So, what about...?" or "So, are you up for...?" and then we talk about it. We also have, standard to our dialogue, greetings and good-byes.

The working title for this project, BTW, is "SPEAKING on ICQ" after the general rule of thumb for doing an ethnography. I found it, oddly enough, on Starrynight's weblog for his communications class. Must look over actual academic sources for this again, but like with other projects, having trouble getting past "pilot study" stuff and thinking abotu related research. Yes, Susan, you can say it outloud--I'm a flake. :)

You know, considering that Starrynight is such an interest, I might as well drop Kaizoku and use Starrynight, so to study his reactions to me. Or I could just use Starrynight and drop everyone else. *bangs head against desk* It's TOO LATE! to REVISE! my DAMN PROJECT! GAH! No. I'll just stay with my four people, and have fun highlighting how he seems to break a LOT of the norms. Yea. That's it!

BTW, info on people, from AT THE TIME the logs were being written to my comp.
Ki: Female, high school senior. Met on a fiction forum. Spoken to on ICQ from November of '99 on.
Ghent: Male, third semester college student. Met in March of '99, through ICQ introduction by a mutual friend. Have met in person over the summer.
Korana: Female pharamacy student, year unsure. (4th, maybe?). Knew from home, keep in touch through ICQ.
Starrynight: Male, college student, 21. Asian studies major.
Ariae: me. Female college student, third semester of college.

I'll add more later, along with my paper introduction, but it's late. Jya ne, little log!

*suddenly scary thought--what if Susan is doing research on the way students use logs for a class? As in, how they address the log, what forms of the language they use...hmm. Now that would be interesting. BUt I'll stop speculating now...
posted by Lady Ariae 10/28/2001

Powered by Blogger


A research and gripe journal for my Anth 401 LPG project.