THIS IS AN INCOMPLETE VERSION OF MY ORIGINAL THESIS WITH THE NOTES AND SOME OTHER PARTS MISSING.
The IRC Vernacular
A Linguistic Study of Internet Relay Chat
Az Internet Relay Chat nyelvi sajátosságainak vizsgálata
SZAKDOLGOZAT
Gelléri Péter
angol nyelv és irodalom szak
Témavezető: Németh Nóra
1998.
Table of contents
Introduction
This study analyses the linguistic characteristics of a new medium of synchronous computer-mediated communication called Internet Relay Chat (IRC). With thousands of people using it from all over the world at any given moment, IRC is undoubtedly one of the most popular conferencing systems in the world today. Despite its amazing popularity and global accessibility, however, IRC has not attracted much attention from a linguistic point of view. Nevertheless, the reason why some scholars are starting to consider the new medium a legitimate field of study is partly because it provides an easily observable and incredibly rich testing ground for linguists and sociologists alike, and partly because interactions on IRC affected by the characteristics and physical constraints (especially its text-based nature) of the medium may reveal an incredibly new approach to language and its special use.
Research questions to be addressed in the disquisition include the following:
· What are the unique characteristics of IRC as a medium, and how do these characteristics affect language use?
· In what particular way and to what extent does language use on IRC differ from other forms of real life (RL) synchronous (face-to-face) and asynchronous communication?
· How is orality achieved in a purely text-based environment?
· What evidence supports the claim that IRC is a medium where personal and socio-emotional interactions can take place, and how are these qualities realised?
· What unique characteristics do IRC interactions exhibit in terms of typography, lexicon, grammar, and discursive features?
· Can the observed linguistic phenomena support the proposition that language use on IRC tends to form a special register of the English language?
The treatise is structured into the following main units:
Questions of Computer-mediated Communication. This section defines computer-mediated communication (CMC), the field to which Internet Relay Chat belongs, and provides a brief overview of the evolution of CMC. The section also deals with some relevant and up-to-date questions as well as different views from the literature to help with the definition of the field under scrutiny.
A Background to Internet Relay Chat. This section describes how Internet Relay Chat works, and argues that some basic knowledge of the structure and functions of the IRC network is necessary even for a purely linguistic analysis of the medium. There is also a discussion of channels, nicknames, users, social behaviour, regulation, and action (or third-person) descriptions.
IRC as a Medium. Besides definitions of the medium, this chapter discusses some of the unique features of IRC, such as its playfulness and its much-debated suitability for personal and emotional interactions.
The IRC Data. This section describes IRC data and how corpora are amassed. It is also mentioned what makes IRC an easy field for scholars to research in terms of collecting data, and how the medium eliminates the so-called "Observer's Paradox".
Deconstructing Dichotomies. Since language use on IRC exhibits many characteristics of speech in spite of its physical written (typed) form, this section is devoted to the discussion of how IRC seems to deconstruct these traditional dichotomies of not only writing and speech, but of synchronous and asynchronous forms of communication as well.
Non-Standard Features. This section is concerned with some of the most important linguistic phenomena of language use on IRC. There are subsections which present and analyse extracts and examples in terms of typography (orthography), lexicon (vocabulary), grammar, and discursive properties.
The analysis was based on three different sources, all of which were predominantly available in electronic format. There were two sources of IRC corpora, a large corpus of recorded IRC conversations, and another corpus from the so-called newbie experiment (see description below). The texts containing more than 500.000 letters were obtained from general chat channels (e.g. #new2irc, #ircnewbies, #newbies, #beginner, #casual, #chataway, #cyberparty, and so forth). The third source was the relevant literature, which provided many different approaches, ranging from utopian contemplation to detailed statistical studies (see references section).
The Newbie Experiment
The present thesis contains some references to a so-called newbie experiment and its subjects. The word newbie means newcomer (Puterman, 1994) in the jargon of synchronous electronic conference systems (e.g. on Internet Relay Chat), referring to the subjects of the experiment. These participants, two men and two women, had very little in common, except for their young age, their voluntary participation in the project, and their lack of experience of communication on IRC (Internet Relay Chat). They were different in all other respect, viz. in terms of personality, typing and reading skills, educational background, and their command of English, from upper intermediate to native. Prior to the experiment, each subject was introduced to the new medium, and they were also shown some basic functions of the IRC client program (e.g. entering a channel, how to describe actions, etc.). Then the subjects were asked to communicate through the medium of IRC without any particular restriction of topic and style. The experiment was conducted in the form of individual sessions during which subjects spent a total of over six hours on-line, observing and having conversations on general "chat" channels of the IRC network. All of these sessions were logged (i.e. saved into log files) and stored in electronic text format for further analysis. Both received and produced utterances were recorded. After each session, subjects were asked to describe their first impressions of the new medium and report any difficulty they might have encountered on-line.
Besides amassing corpus, the main objectives of the newbie experiment were to:
* observe and describe unique characteristics of language use on IRC which are not present in other forms of synchronous face-to-face (FTF) communication;
* reveal and describe linguistic characteristics which are not exclusively endemic to the medium of IRC;
* observe and describe the beginning of the process during which communicants on IRC refine their communicative competence in order to adapt to the new medium;
* find evidence and support claims that text-based computer-mediated communication tends to contradict the traditional dichotomy of oral versus written language;
* corroborate assumptions that language use on IRC is an emerging new register of the English language.
The results of the newbie experiment are incorporated in the sections below in the form of observations and pertaining extracts of IRC interactions. Findings are further summarised in the conclusion of the study.
Some Questions of Computer-Mediated Communication
At the dawn of the digital age, when the magic microchip was conceived, there was hardly anyone who envisaged computers becoming a revolutionary tool in the hands of humankind within just a couple of decades.
Although some early experts might have imagined powerful machines which would be capable of handling thousands, let alone millions of operations per second, an amazing performance at the time, the idea of people using computers for global communication was not even dreamt of.
With the exponential development of computer technology, some of the research began focusing on Artificial Intelligence (AI) in an attempt to examine the extent to which computers could emulate human thinking. It was at this stage that communication and computers became related, even though this meant trying to program the computer to behave like a human interlocutor, an effort which is still limited by the fact that to date no comprehensive theoretical description of language and its natural use has been proposed (Dever & Pennington, 1989).
Computer-mediated communication (CMC) can be viewed as a side-effect of the ancient computer networks which connected a few elite government and academic research institutions in the United States some twenty years ago (Herring, 1996). The primary reason for the construction of these early networks was to provide a convenient way of retrieving electronic information. Using the new medium for personal interactions was a further development.
From a linguistic point of view, CMC has rarely been the subject of any comprehensive theoretical research (González-Bueno, 1998), which is largely attributable to the obvious fact that publicly accessible computer networks are also a recent technological breakthrough, consequently CMC is far too young a field to boast long, dusty shelves of books of relevant literature. As a matter of fact, some of the discussion about CMC is still concerned with weather or not the new field deserves academic research at all. Although, as we shall see later, a close analysis of CMC can evidently refute such claims, yet it is still a widely held belief that computers deprive personal communication of its emotive nature, thereby producing an impersonalising and alienating effect upon interlocutors of the electronic medium. Other linguists reluctant to consider CMC a legitimate object of inquiry may find "digistyle" (Danet, Wachenhauser, Cividalli, Bechar-Israeli, and Rosenbaum-Tamari, 1994) difficult to categorise in relation to traditional channels and conventional registers of language use. As Herring points out, "[r]ather than wondering whether CMC scholarship is legitimate, a more appropriate question now is how scholarship can best keep pace with the continued expansion and diversification of CMC" (Herring, 1996, p. 2).
In addition, scholarship appears to be crippled by the lack of well-suited methods to analyse new trends and to describe emerging linguistic phenomena typical of CMC. Therefore a proportion of current research is aimed at testing both conventional and new methods of analysis in an effort to try and place CMC in one of the existing categories (eg. written-spoken or formal-informal), whereas other studies readily assume the existence of a so-called "electronic language" considering it a new variety of English (eg. Callot and Belmore, 1996).
Another question about CMC is what exactly we mean by it. Much of the available literature considers CMC as purely text-based, excluding any additional channels of information (e.g. sounds, visual cues etc.), however, today's computer technology and increased bandwidth have created the possibility of sending information in the form of computer graphics, digitised pictures, and even live audio and video, so computer-mediated communication no longer means only keyboards and speedy fingers, but microphones and cameras as well. Therefore, even though I adopt the terminology, it should be pointed out that CMC hereafter refers to its original meaning, i.e. to text-only digital environments.
A Background to Internet Relay Chat
How IRC works
In order to approach the linguistic features of Internet Relay Chat (IRC), one should have some basic knowledge of how IRC works (Puterman, 1994). The computer program of Internet Relay Chat was developed by Jarkko Oikarinen in Finland in 1988, and the conferencing system soon became very popular all over the world (Reid, 1991). Technologically speaking, Internet Relay Chat (IRC) originally refers to a number of computer networks within a larger network called the Internet. A computer network by definition consists of one or more servers and many terminals, thus digital information can be sent from any terminal to any other terminal or terminals within the same network.
There are several different IRC networks (eg. EUnet, IRCnet, Undernet, EFnet etc.), which may vary in the number of servers and users but share the same characteristics in all other aspects. Since fully comprehending the notion of "virtual places" seems quite a challenge for most of us, it may be a good idea to draw an analogy by turning to tangible objects and places from real-world (RL). Thus IRC networks can be visualised as offices of a big company. People who are in the same room can talk and listen to each other, as each utterance is heard by everyone present. however, communicants can also exchange information privately if they want to do so. Anyone can leave a room and walk into another one to join a new conversation there. It is also possible to invite people from other rooms, but not from another office building. These buildings in our analogy represent separate IRC networks on the Internet.
Channels
"Rooms" of the above analogy are known as channels on IRC. Channels are the key concept of IRC (Reid, 1991), inasmuch as they enable users to create an infinite number of "chat rooms" where conversations can take place. Therefore joining a conversation means joining one of the thousands of existing channels.
In traditional linguistic terms, "channel" refers to the physical medium through/by which information (a message) is conveyed from the speaker to the listener. Even though applying this terminology to channels on IRC may seem a tempting idea, a closer analysis is likely to call for the need for a more complex definition incorporating several notions which appear far more separable in face-to-face (FTF) communication. Thus defining the IRC term channel should include not only the concept of medium through which messages are sent and received but that of place and topic as well. What proves this proposition lies in the fact that people using CMC tend to reify ideas, not unlike users of IRC, who often refer to their chat channels as "rooms" or other physical places. Below is a list of channels with references to physical places (room, place, house, fortress, etc.). The name of each channel starts with a "#" sign, and is followed by a number which is the number of participants who are on that particular channel. This number is often followed by a topic description:
#chatroom 2
#darkroom 12 Welcome to the Darkroom the home of the DarkCrew
#Finland-Room 7
#lesbiansroom 1
#LunchRoom 1
#MrAllthatsBedroom 2 Welcome to the official MrAllthat® channel!!
#pluto 2 CABLEGUY, I got your room back!!!!
#Zgirl 1 <<<<<<Hello and welcome to my play room>>>>>>> be back in 30
#chatplace 1
#crown 5 The place where KINDRED SPIRITS meet
#dutchplace 15
#Jari's_place 1
#mirc_colors 1 The friendly place to be!
#Speyer 1 this place lost his flavor !!!!! by Official
#chatpalace 1
#Alcazar 1 House of Alcazar, int. chat channel........
#barbiehouse 1 matty is back to his roots... VAX rules!!
#castle 1
#Fortress 1
On IRC no channel can exist without any users on it. When the last person leaves a channel, the channel disappears. Similarly, creating a new channel is achieved (paradoxically) by joining a channel that does not exist. As a result of this, IRC channels are of an extremely flexible and dynamic nature, which is further enhanced by the possibility of changing channel descriptions or topics at any moment of a conversation.
As Rodino (1997) points out, IRC channels vary to a great extent, with some of them being more formally organised than others, therefore scholars need limit claims about IRC to certain channels to avoid erroneous generalisations. Werry (1996), for example, emphasising the extreme diversity of the medium, calls IRC channels "small-scale electronic communities" (p.49). In the present thesis, "chat channels" of no special interest provided the corpus for the analysis, for channels of this kind seem to be the most popular on IRC.
Nicks
Upon logging on to an IRC server, the user has to specify a name also known as nickname or "nick" in the IRC jargon. Unlike real names, nicknames, which help identify users on-line, are chosen by the user themselves, and can be changed anytime, even during a conversation.
Besides identification, nicknames have other important, albeit less apparent, purposes. They can provide users with anonymity and create a resultant liberating effect, without which IRC interactions would bear much less originality and uniqueness. Also, nicknames invite playfulness and virtuosity. Most people decide to choose a nickname related to the self to emphasise some positive or negative quality (e.g. <shydude>, <baddady>, <handsom>) or names relating to flora and fauna (e.g. <froggy>, <tupil>, <the-tiger>), whereas relatively few users keep their real names on IRC (Bechar-Israeli, 1995). Since certain names are traditionally gender-related, IRC creates the possibility of gender-switching (Reid, 1991). While changing one's biological sex would involve a great deal of difficulty in real life, on IRC gender-switching is just a matter of impish thought and only requires a simple IRC command (/NICK) to be issued. In addition, due to their psychological effects and highly emphasised importance in the text-based CMC environment, nicknames may also foster the formation of new communities.
Users of IRC
Given the extreme complexity, flexibility and ephemeral nature of the medium, no exact figures are available as to how many people use IRC on a regular basis. Neither is it possible to verify any estimates about statistical distributions with regard to profession, age, sex, location, social status, and cultural background. It is a generally held opinion that there are far more male users than female, and the majority of IRCers are young adults, mostly members of the academic community and students of computer science (Reid, 1991) or some other fields involving computers and high technology. Nevertheless, with the steady expansion of the Internet and its increasing diversity, even such statements are becoming more and more questionable. What remains a fact is that IRC is very popular and populous a medium. To demonstrate this, consider the information below that was provided by a server of one of the largest IRC networks:
13470 channels formed
Current global users: 26499 Max: 41716
Since no channel can exists without at least one user "sitting" on it, one might conclude that on average each channel has two users on. In fact, an IRC user can join several channels at the same time, and there is a large number of single-user channels with people waiting for somebody else to chat with.
Netiquette and Regulation
Having no written laws, IRC might as well be in the state of total chaos and disorder. Anyone who has an Internet connection and a client program can join almost any channel and say (type) anything, enjoying the absolute freedom of speech and expression. Scholars agree that the apparent uninhibiting effect of text-based CMC stems from the relative anonymity and highly limited social/identifying context cues, which characterise the medium (Reid, 1991; Jaffe, Lee, Huang, Oshagan, 1995). In spite of this, IRC does require participants to adopt certain social behaviour patterns often referred to as "netiquette", coined from "network" and "etiquette" (Wauchope, 1997). Below is a typical list of rules in the form of a reminder for newbies (newcomers):
The rules of the channel are: No flooding, no advertising, no repeating, no (op) begging, no caps and English only please.
This declaration of rules seem to apply to the majority of IRC channels.
Flooding means sending too many lines of (often pre-edited) text at the same time, thus blocking the flow of electronic conversation on a channel.
Advertising includes inviting people to join other channels or visit other places on the Internet.
Repeating is small-scale flooding during which a participant repeats themselves. i.e. he or she keeps sending the same text to the others on the channel.
Begging or OP-begging refers to asking for so-called channel-operator rights. A channel operator (also called as chanop or op) has some special rights to keep control over the channel and its users. The first person to create a new channel automatically becomes the operator, and s/he can give similar rights to others who subsequently enter the channel. The rights of a channel operator include inter alia setting channel properties (e.g. invite-only, mediated), kicking or even banning a user out of the channel.
No caps means "no capitalisation", referring to the use of capitalised characters (e.g.HELLO!) interpreted as shouting in CMC (see section on special typography below).
Enlgish only is a requirement on many an IRC channel, but, of course, there is a plethora of other channels where languages other than English are preferred. Nevertheless, English-only channels generally expect every participant to speak English, even if some of the interlocutors share a different mother-tongue. Thus, English continues to enjoy the status of lingua franca on IRC as well.
Although failing to adhere to any of the above described rules may trigger special retributive measures, interestingly enough, this kind of control does not seem to jeopardise the freedom of IRC, as one might infer, but rather these measures are to ensure the smooth and uninterrupted flow of electronic conversation. As a matter of fact, this kind of control is primarily aimed at securing successful communication, which seems to be the essence of IRC.
As it has been established earlier, flooding and repetition is frowned upon in IRC conversations, let alone sending many lines of text at the same time, thus aggressively monopolising the channel. To keep up with the extremely quick turn-taking technique of IRC, users normally send one or two lines of text at a time, and then wait for the other participants to respond. As opposed to FTF, repetition has no real purpose on CMC, as one can easily recall earlier utterances simply by scrolling back on the computer screen and re-read any given line(s). Besides having no important function and conveying no extra information, repetition is felt to disrupt the conversation, making it dull and unimaginative, hence <Merlyn>'s reaction in the extract below:
<NiceGuy> hello may i chat with nice girl ?? PREFEREDLY FROM AUSTRALIA
<<NiceGuy> hello may i chat with nice girl ?? PREFEREDLY FROM AUSTRALIA
<Merlyn> niceguy, stop repeating
<NiceGuy> hello may i chat with a nice girl ?? PREFEREDLY FROM AUSTRALIA
<Merlyn> niceguy, say that once again and i'll kick you
<bubba`> NiceGuy....yes you may
<Zsazsa> right merlyn , be bossy
<bubba`> thats assuming you can find any nice girls in Australia..hehe
<Lilla> haha
<Merlyn> i hate repeaters
But repetition may result in more severe consequences, such as being kicked out of a channel with only a pithy remark in parentheses for the offender to ponder on. In the following sequence, <tompra> repeats himself/herself and is summarily removed (kicked) from the channel by <XenoX>.
<tompra> any girls?
<tompra> any girls?
<tompra> any girls?
<tompra> any girls?
<tompra> any girls?
<tompra> any girls?
<tompra> any girls?
<tompra> any girls?
*** tompra was kicked by XenoX (flood_)
Perhaps another reason why repetition is mostly disapproved of on IRC is that it is not at all a creative technique. Users of IRC tend to prefer more eloquent and original ways which I shall discuss in subsequent sections.
Actions
The use of action descriptions is a unique aspect of IRC (and other CMC conferencing systems), not present in synchronous face-to-face communication (FTF).
The peculiarity of actions on IRC lies in switching from first-person to third-person singular, whereby users report their own actions as if, to borrow Rodino's (1997) term, "an omniscient narrator" commented on a person's actions or state of being. This linguistic phenomenon is hardly present in any other form of synchronous interpersonal communication (Ruedenberg, Danet, and Rosenbaum-Tamari, 1994) and it provides an abundant source of humour, especially for advanced IRC users who are familiar with the action command and its effects upon the conversation. (Note: displayed actions always start with an asterisk.)
IRC command issued by user <Tompa>: /me is hungry
Effect seen by other users: * <Tompa> is hungry
Action descriptions seen in CMC may be related to those present in the genre of comics where both actions reported in third-person singular and direct first-person utterances are applied to facilitate story-telling, but the ability to enrich IRC interactions with action descriptions obviously demands a great deal of wit, appropriate linguistic competence, awareness of context, and a little bit of technical skill (i.e. familiarity with the action command and its use). In the so-called newbie experiment, even though they had previously been familiarised with it, none of the four participants made any active use of the action facility. One reason for this was perhaps the fact that while attempting to communicate on IRC, they were drawing on their expertise based on RL FTF interactions. This repertoire does not contain the use of third-person action descriptions.
Often in an amazingly tricky and virtuoso way, action descriptions may function to overcome the obvious lack of ordinary channels and non-verbal cues (e.g. tone of voice, intonation, gestures etc.) which, as we shall see, characterises IRC as a medium (Danet, 1996a). The interaction below taken from the corpus of the newbie experiment exemplifies such a use of action descriptions:
<Evetke> How old are you?
<Miles> I'm 28
* Miles nose grows
In the above sequence, <Miles>'s third-person description (of his growing nose) invalidates his previous statement of his age by means of playfully referring to a popular character of a children's book, whose nose grew a little every time he was not telling the truth. In real life face-to-face (FTF) communication, the speaker could choose from a much wider range of metalinguistic cues in order to produce the same effect upon the listener(s), e.g. to invalidate the truth value of a previous utterance. One could, for example, use distinctive intonation patterns and/or special facial expressions, such as winking or smiling, or one could touch one's nose as a sign of not telling the truth. But all these non-verbal tools are simply unavailable in the text-only environment of IRC. Thus <Miles>'s utilising the action description facility of the IRC program not only demonstrates special communicative competence (Aitchison, 1992) and virtuosity, but it also weakens claims about text-based CMC which label the medium as being impersonal, distancing, and unsuitable for personal relationships (e.g. Herring, 1996).
Action descriptions may function to express background or extraneous information without creating a disrupting effect upon the actual conversation. Whereas it might be impolite to ignore a first-person utterance, in the case of action descriptions communicants can decide whether or not they want to respond, but the benefit of any extra information conveyed by the action description is often felt later during the interaction.
Some actions describe certain social gestures, such as hugging, kissing, and other ways of showing affection, while others import RL events from the "real life frame" (Ruedenberg et al., 1994) with no evident connection to the on-going conversation. These RL fragments, however, may function to keep the conversation alive or even trigger a new sequence of interaction (for another playful performance, see Appendix C.). Let us consider the following examples:
1. * Deadcow waves to everyone
2. * neichy1 waves goodnight to jazzzz
3. * Ik4u laughs at dinorex
4. * frankay is happy now :)
5. * SteveC turns his central heating up, mmmmmm nice
6. * DaProphet clears his throat
7. * ^Prophet^ is pissed off coz his computer hanged
8. * Merlyn was on the phone
9. * Kali yawns
10. * Graeling sits in the corner and hopes someone will talk to her
11. * Pixie sneezes againnn
<mysty> bless u
12. * Ox-24 is bored...no one to chat with
<bubba`> Ox-24 look around...
Descriptions (1-4) are examples expressing certain gestures and feelings in the purely text-based IRC medium. These written actions also help break the first-person monotony of IRC, playfully transforming the interaction into a kind of script or play. In this manner, descriptions may be utilised to create the illusion of a physical world where non-verbal cues are also available. What actually happens is that interlocutors verbalise the interpretation of their own speech-acts.
Descriptions (5-8) refer to and import RL events, and do not necessarily intend to be a part of the conversation, yet their ulterior purpose is to create some effect or even attract responses. They may also serve to signal personal presence without the need to carry any additional or relevant information. This is a method of compensating for the lack of physical presence and showing the listener's participation.
Descriptions (9-10) do not relate RL actions, but serve to convey indirect references to the participant's state of mind. Thus, action (9) will not necessarily be interpreted as a RL action, i.e. <Kali> has yawned in front of his/her computer, but rather it implies that <Kali> finds the conversation rather boring. Similarly, in description (10) sitting in the corner is not a RL action but an expression of feeling neglected and lonely. Action (10) also overtly indicates her gender and that she would like somebody else to start a conversation with her.
The exchanges in (11) and (12) are examples of third-person descriptions that have triggered an immediate response. In both (11) and (12), the indication of personal presence evinced through an action description is confirmed by a direct response. As proposed above, the aim of such intermezzo-like interactions is to bridge the obvious lack of physical cues on IRC.
Given their playful attitude to their medium, it is no wonder that IRCers often use action descriptions simply to amuse each other with enthralling performances of language and typography as the following examples demonstrate:
* Leia lays DaMMiT on the ground, crawls on top of him and gives him a slow... long... deep... penetrating... sensual... tonsil tickling... orgasmic... hair straightening... toe curling... organ raising... nipple hardening... spine tingling... life-altering... fantasy causing... did i just see GOD??? I think my clothes are falling off! Why arnt we in bed?? Kind of kiss.
* Starfall plays in the ¡.*°*.¡.*°*.¡.*°*.*°*.¡.SNOW *°*.¡.*°*.¡.*°*.¡.*°*.¡.*°*.¡.WITH *°*.*°*.¡.*°*.¡.*°*.¡.*°*.¡. NUDawn *°*.¡.*°*.¡.*°*.*°*.¡.*°*.¡.AND*°*.¡.*°*.¡.*°*.¡.*°*.¡. HITS *°*.*°*.¡.*°*.¡.*°*.¡.*°*.¡.*°*.¡. NUDawn *°*.¡.*°*.*°*.¡.*°*.¡. SQUARELY *°*.¡.*°*.¡.*°*.¡.*°*.¡ IN THE .*°*.*°*.¡.*°*.¡.*°*.¡.*°*.¡.FACE *°*.¡.*°*.¡.*°*.*°*.¡.*°*.¡.WITH A *°*.¡.*°*.¡.*°*.¡.*°*.¡.*°*.*°*.¡.HUGE*°*.¡.* °*.¡.*°*.¡.*°*.¡.*°*.¡SNOWBALL!
IRC as a medium
Ever since pilot studies began focusing on the special linguistic aspects of both synchronous and asynchronous text-based CMC, scholars have been trying to propose a suitable definition for CMC in an obvious attempt to ascribe the observed linguistic phenomena to the unique characteristics of the medium under scrutiny.
Studies more specifically concerned with IRC have also propounded a number of definitions, emphasising different aspects and approaches. Technically speaking, IRC is a form of text-based CMC, and as such shares all of the physical properties, capabilities and constraints that characterise CMC. Any in-depth comparative analysis of IRC, E-mail, and discussion groups would probably reveal that the only underlying difference between IRC and those other forms of CMC is that IRC tends to be considerably more synchronous, hence more interactive, than either E-mail or discussion groups.
Reid (1991), for example, suggests that IRC is essentially a playground where people can experiment with various forms of communication and self-expression. This view is echoed in many other studies listed below, underlining the playful and highly experimental nature of the medium. In addition to playfulness, Ruedenberg et al. (1994) claim that IRC provides suspension of time and space. According to Werry's (1996) definition, this form of communication is a kind of synchronous textual dialogue that takes place between spatially distant interlocutors. Bechar-Israeli (1995) defines IRC as a "silent", "stripped-down" medium characterised by considerable intimacy, with nicknames providing the objects for certain identity games. December (1993) proposes that IRC best exemplifies the oral culture of CMC, while Puterman's (1994) purely technical definition describes IRC as a multi-user computer-mediated chat program, and according to Reid's (1991) definition, "IRC is a multi-user synchronous communication facility that is available all over the world to people with access to the "Internet" network of computer systems" (sec. preface).
There seems to be a great deal of incongruity between the medium of IRC and its primary purpose. As opposed to FTF communication where messages are simultaneously accompanied by certain verbal and non-verbal cues conveyed on extra parallel channels, IRC only relays typed text, excluding all other auxiliary channels of information. Thus, communicants on IRC cannot make any use of intonation, pitch, gestures, facial expressions or any visual/auditory cues which are intrinsically present in other forms of FTF interactions. From this perspective, one might conclude that IRC is quite an unsuitable medium for interpersonal communication and forming personal relationships.
Its unsuitability notwithstanding, IRC enjoys great popularity among people from all over the world, who use the medium primarily for casual conversations or "chatting". On IRC, communication is often for its own sake, i.e. the sole purpose of an interaction is the interaction itself, unlike in other forms of interactions, such as the telephone, where interlocutors often feel the need to initially pre-select the addressee(s) and provide a traditionally acceptable reason for the interaction. A typical telephone conversation would at some point include the explanation of some purpose or a goal other than communication itself, such as asking for information, making an appointment or inviting someone, whereas IRC users usually do not feel the need to specify the reason for their presence or participation. They just wish to have a chat:
<cedro> sombody wanna chat,please
<Zsazsa> we all do cedro, thats why we are here
It can also be stated that much of the playfulness and originality that characterise language use on IRC comes from the aforementioned "unsuitability" of the medium, in that communicants are forced to develop a virtuoso attitude to language and its typed form which carries all the inferable cues about the self. Language therefore becomes performance (Danet, 1997) in which eccentric spelling, grammar, vocabulary, and punctuation work together to compensate for the deficiency of the medium and to convey socio-emotional content.
IRC is a global village (December, 1993), and its inhabitants seem to take great delight in their magical ability to use high technology for communication without the physical constraints of space. People from different continents can talk to each other as if they were sitting at a round table. Being the member of the same "global village", i.e. sharing the same medium, fosters group identity, which in return shapes language use and creates distinctive features that may function as group markers. Puterman (1994) points out that apart from signalling in-group status, the IRC jargon also increases the efficiency of electronic conversations by speeding up interaction through abbreviations. acronyms, slang, and "technitalk", i.e. the use of the special lexicon related to IRC, the Internet, and other areas of computer technology).
The IRC data
From the researcher's point of view, IRC is not only a new and rich testing ground but a very easily observable one as well. Anyone can join almost any IRC channel at any time of day. Utterances appear on the computer screen in the form of digital written script that can be saved and subsequently printed out or edited at will. Therefore loss of information due to noise, phonetic unintelligibility, or any similar difficulties that traditionally plague research of other fields is hardly ever encountered when studying IRC. Moreover, participants seem to be obsessed by the fact that their written lines can be read by all other observers on a particular channel, which apparently fuels wit and linguistic virtuosity. Being an observer on IRC is a popular activity which is called lurking, a term that has no negative connotations on IRC (Danet et al., 1994). As a matter of fact, interlocutors are bound to spend some time observing. One reason for this is that upon entering a channel one cannot have any advance knowledge of the conversation(s) that is/are taking place there, apart from the topic of the channel, provided it has been set and contains any pertinent information. New participants must therefore "listen in" to find out more about the conversation(s) and the actual topic(s). Another reason for unwanted lurking is that a channel, especially a crowded one, may frequently carry several conversations simultaneously, forcing certain participants to read irrelevant information while waiting for a reply.
Observing is such an integral part of IRC that some may prefer it to active participation. People can be so busy observing that they even overtly refuse to have a chat, as in the extract below:
<Cuddly> Hi Steve1
<steve1> want a chat Cuddly?
<Cuddly> Steve1: Sorry, I am busy lurking...
As I have mentioned earlier, personal presence on IRC is shown and reinforced by means of "saying something", i.e. participating in a conversation by typing lines of texts and sending them off to make the utterances appear on the other participants' screens. Those who do not type anything are mostly ignored and have no influence whatsoever on the conversation(s). This helps avoid the "Observer's Paradox" (Herring, 1996). In other words, the observer's presence does not have any (or very little) effect on the interaction(s) or the way in which participants communicate.
All this obviously makes IRC quite effortless for the daring researcher to observe.
Observed data recorded on IRC are referred to by a variety of names in the literature, viz. script, log or log file, transcript, extract, exchange, sequence, interaction, and so forth. Log file and log are exact, albeit too technical, terms for linguists to refer to IRC data. Transcript, however, seems a rather inappropriate word to describe recorded IRC data inasmuch as it implies copying or writing of a text, whereas IRC data need no transformation of any kind in order to be visible or transferable. The reason for this is that IRC data arrive in an already digitised format and as such require no transcription. Exchange, sequence, interaction are applicable terms, although their use may be questionable in certain cases, especially for system (server) messages giving information on users' entering or leaving a particular channel:
*** Jewel21 has quit IRC (Connection reset by peer_)
*** saltydog (xxxx@xxxxxxxxxx) has left #cyberfriends (saltydog_)
*** Ik4u (xxxx@xxxxxxxxxx has joined #CyberFriends
*** LoKiTo (xxxx@xxxxxxxxxx) has joined #CyberFriends
*** Ik4u (xxxx@xxxxxxxxxx) has left #CyberFriends (Ik4u_)
Since recording would be a misleading term (because it implies storing audio or video signals and not electronic text), I adopt extract as being the most general and applicable for my research purposes, for it expresses that the given IRC corpus is presented in its original but fragmented form.
Deconstructing Dichotomies
Scholars plunging into the pioneer research of Internet Relay Chat (IRC) or other forms of computer-mediated communication (CMC) immediately encounter paradoxes that seem to query the authority of some of our most basic linguistic categories.
A question that initially baffles scholars is whether the language of IRC should be categorised as oral or written. From a purely technical perspective, IRC appears to be a written medium, inasmuch as interlocutors receive and send messages in the form of written letters, numbers, and certain other symbols which are available on an ordinary alphanumeric keyboard. On the other hand, the language of IRC apparently exhibits many features traditionally ascribed to orality. Aitchison (1992, p.107) describes the characteristics of written language as follows:
Single writer
Explicit
Non-repetitive
Full sentences
Elaborate structure
Abstract, less common vocabulary
As the analysis in subsequent sections will show, these characteristics of written language do not fully apply to IRC interactions, even if these interactions are produced and received in a physically written form (viz. typed and read). IRC users themselves consider their communication as being oral rather than written, hence the frequency of verbs such as hear, see, tell, talk, listen, and so forth, emphasising the oral nature of the conversation:
<rockette> say again plz?
<Catfriend> *sigh* I really hate it when people don't say hi back
<Merlyn> Lilla, you didn't want to tell your real name, wanna know mine ?
<Merlyn> i'm trying to talk to you here, Bubba :-)
<bubba`> thanx smurf
<bubba`> ok...im'a listening Merlyn
<Merlyn> quack
<Zsazsa> but hearing nothing
Users of IRC readily accept the obvious contradiction of pretending to have an oral conversation while using a medium that only supports writing and reading. In the following extract, the use of verb say refers to the speaker's action and stresses imagined orality, whereas see expresses a RL action on the listener's part, i.e. receiving written utterances by means of reading them on the screen:
<^pUnK^> syn did u see what i said about your astrology pages?
Furthermore, users call their IRC interactions "chats", which reflects the subjective experience of having a spoken conversation rather than a written one. Since the medium relays textual information, this popular concept of orality is not immediately evident for every newcomer. On her first visit on IRC, <Evetke>, one of the subjects of the newbie experiment, apparently conceptualises language use on IRC as primarily written; for she is the only one on the channel to use the verb write to refer to the production of utterances:
<Evetke> I don't understand much in this IRC thing. Could you write some things about this
<Evetke> How many hours a day do you write a day here?
Most researchers agree that text-based CMC cannot be plausibly defined within the frame of the traditional dichotomy of oral versus written language use. Reid (1991) states that synchronous CMC "defies conventional understandings of the differences between spoken and written language" (sec. Preface), and Danet (1996b) considers CMC both as attenuated speech and as attenuated writing, suggesting the existence of a hybrid language. Daly (1996) contrasts the casual, informal, and spontaneous style of CMC with its carefully and self-consciously composed texts. Conducting a scrupulous statistical comparison of CMC, writing, and speech, Yates (1996) finds that CMC differs from both speech and writing in most aspects. December (1993) goes much further and contends that CMC creates a "tertiary form of orality", comparing the effects of the new medium to those of the Greek alphabet and the invention of the printing press.
IRC is generally labelled as a synchronous form of CMC to differentiate it from other asynchronous modes of CMC such as E-mail or discussion groups. Although Wauchope (1997) points out that IRC breaks down the traditional synchronous-asynchronous dichotomy of language use, she does not elaborate on her observation.
If we interpret synchronous language use as the simultaneous involvement of both the speaker and the listener in the interaction, i.e. messages produced by the speaker should concurrently be received and decoded by the listener, then we may find that communication on IRC does not fully meet this requirement. When using an IRC client program, participants compose and edit their utterances prior to sending them off to the other interlocutors who are present on the same channel. Therefore, the process of the creation of utterances remains invisible, and the traditionally parallel roles of the speaker and the listener, which characterise RL synchronous interactions, become temporally separated. On IRC, however, this gap between the production and reception of utterances does not necessarily jeopardise the efficiency of communication, as it would in most RL situations. Moreover, IRC users can even leave their computer and be away for minutes and still be able to remain listeners; for utterances can subsequently be received and interpreted by "scrolling back" on the screen and read the text, thus "catching up" with the conversation.
Non-standard features
IRC is essentially a text-based environment, which means that all interactions take place in the form of written text. The fact that writing in CMC is actually typing on a computer keyboard ostracises conventional graphology from the inventory of related fields and methods available to CMC researchers. Interlocutors cannot have any influence upon the visual representation of the letters (fonts) that make up computer-mediated texts. While in RL interactions (hand)written and spoken utterances may differ considerably in their visual and phonetic representation, respectively, computer-mediated typed texts do not carry any non-verbal personal markers or physical cues.
Let us now consider the following two utterances:
(A) I'm in love again.
(B) I'm in love again.
Either hand-written or spoken, both utterances would probably show a great number of differences and produce different results. If, for example, (A) is either an elderly or a very young person, the utterance will most likely carry some humorous overtones and may even result in laughter. But the same utterance said by (B) sobbingly would have a different effect upon the listener(s). Conversely, if both (A) and (B) are familiar with standard English and use capitalisation and punctuation accordingly, their utterances will probably look identical in text-based CMC. Therefore, to "personalise" utterances, IRC (and CMC) interlocutors are obliged to break certain rules of standard language use and resort to rebellious and eccentric spelling, non-standard grammar, special vocabulary, and the uniquely CMC-specific emoticons (or smileys). Through their special language use, participants on IRC thus become players (Reid, 1991) on a virtual stage where a good performance means the successful and often virtuoso representation of interpersonal and socio-emotional themes in a purely text-based environment.
Typography
For various reasons described earlier, IRC users have developed certain typographic conventions as part of the IRC jargon. Knowledge and frequent use of such conventions marks the interlocutor as an experienced, in-group member of the IRC society. Since speed is a key factor in IRC interactions, special typography also serves to facilitate, i.e. speed up, the conversations by reducing typing need. Correct spelling is not a prerequisite, neither is it a reliable indication of social status or education, for IRC interlocutors often misspell words and use non-standard punctuation on purpose. Spelling errors that cause no misunderstanding are mostly ignored, but those easily lending themselves to sharp and witty remarks will probably attract more attention as in the following two extracts:
<Dytta> I can do a summersault
<mrTR> mean ?
<angelgrl> yeah, but you can't spell it...heehee
* worried is board
<Caley> worried u r spelling bored wrong
<Caley> or are u a surf board
<Caley> or an ironing board
<Caley> hehehee
Despite their excellent tying skills, IRC users tend to omit certain typographic symbols, especially those of punctuation, such as apostrophes, commas, and full stops. Researchers, however, should not hastily conclude that this is a unique phenomenon on IRC, even though it is undoubtedly a highly characteristic one. While some interlocutors may consciously exclude punctuation marks to speed up typing and/or signal in-group status – i.e. to show familiarity with the special language use of the medium – many others seem to leave out commas, apostrophes, and full stops both on-line and off. (The newbie experiment underlies this assumption, as each subject regardless of their RL knowledge of punctuation sooner or later started omitting full stops at the end of their sentences.) Therefore it seems plausible to posit that – although newcomers on IRC inevitably bring their RL orthography into the medium, showing considerable differences in punctuation and sentence construction – more experienced IRC users are likely to adapt to the typographic style of the medium, and even the most pedant participants tend to develop some non-standard language habits, e.g. missing full stops to mark sentence endings. In addition, frequent full stop omission can also be ascribed to the fact that each utterance on IRC is automatically preceded by the interlocutor's nickname in angle brackets and the utterances rarely consist of complex or compound sentences. As a result, the traditional usage of full stops to end all sentences except direct questions and genuine exclamations (Hacker, 1993) loses its functionality in IRC interactions, in that there is no need to use any markers to separate sentences/utterances. The examples below show various types of missing punctuation on IRC:
<NUDawn> im in dallas
<Airbuster> theyre brown
<{jade}> sorry i don't have it to send
<Anni> Hi all
<{jade}> which did you lose chipper
As seen in the last utterance made by <{jade}>, even question marks tend to be omitted, especially when word order and/or context clearly identifies the particular sentence as a question. The use of question marks (as well as other elements of punctuation) appears to follow imagined intonation rather than conventional rules based on grammar. Consider the uses of the question marks in the following extracts:
<sexy`> just for ontario place
<Cr0wley> ontario place?
<sexy`> it's a theme park in toronto canada
<Cr0wley> Oh right.
<Zsazsa> well I am the best one anyway
<Lilla> yea ??
<^cLueLess> so where abouts in england caz?
<^caz> rochdale
<^cLueLess> ???
<^caz> near manchester
<^cLueLess> ahhhhhh
In the first extract, <Cr0wley> repeats "ontario place" and uses a question mark to ask for a definition. In the second exchange, <Lilla>'s question marks obviously express rising intonation to show surprise and doubt. The three question marks in the last extract, however, work as a very concise textual interpretation of a non-verbal cue, possibly a facial expression of puzzlement.
Capitalisation is no exception to the eccentric typography of IRC. Owing to economy of typing and the lack of need to mark sentence boundaries, capitalisation of first letters is generally ignored in IRC interactions to such an extent that even the first letters of proper names are often spelt in lower case (see example above). This reluctance to use capital letters may partly be due to the fact that on IRC (and CMC) capitalisation is traditionally interpreted as shouting, and excessive use of it is frowned upon. Of the extracts below, the first one demonstrates that some participants are aware of their non-standard use of capitalisation. The other extracts contain typography intended to represent shouting.
<DaMMiT> its buy guns and roses..
<DaMMiT> by even..
<DaMMiT> damn i need grammer check and spell check for mirc
<Meth> whoa
<nudehamster> it would tell you to capitalize
<Ruuuude> COS I'M RANNNNNDY
<joh> bye every1
<moonray> yes!!!! I thought you never asked!
<Ruuuude> WHAT MOONRAY
<Cubby> why do you shout ruuuuude
<DaProphet> WAAAAAAAAAAAAAAAAAAAAAAAAKE UUUUUUUUUUUUUUUUPPPP!
<^Demon^> YEHHHHHHHHHHHHHHHHHHHHHHHHHHHHH
<^cLueLess> U CALLING ME MAD?
Werry's (1996) proposition that "the language produced by users of IRC demands to be read with the simultaneous involvement of the ear and the eye" (p.59) can be further justified by focusing on some other features that IRC typography exhibits. One such characteristic (often combined with non-standard capitalisation) is the use of multiple punctuation and spelling, a technique to express emotional content through the unusual visual representation of typed electronic texts:
hellllllllllo
it is soooooo good
im sooooooooooo sick of genereal ed
* Cierra is sssooooo coolldddd
I'll be back later.. just realised fooooooooooDDDDDDDD!!!!!!!!!!
<ShaDowLin> Hiiiiiii Lise
<mstng_ldy> hi lise!!!!!!!!!!!!!
<bonn^^> lise!!!!!!!!!!!!!!!!!!!
* Nicho^ can handle heat in day but likes it to cool off at night so he can sleep properly :)
<Gerbie> Lovely lise!!!!!!!!!!!!!!!!!!!!!!!11
<alsandor> bonn, chip has BEEN HERE???????????????????????????????????????????????????????????????
<alsandor> bonn, ??????????????????????????????????????????????????????????????????????????????????????????
<alsandor> bonn, ??????????????????????????????????????????????????????????????????????????????????????????
<alsandor> :-)
<oreosmile> als: just let them get you all fired up for your wifey-poo
<bonn^^> als:yes al...calm down
<Buffy> mystyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyy
<mysty> buffffffffffffffffffy!!
<mysty> :)
<mysty> hehe
<MsBehav|n> pppppppppppppppppppppppiiiiiiiiiiiiiiiiiiiiiiiiiiiixxxxxxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxeeeeeeeeeeeeeeeeeeeeee
<pixEEEEE>
MsBeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeee eeeeeeeeeeeeeeeeeeeeeeeeeeeeee :)
Wauchope (1997) notes that a slow typist on IRC is disadvantaged and may even be considered less intelligent. Conversely, a speedy typist with quick, witty responses is always welcome and immediately accepted as an experienced member of the virtual society. To increase typing speed, communicants on IRC are constantly forced to invent ways in which utterances can be shortened without any detriment to meaning and intelligibility. These techniques of "typing economy" include simplified spelling based on pronunciation, single or combined letters or even numbers which can substitute one or more syllables when spelt aloud, as well as other symbols available on the alphanumeric keyboard.
Typed textInterpretation
thot
thought
tho
though
alrite
all right
w/
with
w/o
without
y
why
r
are
u
you
c
see
fa q
fuck you
2 (me 2)
to, too (me too)
every1, some1, no1
everyone, someone, no one
l8er, l8r
later
c u l8r
see you later
m8
mate
age <16
older than 16
tanx, thx
thanks
OIC
Oh, I see!
ne1
anyone
It has already been suggested that users of IRC conceptualise their medium as primarily oral in spite of its physical nature and text-based environment. This fundamental contradiction accounts for the playfully rebellious approach to language, emphasising the importance of orality. Communicants tend to regard writing as a kind of flexible medium to represent speech, and they seem to take particular delight in what normally could only be spoken or "scrawled behind the closed door of a public toilet wall" (Danet et al., 1994). The freedom and liberating effect of IRC stems from this dual attempt at creating a visible language and establishing social relationships. Consequently, the effort to represent speech in the form of typed text – i.e. to present the unpresentable (Reid, 1991) – is a very important part of the IRC "performance". Participants do not think twice to break rules of spelling, punctuation, and even syntax in order to create more speech-like utterances. Although the resulting texts occasionally seems to bear signs of typing economy (e.g. em for them), the main purpose is to accentuate the oral, personal, and more casual nature of IRC interactions. The following extract seems to refute the economy theory, supporting Daly's (1996) observation of the hidden discrepancy between casual, informal style and carefully composed texts:
<Cr0wley> How are you doin' bud?
It is evident that typing doin' requires just as much, if not more, effort and time as typing the full word doing. The speaker even utilises an apostrophe to mark the missing letter g, making it clear for other communicants that he/she is consciously representing phonetic features and not simply omitting the last letter for purpose of economy. Similar efforts can be observed in the extracts below:
NUDawn> well redog has a page for a bunch of em
<Starfall> Welll
<helraizer> been watchin ya
<Deadcow> so whats goin on in here?
<{HeLiOs}> tigs, can you remember what we sposed to be doing today????
<lyser> shudup
<g|rL> lemme see :)
<Kewl-Gal> I like brown eyes... dark n mysterious..
<Hellbound> bah, I'm sick of classes 'n shit
<Kewl-Gal> wassup punk..?
<CpCaveman> worra fecking wanka
<ferngirl> puhleeeeze
<LifeGard> Any gals from or near Atlanta?
<Meth> ordering sumfin
<bubba`> cowboys ride 'em farmers milk 'em
<^SkY^> I duno
<^FieryRed> bout 5'6
<peaches^^> i must push myself cos i have to be fit
<sexy`> that's sorta funny
<^Prophet^> you wanna know where i come from or where i live ?/
<Caley> gimme a sec Crow
* moonray is sad merc doesn't wanna speak to her anymore
<Ruuuude> how da hell did we get into this shit
<alphaUS> moonray: yuh, lemme do my stuff.. don't care about it :)
<CpCaveman> i bet if we work ard enough we`ll give her a complex
Users of IRC have also devised, and are constantly experimenting with, ways to transcribe, as it were, non-verbal sounds and noises through eccentric typography. Some of these elements originate from comics, such as various textual representations of laughter or "zzzzzzz" to indicate sleeping or snoring:
<EvilMoose> hahahahahahahaahahahaha
<CpCaveman> hahahahahahahahahahaahahahaha
<FredDerf> mmmmmmm i'm imagining the tounge action
<^cLueLess> ermmmmm not sure what the date was actually crafty
<Matumbo> I like peanut butter and tuna
<Lise^^> ewwwwwwwwwww matumbo - what a combo
<bonn^^> tuna and peanut butter? ewwwwwwwwwwwwwwww
Marengo> zzzzzzzzzzzzzzz
<Caley> zzzzzzzzzzzzzzzz
<Marengo> heheh
<DaProphet> WAAAAAAAAAAAAAAAAAAAAAAAAKE UUUUUUUUUUUUUUUUPPPP!
Emoticons
The most original by-product of artistic efforts to charge typed electronic texts with emotional cues is the system of emoticons (also called smileys, smiley faces, smiley icons, graphic accents). Emoticons are simple strings of alphanumeric characters and punctuation symbols (Reid, 1991) which, when looked with the head turned sideways, can symbolise various facial expressions. Emoticons are very popular in text-based CMC, because they have the ability to add emotive content to utterances. In this sense, emoticons are transcribed emotions and highly emotive actions. This process involves the speaker's having to disassemble his/her utterance into two distinct elements: the verbal utterance itself and the emotive content transcribed as an emoticon. The listener has to reverse this process and combine the purely verbal (textual) information and the transcribed emotive cue(s).
Some examples of emoticons (for a more comprehensive list, see Appendix A):
:-) smile
:-( frown
;-) wink
;-( angry
: P drooling
<8-) smile with glasses and a hat
The effects of emoticons upon the meaning and perception of positive and negative messages seem far more complex and inconsistent than their simplicity may suggest (Witmer, Katzman, Colman, 1997). Moreover, users of IRC tend to employ emoticons as friendly gestures and often without any obvious reasons. Special uses and variations are also frequently observed, such as applying multiple spelling for emphasis:
they put hot peppers on my sandwich :((((((
An even more virtuoso performance is the recycling of emoticons into one of the original functions of a constituent (Ruedenberg at al., 1994). In the example below, <maoui> recycles a smiling face into a parenthesis:
<maoui> You know, that's what "they" (the astrology gurus :) say
Other even more artistic experiments include depiction of movement through a sequence of emoticons:
<Succubus> :PppPPPpPPpPppP
<Succubus> look at that tongue action
Emoticons are a truly original characteristic of language use in text-based CMC, therefore newcomers need learn how to interpret and utilise these graphic symbols to give a better performance on the virtual stage of IRC. In the following sequence, <Dytta>, one of the subjects of the newbie experiment, takes her courage in both hands and asks for exmplanation:
<Dytta> How old do you want me to be?
<romeo19> 2 year old toddlers
<romeo19> :)
<Dytta> 23, but what's this :)? This is my first IRCing.
<romeo19> this is a smile..look side ways
<romeo19> : the eyes
<romeo19> ) the lips
<Dytta> looks good
To impress her, <romeo 19> decides to show off his knowledge by sending <Dytta> a different kind of emoticon:
<romeo19> @>---@>--@>>--
<Dytta> Is this your age?
<romeo19> thant's 3 roses for the lady!
Apparently, she still fails to successfully interpret this abstract level of keyboard art, so he has to give her a verbal description. The fact that she has already encountered emoticons (on a channel called wasteland) is revealed later during the conversation:
<Dytta> Sorry, I thought you have 19 ears.
<Dytta> :/
<romeo19> he..he..what a monster is that!
<Dytta> I saw it wasteland
Similarly to acquiring the proper use of RL registers, language use of IRC also takes some time to master. Thus, it is only with the more experienced users of IRC that frequent utilisation of emoticons and other typographic features, i.e. familiarity with the medium-specific jargon, can be observed.
Lexicon
The appropriate use of language presupposes good communicative competence, i.e. knowing what to say when (Aitchison, 1992). Given its freer and relaxed style, IRC might seem a medium that does not expect interlocutors to pay particular attention to their choice of vocabulary, since utterances often contain slang and even four-letter swear words. As a matter of fact, the ability to choose the right word is indispensable to successful communication on IRC, where language inevitably becomes the purveyor of personal markers, attitudes, and social cues. This heightened importance of language use together with the expected brevity of utterances demand a more conscious choice of vocabulary on the speaker's part, which, in some respect, seems to contradict the description of the medium as relaxed and free in register. Users of IRC appear to share some covert alertness to the appropriate use of vocabulary, because catching any blunders may provide a chance of witty performances:
* Nicho^ can handle heat in day but likes it to cool off at night so he can sleep properly :)
<mstng_ldy> how do you sleep improperly, Nicho????? haha
Besides colloquial, slang, and swear words, the IRC jargon also consists of three distinguishable sets of lexicons. One such group of words and expressions is often referred to as "technitalk", i.e. terminology related to high technology, electronics, and computer science (e.g. re-boot, mouse, click, server, hang, bug). The second set of lexicon often used on IRC is concerned with CMC and the global computer network called "Internet" (E-mail, URL, ping, bandwidth, cyberspace, spam, IMHO, BTW, emotiocons). Words which belong to the third group are uniquely used on, and in terms of, IRC. The relationship of these three sets of vocabulary items is shown below:
It is especially IRC terminology (i.e. words belonging to the third and smallest group) of which participants need be cognisant if they want to communicate successfully on IRC. As with all social groups, the IRC jargon proper also fulfils two parallel functions: it makes communication more efficient by providing new items of vocabulary, and it creates an explicit marker of in-group status. The jargon is relatively limited, but newcomers often have to ask for definitions.
A proportion of the special IRC lexicon have been coined by programmers and computer experts who also created the system of IRC and wrote the necessary computer programs. Consequently, it is no wonder that the majority of these words are similar to those of computer languages, in that they are mostly abbreviations and acronyms. The two most frequent words of this kind are MSG and DCC, both of which can function as nouns or transitive verbs in the language of IRC:
<kazoo> look, helen, i DCCd you!
<kazoo> why don't you DCC me?
<SINAAL> Anyone wanna chat with a 19/M/Fr ? MSG me please
DCC stands for Direct Chat Connection, and it enables users to establish a direct connection for private conversations or sending/receiving computer files. MSG means private message, i.e. one or more lines of text with a specified addressee.
Below is a list of some frequent abbreviations, acronyms, and other items of the jargon (for a more extensive list, see APPENDIX B.).
IRC term
Meaning
WB
welcome back
RE
hello again, welcome back
LOL
laughing out loud (to express laughter)
ROFL, ROTFL
rolling on the floor laughing (to express laughter)
M
male
F
female
ASL
age, sex, and location (asking for personal information)
DCC
(to have a) direct chat connection
MSG
(to send/receive a ) private message
BRB
(I will) be right back
BBL
(I will) be back later
NP
no problem (in response to "thank you")
LO
hello
OMG
Oh, my God! (exclamation)
WTF
What the fuck! (exclamation)
lag
(to have) a slow connection, delay
kick
to remove a user from the channel
chanop, op
channel operator
ping
to check the speed of network between two computers
ppl
people
k
O.K.
bot
robot
Users of IRC inevitably adopt the love of coining and using acronyms. Despite their frequent use, these words show a considerable amount of flexibility and variation, which is another possible source of playfulness:
<SaBaker> write back?
<MrsCopper> wb=welcome back
<Macy> wb = welcome back
<SaBaker> oh
<Macy> warner brothers?
<MrsCopper> water bottle?
<SaBaker> willing buddy??
<Macy> western britain?
<MrsCopper> weather beacon?
<SaBaker> why buy
<MrsCopper> west berlin?
<Macy> wet basenji?
<SaBaker> what bongo?
*** slip^ is now known as cot^
<cot^> hiyas:)
<MrsCopper> so many options.......how can you choose?
<Macy> :)
<SaBaker> well ill guess it means welcome back
<SaBaker> COT wb
<SaBaker> what became
<SaBaker> really
Truncated or clipped words are also quite frequent but used more inconsistently, such as puter and comp for computer, addy for address, diff for difference, morn for morning, and so forth. This kind of abbreviations hardly cause any problems for newcomers, whereas the interpretation of baffling acronyms usually does (<Evetke> and <Mul> are two subjects of the newbie experiment):
<Evetke> What is LOL?
<Rastababe> laughing out loud
<Miles> evetke lol is large orange lobsters
<Rastababe> miles be good
<Mul> pardon me for asking , but what does LOL mean
<__pat> Mul..... LOL=Laugh Out loud... to make up for the diificulty of expressing appreciation of humour
The newbie experiment has shown that newcomers upon their first visit to the world of IRC need ask for
explanations and definitions of certain frequently employed expressions and words in order to interpret utterances. Therefore it seems plausible to conclude that the lexicon of the language on IRC exhibits features that clearly differentiate it from the kind of vocabulary used in other forms of FTF communication and RL situations.
Grammar
Compared to the predominantly obtrusive phonological differences, grammatical variation of different dialects of English tends to be less extensive (Greenbaum and Quirk, 1973). This observation also seems to hold for the grammatical features of language use on IRC, inasmuch as communicants cannot break basic grammatical rules, as it would most likely result in unintelligibility. Consequently, grammar cannot (and need not) be exposed to such a high level of creativity and virtuosity that characterise other aspects of language use on IRC, viz. typography (orthography) and vocabulary.
Non-standard grammar may be observed both in speech and writing. While in speech non-standard grammar is a sign (and result) of a colloquial and freer style (Pearson, 1977), in writing it is applied mainly for economical purposes, such as in newspaper headlines and notes. On IRC, however, both purposes of non-standard grammar are discernible.
To speed up the conversation, participants of IRC tend to make frequent use of elliptical structures while producing their utterances. Although ellipsis is also present in speech, especially in informal registers, the extent to which elliptical structures can be employed in IRC interactions rather characterise highly condensed written styles such as newspaper headlines or notes. When turning to this kind of ellipsis, IRC interlocutors consciously or subconsciously sacrifice certain oral qualities in order to produce faster utterances, which proves that speed is regarded a more important factor. In most cases, however, ellipsis works as a tool for both economy (speed) and expression of orality. What most frequently occurs is the ellipsis of subject pronouns, and auxiliary verbs. Werry (1996) also observes that subject pronouns are often deleted, because the speaker's nickname automatically precedes his/her utterance (e.g.: <Marlena> can't stand it).
Examples of elliptical structures:
gotta work in 2 hours
sure will
been watchin ya
Anyone know a good channel to hang out in?
why that nick ? :)
<^Lum> thx: Megg :))
<Megg> pleasure Lum :))
why so many ops
<Sluggie> told me she was 8
brb have to get pizza out of the oven
<Kadi> where u from :)
<Zero-Cool> Usa....u
<Kadi> Brazil - where in the US
<Zero-Cool> Dallas
Anything worth looking at?
you going?
anyone like U2?
were u from
<Tassle> you talking to me?
<Nicho^> bonn ... how long you going for?
<Marlena> can't stand it myself
Structures with a higher level of ellipsis are much less frequent, as they tend to deprive the conversation of its oral quality:
brb phone
u from?
how old
so what up
Non-standard grammar also manifests itself in non-standard concord (subject + verb, possessive pronoun + noun phrase), double negation, and the use of ain't, emphasising the oral and colloquial quality of interactions:
and thats why i likes ya
no methinks they are non beleivers
i luvs ya mandy
i changed me colour so ppl would take me more seriously!!
you mean you aint a mental case?
i never get none i'm sad
Ellipsis is not the only grammatical trick that aims at writing in a more speech-like manner and at the same time trying to keep up with the usually fast pace of conversations. Participants usually produce short and simple sentences. Complex and complex-compound sentences are few, for their production is too time-consuming; nevertheless, as Werry (1996) points out, messages tend to become longer (and more complex) when there are fewer participants. The rarity of complex and compound sentences explains the relatively infrequent use of relative pronouns, conjunctive adverbs, and correlative conjunctions, all of which characterise formal written style but definitely not casual conversations. Utterances on IRC containing more than one clause generally utilise parataxis, rather than co-ordination or subordination (December, 1993). Clause boundaries are often visually marked with multiple punctuation or other typographic symbols (e.g. emoticons), or remain unmarked:
<korna> who are you dj do i know you?
<SteveC> kait have you been strutting your stuff on the tennis courts today?
<Kait> steve..i got mybutt kicked this morning...8/8 and 8/2
<`deb> i'm fine shellie how are you?
<Cuddly> Steve1: Sorry, I am busy lurking...and I am male!
<bonn^^> i love my meat....i gotta have meat myself...i could never be a vegetarian
<mrs> i can and will cook anything that contains meat i just wont eat it
bonn^^> dave..how is andrew and his spots?
<DaveH> bonn, seems to be levelling off... haven't seen any new spots in 2 days... everything is scabbing over (afraid for scarring -- kept scratching in his sleep)... Almost no fever now.
<RIP`> I wonder what Inge would think of that :) better not ask her though :)
As the newbie experiment suggests, on IRC multiple punctuation is a special typographic feature that most newcomers do not initially have in their repertoire (none of the subjects made any use of it), while the fact that utterances tend to consist of simple sentences and paratactic clauses is largely attributable to way in which the IRC system works. (Our subjects almost exclusively composed simple and short utterances, probably because they were far too preoccupied trying to become accustomed to the new medium and its challenging use.)
Some discursive features
Although a discourse analysis of interactions on IRC, even it a cursory manner, would certainly exceed the limits of the present study, it is worth examining and pointing out certain characteristics of electronic discourse in order to provide more evidence for our IRC vernacular theory. Moreover, a discourse analysis of IRC interactions also presents the opportunity to scrutinise available extracts in their integrity, thus synthesising all the observations and described features (viz. action descriptions, typography, lexicon, grammar and so forth) which have been dealt with in earlier sections. The extracts below also demonstrate that participation in on-line interactions, i.e. being able to communicate through the new medium, has priority over the topic of the conversation. In this sense, participants join these channels to enjoy membership of a social group, rather than utilising the medium to obtain information for any specific purpose. Anything can be a topic, ranging from personal particulars (age, sex, location) to nail clippers (see extract below). Topics come and go, and they are immediately disposed of as soon as they are no longer felt to stimulate the conversation, or perhaps a more interesting topic is brought forward. Participants tend to follow a kind of stream-of-consciousness technique (Marvin, 1995), and topics are typically elaborated on in a playful manner:
* Nicho^ needs sharper nail clippers ... sheesh !! :)
<bonn^^> those darn old toenails anyway eh nicho?
[...]
<Nicho^> bonn .... fingernails are the prob .... cant get through ones on thumbs ! :)
<bonn^^> then you have good zinc in your system...then
<bonn^^> that's what causes good or poor nails
<Nicho^> bonn ... that what it is? ... very tough nails :)
<bonn^^> yep...and when you get white spots on nails...lack of zinc...
<Nicho^> bonn ... I see .... gonna have to use scissors I think :)
<bonn^^> ewwwwwwwwwwwwwwww
<MsBehav|n> wow learn something new everyday
* ShaDowLin checks her toenails
[...]
* ShaDowLin checks Gerb's toenails
<Nicho^> MsB .. hehehehee
<Gerbie> toenails?
<bonn^^> yep gerby...toe nails
<ShaDowLin> hehehehe
* ShaDowLin paints gerb's toenails ruby red
<MsBehav|n> nicho all i knew was it helped w/ colds and not to get the natural cherry zinc tabs...burn yer tongue and make everything taste funny hahahaha
<Gerbie> LOL
<Gerbie> not my color
When a particular topic becomes worn away, to fill the resulting gap, participants usually start proposing new themes in a various ways, and/or make mostly humorous comments in order to amuse each other, signal personal presence and keep the conversation alive. The following extract is an example of this "gap-filling" activity:
<dinorex> .
<dinorex> ..
<dinorex> ...
<dinorex> ....
<dinorex> ...
<dinorex> ..
<dinorex> .
<dinorex> hehe
<dinorex> modern art..
<^SiD^> hahahaha
Although newcomers seem to find it difficult to keep up with the speed and dynamic nature of topics, they soon learn how they can command attention and playfully participate at the same time. After several futile attempts at joining the conversations by asking questions, all subjects of the newbie experiment discovered a technique of taking advantage of their newcomer status by drawing on and exploiting the other participants' curiosity. <Mul> and <Dytta> even started to play games, which would have been inappropriate in most RL FTF interaction between communicants who are complete strangers to each other:
<Mul> but i'm not in the land of aus at the moment
<Junely> well i wondered cause you addy didnt look like it mul...where are you
<Mul> Junely , i'm somewhere in the northern hemisphear
<Junely> well mul i told you cant you fess up
<Mul> ok a city in eastern europe
<Junely> well thats more of a detail mul :)
<Mul> it starts with a B and ends with an S
<__pat> ends with S?
<__pat> u sure?
<Mul> sorry i fibbed , it ends with a T
<Junely> hmmmm bolivia???
<Junely> ooops thats south america lol
<Junely> wellllll
<__pat> LOL junely
<Junely> another b....lets see
<Mul> think Goulash
<Junely> cant you even give me the country
<Junely> boulash????
<Mul> nem nem
<Mul> goulash
<Mul> Ex commie country
<Junely> but you said b
<Junely> how bout budapest???
<Mul> yes yes , you win the two cent prize
<Junely> thats all i get!!!
<romeo19> u?
<Dytta> guess
<Dytta> help: Europe
<romeo19> wait
<romeo19> 'i'm not good in this guessing game
<romeo19> England
<Dytta> nope
<Dytta> H...
<romeo19> holland
<Dytta> Hungary
I have already stressed participants' playful attitude towards their electronic medium, where everything (even gender and language) can be treated as an object of play (Reid, 1991; Rodino, 1997). Newcomers immediately sense and quickly adopt this attitude. Here is another example of such playful performances:
<dinorex> 1
<star> hi icemanuk
<dinorex> 2
<star> remember me?
<dinorex> 3
<NiceGuy> hello any nice girls wanna chat ??
<dinorex> 4
<dinorex> 5
<hotman1> any girls want hot man?
<star> Hi NiceGuy
<dinorex> 66
<dinorex> ops
<dinorex> 6
<star> Hi HotMan
<dinorex> 7
*** hotman1 was kicked by IcemanUK (I told you not to say that )
<dinorex> 8
<dinorex> 9
<star> boys stop fighting
<dinorex> 10..u wanna know what Im doing?
<dinorex> 11
<NiceGuy> hello any nice girls wanna chat ??
<star> no
<dinorex> 12
<Merlyn> you're being annoying
<delerium> 13
<dinorex> no-....
<delerium> 14
<dinorex> Im counting my iq.....
<IcemanUK> :)
<Merlyn> duh :-)
<delerium> then u must stop now
<Merlyn> he already counted too far :-)
The above extract also illustrates what Werry (1996) calls multidimensional texts, referring to the fact that the IRC system juxtaposes disparate strands of interactions, making it quite difficult – especially for the less experienced – to separate one conversation from the other. Intertwining sequences can occur because no overlaps and interrupts are possible on IRC (Werry, 1996). In other forms of synchronous interactions involving more than two interlocutors, e.g. in amateur radio "nets", where overlaps are possible, participants need apply certain controlling and turn-taking methods (net-control) in order to avoid loss of information (Hildebrand, 1997). On IRC, however, no such measures are necessary, since utterances, even if they have been produced simultaneously, appear in the chronological order in which the IRC system receives them (Werry, 1996). Therefore, interlocutors have to follow and process (i.e. infer the addressee of) each utterance of all of the intertwining conversations if they want to participate successfully. Although the extra work of processing irrelevant information may seem a disadvantage, it also enables participants to follow and join in any other parallel conversation whenever they wish to do so. Applying the traditional model of communications theory to IRC interactions, we may also find that dealing with extraneous information is also present in RL FTF communication, which requires the listener to recognise and filter out noise. In this respect, processing and interpreting intertwined strands of conversations and system messages is similar to the process of filtering out noise on RL channels.
The following example demonstrates "digital noise", i.e. a structure of intertwining exchange. <KewlGal> and <Airbuster> are having a conversation about eyes, while <Strife> and <g|rL> are discussing astrology:
<Kewl-Gal> eyes are the windows to our soles... I love looking into eyes...
<Airbuster> I hate looking in eyes
<Strife> what's the sign for june 15 anyways?
<Airbuster> makes me sick
<g|rL> gemini
<Kewl-Gal> why Airbuster..??
<Strife> gemini?
<Kewl-Gal> what on earth for Airbuster..??
<g|rL> yes
<Airbuster> I feel like my heads about to explode
<Strife> do they go with pisce?
<Kewl-Gal> eyes are so beautiful...
Processing simultaneous strands of conversations present on IRC requires skills which one does not need to possess in other forms of FTF communication, hence the difficulty that newcomers experience while attempting to have a conversation on IRC. Since non-verbal cues (e.g. eye-contact) indicating the addressee of an utterance are not available in the purely text-based environment of IRC (Werry, 1996), participants often specify the addressee by including his/her nickname in the message. The newbie experiment suggests that even this simple technique is one of those to be learnt on IRC, since none of the subjects used it initially. In the extract below, <Mul>, a newcomer, does not specify the addressee of his request, which prompts <__pat>, one of the possible addressees, to ask for reinforcement:
<Mul> tell me something about yourself
[...]
<__pat> me Mul?
<Mul> ah yeah
<Mul> that means you Pat
<__pat> mul im 21, a comp sience student
<Mul> learns the lesson, and subsequently specifies the addressee of his utterances to avoid ambiguity whenever he thinks it is necessary:
<Mul> where are you from Junely?
Another lesson for newcomers to learn is that, unlike in FTF or telephone interactions, receiving messages and composing them seen to occur simultaneously: one may be typing one's text and still be able to receive messages from other participants. Since composing utterances and processing incoming messages cannot be done at the same time, on a busy channel participants have two options: They type their messages as fast as possible and then read any pervious utterances they have missed, or they frequently stop composing their messages to process any new messages which may have arrived. Since neither of these techniques is truly endemic in RL FTF communication, it is no wonder that newcomers tend to type less for fear of missing any utterances addressed to them, or find it quite confusing to type and keep up with the flow of the conversation at the same time. (Cf. the notion of multi-tasking in Ruedenberg, 1994). Follow-up comments (and complaints) made by subjects of the newbie experiment also supports this observation.
Conclusion
Internet Relay Chat (IRC) is an electronic medium for communication between spatially distant interlocutors. In spite of the fact that the medium relays information in a purely written (typed) form, participants have devised certain techniques by means of which written utterances can carry paralinguistic cues. Some of these techniques have been – and are constantly being – conventionalised, which is a result of the apparent effort to create a social group, with written language being the only available carrier of status markers. Interactions on IRC are characterised by a playful, creative, and experimenting attitude towards language, providing an abundant source of virtuosity. Thus, aspects of identity, gender, and even language itself become performance.
IRC as a medium cannot plausibly be categorised according to the traditional dichotomy of written versus spoken forms of communication. Interlocutors have to write (type) and read utterances, yet they refer to their interactions as "chats" and tend to pretend to have synchronous face-to-face (FTF) conversations. Utterances are consciously composed and edited in such a way that they reflect features of spoken language. Participants have at their disposal a set of techniques which they frequently apply to emphasise oral qualities, such as multiple punctuation and typography, non-standard orthography to reflect pronunciation, double negation, ellipsis of subject pronouns, and simple sentences or predominantly paratactic structures instead of co-ordination or subordination.
The newbie experiment revealed that successful and efficient communication on IRC requires newcomers to improve and expand their communicative competence based on real life (RL) registers, asynchronous communication, and synchronous face-to-face (FTF) interactions. Multiple punctuation, multiple typography, and emoticons are some of the special features that interlocutors have to learn on IRC. Another result of the newbie experiment is that pretended orality is not immediately evident for newcomers (their utterances being much better formed and have few, if any, non-standard features). The subjects also had difficulty in interpreting certain vocabulary items, especially acronyms (e.g. LOL), which they had never encountered in RL interactions.
In terms of discursive properties, IRC also shows some unique characteristics which did not initially occur in the utterances produced by the four subjects of the newbie experiment. These features include – inter alia – the use of action descriptions, frequent specification of addressee within a particular utterance, composing and receiving messages simultaneously, and following/processing intertwining messages (reconstructing a multi-dimensional text).
Our results of the analysis of the unique linguistic characteristics on IRC, therefore, appears to support claims that language use within the new medium blends features of both writing and speech, and has some "digital" or "electronic" properties not present in other forms of communication (e.g. emoticons, action descriptions). The observed linguistic phenomena also seems to partly corroborate assumptions proposed by other scholars (e.g. Ruedenberg, 1994; Herring, 1996), describing the silent and visible language of IRC (and that of other forms of computer-mediated communication) as an emerging new register of the English language.
The IRC Vernacular
A Linguistic Study of Internet Relay Chat
Az Internet Relay Chat nyelvi sajátosságainak vizsgálata
SZAKDOLGOZAT
Gelléri Péter
angol nyelv és irodalom szak
Témavezető: Németh Nóra
1998.
Table of contents
Introduction
This study analyses the linguistic characteristics of a new medium of synchronous computer-mediated communication called Internet Relay Chat (IRC). With thousands of people using it from all over the world at any given moment, IRC is undoubtedly one of the most popular conferencing systems in the world today. Despite its amazing popularity and global accessibility, however, IRC has not attracted much attention from a linguistic point of view. Nevertheless, the reason why some scholars are starting to consider the new medium a legitimate field of study is partly because it provides an easily observable and incredibly rich testing ground for linguists and sociologists alike, and partly because interactions on IRC affected by the characteristics and physical constraints (especially its text-based nature) of the medium may reveal an incredibly new approach to language and its special use.
Research questions to be addressed in the disquisition include the following:
· What are the unique characteristics of IRC as a medium, and how do these characteristics affect language use?
· In what particular way and to what extent does language use on IRC differ from other forms of real life (RL) synchronous (face-to-face) and asynchronous communication?
· How is orality achieved in a purely text-based environment?
· What evidence supports the claim that IRC is a medium where personal and socio-emotional interactions can take place, and how are these qualities realised?
· What unique characteristics do IRC interactions exhibit in terms of typography, lexicon, grammar, and discursive features?
· Can the observed linguistic phenomena support the proposition that language use on IRC tends to form a special register of the English language?
The treatise is structured into the following main units:
Questions of Computer-mediated Communication. This section defines computer-mediated communication (CMC), the field to which Internet Relay Chat belongs, and provides a brief overview of the evolution of CMC. The section also deals with some relevant and up-to-date questions as well as different views from the literature to help with the definition of the field under scrutiny.
A Background to Internet Relay Chat. This section describes how Internet Relay Chat works, and argues that some basic knowledge of the structure and functions of the IRC network is necessary even for a purely linguistic analysis of the medium. There is also a discussion of channels, nicknames, users, social behaviour, regulation, and action (or third-person) descriptions.
IRC as a Medium. Besides definitions of the medium, this chapter discusses some of the unique features of IRC, such as its playfulness and its much-debated suitability for personal and emotional interactions.
The IRC Data. This section describes IRC data and how corpora are amassed. It is also mentioned what makes IRC an easy field for scholars to research in terms of collecting data, and how the medium eliminates the so-called "Observer's Paradox".
Deconstructing Dichotomies. Since language use on IRC exhibits many characteristics of speech in spite of its physical written (typed) form, this section is devoted to the discussion of how IRC seems to deconstruct these traditional dichotomies of not only writing and speech, but of synchronous and asynchronous forms of communication as well.
Non-Standard Features. This section is concerned with some of the most important linguistic phenomena of language use on IRC. There are subsections which present and analyse extracts and examples in terms of typography (orthography), lexicon (vocabulary), grammar, and discursive properties.
The analysis was based on three different sources, all of which were predominantly available in electronic format. There were two sources of IRC corpora, a large corpus of recorded IRC conversations, and another corpus from the so-called newbie experiment (see description below). The texts containing more than 500.000 letters were obtained from general chat channels (e.g. #new2irc, #ircnewbies, #newbies, #beginner, #casual, #chataway, #cyberparty, and so forth). The third source was the relevant literature, which provided many different approaches, ranging from utopian contemplation to detailed statistical studies (see references section).
The Newbie Experiment
The present thesis contains some references to a so-called newbie experiment and its subjects. The word newbie means newcomer (Puterman, 1994) in the jargon of synchronous electronic conference systems (e.g. on Internet Relay Chat), referring to the subjects of the experiment. These participants, two men and two women, had very little in common, except for their young age, their voluntary participation in the project, and their lack of experience of communication on IRC (Internet Relay Chat). They were different in all other respect, viz. in terms of personality, typing and reading skills, educational background, and their command of English, from upper intermediate to native. Prior to the experiment, each subject was introduced to the new medium, and they were also shown some basic functions of the IRC client program (e.g. entering a channel, how to describe actions, etc.). Then the subjects were asked to communicate through the medium of IRC without any particular restriction of topic and style. The experiment was conducted in the form of individual sessions during which subjects spent a total of over six hours on-line, observing and having conversations on general "chat" channels of the IRC network. All of these sessions were logged (i.e. saved into log files) and stored in electronic text format for further analysis. Both received and produced utterances were recorded. After each session, subjects were asked to describe their first impressions of the new medium and report any difficulty they might have encountered on-line.
Besides amassing corpus, the main objectives of the newbie experiment were to:
* observe and describe unique characteristics of language use on IRC which are not present in other forms of synchronous face-to-face (FTF) communication;
* reveal and describe linguistic characteristics which are not exclusively endemic to the medium of IRC;
* observe and describe the beginning of the process during which communicants on IRC refine their communicative competence in order to adapt to the new medium;
* find evidence and support claims that text-based computer-mediated communication tends to contradict the traditional dichotomy of oral versus written language;
* corroborate assumptions that language use on IRC is an emerging new register of the English language.
The results of the newbie experiment are incorporated in the sections below in the form of observations and pertaining extracts of IRC interactions. Findings are further summarised in the conclusion of the study.
Some Questions of Computer-Mediated Communication
At the dawn of the digital age, when the magic microchip was conceived, there was hardly anyone who envisaged computers becoming a revolutionary tool in the hands of humankind within just a couple of decades.
Although some early experts might have imagined powerful machines which would be capable of handling thousands, let alone millions of operations per second, an amazing performance at the time, the idea of people using computers for global communication was not even dreamt of.
With the exponential development of computer technology, some of the research began focusing on Artificial Intelligence (AI) in an attempt to examine the extent to which computers could emulate human thinking. It was at this stage that communication and computers became related, even though this meant trying to program the computer to behave like a human interlocutor, an effort which is still limited by the fact that to date no comprehensive theoretical description of language and its natural use has been proposed (Dever & Pennington, 1989).
Computer-mediated communication (CMC) can be viewed as a side-effect of the ancient computer networks which connected a few elite government and academic research institutions in the United States some twenty years ago (Herring, 1996). The primary reason for the construction of these early networks was to provide a convenient way of retrieving electronic information. Using the new medium for personal interactions was a further development.
From a linguistic point of view, CMC has rarely been the subject of any comprehensive theoretical research (González-Bueno, 1998), which is largely attributable to the obvious fact that publicly accessible computer networks are also a recent technological breakthrough, consequently CMC is far too young a field to boast long, dusty shelves of books of relevant literature. As a matter of fact, some of the discussion about CMC is still concerned with weather or not the new field deserves academic research at all. Although, as we shall see later, a close analysis of CMC can evidently refute such claims, yet it is still a widely held belief that computers deprive personal communication of its emotive nature, thereby producing an impersonalising and alienating effect upon interlocutors of the electronic medium. Other linguists reluctant to consider CMC a legitimate object of inquiry may find "digistyle" (Danet, Wachenhauser, Cividalli, Bechar-Israeli, and Rosenbaum-Tamari, 1994) difficult to categorise in relation to traditional channels and conventional registers of language use. As Herring points out, "[r]ather than wondering whether CMC scholarship is legitimate, a more appropriate question now is how scholarship can best keep pace with the continued expansion and diversification of CMC" (Herring, 1996, p. 2).
In addition, scholarship appears to be crippled by the lack of well-suited methods to analyse new trends and to describe emerging linguistic phenomena typical of CMC. Therefore a proportion of current research is aimed at testing both conventional and new methods of analysis in an effort to try and place CMC in one of the existing categories (eg. written-spoken or formal-informal), whereas other studies readily assume the existence of a so-called "electronic language" considering it a new variety of English (eg. Callot and Belmore, 1996).
Another question about CMC is what exactly we mean by it. Much of the available literature considers CMC as purely text-based, excluding any additional channels of information (e.g. sounds, visual cues etc.), however, today's computer technology and increased bandwidth have created the possibility of sending information in the form of computer graphics, digitised pictures, and even live audio and video, so computer-mediated communication no longer means only keyboards and speedy fingers, but microphones and cameras as well. Therefore, even though I adopt the terminology, it should be pointed out that CMC hereafter refers to its original meaning, i.e. to text-only digital environments.
A Background to Internet Relay Chat
How IRC works
In order to approach the linguistic features of Internet Relay Chat (IRC), one should have some basic knowledge of how IRC works (Puterman, 1994). The computer program of Internet Relay Chat was developed by Jarkko Oikarinen in Finland in 1988, and the conferencing system soon became very popular all over the world (Reid, 1991). Technologically speaking, Internet Relay Chat (IRC) originally refers to a number of computer networks within a larger network called the Internet. A computer network by definition consists of one or more servers and many terminals, thus digital information can be sent from any terminal to any other terminal or terminals within the same network.
There are several different IRC networks (eg. EUnet, IRCnet, Undernet, EFnet etc.), which may vary in the number of servers and users but share the same characteristics in all other aspects. Since fully comprehending the notion of "virtual places" seems quite a challenge for most of us, it may be a good idea to draw an analogy by turning to tangible objects and places from real-world (RL). Thus IRC networks can be visualised as offices of a big company. People who are in the same room can talk and listen to each other, as each utterance is heard by everyone present. however, communicants can also exchange information privately if they want to do so. Anyone can leave a room and walk into another one to join a new conversation there. It is also possible to invite people from other rooms, but not from another office building. These buildings in our analogy represent separate IRC networks on the Internet.
Channels
"Rooms" of the above analogy are known as channels on IRC. Channels are the key concept of IRC (Reid, 1991), inasmuch as they enable users to create an infinite number of "chat rooms" where conversations can take place. Therefore joining a conversation means joining one of the thousands of existing channels.
In traditional linguistic terms, "channel" refers to the physical medium through/by which information (a message) is conveyed from the speaker to the listener. Even though applying this terminology to channels on IRC may seem a tempting idea, a closer analysis is likely to call for the need for a more complex definition incorporating several notions which appear far more separable in face-to-face (FTF) communication. Thus defining the IRC term channel should include not only the concept of medium through which messages are sent and received but that of place and topic as well. What proves this proposition lies in the fact that people using CMC tend to reify ideas, not unlike users of IRC, who often refer to their chat channels as "rooms" or other physical places. Below is a list of channels with references to physical places (room, place, house, fortress, etc.). The name of each channel starts with a "#" sign, and is followed by a number which is the number of participants who are on that particular channel. This number is often followed by a topic description:
#chatroom 2
#darkroom 12 Welcome to the Darkroom the home of the DarkCrew
#Finland-Room 7
#lesbiansroom 1
#LunchRoom 1
#MrAllthatsBedroom 2 Welcome to the official MrAllthat® channel!!
#pluto 2 CABLEGUY, I got your room back!!!!
#Zgirl 1 <<<<<<Hello and welcome to my play room>>>>>>> be back in 30
#chatplace 1
#crown 5 The place where KINDRED SPIRITS meet
#dutchplace 15
#Jari's_place 1
#mirc_colors 1 The friendly place to be!
#Speyer 1 this place lost his flavor !!!!! by Official
#chatpalace 1
#Alcazar 1 House of Alcazar, int. chat channel........
#barbiehouse 1 matty is back to his roots... VAX rules!!
#castle 1
#Fortress 1
On IRC no channel can exist without any users on it. When the last person leaves a channel, the channel disappears. Similarly, creating a new channel is achieved (paradoxically) by joining a channel that does not exist. As a result of this, IRC channels are of an extremely flexible and dynamic nature, which is further enhanced by the possibility of changing channel descriptions or topics at any moment of a conversation.
As Rodino (1997) points out, IRC channels vary to a great extent, with some of them being more formally organised than others, therefore scholars need limit claims about IRC to certain channels to avoid erroneous generalisations. Werry (1996), for example, emphasising the extreme diversity of the medium, calls IRC channels "small-scale electronic communities" (p.49). In the present thesis, "chat channels" of no special interest provided the corpus for the analysis, for channels of this kind seem to be the most popular on IRC.
Nicks
Upon logging on to an IRC server, the user has to specify a name also known as nickname or "nick" in the IRC jargon. Unlike real names, nicknames, which help identify users on-line, are chosen by the user themselves, and can be changed anytime, even during a conversation.
Besides identification, nicknames have other important, albeit less apparent, purposes. They can provide users with anonymity and create a resultant liberating effect, without which IRC interactions would bear much less originality and uniqueness. Also, nicknames invite playfulness and virtuosity. Most people decide to choose a nickname related to the self to emphasise some positive or negative quality (e.g. <shydude>, <baddady>, <handsom>) or names relating to flora and fauna (e.g. <froggy>, <tupil>, <the-tiger>), whereas relatively few users keep their real names on IRC (Bechar-Israeli, 1995). Since certain names are traditionally gender-related, IRC creates the possibility of gender-switching (Reid, 1991). While changing one's biological sex would involve a great deal of difficulty in real life, on IRC gender-switching is just a matter of impish thought and only requires a simple IRC command (/NICK) to be issued. In addition, due to their psychological effects and highly emphasised importance in the text-based CMC environment, nicknames may also foster the formation of new communities.
Users of IRC
Given the extreme complexity, flexibility and ephemeral nature of the medium, no exact figures are available as to how many people use IRC on a regular basis. Neither is it possible to verify any estimates about statistical distributions with regard to profession, age, sex, location, social status, and cultural background. It is a generally held opinion that there are far more male users than female, and the majority of IRCers are young adults, mostly members of the academic community and students of computer science (Reid, 1991) or some other fields involving computers and high technology. Nevertheless, with the steady expansion of the Internet and its increasing diversity, even such statements are becoming more and more questionable. What remains a fact is that IRC is very popular and populous a medium. To demonstrate this, consider the information below that was provided by a server of one of the largest IRC networks:
13470 channels formed
Current global users: 26499 Max: 41716
Since no channel can exists without at least one user "sitting" on it, one might conclude that on average each channel has two users on. In fact, an IRC user can join several channels at the same time, and there is a large number of single-user channels with people waiting for somebody else to chat with.
Netiquette and Regulation
Having no written laws, IRC might as well be in the state of total chaos and disorder. Anyone who has an Internet connection and a client program can join almost any channel and say (type) anything, enjoying the absolute freedom of speech and expression. Scholars agree that the apparent uninhibiting effect of text-based CMC stems from the relative anonymity and highly limited social/identifying context cues, which characterise the medium (Reid, 1991; Jaffe, Lee, Huang, Oshagan, 1995). In spite of this, IRC does require participants to adopt certain social behaviour patterns often referred to as "netiquette", coined from "network" and "etiquette" (Wauchope, 1997). Below is a typical list of rules in the form of a reminder for newbies (newcomers):
The rules of the channel are: No flooding, no advertising, no repeating, no (op) begging, no caps and English only please.
This declaration of rules seem to apply to the majority of IRC channels.
Flooding means sending too many lines of (often pre-edited) text at the same time, thus blocking the flow of electronic conversation on a channel.
Advertising includes inviting people to join other channels or visit other places on the Internet.
Repeating is small-scale flooding during which a participant repeats themselves. i.e. he or she keeps sending the same text to the others on the channel.
Begging or OP-begging refers to asking for so-called channel-operator rights. A channel operator (also called as chanop or op) has some special rights to keep control over the channel and its users. The first person to create a new channel automatically becomes the operator, and s/he can give similar rights to others who subsequently enter the channel. The rights of a channel operator include inter alia setting channel properties (e.g. invite-only, mediated), kicking or even banning a user out of the channel.
No caps means "no capitalisation", referring to the use of capitalised characters (e.g.HELLO!) interpreted as shouting in CMC (see section on special typography below).
Enlgish only is a requirement on many an IRC channel, but, of course, there is a plethora of other channels where languages other than English are preferred. Nevertheless, English-only channels generally expect every participant to speak English, even if some of the interlocutors share a different mother-tongue. Thus, English continues to enjoy the status of lingua franca on IRC as well.
Although failing to adhere to any of the above described rules may trigger special retributive measures, interestingly enough, this kind of control does not seem to jeopardise the freedom of IRC, as one might infer, but rather these measures are to ensure the smooth and uninterrupted flow of electronic conversation. As a matter of fact, this kind of control is primarily aimed at securing successful communication, which seems to be the essence of IRC.
As it has been established earlier, flooding and repetition is frowned upon in IRC conversations, let alone sending many lines of text at the same time, thus aggressively monopolising the channel. To keep up with the extremely quick turn-taking technique of IRC, users normally send one or two lines of text at a time, and then wait for the other participants to respond. As opposed to FTF, repetition has no real purpose on CMC, as one can easily recall earlier utterances simply by scrolling back on the computer screen and re-read any given line(s). Besides having no important function and conveying no extra information, repetition is felt to disrupt the conversation, making it dull and unimaginative, hence <Merlyn>'s reaction in the extract below:
<NiceGuy> hello may i chat with nice girl ?? PREFEREDLY FROM AUSTRALIA
<<NiceGuy> hello may i chat with nice girl ?? PREFEREDLY FROM AUSTRALIA
<Merlyn> niceguy, stop repeating
<NiceGuy> hello may i chat with a nice girl ?? PREFEREDLY FROM AUSTRALIA
<Merlyn> niceguy, say that once again and i'll kick you
<bubba`> NiceGuy....yes you may
<Zsazsa> right merlyn , be bossy
<bubba`> thats assuming you can find any nice girls in Australia..hehe
<Lilla> haha
<Merlyn> i hate repeaters
But repetition may result in more severe consequences, such as being kicked out of a channel with only a pithy remark in parentheses for the offender to ponder on. In the following sequence, <tompra> repeats himself/herself and is summarily removed (kicked) from the channel by <XenoX>.
<tompra> any girls?
<tompra> any girls?
<tompra> any girls?
<tompra> any girls?
<tompra> any girls?
<tompra> any girls?
<tompra> any girls?
<tompra> any girls?
*** tompra was kicked by XenoX (flood_)
Perhaps another reason why repetition is mostly disapproved of on IRC is that it is not at all a creative technique. Users of IRC tend to prefer more eloquent and original ways which I shall discuss in subsequent sections.
Actions
The use of action descriptions is a unique aspect of IRC (and other CMC conferencing systems), not present in synchronous face-to-face communication (FTF).
The peculiarity of actions on IRC lies in switching from first-person to third-person singular, whereby users report their own actions as if, to borrow Rodino's (1997) term, "an omniscient narrator" commented on a person's actions or state of being. This linguistic phenomenon is hardly present in any other form of synchronous interpersonal communication (Ruedenberg, Danet, and Rosenbaum-Tamari, 1994) and it provides an abundant source of humour, especially for advanced IRC users who are familiar with the action command and its effects upon the conversation. (Note: displayed actions always start with an asterisk.)
IRC command issued by user <Tompa>: /me is hungry
Effect seen by other users: * <Tompa> is hungry
Action descriptions seen in CMC may be related to those present in the genre of comics where both actions reported in third-person singular and direct first-person utterances are applied to facilitate story-telling, but the ability to enrich IRC interactions with action descriptions obviously demands a great deal of wit, appropriate linguistic competence, awareness of context, and a little bit of technical skill (i.e. familiarity with the action command and its use). In the so-called newbie experiment, even though they had previously been familiarised with it, none of the four participants made any active use of the action facility. One reason for this was perhaps the fact that while attempting to communicate on IRC, they were drawing on their expertise based on RL FTF interactions. This repertoire does not contain the use of third-person action descriptions.
Often in an amazingly tricky and virtuoso way, action descriptions may function to overcome the obvious lack of ordinary channels and non-verbal cues (e.g. tone of voice, intonation, gestures etc.) which, as we shall see, characterises IRC as a medium (Danet, 1996a). The interaction below taken from the corpus of the newbie experiment exemplifies such a use of action descriptions:
<Evetke> How old are you?
<Miles> I'm 28
* Miles nose grows
In the above sequence, <Miles>'s third-person description (of his growing nose) invalidates his previous statement of his age by means of playfully referring to a popular character of a children's book, whose nose grew a little every time he was not telling the truth. In real life face-to-face (FTF) communication, the speaker could choose from a much wider range of metalinguistic cues in order to produce the same effect upon the listener(s), e.g. to invalidate the truth value of a previous utterance. One could, for example, use distinctive intonation patterns and/or special facial expressions, such as winking or smiling, or one could touch one's nose as a sign of not telling the truth. But all these non-verbal tools are simply unavailable in the text-only environment of IRC. Thus <Miles>'s utilising the action description facility of the IRC program not only demonstrates special communicative competence (Aitchison, 1992) and virtuosity, but it also weakens claims about text-based CMC which label the medium as being impersonal, distancing, and unsuitable for personal relationships (e.g. Herring, 1996).
Action descriptions may function to express background or extraneous information without creating a disrupting effect upon the actual conversation. Whereas it might be impolite to ignore a first-person utterance, in the case of action descriptions communicants can decide whether or not they want to respond, but the benefit of any extra information conveyed by the action description is often felt later during the interaction.
Some actions describe certain social gestures, such as hugging, kissing, and other ways of showing affection, while others import RL events from the "real life frame" (Ruedenberg et al., 1994) with no evident connection to the on-going conversation. These RL fragments, however, may function to keep the conversation alive or even trigger a new sequence of interaction (for another playful performance, see Appendix C.). Let us consider the following examples:
1. * Deadcow waves to everyone
2. * neichy1 waves goodnight to jazzzz
3. * Ik4u laughs at dinorex
4. * frankay is happy now :)
5. * SteveC turns his central heating up, mmmmmm nice
6. * DaProphet clears his throat
7. * ^Prophet^ is pissed off coz his computer hanged
8. * Merlyn was on the phone
9. * Kali yawns
10. * Graeling sits in the corner and hopes someone will talk to her
11. * Pixie sneezes againnn
<mysty> bless u
12. * Ox-24 is bored...no one to chat with
<bubba`> Ox-24 look around...
Descriptions (1-4) are examples expressing certain gestures and feelings in the purely text-based IRC medium. These written actions also help break the first-person monotony of IRC, playfully transforming the interaction into a kind of script or play. In this manner, descriptions may be utilised to create the illusion of a physical world where non-verbal cues are also available. What actually happens is that interlocutors verbalise the interpretation of their own speech-acts.
Descriptions (5-8) refer to and import RL events, and do not necessarily intend to be a part of the conversation, yet their ulterior purpose is to create some effect or even attract responses. They may also serve to signal personal presence without the need to carry any additional or relevant information. This is a method of compensating for the lack of physical presence and showing the listener's participation.
Descriptions (9-10) do not relate RL actions, but serve to convey indirect references to the participant's state of mind. Thus, action (9) will not necessarily be interpreted as a RL action, i.e. <Kali> has yawned in front of his/her computer, but rather it implies that <Kali> finds the conversation rather boring. Similarly, in description (10) sitting in the corner is not a RL action but an expression of feeling neglected and lonely. Action (10) also overtly indicates her gender and that she would like somebody else to start a conversation with her.
The exchanges in (11) and (12) are examples of third-person descriptions that have triggered an immediate response. In both (11) and (12), the indication of personal presence evinced through an action description is confirmed by a direct response. As proposed above, the aim of such intermezzo-like interactions is to bridge the obvious lack of physical cues on IRC.
Given their playful attitude to their medium, it is no wonder that IRCers often use action descriptions simply to amuse each other with enthralling performances of language and typography as the following examples demonstrate:
* Leia lays DaMMiT on the ground, crawls on top of him and gives him a slow... long... deep... penetrating... sensual... tonsil tickling... orgasmic... hair straightening... toe curling... organ raising... nipple hardening... spine tingling... life-altering... fantasy causing... did i just see GOD??? I think my clothes are falling off! Why arnt we in bed?? Kind of kiss.
* Starfall plays in the ¡.*°*.¡.*°*.¡.*°*.*°*.¡.SNOW *°*.¡.*°*.¡.*°*.¡.*°*.¡.*°*.¡.WITH *°*.*°*.¡.*°*.¡.*°*.¡.*°*.¡. NUDawn *°*.¡.*°*.¡.*°*.*°*.¡.*°*.¡.AND*°*.¡.*°*.¡.*°*.¡.*°*.¡. HITS *°*.*°*.¡.*°*.¡.*°*.¡.*°*.¡.*°*.¡. NUDawn *°*.¡.*°*.*°*.¡.*°*.¡. SQUARELY *°*.¡.*°*.¡.*°*.¡.*°*.¡ IN THE .*°*.*°*.¡.*°*.¡.*°*.¡.*°*.¡.FACE *°*.¡.*°*.¡.*°*.*°*.¡.*°*.¡.WITH A *°*.¡.*°*.¡.*°*.¡.*°*.¡.*°*.*°*.¡.HUGE*°*.¡.* °*.¡.*°*.¡.*°*.¡.*°*.¡SNOWBALL!
IRC as a medium
Ever since pilot studies began focusing on the special linguistic aspects of both synchronous and asynchronous text-based CMC, scholars have been trying to propose a suitable definition for CMC in an obvious attempt to ascribe the observed linguistic phenomena to the unique characteristics of the medium under scrutiny.
Studies more specifically concerned with IRC have also propounded a number of definitions, emphasising different aspects and approaches. Technically speaking, IRC is a form of text-based CMC, and as such shares all of the physical properties, capabilities and constraints that characterise CMC. Any in-depth comparative analysis of IRC, E-mail, and discussion groups would probably reveal that the only underlying difference between IRC and those other forms of CMC is that IRC tends to be considerably more synchronous, hence more interactive, than either E-mail or discussion groups.
Reid (1991), for example, suggests that IRC is essentially a playground where people can experiment with various forms of communication and self-expression. This view is echoed in many other studies listed below, underlining the playful and highly experimental nature of the medium. In addition to playfulness, Ruedenberg et al. (1994) claim that IRC provides suspension of time and space. According to Werry's (1996) definition, this form of communication is a kind of synchronous textual dialogue that takes place between spatially distant interlocutors. Bechar-Israeli (1995) defines IRC as a "silent", "stripped-down" medium characterised by considerable intimacy, with nicknames providing the objects for certain identity games. December (1993) proposes that IRC best exemplifies the oral culture of CMC, while Puterman's (1994) purely technical definition describes IRC as a multi-user computer-mediated chat program, and according to Reid's (1991) definition, "IRC is a multi-user synchronous communication facility that is available all over the world to people with access to the "Internet" network of computer systems" (sec. preface).
There seems to be a great deal of incongruity between the medium of IRC and its primary purpose. As opposed to FTF communication where messages are simultaneously accompanied by certain verbal and non-verbal cues conveyed on extra parallel channels, IRC only relays typed text, excluding all other auxiliary channels of information. Thus, communicants on IRC cannot make any use of intonation, pitch, gestures, facial expressions or any visual/auditory cues which are intrinsically present in other forms of FTF interactions. From this perspective, one might conclude that IRC is quite an unsuitable medium for interpersonal communication and forming personal relationships.
Its unsuitability notwithstanding, IRC enjoys great popularity among people from all over the world, who use the medium primarily for casual conversations or "chatting". On IRC, communication is often for its own sake, i.e. the sole purpose of an interaction is the interaction itself, unlike in other forms of interactions, such as the telephone, where interlocutors often feel the need to initially pre-select the addressee(s) and provide a traditionally acceptable reason for the interaction. A typical telephone conversation would at some point include the explanation of some purpose or a goal other than communication itself, such as asking for information, making an appointment or inviting someone, whereas IRC users usually do not feel the need to specify the reason for their presence or participation. They just wish to have a chat:
<cedro> sombody wanna chat,please
<Zsazsa> we all do cedro, thats why we are here
It can also be stated that much of the playfulness and originality that characterise language use on IRC comes from the aforementioned "unsuitability" of the medium, in that communicants are forced to develop a virtuoso attitude to language and its typed form which carries all the inferable cues about the self. Language therefore becomes performance (Danet, 1997) in which eccentric spelling, grammar, vocabulary, and punctuation work together to compensate for the deficiency of the medium and to convey socio-emotional content.
IRC is a global village (December, 1993), and its inhabitants seem to take great delight in their magical ability to use high technology for communication without the physical constraints of space. People from different continents can talk to each other as if they were sitting at a round table. Being the member of the same "global village", i.e. sharing the same medium, fosters group identity, which in return shapes language use and creates distinctive features that may function as group markers. Puterman (1994) points out that apart from signalling in-group status, the IRC jargon also increases the efficiency of electronic conversations by speeding up interaction through abbreviations. acronyms, slang, and "technitalk", i.e. the use of the special lexicon related to IRC, the Internet, and other areas of computer technology).
The IRC data
From the researcher's point of view, IRC is not only a new and rich testing ground but a very easily observable one as well. Anyone can join almost any IRC channel at any time of day. Utterances appear on the computer screen in the form of digital written script that can be saved and subsequently printed out or edited at will. Therefore loss of information due to noise, phonetic unintelligibility, or any similar difficulties that traditionally plague research of other fields is hardly ever encountered when studying IRC. Moreover, participants seem to be obsessed by the fact that their written lines can be read by all other observers on a particular channel, which apparently fuels wit and linguistic virtuosity. Being an observer on IRC is a popular activity which is called lurking, a term that has no negative connotations on IRC (Danet et al., 1994). As a matter of fact, interlocutors are bound to spend some time observing. One reason for this is that upon entering a channel one cannot have any advance knowledge of the conversation(s) that is/are taking place there, apart from the topic of the channel, provided it has been set and contains any pertinent information. New participants must therefore "listen in" to find out more about the conversation(s) and the actual topic(s). Another reason for unwanted lurking is that a channel, especially a crowded one, may frequently carry several conversations simultaneously, forcing certain participants to read irrelevant information while waiting for a reply.
Observing is such an integral part of IRC that some may prefer it to active participation. People can be so busy observing that they even overtly refuse to have a chat, as in the extract below:
<Cuddly> Hi Steve1
<steve1> want a chat Cuddly?
<Cuddly> Steve1: Sorry, I am busy lurking...
As I have mentioned earlier, personal presence on IRC is shown and reinforced by means of "saying something", i.e. participating in a conversation by typing lines of texts and sending them off to make the utterances appear on the other participants' screens. Those who do not type anything are mostly ignored and have no influence whatsoever on the conversation(s). This helps avoid the "Observer's Paradox" (Herring, 1996). In other words, the observer's presence does not have any (or very little) effect on the interaction(s) or the way in which participants communicate.
All this obviously makes IRC quite effortless for the daring researcher to observe.
Observed data recorded on IRC are referred to by a variety of names in the literature, viz. script, log or log file, transcript, extract, exchange, sequence, interaction, and so forth. Log file and log are exact, albeit too technical, terms for linguists to refer to IRC data. Transcript, however, seems a rather inappropriate word to describe recorded IRC data inasmuch as it implies copying or writing of a text, whereas IRC data need no transformation of any kind in order to be visible or transferable. The reason for this is that IRC data arrive in an already digitised format and as such require no transcription. Exchange, sequence, interaction are applicable terms, although their use may be questionable in certain cases, especially for system (server) messages giving information on users' entering or leaving a particular channel:
*** Jewel21 has quit IRC (Connection reset by peer_)
*** saltydog (xxxx@xxxxxxxxxx) has left #cyberfriends (saltydog_)
*** Ik4u (xxxx@xxxxxxxxxx has joined #CyberFriends
*** LoKiTo (xxxx@xxxxxxxxxx) has joined #CyberFriends
*** Ik4u (xxxx@xxxxxxxxxx) has left #CyberFriends (Ik4u_)
Since recording would be a misleading term (because it implies storing audio or video signals and not electronic text), I adopt extract as being the most general and applicable for my research purposes, for it expresses that the given IRC corpus is presented in its original but fragmented form.
Deconstructing Dichotomies
Scholars plunging into the pioneer research of Internet Relay Chat (IRC) or other forms of computer-mediated communication (CMC) immediately encounter paradoxes that seem to query the authority of some of our most basic linguistic categories.
A question that initially baffles scholars is whether the language of IRC should be categorised as oral or written. From a purely technical perspective, IRC appears to be a written medium, inasmuch as interlocutors receive and send messages in the form of written letters, numbers, and certain other symbols which are available on an ordinary alphanumeric keyboard. On the other hand, the language of IRC apparently exhibits many features traditionally ascribed to orality. Aitchison (1992, p.107) describes the characteristics of written language as follows:
Single writer
Explicit
Non-repetitive
Full sentences
Elaborate structure
Abstract, less common vocabulary
As the analysis in subsequent sections will show, these characteristics of written language do not fully apply to IRC interactions, even if these interactions are produced and received in a physically written form (viz. typed and read). IRC users themselves consider their communication as being oral rather than written, hence the frequency of verbs such as hear, see, tell, talk, listen, and so forth, emphasising the oral nature of the conversation:
<rockette> say again plz?
<Catfriend> *sigh* I really hate it when people don't say hi back
<Merlyn> Lilla, you didn't want to tell your real name, wanna know mine ?
<Merlyn> i'm trying to talk to you here, Bubba :-)
<bubba`> thanx smurf
<bubba`> ok...im'a listening Merlyn
<Merlyn> quack
<Zsazsa> but hearing nothing
Users of IRC readily accept the obvious contradiction of pretending to have an oral conversation while using a medium that only supports writing and reading. In the following extract, the use of verb say refers to the speaker's action and stresses imagined orality, whereas see expresses a RL action on the listener's part, i.e. receiving written utterances by means of reading them on the screen:
<^pUnK^> syn did u see what i said about your astrology pages?
Furthermore, users call their IRC interactions "chats", which reflects the subjective experience of having a spoken conversation rather than a written one. Since the medium relays textual information, this popular concept of orality is not immediately evident for every newcomer. On her first visit on IRC, <Evetke>, one of the subjects of the newbie experiment, apparently conceptualises language use on IRC as primarily written; for she is the only one on the channel to use the verb write to refer to the production of utterances:
<Evetke> I don't understand much in this IRC thing. Could you write some things about this
<Evetke> How many hours a day do you write a day here?
Most researchers agree that text-based CMC cannot be plausibly defined within the frame of the traditional dichotomy of oral versus written language use. Reid (1991) states that synchronous CMC "defies conventional understandings of the differences between spoken and written language" (sec. Preface), and Danet (1996b) considers CMC both as attenuated speech and as attenuated writing, suggesting the existence of a hybrid language. Daly (1996) contrasts the casual, informal, and spontaneous style of CMC with its carefully and self-consciously composed texts. Conducting a scrupulous statistical comparison of CMC, writing, and speech, Yates (1996) finds that CMC differs from both speech and writing in most aspects. December (1993) goes much further and contends that CMC creates a "tertiary form of orality", comparing the effects of the new medium to those of the Greek alphabet and the invention of the printing press.
IRC is generally labelled as a synchronous form of CMC to differentiate it from other asynchronous modes of CMC such as E-mail or discussion groups. Although Wauchope (1997) points out that IRC breaks down the traditional synchronous-asynchronous dichotomy of language use, she does not elaborate on her observation.
If we interpret synchronous language use as the simultaneous involvement of both the speaker and the listener in the interaction, i.e. messages produced by the speaker should concurrently be received and decoded by the listener, then we may find that communication on IRC does not fully meet this requirement. When using an IRC client program, participants compose and edit their utterances prior to sending them off to the other interlocutors who are present on the same channel. Therefore, the process of the creation of utterances remains invisible, and the traditionally parallel roles of the speaker and the listener, which characterise RL synchronous interactions, become temporally separated. On IRC, however, this gap between the production and reception of utterances does not necessarily jeopardise the efficiency of communication, as it would in most RL situations. Moreover, IRC users can even leave their computer and be away for minutes and still be able to remain listeners; for utterances can subsequently be received and interpreted by "scrolling back" on the screen and read the text, thus "catching up" with the conversation.
Non-standard features
IRC is essentially a text-based environment, which means that all interactions take place in the form of written text. The fact that writing in CMC is actually typing on a computer keyboard ostracises conventional graphology from the inventory of related fields and methods available to CMC researchers. Interlocutors cannot have any influence upon the visual representation of the letters (fonts) that make up computer-mediated texts. While in RL interactions (hand)written and spoken utterances may differ considerably in their visual and phonetic representation, respectively, computer-mediated typed texts do not carry any non-verbal personal markers or physical cues.
Let us now consider the following two utterances:
(A) I'm in love again.
(B) I'm in love again.
Either hand-written or spoken, both utterances would probably show a great number of differences and produce different results. If, for example, (A) is either an elderly or a very young person, the utterance will most likely carry some humorous overtones and may even result in laughter. But the same utterance said by (B) sobbingly would have a different effect upon the listener(s). Conversely, if both (A) and (B) are familiar with standard English and use capitalisation and punctuation accordingly, their utterances will probably look identical in text-based CMC. Therefore, to "personalise" utterances, IRC (and CMC) interlocutors are obliged to break certain rules of standard language use and resort to rebellious and eccentric spelling, non-standard grammar, special vocabulary, and the uniquely CMC-specific emoticons (or smileys). Through their special language use, participants on IRC thus become players (Reid, 1991) on a virtual stage where a good performance means the successful and often virtuoso representation of interpersonal and socio-emotional themes in a purely text-based environment.
Typography
For various reasons described earlier, IRC users have developed certain typographic conventions as part of the IRC jargon. Knowledge and frequent use of such conventions marks the interlocutor as an experienced, in-group member of the IRC society. Since speed is a key factor in IRC interactions, special typography also serves to facilitate, i.e. speed up, the conversations by reducing typing need. Correct spelling is not a prerequisite, neither is it a reliable indication of social status or education, for IRC interlocutors often misspell words and use non-standard punctuation on purpose. Spelling errors that cause no misunderstanding are mostly ignored, but those easily lending themselves to sharp and witty remarks will probably attract more attention as in the following two extracts:
<Dytta> I can do a summersault
<mrTR> mean ?
<angelgrl> yeah, but you can't spell it...heehee
* worried is board
<Caley> worried u r spelling bored wrong
<Caley> or are u a surf board
<Caley> or an ironing board
<Caley> hehehee
Despite their excellent tying skills, IRC users tend to omit certain typographic symbols, especially those of punctuation, such as apostrophes, commas, and full stops. Researchers, however, should not hastily conclude that this is a unique phenomenon on IRC, even though it is undoubtedly a highly characteristic one. While some interlocutors may consciously exclude punctuation marks to speed up typing and/or signal in-group status – i.e. to show familiarity with the special language use of the medium – many others seem to leave out commas, apostrophes, and full stops both on-line and off. (The newbie experiment underlies this assumption, as each subject regardless of their RL knowledge of punctuation sooner or later started omitting full stops at the end of their sentences.) Therefore it seems plausible to posit that – although newcomers on IRC inevitably bring their RL orthography into the medium, showing considerable differences in punctuation and sentence construction – more experienced IRC users are likely to adapt to the typographic style of the medium, and even the most pedant participants tend to develop some non-standard language habits, e.g. missing full stops to mark sentence endings. In addition, frequent full stop omission can also be ascribed to the fact that each utterance on IRC is automatically preceded by the interlocutor's nickname in angle brackets and the utterances rarely consist of complex or compound sentences. As a result, the traditional usage of full stops to end all sentences except direct questions and genuine exclamations (Hacker, 1993) loses its functionality in IRC interactions, in that there is no need to use any markers to separate sentences/utterances. The examples below show various types of missing punctuation on IRC:
<NUDawn> im in dallas
<Airbuster> theyre brown
<{jade}> sorry i don't have it to send
<Anni> Hi all
<{jade}> which did you lose chipper
As seen in the last utterance made by <{jade}>, even question marks tend to be omitted, especially when word order and/or context clearly identifies the particular sentence as a question. The use of question marks (as well as other elements of punctuation) appears to follow imagined intonation rather than conventional rules based on grammar. Consider the uses of the question marks in the following extracts:
<sexy`> just for ontario place
<Cr0wley> ontario place?
<sexy`> it's a theme park in toronto canada
<Cr0wley> Oh right.
<Zsazsa> well I am the best one anyway
<Lilla> yea ??
<^cLueLess> so where abouts in england caz?
<^caz> rochdale
<^cLueLess> ???
<^caz> near manchester
<^cLueLess> ahhhhhh
In the first extract, <Cr0wley> repeats "ontario place" and uses a question mark to ask for a definition. In the second exchange, <Lilla>'s question marks obviously express rising intonation to show surprise and doubt. The three question marks in the last extract, however, work as a very concise textual interpretation of a non-verbal cue, possibly a facial expression of puzzlement.
Capitalisation is no exception to the eccentric typography of IRC. Owing to economy of typing and the lack of need to mark sentence boundaries, capitalisation of first letters is generally ignored in IRC interactions to such an extent that even the first letters of proper names are often spelt in lower case (see example above). This reluctance to use capital letters may partly be due to the fact that on IRC (and CMC) capitalisation is traditionally interpreted as shouting, and excessive use of it is frowned upon. Of the extracts below, the first one demonstrates that some participants are aware of their non-standard use of capitalisation. The other extracts contain typography intended to represent shouting.
<DaMMiT> its buy guns and roses..
<DaMMiT> by even..
<DaMMiT> damn i need grammer check and spell check for mirc
<Meth> whoa
<nudehamster> it would tell you to capitalize
<Ruuuude> COS I'M RANNNNNDY
<joh> bye every1
<moonray> yes!!!! I thought you never asked!
<Ruuuude> WHAT MOONRAY
<Cubby> why do you shout ruuuuude
<DaProphet> WAAAAAAAAAAAAAAAAAAAAAAAAKE UUUUUUUUUUUUUUUUPPPP!
<^Demon^> YEHHHHHHHHHHHHHHHHHHHHHHHHHHHHH
<^cLueLess> U CALLING ME MAD?
Werry's (1996) proposition that "the language produced by users of IRC demands to be read with the simultaneous involvement of the ear and the eye" (p.59) can be further justified by focusing on some other features that IRC typography exhibits. One such characteristic (often combined with non-standard capitalisation) is the use of multiple punctuation and spelling, a technique to express emotional content through the unusual visual representation of typed electronic texts:
hellllllllllo
it is soooooo good
im sooooooooooo sick of genereal ed
* Cierra is sssooooo coolldddd
I'll be back later.. just realised fooooooooooDDDDDDDD!!!!!!!!!!
<ShaDowLin> Hiiiiiii Lise
<mstng_ldy> hi lise!!!!!!!!!!!!!
<bonn^^> lise!!!!!!!!!!!!!!!!!!!
* Nicho^ can handle heat in day but likes it to cool off at night so he can sleep properly :)
<Gerbie> Lovely lise!!!!!!!!!!!!!!!!!!!!!!!11
<alsandor> bonn, chip has BEEN HERE???????????????????????????????????????????????????????????????
<alsandor> bonn, ??????????????????????????????????????????????????????????????????????????????????????????
<alsandor> bonn, ??????????????????????????????????????????????????????????????????????????????????????????
<alsandor> :-)
<oreosmile> als: just let them get you all fired up for your wifey-poo
<bonn^^> als:yes al...calm down
<Buffy> mystyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyy
<mysty> buffffffffffffffffffy!!
<mysty> :)
<mysty> hehe
<MsBehav|n> pppppppppppppppppppppppiiiiiiiiiiiiiiiiiiiiiiiiiiiixxxxxxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxeeeeeeeeeeeeeeeeeeeeee
<pixEEEEE>
MsBeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeee eeeeeeeeeeeeeeeeeeeeeeeeeeeeee :)
Wauchope (1997) notes that a slow typist on IRC is disadvantaged and may even be considered less intelligent. Conversely, a speedy typist with quick, witty responses is always welcome and immediately accepted as an experienced member of the virtual society. To increase typing speed, communicants on IRC are constantly forced to invent ways in which utterances can be shortened without any detriment to meaning and intelligibility. These techniques of "typing economy" include simplified spelling based on pronunciation, single or combined letters or even numbers which can substitute one or more syllables when spelt aloud, as well as other symbols available on the alphanumeric keyboard.
Typed textInterpretation
thot
thought
tho
though
alrite
all right
w/
with
w/o
without
y
why
r
are
u
you
c
see
fa q
fuck you
2 (me 2)
to, too (me too)
every1, some1, no1
everyone, someone, no one
l8er, l8r
later
c u l8r
see you later
m8
mate
age <16
older than 16
tanx, thx
thanks
OIC
Oh, I see!
ne1
anyone
It has already been suggested that users of IRC conceptualise their medium as primarily oral in spite of its physical nature and text-based environment. This fundamental contradiction accounts for the playfully rebellious approach to language, emphasising the importance of orality. Communicants tend to regard writing as a kind of flexible medium to represent speech, and they seem to take particular delight in what normally could only be spoken or "scrawled behind the closed door of a public toilet wall" (Danet et al., 1994). The freedom and liberating effect of IRC stems from this dual attempt at creating a visible language and establishing social relationships. Consequently, the effort to represent speech in the form of typed text – i.e. to present the unpresentable (Reid, 1991) – is a very important part of the IRC "performance". Participants do not think twice to break rules of spelling, punctuation, and even syntax in order to create more speech-like utterances. Although the resulting texts occasionally seems to bear signs of typing economy (e.g. em for them), the main purpose is to accentuate the oral, personal, and more casual nature of IRC interactions. The following extract seems to refute the economy theory, supporting Daly's (1996) observation of the hidden discrepancy between casual, informal style and carefully composed texts:
<Cr0wley> How are you doin' bud?
It is evident that typing doin' requires just as much, if not more, effort and time as typing the full word doing. The speaker even utilises an apostrophe to mark the missing letter g, making it clear for other communicants that he/she is consciously representing phonetic features and not simply omitting the last letter for purpose of economy. Similar efforts can be observed in the extracts below:
NUDawn> well redog has a page for a bunch of em
<Starfall> Welll
<helraizer> been watchin ya
<Deadcow> so whats goin on in here?
<{HeLiOs}> tigs, can you remember what we sposed to be doing today????
<lyser> shudup
<g|rL> lemme see :)
<Kewl-Gal> I like brown eyes... dark n mysterious..
<Hellbound> bah, I'm sick of classes 'n shit
<Kewl-Gal> wassup punk..?
<CpCaveman> worra fecking wanka
<ferngirl> puhleeeeze
<LifeGard> Any gals from or near Atlanta?
<Meth> ordering sumfin
<bubba`> cowboys ride 'em farmers milk 'em
<^SkY^> I duno
<^FieryRed> bout 5'6
<peaches^^> i must push myself cos i have to be fit
<sexy`> that's sorta funny
<^Prophet^> you wanna know where i come from or where i live ?/
<Caley> gimme a sec Crow
* moonray is sad merc doesn't wanna speak to her anymore
<Ruuuude> how da hell did we get into this shit
<alphaUS> moonray: yuh, lemme do my stuff.. don't care about it :)
<CpCaveman> i bet if we work ard enough we`ll give her a complex
Users of IRC have also devised, and are constantly experimenting with, ways to transcribe, as it were, non-verbal sounds and noises through eccentric typography. Some of these elements originate from comics, such as various textual representations of laughter or "zzzzzzz" to indicate sleeping or snoring:
<EvilMoose> hahahahahahahaahahahaha
<CpCaveman> hahahahahahahahahahaahahahaha
<FredDerf> mmmmmmm i'm imagining the tounge action
<^cLueLess> ermmmmm not sure what the date was actually crafty
<Matumbo> I like peanut butter and tuna
<Lise^^> ewwwwwwwwwww matumbo - what a combo
<bonn^^> tuna and peanut butter? ewwwwwwwwwwwwwwww
Marengo> zzzzzzzzzzzzzzz
<Caley> zzzzzzzzzzzzzzzz
<Marengo> heheh
<DaProphet> WAAAAAAAAAAAAAAAAAAAAAAAAKE UUUUUUUUUUUUUUUUPPPP!
Emoticons
The most original by-product of artistic efforts to charge typed electronic texts with emotional cues is the system of emoticons (also called smileys, smiley faces, smiley icons, graphic accents). Emoticons are simple strings of alphanumeric characters and punctuation symbols (Reid, 1991) which, when looked with the head turned sideways, can symbolise various facial expressions. Emoticons are very popular in text-based CMC, because they have the ability to add emotive content to utterances. In this sense, emoticons are transcribed emotions and highly emotive actions. This process involves the speaker's having to disassemble his/her utterance into two distinct elements: the verbal utterance itself and the emotive content transcribed as an emoticon. The listener has to reverse this process and combine the purely verbal (textual) information and the transcribed emotive cue(s).
Some examples of emoticons (for a more comprehensive list, see Appendix A):
:-) smile
:-( frown
;-) wink
;-( angry
: P drooling
<8-) smile with glasses and a hat
The effects of emoticons upon the meaning and perception of positive and negative messages seem far more complex and inconsistent than their simplicity may suggest (Witmer, Katzman, Colman, 1997). Moreover, users of IRC tend to employ emoticons as friendly gestures and often without any obvious reasons. Special uses and variations are also frequently observed, such as applying multiple spelling for emphasis:
they put hot peppers on my sandwich :((((((
An even more virtuoso performance is the recycling of emoticons into one of the original functions of a constituent (Ruedenberg at al., 1994). In the example below, <maoui> recycles a smiling face into a parenthesis:
<maoui> You know, that's what "they" (the astrology gurus :) say
Other even more artistic experiments include depiction of movement through a sequence of emoticons:
<Succubus> :PppPPPpPPpPppP
<Succubus> look at that tongue action
Emoticons are a truly original characteristic of language use in text-based CMC, therefore newcomers need learn how to interpret and utilise these graphic symbols to give a better performance on the virtual stage of IRC. In the following sequence, <Dytta>, one of the subjects of the newbie experiment, takes her courage in both hands and asks for exmplanation:
<Dytta> How old do you want me to be?
<romeo19> 2 year old toddlers
<romeo19> :)
<Dytta> 23, but what's this :)? This is my first IRCing.
<romeo19> this is a smile..look side ways
<romeo19> : the eyes
<romeo19> ) the lips
<Dytta> looks good
To impress her, <romeo 19> decides to show off his knowledge by sending <Dytta> a different kind of emoticon:
<romeo19> @>---@>--@>>--
<Dytta> Is this your age?
<romeo19> thant's 3 roses for the lady!
Apparently, she still fails to successfully interpret this abstract level of keyboard art, so he has to give her a verbal description. The fact that she has already encountered emoticons (on a channel called wasteland) is revealed later during the conversation:
<Dytta> Sorry, I thought you have 19 ears.
<Dytta> :/
<romeo19> he..he..what a monster is that!
<Dytta> I saw it wasteland
Similarly to acquiring the proper use of RL registers, language use of IRC also takes some time to master. Thus, it is only with the more experienced users of IRC that frequent utilisation of emoticons and other typographic features, i.e. familiarity with the medium-specific jargon, can be observed.
Lexicon
The appropriate use of language presupposes good communicative competence, i.e. knowing what to say when (Aitchison, 1992). Given its freer and relaxed style, IRC might seem a medium that does not expect interlocutors to pay particular attention to their choice of vocabulary, since utterances often contain slang and even four-letter swear words. As a matter of fact, the ability to choose the right word is indispensable to successful communication on IRC, where language inevitably becomes the purveyor of personal markers, attitudes, and social cues. This heightened importance of language use together with the expected brevity of utterances demand a more conscious choice of vocabulary on the speaker's part, which, in some respect, seems to contradict the description of the medium as relaxed and free in register. Users of IRC appear to share some covert alertness to the appropriate use of vocabulary, because catching any blunders may provide a chance of witty performances:
* Nicho^ can handle heat in day but likes it to cool off at night so he can sleep properly :)
<mstng_ldy> how do you sleep improperly, Nicho????? haha
Besides colloquial, slang, and swear words, the IRC jargon also consists of three distinguishable sets of lexicons. One such group of words and expressions is often referred to as "technitalk", i.e. terminology related to high technology, electronics, and computer science (e.g. re-boot, mouse, click, server, hang, bug). The second set of lexicon often used on IRC is concerned with CMC and the global computer network called "Internet" (E-mail, URL, ping, bandwidth, cyberspace, spam, IMHO, BTW, emotiocons). Words which belong to the third group are uniquely used on, and in terms of, IRC. The relationship of these three sets of vocabulary items is shown below:
It is especially IRC terminology (i.e. words belonging to the third and smallest group) of which participants need be cognisant if they want to communicate successfully on IRC. As with all social groups, the IRC jargon proper also fulfils two parallel functions: it makes communication more efficient by providing new items of vocabulary, and it creates an explicit marker of in-group status. The jargon is relatively limited, but newcomers often have to ask for definitions.
A proportion of the special IRC lexicon have been coined by programmers and computer experts who also created the system of IRC and wrote the necessary computer programs. Consequently, it is no wonder that the majority of these words are similar to those of computer languages, in that they are mostly abbreviations and acronyms. The two most frequent words of this kind are MSG and DCC, both of which can function as nouns or transitive verbs in the language of IRC:
<kazoo> look, helen, i DCCd you!
<kazoo> why don't you DCC me?
<SINAAL> Anyone wanna chat with a 19/M/Fr ? MSG me please
DCC stands for Direct Chat Connection, and it enables users to establish a direct connection for private conversations or sending/receiving computer files. MSG means private message, i.e. one or more lines of text with a specified addressee.
Below is a list of some frequent abbreviations, acronyms, and other items of the jargon (for a more extensive list, see APPENDIX B.).
IRC term
Meaning
WB
welcome back
RE
hello again, welcome back
LOL
laughing out loud (to express laughter)
ROFL, ROTFL
rolling on the floor laughing (to express laughter)
M
male
F
female
ASL
age, sex, and location (asking for personal information)
DCC
(to have a) direct chat connection
MSG
(to send/receive a ) private message
BRB
(I will) be right back
BBL
(I will) be back later
NP
no problem (in response to "thank you")
LO
hello
OMG
Oh, my God! (exclamation)
WTF
What the fuck! (exclamation)
lag
(to have) a slow connection, delay
kick
to remove a user from the channel
chanop, op
channel operator
ping
to check the speed of network between two computers
ppl
people
k
O.K.
bot
robot
Users of IRC inevitably adopt the love of coining and using acronyms. Despite their frequent use, these words show a considerable amount of flexibility and variation, which is another possible source of playfulness:
<SaBaker> write back?
<MrsCopper> wb=welcome back
<Macy> wb = welcome back
<SaBaker> oh
<Macy> warner brothers?
<MrsCopper> water bottle?
<SaBaker> willing buddy??
<Macy> western britain?
<MrsCopper> weather beacon?
<SaBaker> why buy
<MrsCopper> west berlin?
<Macy> wet basenji?
<SaBaker> what bongo?
*** slip^ is now known as cot^
<cot^> hiyas:)
<MrsCopper> so many options.......how can you choose?
<Macy> :)
<SaBaker> well ill guess it means welcome back
<SaBaker> COT wb
<SaBaker> what became
<SaBaker> really
Truncated or clipped words are also quite frequent but used more inconsistently, such as puter and comp for computer, addy for address, diff for difference, morn for morning, and so forth. This kind of abbreviations hardly cause any problems for newcomers, whereas the interpretation of baffling acronyms usually does (<Evetke> and <Mul> are two subjects of the newbie experiment):
<Evetke> What is LOL?
<Rastababe> laughing out loud
<Miles> evetke lol is large orange lobsters
<Rastababe> miles be good
<Mul> pardon me for asking , but what does LOL mean
<__pat> Mul..... LOL=Laugh Out loud... to make up for the diificulty of expressing appreciation of humour
The newbie experiment has shown that newcomers upon their first visit to the world of IRC need ask for
explanations and definitions of certain frequently employed expressions and words in order to interpret utterances. Therefore it seems plausible to conclude that the lexicon of the language on IRC exhibits features that clearly differentiate it from the kind of vocabulary used in other forms of FTF communication and RL situations.
Grammar
Compared to the predominantly obtrusive phonological differences, grammatical variation of different dialects of English tends to be less extensive (Greenbaum and Quirk, 1973). This observation also seems to hold for the grammatical features of language use on IRC, inasmuch as communicants cannot break basic grammatical rules, as it would most likely result in unintelligibility. Consequently, grammar cannot (and need not) be exposed to such a high level of creativity and virtuosity that characterise other aspects of language use on IRC, viz. typography (orthography) and vocabulary.
Non-standard grammar may be observed both in speech and writing. While in speech non-standard grammar is a sign (and result) of a colloquial and freer style (Pearson, 1977), in writing it is applied mainly for economical purposes, such as in newspaper headlines and notes. On IRC, however, both purposes of non-standard grammar are discernible.
To speed up the conversation, participants of IRC tend to make frequent use of elliptical structures while producing their utterances. Although ellipsis is also present in speech, especially in informal registers, the extent to which elliptical structures can be employed in IRC interactions rather characterise highly condensed written styles such as newspaper headlines or notes. When turning to this kind of ellipsis, IRC interlocutors consciously or subconsciously sacrifice certain oral qualities in order to produce faster utterances, which proves that speed is regarded a more important factor. In most cases, however, ellipsis works as a tool for both economy (speed) and expression of orality. What most frequently occurs is the ellipsis of subject pronouns, and auxiliary verbs. Werry (1996) also observes that subject pronouns are often deleted, because the speaker's nickname automatically precedes his/her utterance (e.g.: <Marlena> can't stand it).
Examples of elliptical structures:
gotta work in 2 hours
sure will
been watchin ya
Anyone know a good channel to hang out in?
why that nick ? :)
<^Lum> thx: Megg :))
<Megg> pleasure Lum :))
why so many ops
<Sluggie> told me she was 8
brb have to get pizza out of the oven
<Kadi> where u from :)
<Zero-Cool> Usa....u
<Kadi> Brazil - where in the US
<Zero-Cool> Dallas
Anything worth looking at?
you going?
anyone like U2?
were u from
<Tassle> you talking to me?
<Nicho^> bonn ... how long you going for?
<Marlena> can't stand it myself
Structures with a higher level of ellipsis are much less frequent, as they tend to deprive the conversation of its oral quality:
brb phone
u from?
how old
so what up
Non-standard grammar also manifests itself in non-standard concord (subject + verb, possessive pronoun + noun phrase), double negation, and the use of ain't, emphasising the oral and colloquial quality of interactions:
and thats why i likes ya
no methinks they are non beleivers
i luvs ya mandy
i changed me colour so ppl would take me more seriously!!
you mean you aint a mental case?
i never get none i'm sad
Ellipsis is not the only grammatical trick that aims at writing in a more speech-like manner and at the same time trying to keep up with the usually fast pace of conversations. Participants usually produce short and simple sentences. Complex and complex-compound sentences are few, for their production is too time-consuming; nevertheless, as Werry (1996) points out, messages tend to become longer (and more complex) when there are fewer participants. The rarity of complex and compound sentences explains the relatively infrequent use of relative pronouns, conjunctive adverbs, and correlative conjunctions, all of which characterise formal written style but definitely not casual conversations. Utterances on IRC containing more than one clause generally utilise parataxis, rather than co-ordination or subordination (December, 1993). Clause boundaries are often visually marked with multiple punctuation or other typographic symbols (e.g. emoticons), or remain unmarked:
<korna> who are you dj do i know you?
<SteveC> kait have you been strutting your stuff on the tennis courts today?
<Kait> steve..i got mybutt kicked this morning...8/8 and 8/2
<`deb> i'm fine shellie how are you?
<Cuddly> Steve1: Sorry, I am busy lurking...and I am male!
<bonn^^> i love my meat....i gotta have meat myself...i could never be a vegetarian
<mrs> i can and will cook anything that contains meat i just wont eat it
bonn^^> dave..how is andrew and his spots?
<DaveH> bonn, seems to be levelling off... haven't seen any new spots in 2 days... everything is scabbing over (afraid for scarring -- kept scratching in his sleep)... Almost no fever now.
<RIP`> I wonder what Inge would think of that :) better not ask her though :)
As the newbie experiment suggests, on IRC multiple punctuation is a special typographic feature that most newcomers do not initially have in their repertoire (none of the subjects made any use of it), while the fact that utterances tend to consist of simple sentences and paratactic clauses is largely attributable to way in which the IRC system works. (Our subjects almost exclusively composed simple and short utterances, probably because they were far too preoccupied trying to become accustomed to the new medium and its challenging use.)
Some discursive features
Although a discourse analysis of interactions on IRC, even it a cursory manner, would certainly exceed the limits of the present study, it is worth examining and pointing out certain characteristics of electronic discourse in order to provide more evidence for our IRC vernacular theory. Moreover, a discourse analysis of IRC interactions also presents the opportunity to scrutinise available extracts in their integrity, thus synthesising all the observations and described features (viz. action descriptions, typography, lexicon, grammar and so forth) which have been dealt with in earlier sections. The extracts below also demonstrate that participation in on-line interactions, i.e. being able to communicate through the new medium, has priority over the topic of the conversation. In this sense, participants join these channels to enjoy membership of a social group, rather than utilising the medium to obtain information for any specific purpose. Anything can be a topic, ranging from personal particulars (age, sex, location) to nail clippers (see extract below). Topics come and go, and they are immediately disposed of as soon as they are no longer felt to stimulate the conversation, or perhaps a more interesting topic is brought forward. Participants tend to follow a kind of stream-of-consciousness technique (Marvin, 1995), and topics are typically elaborated on in a playful manner:
* Nicho^ needs sharper nail clippers ... sheesh !! :)
<bonn^^> those darn old toenails anyway eh nicho?
[...]
<Nicho^> bonn .... fingernails are the prob .... cant get through ones on thumbs ! :)
<bonn^^> then you have good zinc in your system...then
<bonn^^> that's what causes good or poor nails
<Nicho^> bonn ... that what it is? ... very tough nails :)
<bonn^^> yep...and when you get white spots on nails...lack of zinc...
<Nicho^> bonn ... I see .... gonna have to use scissors I think :)
<bonn^^> ewwwwwwwwwwwwwwww
<MsBehav|n> wow learn something new everyday
* ShaDowLin checks her toenails
[...]
* ShaDowLin checks Gerb's toenails
<Nicho^> MsB .. hehehehee
<Gerbie> toenails?
<bonn^^> yep gerby...toe nails
<ShaDowLin> hehehehe
* ShaDowLin paints gerb's toenails ruby red
<MsBehav|n> nicho all i knew was it helped w/ colds and not to get the natural cherry zinc tabs...burn yer tongue and make everything taste funny hahahaha
<Gerbie> LOL
<Gerbie> not my color
When a particular topic becomes worn away, to fill the resulting gap, participants usually start proposing new themes in a various ways, and/or make mostly humorous comments in order to amuse each other, signal personal presence and keep the conversation alive. The following extract is an example of this "gap-filling" activity:
<dinorex> .
<dinorex> ..
<dinorex> ...
<dinorex> ....
<dinorex> ...
<dinorex> ..
<dinorex> .
<dinorex> hehe
<dinorex> modern art..
<^SiD^> hahahaha
Although newcomers seem to find it difficult to keep up with the speed and dynamic nature of topics, they soon learn how they can command attention and playfully participate at the same time. After several futile attempts at joining the conversations by asking questions, all subjects of the newbie experiment discovered a technique of taking advantage of their newcomer status by drawing on and exploiting the other participants' curiosity. <Mul> and <Dytta> even started to play games, which would have been inappropriate in most RL FTF interaction between communicants who are complete strangers to each other:
<Mul> but i'm not in the land of aus at the moment
<Junely> well i wondered cause you addy didnt look like it mul...where are you
<Mul> Junely , i'm somewhere in the northern hemisphear
<Junely> well mul i told you cant you fess up
<Mul> ok a city in eastern europe
<Junely> well thats more of a detail mul :)
<Mul> it starts with a B and ends with an S
<__pat> ends with S?
<__pat> u sure?
<Mul> sorry i fibbed , it ends with a T
<Junely> hmmmm bolivia???
<Junely> ooops thats south america lol
<Junely> wellllll
<__pat> LOL junely
<Junely> another b....lets see
<Mul> think Goulash
<Junely> cant you even give me the country
<Junely> boulash????
<Mul> nem nem
<Mul> goulash
<Mul> Ex commie country
<Junely> but you said b
<Junely> how bout budapest???
<Mul> yes yes , you win the two cent prize
<Junely> thats all i get!!!
<romeo19> u?
<Dytta> guess
<Dytta> help: Europe
<romeo19> wait
<romeo19> 'i'm not good in this guessing game
<romeo19> England
<Dytta> nope
<Dytta> H...
<romeo19> holland
<Dytta> Hungary
I have already stressed participants' playful attitude towards their electronic medium, where everything (even gender and language) can be treated as an object of play (Reid, 1991; Rodino, 1997). Newcomers immediately sense and quickly adopt this attitude. Here is another example of such playful performances:
<dinorex> 1
<star> hi icemanuk
<dinorex> 2
<star> remember me?
<dinorex> 3
<NiceGuy> hello any nice girls wanna chat ??
<dinorex> 4
<dinorex> 5
<hotman1> any girls want hot man?
<star> Hi NiceGuy
<dinorex> 66
<dinorex> ops
<dinorex> 6
<star> Hi HotMan
<dinorex> 7
*** hotman1 was kicked by IcemanUK (I told you not to say that )
<dinorex> 8
<dinorex> 9
<star> boys stop fighting
<dinorex> 10..u wanna know what Im doing?
<dinorex> 11
<NiceGuy> hello any nice girls wanna chat ??
<star> no
<dinorex> 12
<Merlyn> you're being annoying
<delerium> 13
<dinorex> no-....
<delerium> 14
<dinorex> Im counting my iq.....
<IcemanUK> :)
<Merlyn> duh :-)
<delerium> then u must stop now
<Merlyn> he already counted too far :-)
The above extract also illustrates what Werry (1996) calls multidimensional texts, referring to the fact that the IRC system juxtaposes disparate strands of interactions, making it quite difficult – especially for the less experienced – to separate one conversation from the other. Intertwining sequences can occur because no overlaps and interrupts are possible on IRC (Werry, 1996). In other forms of synchronous interactions involving more than two interlocutors, e.g. in amateur radio "nets", where overlaps are possible, participants need apply certain controlling and turn-taking methods (net-control) in order to avoid loss of information (Hildebrand, 1997). On IRC, however, no such measures are necessary, since utterances, even if they have been produced simultaneously, appear in the chronological order in which the IRC system receives them (Werry, 1996). Therefore, interlocutors have to follow and process (i.e. infer the addressee of) each utterance of all of the intertwining conversations if they want to participate successfully. Although the extra work of processing irrelevant information may seem a disadvantage, it also enables participants to follow and join in any other parallel conversation whenever they wish to do so. Applying the traditional model of communications theory to IRC interactions, we may also find that dealing with extraneous information is also present in RL FTF communication, which requires the listener to recognise and filter out noise. In this respect, processing and interpreting intertwined strands of conversations and system messages is similar to the process of filtering out noise on RL channels.
The following example demonstrates "digital noise", i.e. a structure of intertwining exchange. <KewlGal> and <Airbuster> are having a conversation about eyes, while <Strife> and <g|rL> are discussing astrology:
<Kewl-Gal> eyes are the windows to our soles... I love looking into eyes...
<Airbuster> I hate looking in eyes
<Strife> what's the sign for june 15 anyways?
<Airbuster> makes me sick
<g|rL> gemini
<Kewl-Gal> why Airbuster..??
<Strife> gemini?
<Kewl-Gal> what on earth for Airbuster..??
<g|rL> yes
<Airbuster> I feel like my heads about to explode
<Strife> do they go with pisce?
<Kewl-Gal> eyes are so beautiful...
Processing simultaneous strands of conversations present on IRC requires skills which one does not need to possess in other forms of FTF communication, hence the difficulty that newcomers experience while attempting to have a conversation on IRC. Since non-verbal cues (e.g. eye-contact) indicating the addressee of an utterance are not available in the purely text-based environment of IRC (Werry, 1996), participants often specify the addressee by including his/her nickname in the message. The newbie experiment suggests that even this simple technique is one of those to be learnt on IRC, since none of the subjects used it initially. In the extract below, <Mul>, a newcomer, does not specify the addressee of his request, which prompts <__pat>, one of the possible addressees, to ask for reinforcement:
<Mul> tell me something about yourself
[...]
<__pat> me Mul?
<Mul> ah yeah
<Mul> that means you Pat
<__pat> mul im 21, a comp sience student
<Mul> learns the lesson, and subsequently specifies the addressee of his utterances to avoid ambiguity whenever he thinks it is necessary:
<Mul> where are you from Junely?
Another lesson for newcomers to learn is that, unlike in FTF or telephone interactions, receiving messages and composing them seen to occur simultaneously: one may be typing one's text and still be able to receive messages from other participants. Since composing utterances and processing incoming messages cannot be done at the same time, on a busy channel participants have two options: They type their messages as fast as possible and then read any pervious utterances they have missed, or they frequently stop composing their messages to process any new messages which may have arrived. Since neither of these techniques is truly endemic in RL FTF communication, it is no wonder that newcomers tend to type less for fear of missing any utterances addressed to them, or find it quite confusing to type and keep up with the flow of the conversation at the same time. (Cf. the notion of multi-tasking in Ruedenberg, 1994). Follow-up comments (and complaints) made by subjects of the newbie experiment also supports this observation.
Conclusion
Internet Relay Chat (IRC) is an electronic medium for communication between spatially distant interlocutors. In spite of the fact that the medium relays information in a purely written (typed) form, participants have devised certain techniques by means of which written utterances can carry paralinguistic cues. Some of these techniques have been – and are constantly being – conventionalised, which is a result of the apparent effort to create a social group, with written language being the only available carrier of status markers. Interactions on IRC are characterised by a playful, creative, and experimenting attitude towards language, providing an abundant source of virtuosity. Thus, aspects of identity, gender, and even language itself become performance.
IRC as a medium cannot plausibly be categorised according to the traditional dichotomy of written versus spoken forms of communication. Interlocutors have to write (type) and read utterances, yet they refer to their interactions as "chats" and tend to pretend to have synchronous face-to-face (FTF) conversations. Utterances are consciously composed and edited in such a way that they reflect features of spoken language. Participants have at their disposal a set of techniques which they frequently apply to emphasise oral qualities, such as multiple punctuation and typography, non-standard orthography to reflect pronunciation, double negation, ellipsis of subject pronouns, and simple sentences or predominantly paratactic structures instead of co-ordination or subordination.
The newbie experiment revealed that successful and efficient communication on IRC requires newcomers to improve and expand their communicative competence based on real life (RL) registers, asynchronous communication, and synchronous face-to-face (FTF) interactions. Multiple punctuation, multiple typography, and emoticons are some of the special features that interlocutors have to learn on IRC. Another result of the newbie experiment is that pretended orality is not immediately evident for newcomers (their utterances being much better formed and have few, if any, non-standard features). The subjects also had difficulty in interpreting certain vocabulary items, especially acronyms (e.g. LOL), which they had never encountered in RL interactions.
In terms of discursive properties, IRC also shows some unique characteristics which did not initially occur in the utterances produced by the four subjects of the newbie experiment. These features include – inter alia – the use of action descriptions, frequent specification of addressee within a particular utterance, composing and receiving messages simultaneously, and following/processing intertwining messages (reconstructing a multi-dimensional text).
Our results of the analysis of the unique linguistic characteristics on IRC, therefore, appears to support claims that language use within the new medium blends features of both writing and speech, and has some "digital" or "electronic" properties not present in other forms of communication (e.g. emoticons, action descriptions). The observed linguistic phenomena also seems to partly corroborate assumptions proposed by other scholars (e.g. Ruedenberg, 1994; Herring, 1996), describing the silent and visible language of IRC (and that of other forms of computer-mediated communication) as an emerging new register of the English language.