Home / Articles / Volume 8 (2011) / Conversation in a multimodal 3D virtual environment
Document Actions

 


Introduction

In conversation analysis, the main focus is talk-in-interaction—language in use, rather than the language system per se. However, when language use is mediated by computer technology, ‘system properties’ take on new significance. Technical constraints can affect communication, although they do not predetermine it: “While the case for the deterministic influence of the computer medium on language use is often overstated, properties of computer messaging systems nonetheless play a significant role in shaping [computer-mediated discourse]” (Herring, 2001, p. 614). This study explores the semiotic potential inherent in the visual properties of the computer medium through an example of a virtual world.

The presentation of a virtual world through graphical means is an example of meaning representation through semiotic systems other than language, namely three-dimensional (3D) computer graphics. 3D graphics are a virtual world’s visual representation, in which communication and other activities involving production and negotiation of meaning are carried out. Presentation of self—appearance, gestures, etc.—is accomplished by avatars, or cartoon-like characters, in graphical worlds. However, just because verbal language is not the chief mode of communication does not mean that graphical worlds are of no interest to discourse analysts. By adjusting one’s analytical perspective, one can find meaning potential outside the narrow scope of language as it is traditionally conceived (Kress & van Leeuwen, 1996, 2000, 2001). The examination of textual conversation in combination with graphics and sound in this study will demonstrate such meaning potentials with respect to a 3D virtual world.

Specifically, this article closely examines a conversation in which guest users and experienced users talk about the differences in appearance and privileges of users—all visual differences that have an effect on communication. Also considered is an instance of narrative play, a common phenomenon in computer-mediated discourse (e.g., Cherny, 1999; Kolko, 1995; Virtanen, in press). Narrative play in this 3D virtual world is accomplished through textual and visual (graphical) means of communication. Graphical objects such as avatars, imaginary objects (or imaginary modifications of graphical objects) represented by textual means, and textual narration that makes reference to graphical reality, virtual reality and reality itself, together constitute what is here introduced as multimodal narrative play.

It is argued that 3D virtual environments—in addition to being amenable to linguistic and ethnographic analyses—consist of collaborative, constantly developing, modes of communication in addition to language. In such socio-cultural settings, the researcher may look for institutionalized uses of the multimodal sign systems to gain knowledge about how and why people in a particular social setting communicate with and understand one another. However, merely emphasizing the meaning potential that resides in the virtual world is not enough. It is necessary to broaden one’s analytical scope and look for a framework in which the use and evaluation of language (and other symbol systems) and the social nature of language are incorporated. Dialogism, as presented in Linell (1998), is one such framework.

In dialogism, the social nature of language is foregrounded. Communication is viewed as a collaborative process in which readers and listeners play important roles, taking an active part in the communication. Terms such as coordination, cooperation, and negotiation are used to describe the social character of communication processes, emphasizing an orientation towards the “other.” Other-orientation in computer-mediated discourse (CMD) is probably no more frequent than in oral conversation. However, other-orientation in CMD is more explicit. Users often address each other in synchronous chat environments such as virtual worlds by initiating messages with a nickname followed by a colon; this feature is called addressivity (Werry, 1996).

Addressivity in textual CMD has been described as a strategy that helps compensate for the lack of visual cues that establish inter-turn coherence in face-to-face communication (Herring, 1999). This study looks at how addressivity is managed when nonverbal visual means of communication are also available. The analysis of a 3D virtual world known as Patagonia reveals that placing avatars representing interlocutors in proximity to each other simulating a face-to-face encounter is preferred to (typed) nickname + colon, suggesting an inclination towards a visual addressivity similar to that in oral conversation.

“Welcome to Patagonia—the Scandinavian World”

The Microsoft Windows based ActiveWorlds,1 developed in 1995 by Worlds Inc., is a 3D graphical environment that allows users to participate in recreational social interaction and to surf the Internet simultaneously.2 Like MUDs and MOOs which are textual virtual worlds (cf. Cherny, 1999), ActiveWorlds also supports 3D building: Users are allowed to design their own “virtual homes” by “building” on virtual properties. At the time of this study, one of the largest and most popular worlds in ActiveWorlds was AlphaWorld (AW), established in 1995. Norwegians started up their own world, Patagonia (Pata), in 1997.

Figure 1 shows the ActiveWorlds interface for Patagonia as it was displayed on screen at the time of this study. There are three main windows: a graphics window, a chat window, and a browser window. Interlocutors are represented in the graphical environment as avatars. Since the ActiveWorlds software is shared by all worlds, system features are in principle the same for Patagonia as for other worlds, although virtual geography—graphic design—and the assortment of avatars may vary. ActiveWorlds enable sound files, which are usually activated by the system when an avatar is detected within a given range or by a user with privileges to play music.

Figure 1. ActiveWorlds interface: Patagonia Ground Zero

The Ground Zero, or point of entry into Patagonia, is designed as a (town) square surrounded by signs. From Ground Zero, four “roads” lead to the peripheral geography. There are three basic ways of moving around or transferring to other places: walking, flying, and teleporting. Teleporting is mainly performed by inserting coordinates into a teleport prompt, by (mouse-)clicking a pre-programmed teleport object (such as a sign), or by joining other registered users through a join function. Some worlds are programmed to support the perception of gravity. In Patagonia, if an avatar walks off the roof of a building, for example, it will fall down to the ground. However, when in flying mode, avatars are usually not influenced by gravity. Online users often place their avatars in opposite corners of the square of Ground Zero and in the airspace above it. Moreover, regular users tend to establish regular spots for their avatars at Ground Zero.

Guests and Citizens of Patagonia

Registered users of the ActiveWorlds are called citizens, and the registration process is called immigration. Registered users tend to be a relatively stable group, because they have paid for a registration license.3 Unregistered users gain tourist user status for free. Several guests have frequented Patagonia over the years, but only a handful of the original crowd from the world opening in 1997 remained two years later. Registered users consequently comprise the majority group present. Patagonia presents itself as a Scandinavian world. Some non-regular users or tourists in Patagonia speak English or Spanish, but most conversations take place in Norwegian or other Scandinavian languages, such as Danish or Swedish.

From photographs of people from real life gatherings in the Patagonia Gallery, it is possible to draw some conclusions with regard to people’s age and gender. According to one estimate, the age average is approximately 27 years (Hansen, 2001), although the citizens of Patagonia range in age roughly from adolescence to 60. Based upon her experience in Patagonia, the author believes that the choice of avatars corresponds with gender in real life, and that there is a fairly equal distribution between the genders. However, from time to time some users seem to prefer gender neutral avatars, such as an alien or a bird. Moreover, the default avatar for guest users is a male avatar, and thus it may appear that the majority of tourists are male.

A handful of people with full-featured rights reside at the top of the social hierarchy. In MUDs and MOOs, system administrators are called gods or wizards (Cherny, 1999, p. 46); in ActiveWorlds, they are known as caretakers, peacekeepers, or world owner(s). This status provides them with special privileges, such as the ability to alter virtual property anywhere, to run bots, and to use boldface writing. In Patagonia, there were social conventions (and technical limitations) restricting the use of boldface writing to speeches and selected online events. At the time of this study in 1999, only public speakers (or peacekeepers) could employ bold black fonts (they also had the choice to use normal black writing).

Analytic Approach

While Conversation Analysis is based in oral face-to-face conversation, computer-mediated discourse analysis has so far evolved on the basis of text-only systems. However, communication in the ActiveWorlds is neither oral face-to-face nor text-only. Research in 3D virtual environments thus calls into question the theoretical bases of CMD methods. The next section describes work that has been crucial for the present research.

Computer-Mediated Discourse

A computer program may serve as a foundation, but it has no particular interest for the discourse analyst unless it is utilized by people in a socio-cultural context. As Linell (1998) writes:

Social construction and reconstruction do not take place in vacuo, but are (inter)dependent with/on affordances from the material environment, and with/on the socio-economic conditions and the socio-cultural traditions within which they are embedded. (p. 273)

One approach that considers socio-cultural context in computer-mediated communication (CMC) is computer-mediated discourse analysis (CMDA), defined by Herring (2001) as:

a specialization within the broader interdisciplinary study of computer-mediated communication (CMC), distinguished by its focus on language and language use in computer networked environments, and by its use of methods of discourse analysis to address that focus. (p. 612)

CMDA emphasizes the importance of taking into account contextual factors, which Herring (2007) characterizes as being of two types: medium variables, which have to do with system features such as synchronicity and available channels of communication, and situational variables, which have to do with features of the social context such as user demographics, purpose, and topic of discussion. A situational variable relevant to the present context is the cultural knowledge that users have of the offline world (for example, relating to their national culture, or their membership in particular sub-cultures) and in the online environment (for example, that distinguishes experienced users from new users).

The CMDA approach is illustrated by Cherny’s (1999) comprehensive ethnography of a virtual world, a text-based social MOO4 populated mostly by computer programmers living in the United States. Cherny studied the MOO for two years as a participant observer, focusing on the linguistic register, or “the special linguistic characteristics of speech in the MOO…” (1999, p. 27).

Cherny’s observations are useful when comparing the ActiveWorlds with text-based MOOs. According to Cherny, knowledge of the typing of certain commands is required in order to communicate in a MOO. These commands include different ways of “speaking,” as well as ways of performing textual actions. Cherny also notes the existence of cooperative behavior known as “byplay” in her social MOO—interaction with objects and imaginary objects, along with reference to different reality strata in play. All objects in the MOO, programmed and imagined, are purely textual.

Cherny’s social MOO, like Patagonia, is a real-time social chat environment with a regular community of participants organized around a geographical theme. In the case of the social MOO, the theme is a house, and basic navigation from room to room involves typing commands such as “go north” and “open door” (although it is also possible to teleport from room to room) and then typing “look” to get a textual description of where one is. We will see that Patagonia preserves a number of features of textual MOOs, while adding to them graphical semiotic resources to create an environment that is visually much richer.

Multimodality

The analytical approach in the present work is inspired by the work on multimodality of scholars such as Kress and Leeuwen (1996). According to Kress, to be multimodal is “to be constituted by a number of modes of representation” (2000, p. 184). In his chapter on multimodality, Kress (2000, p. 183) points out how modes of communication such as visual and music are “untheorised, or at least undertheorised,” in communication theory. Kress further advocates the multimodality of all texts, and of language itself. For example, Kress and van Leeuwen (1996, p. 39) argue that “the age of multimedia” can make way for perceiving what they call “the multimodality of written texts”:

A spoken text is not just verbal but also visual, combining with ‘non-verbal’ modes of communication such as facial expression, gesture, posture and other forms of self-presentation. A written text, similarly, involves more than language: [I]t is written on something, on some material and it is written with something with letters formed in systems influenced by aestethic, psychological, pragmatic and other considerations; and with a layout imposed on the material substance, whether on the page, the computer screen or a polished brass plaque. (p. 39)

In this article, we shall see that what is communicated by textual means in MOOs and other forms of text-only CMC can be mediated graphically in the ActiveWorlds. In Patagonia, for example, text color is assigned to certain types of messages. Color information is crucial in order to understand what kind of message it is, who sent it, and who is able to read it. Thus a shift in analytical focus from text content to text color is required to recognize the meaning potential of the visual appearance of such messages. Previous research on text-only CMD has paid little attention to the question of visual communication aside from the acknowledgement of text-based visual elements, such as emoticons and ASCII character play (e.g., Danet, 2001; Danet, Ruedenberg-Wright, & Rosenbaum-Tamari, 1997; Dresner & Herring, 2010), although this is starting to change in recent times.5 Methodological frustration can result when looking at communication in a multimodal interface from a framework founded in studies of text-only systems. What is needed in the case of ActiveWorlds is an approach that takes account of multimodality.

Social Semiotics of Visual Communication

The framework of social semiotics of visual communication as introduced by Kress and van Leeuwen (1996) and by Jewitt and Oyama (2001) is proposed here as a framework to supplement existing discourse studies of computer-mediated communication. According to Jewitt and Oyama (2001, p. 134), “social semiotics of visual communication involves the description of semiotic resources, what can be said and done with images (and other visual means of communication) and how the things people say and do with images can be interpreted).”

Whereas in 1996, Kress and van Leeuwen take the view that language and visual communication “both realize the same more fundamental and far-reaching systems of meaning that constitute our cultures, but […] each does so by means of its own specific forms, and independently” (p. 14), by 2001, they perceive that a “semiotic shift” has taken place, such that the different systems of meaning have become interdependent:

[I]n the age of digitization, the different modes have technically become the same at some level of representation, and they can be operated by one multi-skilled person, using one interface, one mode of physical manipulation, so that he or she can ask, at every point: ‘Shall I express this with sound or music?’, ‘Shall I say this visually or verbally?’, and so on. (Kress & van Leeuwen, 2001, p. 2)

The ActiveWorlds’ system of assigning different colors to different textual communication modes is one example of text carrying one level of meaning, while its visual appearance provides additional information that is crucial to the interpretation of the message. In the 3D virtual environment, the different modes of representation have merged into a single, multimodal communication context. In Patagonia one is not required to choose between saying something visually or verbally, one may use one interface to say it using text, graphics, and even sound.

Dialogism

In the dialogistic tradition as introduced, for example, by Linell (1998), communication is recognized as an ongoing negotiation process where production of meaning is developed in a socio-cultural context. Dialogue is understood as: “any interaction through language (or other symbolic means) between two or several individuals who are mutually co-present” (1998, p. 13). Dialogism contrasts with what is characterized as monologism, the sender-receiver dichotomy of the (idealized) transfer-model of communication. The listener/receiver takes part in the sense-making; he or she is not merely a passive receiver. In dialogism, the reflexivity between language and context is emphasized. As Duranti and Goodwin (1992) point out:

The characteristics of language as an interactive phenomenon have challenged traditional notions of linguistic structure and linguistic rules, suggesting a view of the relationship between language and context as a process that emerges and changes through time and space. (p. 31)

A focus on language as interaction is familiar in the field of Conversation Analysis. However, dialogism is a theory of all discourse including written texts and an attempt to create a “general framework for understanding discourse, cognition and communication” (Linell, 1998, p. 67). As such, dialogism is an interdisciplinary approach. Within this broadened framework focusing on the social character of interaction through any symbol system(s), a foundation can be found for approaching communication in a multimodal environment.

The approach to computer-mediated conversation adopted here applies some key terms of the dialogistic framework as introduced by Linell (1998): sequentiality, joint construction, and act-activity interdependence. In dialogism,

the positioning in the sequence is built into (the interactional status of) the single contribution (turn) in dialogue, and conversely, the sequence is constituted by the contributions embedded within it. The contributions by the different interlocutors reflect each other, and jointly produce meaning. Finally, of course, the activity and the elementary acts are mutually constitutive; they are reflexively related too. (Linell, 1998, p. 88)

By applying a dialogistic framework to a conversation, the researcher is able to focus on interactional aspects, each utterance being viewed as a communicative, other-oriented, act. In Linell (1998, p. 159), turns are classified as basic elements of dialogue. Sequentiality refers to the fact that from a dialogical viewpoint, all communicative acts can be seen as a response to something; an utterance cannot appear out of “nowhere.” Joint Construction refers to the fact that at least two persons are required to establish, develop, and close a topic. Linell (1998, p. 183) argues that topic development is a typical example of joint accomplishment. Act-activity interdependence refers to the close interdependency between understanding and responding, and that “both speaking and listening involve sense-making” (Linell (1998, p. 80). We might add that visual appearance involves sense-making, as well.

In dialogism, it is assumed that no understanding is complete and that intersubjectivity can be obtained only partially. This is the reason why the dynamic term meaning potential is preferred over meaning, along with the notion of meaning-making as a continuous negotiating process (cf. Linell, 1998, pp. 111-115). A discourse analysis within the dialogistic tradition focuses on larger units than what is traditionally done within Conversation Analysis. Notions such as intersubjectivity and interactional processes are important.

In graphical virtual worlds, the graphical environment serves as a semiotic system that is as changeable and negotiable as is verbal language use. How is intersubjectivity achieved in this particular communication medium? What properties of the system regarding communication are available to participants, and how are they used? To what degree is the graphic interface important to understanding conversation in Patagonia, and what methodological and ethical practices should be followed in studying such conversations?

Data and Methodology

For this article,6 I studied conversations in the most public or official virtual premises of the 3D world Patagonia, namely the Ground Zero and official meetings. When doing research in text-only CMC systems, chat session transcripts comprise the main data for analysis. With Patagonia, I found text logs to be of particular interest at the point where conversation transcripts alone are no longer comprehensible to the reader due to participants’ interaction with the graphical surroundings. Digital video recordings were also used to support this research.

One must be logged on in order to gain access to public conversations. Although they are an important part of the chat experience, private communication modes, such as whisper messages and telegrams, were not available to me, since in order to access them, I would have had to make prior arrangements with the involved parties. The texts analyzed in this study are all drawn from public chat logs accessed through passive observation. Cherny (1999, p. 303) claims that “lurking without participating is noticeable and ultimately difficult on MUDs.” In Patagonia, the situation is different: I placed a tourist avatar (with privileges of a registered user) on various spots at Ground Zero. Since tourists are often inactive, my passive presence was mostly unnoticed by others.

The observation period was set at one week and approved by the Union of Norwegian ActiveWorlds Users. Since the UNA represents more than 400 members, I did not attempt to acquire the informed consent of each user. Prior to the fieldwork period, I frequently visited Patagonia as a private person. I have made use of the knowledge gained in this way, especially when selecting representative text samples and describing the technical and socio-cultural features of the site. Data from public meetings were partially collected from meeting logs published on the Web and partially from my own experiences as a meeting participant. A meeting minutes book on the Web presents information about the UNA having permitted a researcher to carry out a research project involving logging chat in Ground Zero Patagonia. The board decided that nicknames were to be changed when publishing logs.

Ethical concerns affected data selection in other ways, as well. The analysis material consists of text-only chat logs. The original material also includes pictures and digital video recordings of chat sessions that display multimodal, dynamic, and temporal features. However, public use of these materials is heavily restricted due to their insufficient ability to protect users’ identities. The decision to focus on textual conversations while emphasizing multimodality was necessary to protect users’ identities, even though the users were in a public forum.

Figure 2. Patagonia average number of users during one week

Participation statistics were compiled for the week of data collection. These are shown in Figure 2, which presents user averages—number of users logged on—by time of day based on statistics gathered during one week in March 2000. The averages are broken down into a “world total rating” and a “Ground Zero rating” according to registered/tourist user status. The number of users connected to the world (the world total rating) is available at all times through the system. In addition, I counted the number of active users in Ground Zero (i.e., users who had “spoken” one message or more, such that their nickname appeared in the chat window). In the fieldwork I included registration of how many users were logged on to Patagonia and in Ground Zero every 30 minutes from 4 pm—1 am each day; this, I discovered through preliminary observations, was the most active period of any day in Patagonia. User averages vary within the 30 minute intervals, and an investigation at another time of year would perhaps indicate different ratings; however, I believe this week was typical. Even though Figure 2 indicates a relatively small group of users (an average of 4.4 active users in Ground Zero), approximately 350 text log pages (A4, standard font size, with most messages being only one line) were collected during the fieldwork period.

Medium variables

Computer systems that support communication can be characterized by “medium variables” (Herring, 2007), such as synchronous/asynchronous or one-way/two-way message transfer protocols. This study focuses on those properties of a specific computer program that affect communication and does not attempt to account for all the system features available. The discourse of Patagonia corresponds with what Herring (2007) has classified as one-way synchronous computer-mediated discourse. This classification distinguishes ActiveWorlds from two-way systems (such as UNIX Talk or ICQ) and from asynchronous systems, such as email. Graphic representation of avatars in a virtual environment, together with different textual communication modes, sound, and the web browser, further distinguish ActiveWorlds from text-based virtual realities such as MOOs. These representations and other multimodal affordances of ActiveWorlds are discussed as semiotic features in what follows. Finally, being based on Render Ware, it differs from VRML-based systems: While Blaxxun (VRML), for example, is a plug-in system, ActiveWorlds is not.

Semiotic Features of Patagonia

Some visual elements that carry a meaning potential in the particular community of Patagonia are avatars, different text features including color, the notion of place, different graphical textures, and objects, such as signboards and buildings. Even sound may be added to this compilation. The following paragraphs attempt to shed light on what meaning potentials the different semiotic features represent. One consequence of using the notion “meaning potential” is that observations are dynamic and negotiable rather than static and fixed. The interpretation is subjective, based on social conventions, and closely related to knowledge of the system.

Avatars

Avatars encode meaning about user status and gender. There is large difference in the number of avatars available for tourists and for registered users: only two as compared to 57.7 Tourist avatars in Patagonia are male by default. Tourists may select between only one male and one female avatar (see Figure 4); however, knowledge about how to change avatars is required in order to perform this selection. An example of an avatar available only to registered users is shown in Figure 3. There are a few gender-neutral avatars to choose from for registered users, but none for tourists.

Figure 3. ActiveWorlds’ avatar “Venus,” available for registered users


Figure 4. Female tourist avatar “Kelly” and male tourist avatar “Helmut”

Avatars are semiotic resources. First, avatars are a (more or less conscious) presentation of self, and thus they carry meaning. Second, the (albeit limited) communicative behaviors avatars are able to perform also constitute a type of meaning potential. Action buttons are pre-programmed gestures/actions that move the avatar in a certain pattern, such as fight (without actually hitting or hurting other avatars), bow, wave, or dance. Most avatars available for registered users have a wider selection of action buttons than tourist avatars. Compared to contemporary computer or video games, however, avatars enjoy a limited selection of character moves. Some of these pre-programmed moves, such as the wave command and bow, may be communicative.

Text Features

The “medium variables” available in Patagonia, such as synchronicity and one-way message transfer protocols, are the same for tourists as for registered users, with one small exception: Tourists are not allowed to send telegrams (asynchronous page messages). In addition, tourists and regular users differ in other privileges, as summarized in Table 1. These features are important basic conditions for the computer program per se, and important conditions for enabling interaction to take place.

TOURIST

REGISTERED

Text color:

Grey

Black

Avatars:

2

57 (+2)

Telegrams:

No

Yes

Contacts list:

No

Yes

Join contacts:

No

Yes

Exchange files:

No

Yes

Costs:

None

$19.95 p.a.

Save personal nickname:

No

Yes

Immigration officer infomercials:

Yes

No

Building:

Yes

Yes, password protected

Table 1. Appearance and privileges of different users

Textual interaction is possible through various communication modes, such as “normal” chat, whisper, and telegrams displayed in different colors (Table 2).8 The default setting displays text over avatars’ heads. Tourists receive recurring infomercial messages in green. These messages provide information about features of the computer program, and their main purpose is to persuade tourists to “immigrate.” Conversations within a defined radius appear in the chat window. However, sending telegrams between registered users is possible whether located in the same place, outside hearing radius, or in another world. Hypertextual links appear either written in the chat window or pre-programmed on a (clickable) sign or picture linked to a web page. In the 3.0 browser version analyzed here, hypertextual links (or what appear to be links) are automatically detected.

All users are able to ignore/mute others. The chat is otherwise moderated through a system of “Public Speakers”: Some regular users are entrusted with additional features (privileges), such as the ability to see all other users’ nicknames and IP addresses. As Public Speakers they are able to kick/ban users who do not act according to a Code of Conduct set up by the Union of Norwegian ActiveWorlds Users.

COLOR

COMMUNICATION MODE

PUBLIC/PRIVATE

Grey:

Tourist user

Public

Black:

Registered user

Public

Bold black:

Extra featured users,

restricted use

Public

Blue italics:

Whisper message

Private

Red:

Telegram message

Private

Green:

Public/infomercial

Public

Underlined blue:

Hypertext

Both

Teal:

Building inspector

Private

Table 2. Text color codes of Patagonia

The chat window has a limit of approximately 900 lines of text (when copied and pasted directly into Microsoft Word, this is about 12 pages A4 with standard font setting) before deleting from the top is performed automatically as new text appears, as in other multi-participant chat systems (e.g., Herring, 1999). Scrolling backwards is possible, and the last message sent is always at the bottom line. A default setting stores the chat log in a text file on the computer’s hard disk. Both tourists and registered users are able to store chat sessions locally on their own computers, and are allowed the same degree of persistence regarding textual settings, except for telegrams. Telegrams are the only messages that are stored until deleted by a recipient. At the time of this study, there was no setting for logging synchronous graphics action.

Understanding the Meaning of Text Color

In ActiveWorlds, coexistent and mutually dependent modes of communication constitute multimodality. One of the most striking differences between a conversation transcript extracted from a MOO and ActiveWorlds is the use of a standard set of message colors. One of the chief visual semiotic features of Patagonia is text color and the decoding that users do of such colors. For example, when a user sees underlined blue text in the chat window, he or she recognizes it as hypertext.

Grey text, the guest user color, has the lowest status. Bold black has the highest rank and is usually employed for speeches, while black is considered normal. Text colors differentiate among a complex set of communication modes, which is not likely to be comprehended by uninitiated users, even if they are experienced with CMC in general. The fact that the set of modes is based upon a color system differentiates ActiveWorlds communication commands from those described for text-based MOO communication by Cherny (1995, 1999), as summarized in Table 3.

COLOR

COMMUNICATION MODE

MOO MODE

Grey:

Tourist user message

--

Black:

Registered user message

“Say,” “emote”

Bold black:

Extra featured users, restricted use

--

Blue italics:

Whisper/private message

“Whisper”

Red:

Telegram message (asynchronous)

“Page”

Green:

Public/infomercial message

Welcome message

Underlined blue:

Hypertext

--

Table 3. Text color codes of Patagonia compared to MOO communication modes

Moreover, Cherny (1999) describes requirements for sophisticated communication command typing skills in MUDs or MOOs, for example the Carpal Tunnel Syndrome Feature, CTS. There is no equivalent to the enabling of different modes of communication by typing in commands in the ActiveWorlds.

The Meaning Potential of Place

Patagonia is a virtual reality world simulating the Norwegian landscape and seasons. Computer graphics as a symbolic system are utilized to support the perception of Patagonia as a geographical place: Snowy mountains on the horizon are displayed all year long. What would be green grass in other ActiveWorlds is replaced with a white, snowy texture, corresponding to the Norwegian winter. Stability is necessary for maintaining a virtual world’s identity as such; hence, the graphic design of Patagonia’s Ground Zero is coherent, recognizable, and familiar, even for less regular users.

Ground Zero: Home

Ground Zero is the place visited most by users during an online session. It is the name of the geographical center of each world. The role of Ground Zero in virtual worlds’ information structure might be compared to a Web portal, such as AOL or CompuServe. The ActiveWorlds program can be set to enter a world or a particular “place” in a world as a default “home.” In Patagonia, Ground Zero is the default, “central” in relation to other “peripheral” areas where private 3D houses, castles, etc. are built. In December a 3D Christmas tree is placed at Ground Zero to serve as Christmas decoration.

The virtual geography is also a semiotic system that acts as an intermediary for important information and commercials. Information signboards are displayed close to Ground Zero: When users descend in Ground Zero the signs appear in the graphics window (see Figure 1).

Tingvollen: The meeting place

Another example of the semiotics of place involves meetings of the Union of Norwegian ActiveWorlds Users.9 UNA meetings are situated in the virtual geography at a permanent meeting place, Tingvollen (Figure 5). Several means are utilized to implement online meetings: re-location in the virtual geography, use of signs, chat, and websites before and after meetings. (Logs from most meetings are published on the official UNA web site.) The virtual environment—Tingvollen—plays an important role in providing the relevant situational context of a meeting. A teleport signboard is placed in Ground Zero for easy access to Tingvollen.

Figure 5. Tingvollen, Patagonia

Beyond the graphics defining the situational context, references to virtual surroundings or play with graphic objects are otherwise sparse in meetings. The chairman usually narrates an imaginary gavel to carry motions or keep the boardroom in order, followed by sound descriptions, such as *thud*, when the gavel is in use. Similar narrative play practices have been observed in purely textual CMC environments (Virtanen, in press). In Patagonia, however, the textual communication and virtual surroundings are interdependent, and the virtual design of Tingvollen serves as a frame of understanding that differentiates the activities that take place there from those at Ground Zero.

Graphical Objects: Signboards at Ground Zero Patagonia

The most salient signs at Ground Zero contain the most important information, such as netiquette tips, list of Public Speakers, and commercial messages. These signs define the borders between what is Ground Zero and what is not. Consistent with Ground Zero being an “official” geographical place, signs at Ground Zero are assigned an official status. In addition to its message, a sign’s location as part of the geographical center of the 3D environment is significant and carries meaning for Patagonia’s users.

Users may interact with each other by acting with avatars or placing or creating graphical objects in a certain manner. A signboard in the graphical environment is such an object. Once displayed in a world, a signboard will still be there when the user who put it there is gone. The signboard is thus of a more persistent character than, e.g., a chat message. It also has a different communicative meaning from most chat messages, in that it usually displays some kind of information. The graphical environment is changed when a signboard is created, and when it is removed, the environment is changed again.

Sounds

To some extent, the virtual world is also constituted by sound effects, although Patagonia users seem to prefer listening to music of their own choice rather than pre-selected audio (MIDI) files. Likewise, WAV files of, for example, birds singing or waterfalls can become monotonous during a long chat session. However, music files are used actively during online events, such as parties and concerts. In Patagonia special music files have been composed to serve as fanfares on Election Day. The world Mars, the Red Planet has the sound of Martian winds blowing to support the simulation of physical presence on the planet Mars.

Conversation in Patagonia

The remainder of this article consists of two sample analyses whereby conversations are viewed through the lens of the theoretical framework presented earlier. Conversation in Patagonia is of a recreational, social character. In this respect, the text excerpts are not that different from conversation in other electronic synchronous one-way CMC systems. What distinguishes them is that they crucially involve meanings communicated by graphical means. The criteria for choosing the two samples below were, first, that the samples contain textual references to the graphics window; in this respect, they are representative of typical Patagonia conversations. In “the hit and run” example, language is reduced to a supplement to the “real” action, which is carried out by avatars. (Having captured the conversations on video is crucial to an understanding of this conversation.) I also selected examples of wholly nonverbal (non-textual) communication in Patagonia, which are not at all difficult to find during a chat session. Even novice users who possess some video- or computer-gaming experience are able to explore these features of the ActiveWorlds on their first encounter.

Sample Analyses

The first example is an encounter involving several registered and tourist users, describing the consequences of differences in available system features. Different system features are also a topic in the “Hit and Run” example, a conversation between a registered and a tourist user.

A Part of the Crowd… or Not?

The establishment of online communities suggests new lines of social demarcation. One such demarcation is that between experienced or regular participants and new participants. New users appear different from registered users, and their writing is a faint color. Moreover, not all of them know the conventional behavior (netiquette) of Patagonia. Regular users occasionally imitate “tourist behavior” for fun; this is performed by exaggerated use of action buttons, foul language, spamming, and sexual harassment of others. In response to reports by tourists who claim to have been ignored or harassed by registered users, citizens of Patagonia have established a forum called the Welcoming Committee. The Committee is also supposed to provide a somewhat organized socialization of new users (cf. the article by Weber in this issue describing the socialization of new users in a textual online forum), but its members are not always present.

Users of Patagonia have assigned symbolic value to system features and thus have given meaning to the term virtual social status. The section below is a typical encounter between registered members and tourists. My observations describing actions carried out in the graphics window are indicated with black italics. For ease of reference, numbers have been added to each utterance. (Translation from Norwegian is by the author.)

Immigration Officer: Welcome to Patagonia - the Scandinavian world
1. Sylvester: nothing’s happenin’
2. Sylvester: nothing but tourists here
3. Vixen: nope, very little
4. Vixen: yes
5. “MsMcDonald”: Nothing but tourists............
6. “three”: what’s wrong with tourists?? : )
7. “Morgan Kane”: hi three... how are you?

8. Vixen: nothing
9. Sylvester: join me
10. Vixen: oki
11. Sylvester: all citizens
Sylvester teleports from Ground Zero
12. “three”: I’m fine thanks ?
13. “Morgan Kane”: .those who don’t bother paying $19.95 to be a part of the crowd, MsMcDonald : )

14. Vixen: joining sylvester........
Vixen joins Sylvester
15. “Morgan Kane”: ok… what can we talk about?
16. “Morgan Kane”: by the way, I’m not so bad either...
17. “three”: .how can you tell the difference between tourist andnot?
18. “three”: hmm.. gone smoking
19. “Morgan Kane”: .tourists
[sic] have all in all more possibilities, besides, their writing turns black while ours
is in grey and they get to keep their own nick, sort of…

In Patagonia, social status among users is to some extent proportional to access to features and privileges. Text color and selection (assortment) of avatars are indicators of status. The two avatars available for tourists are considered unattractive, and as speaking with a color that is hard to see. In the excerpt above, the tourists are left behind at Ground Zero when the registered users join each other, since the join function is not available to tourists. There is little or no communicative behavior in the graphics window in this short session, and the text carries no references to the graphics besides the fact that tourists are present.

In this session, the registered users Sylvester and Vixen relocate their avatars in the virtual geography because they want to get away from the tourists at Ground Zero. There is no way of determining exactly where they went without joining them. (Tourists are able to teleport, but in order to perform this action one has to know where to teleport to.) Thus, the Ground Zero as place is given new meaning as a place for those who are not able to join others (i.e., tourists). Ground Zero is a place where nothing happens (lines 1-2), a place where registered users do not want to be located. Vixen does not want to offend the tourists present, so she says that there is “nothing” (8) wrong with tourists. Nevertheless, she prefers joining Sylvester to staying with the tourists. Note that (8) may also be interpreted as a confirmation of Sylvester’s “nothing” in (1-2).

Of particular interest in this excerpt is the process of intersubjectivity between Vixen and Sylvester. There is a sequential distribution between Sylvester and Vixen’s turns, which form a coherent and understandable exchange if the grey in between the writing in black (1-14) is ignored. Sylvester implicitly addresses citizens only. His utterance in (2) may be rhetorical, or it may be directed to any other present citizens. However, the utterance is nevertheless other-oriented. Through a few utterances Sylvester and Vixen are able to confirm and develop their mutual understanding of tourists. In (1-4) there is explicit collaboration through Vixen’s confirming responses (3-4) respectively to Sylvester’s (1-2). The intersubjective understanding between Vixen and Sylvester can be interpreted as follows: “we do not like tourists, we are better than tourists, there are a lot of them here, let us go somewhere else where they cannot reach us. And additionally: we do not mind if tourists are offended by our ignoring them or leaving them behind.” In (11) the exclusive privilege of being a registered citizen is revealed: the ability to join each other, and to get away from the crowd. When Sylvester invites Vixen to join him (9), in fact, he invites all citizens to join him (11), thus excluding all non-citizens. Vixen soon acts upon his request, and subsequent to (14) they are both gone. The rest of the conversation then dies slowly.

“Morgan Kane” is the most active coherence maker during this session. He responds to all the other initiations by tourists: (7) is a response to (6), followed up by (15). (13) refers to MsMcDonald’s question in (5), and 18 is an answer to the question in (16); however, by that time “Three” has already “gone smoking” (17). This may be due to the fact that approximately 85 seconds go by before “Morgan Kane” is able to produce an answer to her question, although it is likely that he intended to write “registered citizens,” rather than “tourists” in (18).

“Morgan Kane” appears to try his best to keep the conversation going. He moves his avatar around Ground Zero addressing other tourists and trying to initiate conversation. However, he receives no verbal response to (20-26), and the chat log looks like a monologue:

20. “Morgan Kane”: Hey
21. “Morgan Kane”: well… why that grumpy mood
22. “Morgan Kane”: :)
23. “Morgan Kane”: I’m off to chat with someone else
24. “Morgan Kane”: beginning to think you were all dumb
25. “Morgan Kane”: Hmm… over there at least I see movement… Hey ppl – please answer me!
26. “Morgan Kane”: boo

The “Hit and Run”

In the following conversation, my research avatar (and recording device) was placed on Ground Zero for the entire session, with the result that the two interlocutors can not be seen from segments (41) to (65), although their conversation was still within the Ground Zero “hearing” range, meaning that their chat was visible in the text chat window. Thomas is a regular registered user, and he has selected one of the male avatars available for registered users only. Peach pie is represented by the male tourist avatar, Helmut, when we enter the conversation. Thomas is a Norwegian, although he is typing in English in this session. Peach pie is also English speaking, thus the excerpt is not translated.

1. “Peach pie”: Do you speak any English?
2. Thomas: Yes....
3. “Peach pie”: Can I ask you something then?
4. Thomas: Ask!
5. “Peach pie”: When you look at me...what do you see?
6. Thomas: Come over here!
Peach pie moves her avatar closer to Thomas
7. Thomas: I see a helmut..
8. “Peach pie”: I have never done this before....I am right here!
9. “Peach pie”: A what?
10. Thomas: A green man with black pants...
11. “Peach pie”: oh great.
12. Thomas: And sunglases
13. Thomas: no hear
14. “Peach pie”: Are you joking?
15. Thomas: No...
Peach pie changes her avatar to female tourist avatar “Kelly”
16. “Peach pie”: There now at least I am female...lol
17. Thomas: Now you are a girl like this
Thomas also changes his avatar to “Kelly”
18. “Peach pie”: I am not?
19. Thomas: Yes you are!
20. “Peach pie”: lol...yep.....she is prissy I think...I need to register soon so I can look decent.
21. Thomas: ....
Thomas changes his avatar to “Helmut”
22. “Peach pie”: ok...thanks for your time.
23. “Peach pie”: ewwwww
24. Thomas: like this you look for 5 min..
25. “Peach pie”: that’s what I did look like.....yuck
26. Thomas: I like this...
Thomas changes back to his favorite avatar
27. “Peach pie”: yes...much better
28. Thomas: you can´t bee like this... Sorry
29. Thomas: or this
Thomas switches between a couple of his other favorite male avatars
30. “Peach pie”: So I will be able to pick a better look if I register?
31. “Peach pie”: what is up with that?

32. Thomas: Yes it right
33. “Peach pie”: lolol
34. “Peach pie”: You are a nutbar!
35. “Peach pie”: lol

Thomas switches between several female avatars and recurrently uses “action buttons”
36. “Peach pie”: woah!
Thomas settles for an avatar equipped with a vehicle. He goes for a ride around Ground Zero and then back to “Peach pie”
37. “Peach pie”: lol
38. “Peach pie”: Gonna run me over?

39. Thomas: Yes!
40. “Peach pie”: AAAAAAAA
Thomas and Peach pie now move away from Ground Zero to a virtual fountain (which makes sounds of water running when the avatars approach)
41. Thomas: *splas*
42. “Peach pie”: ouch
43. Thomas: lol
44. Thomas: lol
45. Thomas: hold stil so i get a feet on you
46. Thomas: hehe
47. “Peach pie”: ok...you will wreck if you hi this fountain
48. Thomas: pihaeryhthu
49. “Peach pie”: hookkadddiek,s’f
50. Thomas: You rigth... Now i have to get a nwu one!
51. Thomas: hehe
52. Thomas: here it is!
53. “Peach pie”: ok...I’ll stand and fight
54. “Peach pie”: waaaaa

55. Thomas: Get out of the fuotnasione!
56. “Peach pie”: hia cha
57. “Peach pie”: ha ha ha
58. “Peach pie”: tbbbb

59. Thomas: no this is wather resist
60. Thomas: *splas*
61. 007: hei
62. Thomas: Hey!
63. “Peach pie”: hello
64. “AgentMulder”: hello

65. “Peach pie”: I have been involved in a hit and run....

Even though Peach pie says she has never tried using ActiveWorlds before, she seems familiar with conventions of (text based) computer-mediated communication (Werry, 1996): Peach pie makes use of, e.g., the “LOL” acronym (35), (even “lolol” in 33), back channels (11, 33-37) and graphic simulation of sounds (23, 40, 54, and 58). Kolko (1995) observes that, “longtime MOOers engage in such narrative progressions almost unconsciously.”

This excerpt is an example of play. Cherny (1999, p. 208) describes play in MOOs as a complex collaborative activity: “Play in the MOO is a highly cooperative behavior, usually happening during multi-person conversation and often triggered by or directed at objects in the environment.” Similarly, in the above excerpt, the interplay between communicative activities in 3D graphics and text are important components of the interaction. Indeed, the interlocutors refer to and change the situational context of their session constantly. An ongoing reflexivity of discourse and context penetrates the entire conversation, as meaning is co-produced by participants responsible for the progression of the conversation.

Objects, avatars, and graphical surroundings are included in narrative play in Patagonia. Specifically, the “Hit and Run” episode (38-65) is an interaction with graphical objects (avatars and water) combined with imaginary objects, such as in (41) *splas*—splashing of water—that cannot be performed with graphical objects in Patagonia. These two types of interacting—narration of objects together with pre-programmed objects—also occur in MOOs. They “require different uses of the imagination; with one type the participant must contribute almost all the details of the space and scene, and with another, programmable objects provide a starting point from which the narration can be built” (Kolko, 1995, p. 3). As such, the conversation takes place on the borderlines between the 3D virtual reality, a narrated virtual reality, and references to actual reality.

Sequentiality

Turn taking in “Hit and Run” is distributed quite symmetrically in segments (1-32): The interlocutors each contribute about half until Thomas starts his “avatar show” (32), following which the number of Peach pie’s utterances exceeds Thomas’, because he is busy changing avatars. However, his non-verbal responses contribute to the sequential construction of the conversation, and the two participants are mutually oriented to one another throughout the chat session. Although the conversation might seem disorganized because of asymmetric distribution of turns, it is nevertheless sequentially structured. The reasons for what seems to be the asymmetrical distribution of explicit verbal turns lie in the graphic interface.

Joint construction

Changing of avatars is the organizing topic of the “Hit and Run” conversation. The conversation has three main parts. The over-arching topic for (1-21) is how “Peach pie’s” avatar looked before and how it looks now. In segment (22) it appears that Peach pie thinks the chat is over, but in (23-37) the topic shifts to Thomas’ “avatar show.” Thomas initiates a topic change nonverbally by changing his avatar. The topic in segments (38-65) is the “by the fountain” session. Peach pie describes the experience herself in (65) “I have been involved in a hit and run.”

As Kolko (1995) points out, when a MOOer narrates, an answer is called forth. This dynamic—a Saying that forces a response from another—creates a web of connections, the shape of which is determined by the choices of the narrator. Both parties are responsible for how the discourse turns out, and simultaneously they attempt to induce the dialogue to take a particular direction. Thomas replaces the narrated “wrecked” avatar with a new, narrated one (50)—even increasing its qualities (59)—and Peach pie plays along.

Act-activity Interdependence

In (18), Peach pie asks for a confirmation that Thomas’ prior utterance, and changing his visual appearance to be the same as hers, is actually valid. Thomas confirms that she has understood his utterance, and his act, correctly in (19). Thomas is able to tell Peach pie exactly how she looks by visual and verbal means combined (17). At this point Peach pie has understood what her avatar looks like and has even confirmed to Thomas that she has understood what he has just explained verbally and visually. While Thomas is busy changing avatars Peach pie replies or responds textually in (33-35), (36), and (37-38) to his nonverbal, but nevertheless communicative, actions.

Visual Addressivity

A video clip of the sequence above shows that utterances 20-26 are directed to the other tourists present10 in Ground Zero at that time. According to dialogistic theory, all utterances are addressed to someone. To determine who is talking to whom in face-to-face conversation is relatively easy when one is able to see the faces and bodies of the interlocutors. In text-based CMC the physical proximity of participants is absent. A question of general interest in CMC research is how participants address one another, including how they keep track of who is addressing whom in multi-participant interaction. Werry (1996) observes that a high degree of addressivity is common in Internet Relay Chat (IRC), since the addressee’s attention must be recaptured anew with each utterance. He describes a convention whereby speakers “indicate the intended addressee by putting that person’s name at the start of an utterance, followed by a colon” (Werry, 1996, p. 52). In Patagonia, in contrast, the text may not encode any explicit indication of who is addressing whom, but the placing of avatars does. Similarly, a study of the 2D graphical environment “The Palace” (Krikorian, Lee, Chock, & Harms, 2000) shows that, when measured, distance between participants’ matters in a virtual computer-mediated environment.

In cases where Patagonian interlocutors address each other verbally, the nicknames tend to be located in utterance-final position, rather than at the start of utterances. In the following conversation excerpts, nicknames are underlined. Messages are reproduced in the original languages used:

MRS: Du ser ikke hele pata her Ollis :-)
---
“AgentMulder”: i dont understand swedish good
“AgentMulder”: i know only ew words
MRS: This is Norwegian AgentMulder :-)
---
ZZtop: jøss....er du her oxo nå Stephen :o)
---
Thomas: Har du det bra MRS
MRS: jatakk Thomas... du da?

The (very) few instances I found of nicknames followed by a colon were used by English-speaking tourists, not by regular Norwegian users.

Werry’s addressivity features are, with few exceptions, absent from conversations in Patagonia for several reasons. First, there are generally fewer participants involved than on the IRC, and thus there is less ambiguity as to who is addressing whom at any given moment. Given the recreational character of conversation in ActiveWorlds, it is also possible that users find addressing someone using nicknames followed by a colon to be a more formal practice than putting this in a final position, although both formats are acceptable in Norwegian. Finally, graphic nonverbal means of addressivity correspond more closely with how addressivity is carried out in face-to-face conversation. If another user’s avatar is displayed in your graphics window—i.e., is in your field of vision—and he or she starts talking, you are likely to assume that he is addressing you, even if he does not know your nickname. An avatar responding by waving to another avatar waving is an adequate, conventional, nonverbal response to a nonverbal, albeit communicatively meaningful, act.

When communicating through avatars nonverbal communication is achieved. What may have been thought to be the exclusive property of face-to-face communication is found also in conversation in a multimodal interface. In this sense, the properties of the graphical environment invite users to perform in a more face-to-face conversation-like manner than does text-based CMC.

Nevertheless, after the novelty wears off, experienced users seem to grow tired of moving the avatars around constantly. Vilhjálmsson (1997) writes about avatars in the ActiveWorlds: “one soon realizes that the animation displayed is not reflecting the actual events and conversations taking place, as transcribed by the scrolling text window beneath the world view” (p. 14). There is a technical constraint to the moving of avatars by using arrow keys or mouse, as fingers are not free while conversation is typed. Some correspondence between animated avatar movements and text chat does occur, but it is only a vague imitation of face-to-face conversation.

Discussion and Conclusion

Substantial changes in the properties of CMC systems call for new perspectives in data gathering, ethics, and methods of analysis that move beyond solely text-based analysis. This article has examined the semiotic potential embedded in a Scandinavian 3D graphical world, Patagonia, based on dialogism, an over-arching interdiciplinary theoretical approach to discourse, cognition, and communication. Focusing on social interaction, dialogism embraces language and other symbol systems, making it well-suited to the analysis of complex environments. The approach brings forth the notion of multimodality, a key term in the understanding of a praxis whereby users combine different modes of communication when interacting. Examples are analyzed that show how multimodal resources such as visual modes of communication in Patagonia carry interactional meaning. I have argued for the legitimacy of this approach, emphasizing the resources one misses by neglecting the visual modes of representation.

While strides have been made in CMDA since the data for this study were collected, there is still a long way to go as regards the analysis of multimodal online discourse. Discourse analysis in general and conversation analysis in particular would benefit from incorporating visual and other means of communication into analysis. At the same time, discourse analysis faces a wide expansion of perspective if visual communicative acts, and their sequential positioning, are to become objects of analyses. Included in the changes are challenges related to data gathering, transcription, and the reporting of such data (for a treatment of such challenges in World of Warcraft, a multiplayer online game environment, see Newon, 2011). The ethical considerations limiting the use of video material also need to be explored further. Finally, there is a paradox in writing a text, such as the present article, on the subject of multimedia features of a virtual environment, without being able to show video recordings. Is the research paper as activity genre open to entering the world of multimedia?

Towards Realism?

Multimedia features increasingly make computer/video games of recent years appear more “real” than in earlier stages. Avatars of the ActiveWorlds tend to look like cartoon stereotypical fantasy figures (Lara Croft from the PlayStation game Tomb Rider perhaps being the most famous example; see also the female avatar in Figure 3) rather than like real people. However, newer avatars and new system updates typically provide enhanced 3D effects and higher resolution. Even graphical surroundings may be rendered more “real” in terms of resolution, etc. and still be a fantasy environment, such as StarWars or the lost city of Atlantis.

The work of Vilhjálmsson (1997) indicates how avatars may be further developed to perform communicatively in a more human-like manner by rendering nonverbal cues. Vilhjálmsson (1997) claims that lip synchronization, such as in the OnLive world, “greatly enhances the experience.” To the best of this author’s knowledge, there are no technical obstacles to establishing richer or more complex action buttons in the design of new system versions of the ActiveWorlds. At the same time, there is the question of to what extent the ActiveWorlds users wish their avatars to appear and/or behave in a more human-like manner. Whether avatars are to be more than “presence indicators” in the future is related to a fundamental question regarding Virtual Reality, namely how real one prefers it to be. It is not clear that online Virtual Reality necessarily profits from a high level of reality simulation.

Conversations in Patagonia suggest that the graphical environments are a resource for play activities, including play in which behaviors and properties of the physical world are exploited. How activities of this type might be affected in more realistic systems is an interesting question for future research.

Notes

  1. The URL for the ActiveWorlds.com Inc. official website is: http://www.activeworlds.com.

  2. The ActiveWorlds program displays an Internet Explorer-based browser. This study is based on the 3.0 version of the ActiveWorlds browser.

  3. Tourists and registered users were both present at the time of the fieldwork for this paper. However, in February 2002, a new charging policy was introduced. Tourist user status is a limited (short) free trial period, which may only be obtained by providing a credit card number, and world owners are charged with an annual tourist access fee. In February 2002 tourists were denied access to Patagonia. However, this policy was changed a couple of months later. Further information is available about Patagonia’s citizenship policy at http://www.activeworlds.com.

  4. The acronym MOO stands for MUD Object Oriented, a subcategory of MUDs, Multi-User Dungeons/Dimensions.

  5. See, e.g., McDonald (2007); Sindoni (in press), Herring and Demarest (under review).

  6. Part of a dissertation project at the University of Oslo (Naper 2001).

  7. As of August 2002, there were 107 avatars available for registered users.

  8. For a detailed description of the communication modes available in a textual MOO, see Cherny (1995).

  9. The Union of Norwegian ActiveWorlds Users is an organization developed by citizens of Patagonia and some other Norwegian worlds. It deals with issues such as the Public Speaker system and the Welcoming Committee. All tourists and registered avatars are allowed to attend meetings, but only board members are allowed to speak and vote. Tourists are not allowed membership.

  10. Since I had to move my avatar around the Ground Zero in order to capture this particular conversation from a good camera angle, attention was drawn to my avatar as well as the others.’ Utterances (20-22) appear to be addressing my observation avatar.

References

Cherny, L. (1995). The modal complexity of speech events in a social MUD. The Electronic Journal of Communication, 5(4). http://www.cios.org/EJCPUBLIC/005/4/00546.HTML

Cherny, L. (1999). Conversation and community: Chat in a virtual world. Stanford, CA: CSLI Publications.

Danet, B. (2001). Cyberpl@y: Communicating online. Oxford, UK: Berg Publishers. Danet, B., Ruedenberg-Wright, L., & Rosenbaum-Tamari, Y. (1997). Smoking dope at a virtual party: Writing, play and performance on Internet Relay Chat. In S. Rafaeli, F. Sudweeks, & M. McLaughlin (Eds.), Network and netplay: Virtual groups on the Internet. Cambridge, MA: AAAI/MIT Press.

Dresner, E., & Herring, S. C. (2010). Functions of the non-verbal in CMC: Emoticons and illocutionary force. Communication Theory, 20, 249-268.

Duranti, A., & Goodwin, C. (1992). Rethinking context: An introduction. In A. Duranti, and C. Goodwin (Eds.), Rethinking context: Language as an interactive phenomenon (pp. 1-42). Cambridge, UK: Cambridge University Press.

Hansen, K. (2001). Virtuel Kultur. Ph.D. dissertation, Department of Communication, Aalborg University, Denmark.

Herring, S. C., ed. (1996). Computer-mediated communication: Linguistic, social, and cross-cultural perspectives. Amsterdam/Philadelphia: John Benjamins.

Herring, S. C. (1999). Interactional coherence in CMC. Journal of Computer-Mediated communication, 4(4). http://jcmc.indiana.edu/vol4/issue4/herring.html

Herring, S. C. (2001). Computer-mediated discourse. In D. Tannen, D. Schiffrin, & H. Hamilton (Eds.), Handbook of discourse analysis (pp. 612-634). Oxford, UK: Blackwell.

Herring, S. C. (2004). Computer-mediated discourse analysis: An approach to researching online behavior. In S. A. Barab, R. Kling, and J. H. Gray (Eds.), Designing for virtual communities in the service of learning (pp. 338-376). New York: Cambridge University Press.

Herring, S. C. (2007). A faceted classification scheme for computer-mediated discourse. Language@Internet, 4, article 1. http://www.languageatinternet.org/articles/2007/761

Herring, S. C., & Demarest, B. (under review). Mode choice in multimodal comment threads: Effects on participation, sociability, and attitude.

Jewitt, C., & Oyama, R. (2001). Visual meaning: A social semiotic approach. In T. van Leeuwen & C. Jewitt (Eds), Handbook of visual analysis (pp. 134–156). London: Sage.

Kolko, B. (1995). Building a world with words: The narrative reality of virtual communities. Works and Days 25/26, 13(1&2). http://bethkolko.com/includes/pdfs/buildingaworldwithwords.pdf

Kress, G. (2000). Multimodality. In B. Cope & M. Kalantzis (Eds.), Multiliteracies (pp. 182-202). London: Routledge.

Kress, G., & van Leeuwen, T. (1996). Reading images: The grammar of visual design. London: Routledge.

Kress, G., & van Leeuwen, T. (2001). Multimodal discourse: The modes and media of contemporary communication. London: Arnold.

Krikorian, D. H., J.-S. Lee, T. M. Chock, & C. Harms (2000). Isn’t that spatial?: Distance and communication in a 2-D virtual environment. Journal of Computer-Mediated Communication, 5(4). http://jcmc.indiana.edu/vol5/issue4/krikorian.html

McDonald, D. (2007). Visual conversation styles in web communities. Proceedings of the Fortieth Hawaii International Conference on System Sciences. Los Alamitos, CA: IEEE Press. http://www.pensivepuffin.com/dwmcphd/papers/McDonald.HICSS-40.preprint.pdf

Newon, L. (2011). Multimodal creativity and identities of expertise in the digital ecology of a World of Warcraft guild. In C. Thurlow & K. Mroczec (Eds.), Digital discourse: Language in the new media (pp. 131-153). New York: Oxford University Press.

Linell, P. (1998). Approaching dialogue: Talk, interaction and contexts in dialogical perspectives. Amsterdam/Philadelphia: John Benjamins.

Sindoni, M. G. (in press). Through the looking glass: A social semiotic introduction to the study of videochats. In G. Kress, E. Adami, & R. Facchinetti (Eds.), The challenge of multimodal analysis to notions of text [special issue of Text & Talk].

Vilhjálmsson, H. H. (1997). Autonomous communicative behaviors in avatars. Masters thesis, Media Arts and Sciences, Massachusetts Institute of Technology. http://web.media.mit.edu/~hannes/msthesis

Virtanen, T. (in press). Performativity in CMC. In S. C. Herring, D. Stein, & T. Virtanen (Eds.), The handbook of pragmatics of computer-mediated communication. Berlin: Mouton de Gruyter.

Werry, C. C. (1996). Linguistic and interactional features of Internet Relay Chat. In S. C. Herring (Ed), Computer-mediated communication: Linguistic, social, and cross-cultural perspectives (pp. 47-65). Amsterdam/Philadelphia: John Benjamins.

Biographical Note

Ida Naper[ idanaper "at" online "dot" no] holds a Cand.Philol. degree from the Department of Scandinavian Studies at the University of Oslo, Norway. Her areas of interest are digital discourse, language and visual communication and, most recently, social media in customer services.


License

Any party may pass on this Work by electronic means and make it available for download under the terms and conditions of the Digital Peer Publishing License. The text of the license may be accessed and retrieved at http://www.dipp.nrw.de/lizenzen/dppl/dppl/DPPL_v2_en_06-2004.html.

Fulltext