Computer-Mediated Chat:

Ways of Finding Chat Partners

The technical apparatus is, then, being made at home with the rest of our world. And that's a thing that's routinely being done, and it's the source of the failure of technocratic dreams that if only we introduced some fantastic new communication machine, the world will be transformed. Where what happens is that the object is made at home in the world that has whatever organisation it already has. -- Harvey Sacks (Lectures on Conversation Vol. 2., 548-9)

1Chatting, or having a conversation, has long been a favourite activity for people. It seemed so ordinary, if not to say trivial, that it has for almost equally long not been studied in any dedicated way. It was only when Harvey Sacks and his early collaborators started using the tape recorder to study telephone conversations that 'conversation' as a topic has become established (cf. Sacks, Lectures Vol. 1). Inspired by Harold Garfinkel, the perspective chosen was a procedural one: they wanted to analyse how conversations are organised on the spot. As Sacks once said:

2The gross aim of the work I am doing is to see how finely the details of actual, naturally occurring conversation can be subjected to analysis that will yield the technology of conversation. (Sacks, "On Doing 'Being Ordinary'" 411)

3Later, Sacks also started using data from audio-recorded face-to-face encounters. Most of the phenomena that the research on telephone conversation unearthed could also be found in face-to-face data. Whether something was lost by relying on just audio materials was not clear at the beginning. But with video-based research, as initiated by Charles Goodwin in the 1970s, one was later able to demonstrate that visual exchanges did play an essential role the actual organisation of face-to-face conduct. When using telephone technology, people seemed to rely on a restricted set of the interactional procedures used in face-to-face settings. But new ways to deal with both general and setting-specific problems, such as mutual identification, were also developed. Now that an increasing number of people spend various amounts of their time 'online', chatting with friends or whoever is available, it is time to study Computer-Mediated Conversation (CMC), as we previously studied face-to-face conversation and Telephone (Mediated) Conversation, using the same procedural perspective. We may expect that we will encounter many phenomena that have become familiar to us, and that we will be able to use many of the same concepts. But we will probably also see that people have developed new technical variations of familiar themes as they adapt the technology of conversation to the possibilities and limitations of this new technology of communicative mediation. In so doing, they will make the new technology 'at home in the world that has whatever organisation it already has.'

4Space does not allow a full discussion of the properties of text-based CMC as instantiated in 'chat' environments, but comparing CMC with face-to-face communication and telephone conversations, it is obvious that the means to convey meanings are severely restricted. In face-to-face encounters, many of the more subtle aspects of the conversation rely on visual and vocal productions and perceptions, which are more or less distinguishable from the 'text' that has been uttered. Following the early work of Gregory Bateson, these aspects are mostly conceived of as a kind of commentary on the core communication available in the 'text', that is as 'meta-communication'. While the 'separation' between 'levels' of communication, that these conceptualisations imply may distort what actually goes on in face-to-face encounters, there is no doubt that telephone conversations, in which the visual 'channel' is not available, and text-based CMC, which in addition lacks access to voice qualities, do confront participants with important communicative restrictions.

5An important aspect of text-based computer-mediated chatting is that it offers users an unprecedented anonymity, and therefore an unprecedented licence for unaccountable action, ranging from bland banality to criminal threat, while passing through all imaginable sexual 'perversities'. One upshot of this is that they can present themselves as belonging to any plausible category they may choose, but they will -- in the chat context -- never be sure whether the other participants 'really' are legitimate members of the categories they claim for themselves. In various other formats for CMC, like MUDs and MOOs, the looseness of the connections between the people who type messages and the identities they project in the chat environment seems often to be accepted as an inescapable fact, which adds to the fascination of participation1. The typists can then be called 'players' and the projected identities 'characters', while the interaction can be seen as a game of role-playing. In general chat environments, as the one I will discuss later, such a game-like quality seems not to be openly admitted, although quite often hinted at. Rather, the participants stick to playing who they claim they are. In my own text, however, I will use 'player' and 'character' to indicate the two faces of participation in computer-mediated, text-based chats.

6In the following sections, I will discuss the organised ways in which one particular problem that chat-players have is dealt with. That problem can be glossed as: how do people wanting to 'chat' on the Internet find suitable partners for that activity? The solution to that problem lies in the explicit naming or implicit suggestion of various kinds of social categories, like 'age', 'sex' and 'location'. Chat players very often initiate a chat with a question like: "hi, a/s/l please?", which asks the other party to self-identify in those terms, as, for instance "frits/m/amsterdam", if that fits the character the player wants to project. But, as I will explain, categorisation plays its role both earlier and later in the chat process.

'Membership Categorisation' in Finding Chat Partners

7The following exploration is, then, an exercise in Membership Categorisation Analysis (MCA; Hester & Eglin) as based on the ideas developed by Harvey Sacks in the 1960s (Sacks, "An Initial Investigation", "On the Analyzability of Stories", Lectures on Conversation Vol. 1). An immense part of the mundane knowledge that people use in living their everyday lives is organised in terms of categories that label members of some population as being of certain types. These categories are organised in sets, called Membership Categorisation Devices (MCDs). The MCD 'sex' (or 'gender'), for instance, consists of the two categories of 'male' and 'female'. Labelling a person as being male or female carries with it an enormous amount of implied properties, so called 'category-predicates', such as expectable or required behaviours, capacities, values, etc. My overall thesis is that people who want to chat rely mostly on categorical predications to find suitable chat partners.

8Finding a chat partner or chat partners is an interactive process between at least two parties. Their job involves a combination of presenting themselves and reading others' self presentations. For each, the job has a structure like 'find an X who wants a Y as a partner', where X is the desired chat character and Y is the character you yourself want to play. The set of XY-combinations varies in scope, of course, from very wide, say any male/female combination, to rather narrow, as we will see.

9The partner finding process for chats can be loosely compared with partly similar processes in other environments, such as cocktail parties, poster sessions at conferences, and telephone calls. The openings of telephone calls have been researched extensively by conversation analysts, especially Schegloff ("Sequencing", "Identification", "Routine"; also Hopper). An interesting idea from this work is that a call opening tends to follow a loosely defined pattern, called the canonical model for telephone openings. This involves making contact, mutual identification/recognition, greetings and 'how-are-you?'s, before the actual business of the call is tackled. When logging on to a chat environment, one enters a market of sorts, where the participants are both buyers and sellers: a general sociability-market like a cocktail party. And indeed some writers have characterised chat rooms as 'virtual cocktail parties'. Some participants in a cocktail party may, of course, have quite specific purposes in mind, like wanting to meet a particular kind of person, or a particular individual, or even being open to starting a relationship which may endure for some time after the event. The same is true for CMC chats.

10The trajectory that the partner-finding process will take is partly pre-structured by the technology used. I have limited my explorations to one particular chat environment (Microsoft Chat). In that program, the actual partner-finding starts even before logging on, as one is required to fill in certain information slots when setting up the program, such as Real Name and Nickname and optional slots like Email Address and Profile. When you click on the Chat Room List icon, you are presented with a list of over a thousand rooms, alphabetically arranged, with the number of participants. You can select a Room and click a button to enter it. When you do, you get a new screen, which has three windows, one that represents the ongoing general conversation, one with a list of the participants' nicks, and a window to type your contributions in. When you right-click on a name in the participant list, you get a number of options, including Get Profile. Get Profile allows you to get more information on that person, if he/she has filled in that part of the form, but often you get "This person is too lazy to create a profile entry."

Categorisation in Room Names

11When you log in to the chat server, you can search either the Chat Room List or the Users List. Let us take the Chat Room List first. Some room names seem to be designed to come early in the alphabetically ordered list, by starting with one or more A's, as in A!!!!!!!!!FriendlyChat, while others rely on certain key words. Scanning over a thousand names for those words by scrolling the list might take a lot of time, but the Chat Room List has a search facility. You can type a string and the list will be shortened to only those with that string in their name. Many room names seem to be designed for being found this way, by containing a number of more or less redundant strings that people might use in a search. Some examples of room names are: A!!!!!!!!!FriendlyChat, Animal&Girls, Australia_Sydney_Chat_Room, christian evening post, desert_and_cactus_only, engineer, francais_saloppes, francais_soumise_sub_slave, german_deutsch_rollenspiele, hayatherseyeragmensürüyor, holland_babbel, italia_14_19anni, italia_padania_e_basta, L@Ros@deiVenti, nederlandse_chat, sex_tr, subslavespankbondage, Sweet_Girl_From_Alabama, #BI_LES_FEM_ONLY, #Chinese_Chat, #France, #LesbiansBiTeenGirls_Cam_NetMeeting, #polska_do_flirtowania, #russian_Virtual_Bar?, #tr_%izmir, #ukphonefantasy.

12A first look at this collection of room names suggests two broad classes of categorisation: first a local/national/cultural/ethnic class, and second one oriented to topics, with a large dose of sexual ones. For the first class, different kinds of indicators are available, such as naming as in Australia_Sydney_Chat_Room, and the use of a local language as in hayatherseyeragmensürüyor, or in combination: german_deutsch_rollenspiele. When you enter this type of room, a first function of such categorisations becomes apparent in that non-English categorisations suggest a different language practice. While English is the default language, quite a few people prefer using their own local language. Some rooms even suggest a more restricted area, as in Australia_Sydney_Chat_Room, for those who are interested in chatting with people not too far off. This seems a bit paradoxical, as chatting in a world-wide network allows contacts between people who are physically distant, as is often mentioned in chats. Rooms with such local restrictions may be designed, however, to facilitate possible subsequent face-to-face meetings or telephone contacts, as is suggested by names like Fr@nce_P@ris_Rencontre and #ukphonefantasy.

13The collection of sexually suggestive names is not only large, but also indicative of a large variety of interests, including just (probably heterosexual) sex, male gay sex, female lesbian or bi-sexuality. Some names invoke some more specialized practices like BDSM, and a collection of other 'perversities', as in names like 'francais_soumcateise_sub_slave', 'subslavespankbondage', 'golden_shower' or 'family_secrets'. But quite often sexual interest are only revealed in subsequent stages of contact.

14Non-sexual interests are, of course, also apparent, including religious, professional, political or commercial ones, as in 'christian evening post', or 'culturecrossing', 'holland_paranormaal', 'jesussaves', 'Pokemon_Chat', 'francais_informatique', and '#Russian_Philosophy_2918'.

Categorisation through Nicknames

15Having selected a room, your next step is to see who is there. As chatting ultimately concerns exchanges between (virtual) persons, it is no surprise that nicknames are used as concise 'labels' to announce who is available on the chat network or in a particular room. Consider some examples: ^P0371G , amanda14, anneke, banana81, Dream_Girl, emma69, ericdraven, latex_bi_tch1 , Leeroy, LuCho1, Mary15, Miguelo, SomeFun, Steffi, teaser.

16Some of these are rather opaque, at least at first, while others seem quite ordinary. Anneke, for instance, is an ordinary Dutch name for girls. So, by using this nick name, a person at the same time categorises herself in two Membership Categorisation Devices: gender: 'female' and language: 'Dutch'. When using this type of nick, you will quite often be addressed in Dutch, for instance with the typically Dutch chat-greeting "hoi" and/or by a question like "ben jij Nederlandse?" ("are you Dutch?" -- female form). This question asks you to categorise yourself, using the nationality device 'Dutch/Belgian', within the language category 'speaker of Dutch'. Many other first names like 'amanda' and 'emma', do not have such a language specificity and so do not 'project' a specific European language/nationality as 'anneke' does. Some French names, like 'nathalie' are a bit ambiguous in that respect, as they are used in quite a number of other language communities, so you may get a more open question like "bonjour, tu parle francais?" ("hi, do you speak French?"). A name like 'Miguelo' suggests a roman language, of course, while 'LuCho1' or 'Konusmaz' indicate non-European languages (here Chinese and Turkish, respectively).

17Quite often, a first name nick also carries an attached number, as in 'Mary15'. One reason for such attachments is that a nick has to be unique, so if you join the channel with a nick like 'Mary', there will mostly be another who has already claimed that particular name. An error message will appear suggesting that you take another nick. The easiest solution, then, is to add an 'identifying detail', like a number. Technically, any number, letter or other character will do, so you can take Mary1, or Mary~, or Mary_m. Quite often, numbers are used in accord with the nick's age, as is probably the case in our examples 'Mary15' and 'amanda14', but not in 'emma69', which suggests an 'activity preference' rather than an age category. Some of the other nicks in our examples suggest other aspects, claims or interests, as in Dream_Girl, latex_bi_tch1, SomeFun, or teaser. Other examples are: 'machomadness', 'daddyishere', 'LadySusan28', 'maleslave', 'curieuse33', 'patrickcam', or 'YOUNG_GAY_BOY'. More elaborate information about a character can sometimes be collected from his or her profile, but for reasons of space, I will not discuss its use here.

18This paper's interest is not only in finding out which categories and MCDs are actually used, but also how they are used, what kind of function they can be seen to have. How do chat participants organise their way to 'the anchor point' (Schegloff, "Routine"), at which they start their actual chat 'business'? For the chatting environment that I have observed, there seems to be two major purposes, one may be called social, i.e. 'just chatting', as under the rubric 'friendly chat', and the other is sexual. These purposes may be mixed, of course, in that the first may lead to the second, or the second accompanied by the first. Apart from those two major purposes, a number of others can be inferred from the room titles, including the discussion of political, religious, and technical topics.

19Sexual chats can take various forms, most prominently 'pic trading' and 'cybersex'. As becomes clear from research by Don Slater, an enormous 'market' for 'pic trading' has emerged, with a quite explicit normative structure of 'fair trading', i.e. if one receives something, one should reciprocate in kind. When one is in an appropriate room, and especially if one plays a female character, other participants quite often try to initiate pic trading. This can have the form of sending a pic, without any verbal exchange, possibly followed by a request like 'send also'. But you may also get a verbal request first, like "do you have a (self) pic?" If you reply in a negative way, you often do not get any further reaction, or just "ok." A 'pic request' can also be preceded by some verbal exchanges; social, sexual or both. That question -- "have a pic?" or "wanna trade" -- can then be considered the real starting point for that particular encounter, or it can be part of a process of getting to know each other: "can i c u?"

20The second form of sexual chats involves cyber sex. This may be characterised as interactionally improvised pornography, the exchange of sexually explicit messages enacting a sexual fantasy or a shared masturbation session. There is a repertoire of opening moves for these kinds of games, including "wanna cyber?", "are you alone?" and "what are you wearing now?"

Functions of Categorisations

21Categorisations in room names, nicks and profiles has two major functions: guiding the selection of suitable chat partners and suggesting topics. Location information has quite diverse implications in different contexts, e.g. linguistic, cultural, national and geographical. Language is a primordial parameter in any text-based activity, and chatting offers numerous illustrations for this. Cultural implications seem to be more diffuse, but probably important for some (classes of?) participants. Nationality is important in various ways, for instance as an 'identity anchor'. So when you use a typically Dutch nick, like 'frits' or 'anneke', you may get first questions asking whether you are from the Netherlands or from Belgium and subsequently from which region or town. This may be important for indicating reachability, either in person or over the phone. Location information can also be used as topic opener. So when you mention that you live in Amsterdam, you often get positive remarks about the city, like "I visited Amsterdam last June and I liked it very much", or "I would die to live there" (sic) from a pot-smoking U.S. student.

22After language, age and gender seem to be the most important points in exploring mutual suitability. When possible partners differ in age or gender category, this quite often leads to questions like "Am I not too old/young for you?" Of course, age and gender are basic parameters for sexual selection, as people differ in their range of sexual preferences along the lines of these categories, i.e. same sex or opposite sex, and roughly the same age or older/younger age. Such preferences intersect with straight or kinky ones, of which a large variety can be found. Many rooms are organised around one or another combination, as announced in names like '#LesbiansBiTeenGirls_Cam_NetMeeting', 'Hollandlolita' or '#Lesbian_Domination'. In some of these, the host makes efforts to keep to a more or less strict 'regime', for instance by banning obvious males from a room like '#BI_LES_FEM_ONLY'. In others, an automated welcome message is used to lay out the participation rules.


23To sum up, categorisation plays an essential role in a sorting-out process leading, ideally, to small-group or dyadic suitability. A/S/L, age, sex and location, are obvious starting points, but other differentiations, as in sexual preferences which are themselves partly rooted in age/gender combinations, also play a role. In this process, suitability explorations and topic initiations are intimately related. Chatting, then, is text-based categorisation. New communication technologies are invented with rather limited purposes in mind, but they are quite often adopted by masses of users in unexpected ways. In this process, pre-existing communicational purposes and procedures are adapted to the new environment, but basically there does not seem to be any radical change. Comparing mutual categorisation in face-to-face encounters, telephone calls, and text-based CMC as in online chatting, one can see that similar procedures are being used, although in a more and more explicit manner, as in the question: "a/s/l please?"