A4: I, We, You

From reading the Monk et al piece and our discussion in class I found myself very weary of some of the claims that were made. In particular, the claim that the use of first and second person pronouns renders queries from direct to indirect. Further, the claim that their use makes a statement more polite. I decided to take a look at a number of email conversations I've had with different people: my mom, my boss, and a potential landlord. The email conversations with my mom contained very few first person pronouns but a decent amount of second person pronouns. This makes sense because my mother and I are quite straightforward with one another and politeness is not necessary. The email conversation with my boss was quite different. I found that I used a lot of first and second person pronouns but he barely used first person and a few second person pronouns. The relationship between a boss and employee also supports Monk et al in this case. Finally, an email conversation with a potential landlord as I have been trying to find a summer sublet in New York City was examined. In this, I found a great amount of both first and second person pronouns. I, as a potential tenant was trying to manage my impression in order to have a good chance at landing the apartment. He, on the other hand, wanted to make sure to attract as many people to his listing to be able to be more choosey. My findings actually ended up supporting Monk et al's claims but I do not believe there is a strict causality here. There is great possibility for context to completely undermine this claim, as well as tone. To conclude, I found that this exercise helped me think a bit more about correlation and research claims that may or may not be hiding behind a lurking variable.

Conversation Structure: Why its important....

Monk et al. “Measure of Processes” provides one with an outlay of ideas pertinent to understanding conversation amongst partners. The paper provides insight into how dialogue may be hindered or facilitated by perceived common ground or any system that structures conversation. CMC doesn’t provide a bevy of communication cues for individuals to rely on, so at times it may be difficult for individuals to relay everything they have to say to their partners. We’ve all experienced times when we’ve been cut off by friends while on AOL instant messenger. You may be typing a phrase that is part of a string of sentences, only to be interrupted by your partner who wants to chime in about whatever is going on inside his mind. Measure of Processes proclaimed that the conversation structure of an online dialogue is much less structured than that of a face to face conversation. Its not unusual to have multiple online conversations, a rarity in face to face interactions. Face to face interaction usually involve a sequence of topics, one after another. This leads to Monk’s hypothesis of “Topic Mention”; its difficult to recall information about the previous topic one was talking about in face to face situations compared to the ease of remembering a topic via online conversation. Monk believed that this would hold true if one observed the idea of distance between topic mentioning in CMC vs FTF. I decided to test this theory using an aim conversation and a FTF conversation.

Hypothesis: More topics covered in two minutes of online conversation than in FTF conversation ( I will use an old aim conversation and FTF conversation from a past assignment)

IM Conversation with a Friend

stillBETTAthanYa: sup
Prophecy620: wut up man
stillBETTAthanYa: ur icon is stupid
stillBETTAthanYa: lol
Prophecy620: w.e man
Prophecy620: whats godo wit u
Prophecy620: do u know if yo parents got the card i sent a while back)
stillBETTAthanYa: o yea they got it
stillBETTAthanYa: they said thanks but u didnt need to send a gift
stillBETTAthanYa: all good
Prophecy620: ight
stillBETTAthanYa: how is school
Prophecy620: crazyy
Prophecy620: dis semester is hard
Prophecy620: wit 5 calsses n a discussion
stillBETTAthanYa: discussion?
Prophecy620: its paort o a history class
Prophecy620: the calsses r hard though
Prophecy620: my 4000 class is probably the easist rite now
Prophecy620: how many classes u taking this semester
stillBETTAthanYa: I have eight
Prophecy620: got damn
stillBETTAthanYa: craziness
Prophecy620: hell yea
stillBETTAthanYa: three evening classes
Prophecy620: least u aint workin 32 hours a week
stillBETTAthanYa: your working 32 hours for radioshack?
stillBETTAthanYa: or for class?
Prophecy620: radioshack
Prophecy620: i was during skool
Prophecy620: now its 26 i cut it down
Prophecy620: i was full time
stillBETTAthanYa: tough stuff
stillBETTAthanYa: I work but only for about five hours a week
stillBETTAthanYa: do homework mostly
Prophecy620: dats it
Prophecy620: oo ight
Prophecy620: so its not dat bad then
stillBETTAthanYa: yea
Prophecy620: but still 8 classes
stillBETTAthanYa: i usually do homework for 4 and a half of those hours
stillBETTAthanYa: not all are hard work
Prophecy620: i think u nee to have that cleared by the dean dont you
stillBETTAthanYa: yup
stillBETTAthanYa: had to have its signed
Prophecy620: wow
Prophecy620: u killin yo self kene
Prophecy620: how u do good in all the classes
stillBETTAthanYa: i'll handle it
Prophecy620: ight bet
Prophecy620: im waitin for jay 2 give me a date on the wedding
Prophecy620: so i can get the tickets to fly back
stillBETTAthanYa: cool
stillBETTAthanYa: should be nice
stillBETTAthanYa: u hear the chargers fired schottenheimer?
Prophecy 620: Yup

Topics covered: four

FTF Conversation

Kene: What’s Up Matt?
Matt: I’m doing fine
Kene: How is Priscilla?
Matt: Priscilla the dog or Priscilla my cousin?
Kene: Your dog, you know I don’t like your cousin.
Matt: She’s doing fine, finally house trained
Kene: Ok that’s cool. How many classes do you have this semester?
Matt: I got 6, 18 credits. Trying to get up to speed for Bio
Kene: Ok. Well I’ve gotta go. Get at me later.
Matt: Yea, I will. Saturday 10pm still on?
Kene: Yea, I’ll be ready for the madden tournament. Say hi to the dog for me.
Matt: Will do. Later.

Topics covered: Two

It makes sense that more topics would be covered in an online conversation than in a FTF conversation. CMC is often recorded and its easy to keep track of information this is being exchanged between partners. FTF conversations consists of less topics because its hard to keep track of all the information that has been exchanged. Understanding this notion can allow one to determine what type of messages are best exchanged over which mediums.

Conversation Structure: Topic Mention Distances

In Monk’s article, there is a discussion about Conversation Structure, and more specifically, Topic Mention. According to the article, McCarthy (1993) noted that text-based communication had much less order than face-to-face spoken conversation. Specifically, new topics would begin before the last topic was resolved, causing disorder to text-based communication. McCarthy and Monk suggest that text-based conversation can sustain multiple topics at a time because there is a written account to review past conversation, whereas in face-to-face, there is no record. Once spoken, a face-to-face comment is rarely saved. Thus, recalling one topic’s comments while engaging in another topic simultaneously can be confusing in a face-to-face context. To measure orderliness, McCarthy and Monk measured the distances between references to the same topic. For my assignment, I decided to analyze two brief conversations - text-based communication and face-to-face conversations – and compare for topic distance differences. I hypothesized that, if McCarthy and Monk’s measure is correct, I should see greater differences between topic references in text-based communication compared to face-to-face communication. It was difficult finding technology-media examples, so I chose to examine AIM and face-to face conversations discussing technology.

Text-Based Conversation

1. Person A: He hasn’t been on ICQ this week, do you know why?
2. Person B: I wouldn’t know, I only use AIM
3. A: I think he’s avoiding me
4. B: Why would he do that?
5. A: Well, he hasn’t answered my text either
6. B: People still use ICQ?
7. A: Yeah, some people still use ICQ…out of the US
8. B: Why would he be avoiding you?

Lines 1,3,4,5, and 8 deal with one topic, while a second topic, using AIM vs ICQ is seem in 2, 6,and 7. This is a large distance from 2 to 6 – large than in the face-to-face conversation. This supports McCarthy and Monk’s idea that distances will be greater in text-based conversation

Face-to-Face Conversation

1. Person A: Have you ever heard of Second Life?
2. Person B: No, what’s that?
3. A: It’s a virtual community with avatars…similar to a game except you can’t win

(laughter)

4. B: How’d you get into that?
5. A: I didn’t really…I heard about it though a class and I ended up doing a project on it
6. B: I have so many projects this semester
7. A: how many?
8. B: more than you can imagine

Lines 1,2,3,4, and 5 deal with one topic, and then the conversation shifts in 7 and 8 to a new topic and continue from there. This supports McCarthy and Monk for two reasons. First, it confirms my hypothesis that distance between topics is greater in text-based compared to face-to-face communication. Additionally, this conversation exemplifies McCarthy and Monk’s idea that face-to-face conversations have my topic organization, meaning one topic is discussed, and when it is resolved, a new topic begins

Assignment #4: Fun with Chatterbots

For my assignment, I decided to do analyze transcripts from the Loebner Prize in Artificial Intelligence’s homepage. The Loebner Prize is a competition held each year where computer programs aiming to impersonate a human in chat are pit against humans whose job it is to judge whether or not the entity they are conversing with in an instant messenger interface is a human confederate a computer contestant. A quick perusal of the transcripts at http://loebner.net/Prizef/2005_Contest/Transcripts.html from the 2005 contest reveals that most of them, well, suck. However, I wanted to find out whether or not we could quantify the results; looking at the Monk piece, it is suggested that co-referring expressions such as “it,” “that,” “they,” “he,” and “she” are evidence that the two speakers believe they have achieved common ground. Although the paper described them as being generally infrequent in text-based communication, I wanted to see whether or not they would appear more often in dialogue with a human confederate or with Jabberwacky, the 2005 contest winner (its transcripts are at the bottom of the page linked to above). My hypothesis was that because the computers would be unable to achieve common ground with the human judges and are not smart enough to stay on-topic across a chain of messages, there would be fewer co-referring expressions in the transcripts of the judge-computer conversations than in the judge-confederate conversations.

My hypothesis was actually wrong:

	Confederate		Jabberwacky
it	2		7
that	3		6
they	0		4
he	0		1
she	2		0
total	7		18
	Confederate		Jabberwacky
it	16		3
that	6		6
they	1		0
he	0		2
she	0		0
total	23		11
	Confederate		Jabberwacky
it	5		4
that	1		8
they	0		0
he	0		0
she	0		0
total	6		12
	Confederate		Jabberwacky
it	3		9
that	5		8
they	1		0
he	0		0
she	0		0
total	9		17
	Confederate		Jabberwacky
Average total	11.25		14.5

There are a few explanations for this result; for one, I suspect that the creator of Jabberwacky programmed it to use co-referring expressions precisely for the reason I formed my hypothesis. In order for the computer to stand a chance of seeming like a human, it had to be able to refer to previously discussed topics in a natural way, so it’s possible that Jabberwacky is programmed to use them in an almost exaggerated way. Also, the human judges seemed to become confused by Jabberwacky because of its bizarre sentences and would ask things about what it had just said. Finally, most of the conversations with the confederates are shorter; the judges realized they were speaking to a human and decided to converse more with the other entity to study it.

However, I found it interesting that there was a huge spike in the use of “it” in the second conversation between a judge and a confederate; looking at the transcript, they became interested in talking about the copyright policies of the RIAA and that fueled the rest of the discussion. This is something that could not possibly have happened with Jabberwacky; never does the frequency count of any single word in the Jabberwacky transcript rise above 9. Again, I suspect this is because Jabberwacky cannot stay on-topic.

Assignment #4

The readings for this week focus on the methods of looking at language in conversation. Everyone should read the Monk et al. (1996) piece on process and outcome variables. If you will be transcribing language at all (transcribing audio or video) then you should read Bavelas et al. piece (it is really a lovely summary on how to approach language). Finally, if you are interested in how track 2 type analyses work, read the Hancock and Dunham piece.

This week's assignment is about practicing some analysis and measurement techniques. You should draw on a measurement issue from one of the 3 readings, and then do some kind of language analysis on technology-related media. It will be fun to look at how technology affects language used in media coverage of tech topics. The options are wide-open. For example, you could analyze how often “technology” or the actual technology (video games, TV, movie, etc.) is mentioned in the piece you are analyzing. You could analyze 1 minute of an interview for positive or negative face-related language. You could do an analysis of how media use some language processes in reports on technology-related subjects. You could analyze interviews for track 2 devices (e.g., does the collocation of the interviewer and interviewee matter? would the background noise (studio interviews vs. field interviews) lead to increased track 2 signals at level 1?). The main point is to take some small idea or question, and then measure it in the context of the media coverage of technology.

Your post, therefore, should have a brief summary of the question, a brief description of the method of measurement you used (drawing on one of the readings), and a brief report of your findings. Have fun with it; the point is to get you to practice analyzing real language.

Common Ground and Trust in CMC

Much analysis has already taken place in the low-cues CMC environment, particularly about how any sort of relationship can form in such a medium, but more study needs yet to be done on how language itself plays a role in establishing those relationships. As a small amount of trust and familiarity is gained in any medium, language changes drastically and in a variety of ways, and that very language change may prompt more trust to be extended. Barrett and I are researching common ground and trust in a CMC environment, preferably IM. We aim to see if a difference exists in language use when a subject has a minimal level of trust for another and when they don’t, and we predict that the difference between the two conditions is simply common ground.We have two potential ideas for the experiment. First, we could hint or overtly state to a subject that the other either should or should not be trusted, giving them a legitimate reason why. If both parties are working to get to know each other, for example, each will extend a certain amount of trust. If a seed of doubt has been planted that the other’s objective may be different or even contrary to the subject’s, they may act differently – more guarded? – toward the other. Alternately, we could set up two conditions. In one condition, two students meet briefly in person before being escorted to separate CMC rooms to talk online. The common ground: they’re both students, they’re both participants, and they both are probably doing the experiment for the same reason…credit. This is immediately established and a small amount of trust is extended. The other condition is where one person is led alone into a CMC conversation with someone they haven’t met, either a confederate or another subject. At the outset, they’re unsure if the other is a fellow student subject, a confederate experimenter, or even some unknown third party. Without the original condition’s common ground, no trust is initially extended. Language used between the two conditions may change drastically. One possible change in language (though this is far from a working hypothesis) is that explicit signaling may become implicit as trust is gained. For example, in a non-trust condition, a subject may say “Shall we begin?” or “So we’re supposed to get to know each other,” both signals for the conversation to proceed that trusting individuals probably aren’t as likely to use. This is only one example of how language may vary between the two conditions, and many more will be considered before a hypothesis is decided upon and an experiment designed.

Barrett Amos
Sam Warren

Influence of Gender and Topic on Language

Our group is interested in examining gender roles in online communication.
We will study the differences between communication between same-sex
pairs and opposite-sex pairs, and between male and female language use.

Hypothesis: The occurrence of idioms specific to computer mediated
communication in CMC settings is correlated with the gender of the
communication participants and the 'seriousness' of the topic of
conversation.

We are researching the different use of Language between men and women
when using the computer as the communication medium. We feel that people
tend to deviate in their language use when interacting with others via the
internet. For example, when people speak with friends and peers over the
internet, there tends to be more grammatical errors and increased use of
"Internet-isms", such as emoticons and abbreviations, in the message.
Through getting a group of people to interact via a chat room or through
AIM, we will try to understand if our hypothesis holds and what factors
will affect it. Similarly, we will be analyzing how gender affects the
type of language use over the internet. In order to get unbiased results,
we will pick a specific topic and have our participants engage in online
discourse. This will help us understand if gender is analogous to specific
types of language use. Also, to reaffirm our hypothesis, we will use two
discussion topics to see if the seriousness of a discussion alters the
amount of internet-isms used and if it deviates between genders.

Danny Duran
Henry Mason
Sarah Perkins
Greg Vixama