courses :: mas 963   the virtual society :: judith donath   graphical conversations The two graphical environments I choose to look at were OnChat and ActiveWorlds. I had never used this type of portal for communicating before and I'm not sure if I'll be using it too much again in the future. Whilst playing around in both of these environments I encountered inane "conversations" which seemed to solely consist of the banalist of greetings and little else. Although I had some fun moving around in the space - flying in particular in the case of ActiveWorlds - this wasn't sufficiently distracting to make up for the disappointing conversation... Describe the differences between the graphical and text-based environments you have observed. The text-based newsgroups I discussed last week displayed a richness in conversational content that I found wholly lacking in the graphical environments. Constrained by the fact that the text in the graphical environments I studied was displayed above your avatar's head, meaningful exchanges generally could only consist of a phrase or short sentence which I found rather confining. What are the entrance requirements to the graphical sites? OnChat is launched from the OnChat homepage as a java applet. As a guest user you simply choose a nickname and then select the room where you wish to chat. You are then represented by the most basic avatar, in this case a smiley face, which obviously gives off clues as to your newbie status. If you register(which is free) you have the option of customising your avatar, creating your own webpage and are also given permission to build your own chatworld. To communicate using ActiveWorlds, you must first download the necessary software client and install it on your machine. When you launch the browser you must choose a nickname or handle in order to obtain your tourist visa. You must pay($19.95) if you wish to become a citizen of ActiveWorlds, a qualification that allows you to choose your own avatar, build your own worlds and send files as well as some other basic features. Are there specifically graphical forms of communicative competence? Maintaining an appropriate spatial distance from other avatars is not only respectful of common real-world social mores regarding 'personal space', but is also often a necessity in faciliating the reading of textual messages. I found it somewhat discomforting when people moved through my body in ActiveWorlds and in one instance, one avatar kept trying to stroke people which was greeted with both appreciation and annoyance in almost equal measures. From my experience as a newbie, the ability to move around in an apparantly purposeful manner and not through buildings or at random angles signaled proficiency in the environment and made people more inclined to talk to you. Certain avatars were able to zip around a la a social butterfly at a party which made them seem like old hands at this graphical conversation experience. How are the graphics used to establish identity? to interact and communicate? As an unregistered citizen in both of these environments, my graphical default avatar merely signaled that I was new to this game. People were willing to help me and explain how to move/fly/speak in whispers as a result of my very basic avatar. Most people in OnChat seemed to use one of the cool hip-hop avatars from the database of avatars available, whilst a few used pictures of their favorite rockstars/moviestars as their graphical representation. Your nickname appears above(ActiveWorlds) or below(OnChat) your avatar in a standard default font. In OnChat you could choose to attach a number of graphic symbols/emoticons to your avatar, such as lovehearts, exclamation marks, or snooze 'z's' which could be used either to emphasize a specific point you were making, or could be used throughout your conversations to indicate you were looking for love, feeling rather angry or feeling extraordinarily happy(two smiley faces). As a member in OnChat you also had the option to spraypaint tags on walls or leave messages in books. You could also change the color of your text bubble/text which could add some rudimentary feeling of individuality to your avatar by making it stand out more. There was also an option to emphasize or deemphasize your chat message, by using the think or shout buttons which would display your message differently. Some 'cool' avatars in ActiveWorlds arrived into the environment on specially made cars which helped establish them as experienced users to everyone else in the environment. What is the role of the background? The background in both of these environments is fairly rudimentary, particularly in the 2D OnChat where your avatar always sits on top of objects and cannot move behind them. In Active Worlds you get some sense of depth and distance owing to its 3D nature and the relative solidity of most of the objects in the worlds. There are numerous areas of the background reserved either as 'windows' into other worlds/environments or for advertising links. How is motion used? In OnChat, motion is simply a matter of clicking on the background space where you want your avatar to be. It is not especially smooth and there is little subtlety to it. In Active Worlds, you have more degrees of freedom regarding how you manipulate your avatar, such as being able to turn to face specific directions/people, utilize certain pre-canned actions(wave/maca/dance/happy) and best of all you can fly up and down which is most enjoyable. What do you think works well in the systems you observed and what does not? I think both of these hybrid environments, whilst offering a certain level of playfulness via the graphical elements, overall provide for a disappointing experience. Perhaps I am just too used to the overtly wordy ways of newsgroups and other forms of textual online conversation, but being confined to such short sentences made for rather boring interaction. I found it somewhat distracting trying to follow conversations using the word bubbles over the avatars heads, particularly if everyone was standing near each other, thus making the words appear on top of each other. As an alternative, you can follow the flow of conversation on the chat log, but that includes all text being contributed in the environment which jars with received notions of spatial proximity. ie. I can 'hear' people although I have no idea where they are. It can also be rather annoying, as in one instance I was waiting for a reply from another person whilst monitoring the chat log, only to discover after quite a while that they had simply walked away from me, and as they hadn't actually said they were leaving, I hadn't noticed. How would you design a better graphical environment? Would you use representational or abstract graphics? How would the space be used - would different areas serve different roles or functions? How would they be delineated? Kurlander et al mention the importance of history in chat programs, in their description of Comic Chat. The temporal nature of the chat(speech bubbles over the avatar's head) exemplified by the two environments outlined above means that the only evidence left behind from a conversation is a textual chat log. Kurlander explains how "since there is no graphical history of the conversation, it is difficult to share the experience with non-participants, or to review the chat at a later time." This is a moot point, although judging from the kind of conversations that took place whilst I was reviewing these environments, I would be at a bit of a loss as to why you would want to keep records of what was actually said. Nonetheless, I think I would find it intriguing to be part of a conversation that could be instantiated and represented in the contextually rich way as described by this paper. In their conclusions, Kurlander mentions how he would like to allow the system to experiment more with layout. I think this is a fundamental point. For example, someone who starts a conversation or who launches a new topic of discussion could be framed differently as in: Perhaps another way of representing the contributor dynamic of a conversation environment might be to expand and contract frames depending on the amount of people present: Disagreement/contention or desire to leave a conversation could be playfully depicted by destroying the boundary frame or breaking through it. The use of acronyms, and in particular "lol", is very common in the graphical environments described above. This form of phatic communication might be nicely represented by keeping the protagonist/avatar off screen, but putting their words in the frame. If they want to appear in a visual sequence, they must contribute something more constructive. They same system could be used introducing new characters. Instead of thrusting them into the visual environment just for saying "hello", they might be initially introduced using text and then giving a visual presence if others respond or if they contribute further. readings: rossney, R. 1996. Metaworlds. Wired 4.06 (June 1996). durlander, D., T. Skelly and D. Salesin. 1996. .Comic Chat In Proceedings of ACM SIGGRAPH. diegas, F. and J. Donath. 1999. Chat Circles. In Proceedings of CHI 99. eisner, Will. Comics and Sequential Art. Selections from chapter 5.   back