This issue of the EJC/REC focuses on discourse analysis -- defined here broadly as the identification of patterns of structure and meaning in language use -- as applied to the analysis of computer-mediated communication (CMC). The impetus for this focus is the growing conviction of many researchers that computer-mediated communication systems are a boon to the study of language-in-use, indeed possibly the greatest boon since the invention of the portable tape recorder in the 1950's.

Where the tape recorder expanded the limits of human memory by enabling verbatim accounts of continuous speech to be recorded and transcribed, text-based computer mediation systems take us beyond the physical limitations of transcription by generating pre-transcribed data which are easily downloaded for analysis. The lack of visual cues in most CMC means that the presence of an observer need not be known to the people whose communication is being studied, thereby allowing for the collection of more authentic, unselfconscious data. Experimental researchers also benefit from the ease with which it is possible to vary aspects of the interaction to test for hypotheses about computers and communication, simply by programming the system to modify electronic messages between the time they are produced by one subject and the time they are received by another.

As if all that were not enough to tempt researchers to investigate computer-mediated discourse, CMC such as takes place on the Internet is also of considerable theoretical interest, as an increasingly popular means of social interaction. By virtue of being carried out exclusively via language, text-based CMC makes transparent as never before the role of language in the presentation of self, and in the genesis and organization of social practices.

Despite this enormous potential, computer-mediated discourse has as yet received little serious attention from language-focused scholars, including those from my own discipline, linguistics. There are various reasons for this, not the least of which is the current fashion within the linguistics discipline for analyzing invented, decontextualized sentences, a trend which has tended to marginalize research on actual language use. Consequently, much of the work that has been published on computer- mediated language to date (e.g., Herring, Ed., 1996) has been done by scholars on the margins of, or entirely outside of, mainstream linguistics. In part because of this, and in part because their numbers are small, researchers studying computer-mediated language have generally had to do so on their own, in relative isolation from one another. Some exciting work has been produced in this manner. However, if the full potential of computer-mediated language studies is eventually to be realized, it is important that available work be brought together, so that the findings of previously isolated researchers can inform the research of others, insights can be consolidated, and the state of our current understanding can more readily be assessed.

The present issue is the first publication, to the best of my knowledge, devoted exclusively to discourse analysis of computer-mediated communication. Three out of the four articles -- those by Paolillo, Condon & Cech, and Miller et al. -- were first presented at a symposium entitled "Computer-Mediated Discourse Analysis" that I organized in March 1995, as a presession of the Georgetown University Round Table on Languages and Linguistics (GURT). The fourth contribution, by Kwang-Kyu Ko, was added in response to a call for papers distributed subsequently on the Internet.

The four articles have in common that they report empirical studies of computer-mediated language. All of the studies also have self-evident implications for issues of practical concern, whether it be the use of CMC in educational settings, for decision-making in organizations, for machine translation, or for the status of languages other than English on the Internet. Beyond these similarities, the four articles analyze different types of CMC -- synchronous chat in natural and experimental settings, asynchronous postings to a Usenet newsgroup, and voice-only computer-mediated transmission. They also represent strikingly diverse approaches to discourse analysis. Ko's text linguistic study considers the distribution of grammatical and lexical markers in a computer-mediated corpus, in an attempt to situate synchronous CMC with respect to spoken and written language. Condon and Cech are psycholinguists concerned with discourse processing, specifically with the differences between computer-mediated and face-to-face negotiation of decision- making tasks. Miller, LuperFoy, Kim and Duff are computer- human-interface researchers designing a discourse module for dialogue management in a machine translation system. Finally, Paolillo brings a sociolinguistic perspective to bear on the analysis of code-switching in a bilingual Usenet group. While the collection is modest in size, and does not represent all approaches to discourse analysis, it nevertheless gives a good indication of the theoretical and methodological diversity that exists within the study of language use, and the fruitfulness of such approaches for the study of computer-mediated language.

Needless to say, this collection is only a beginning. Many other questions about computer-mediated language cry out to be addressed, especially as CMC technology evolves to permit multiple modalities of message transmission. And other methods of language analysis -- from discourse analysis, sociolinguistics, rhetoric, psycholinguistics, corpus linguistics and elsewhere -- await application to computer-mediated data, with the promise of results and new understandings we can as yet only imagine. With open- mindedness and hard work, we may yet realize the boon to language studies that is computer-mediated communication.

Credit is due to the following people for their high- quality professional efforts in producing this issue: the authors of the articles, 18 anonymous reviewers, and the managing editor of the EJC/REC, Teresa Harrison. I extend to them all my sincere thanks.

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

Susan Herring
Program in Linguistics
University of Texas
Arlington, TX 76019 USA

Phone: (817) 272-3133
Fax: (817) 272-2731
E-mail: susan@utafll.uta.edu

Ce numero de la EJC/REC tente de faire le point sur l'analyse du discours - definie ci-apres en tant que identification de formes de structures et de significations dans l'usage du langage - telle qu'elle s'applique dans l'analyse de la communication assistee par ordinateur. Ce qui galvanise bien des chercheurs a se concentrer sur ce sujet c'est leur conviction sans cesse croissante que les systemes de communication assistes par ordinateur constituent une aubaine inesperee pour l'etude du langage parle depassant de loin le profit que l'on a pu tirer de l'emploi du magnetophone portatif depuis son invention dans les annees cinquante.

La ou le magnetophone a fait reculer les limites de la memoire humaine en rendant possible l'enregistrement et la retranscription, mot pour mot et sans interruption de conversation, le systeme de langage base sur des textes et assiste par ordinateur nous permet de depasser les limitations imposees par la retranscription en produisant des donnees retranscrites d'avance qui peuvent facilement etre telechargees dans le but de les analyser. L'absence de signaux visuels dans la plupart des discours assistes par ordinateur (DAPO) signifie que les gens dont on etudie le discours n'ont pas besoin de savoir qu'on les observe, ce qui permet de reunir aisement des donnees moins intimidantes et plus authentiques. La possibilite de varier les aspects de l'interaction qui servent a faire une recherche d'hypotheses concernant les ordinateurs et la communication serait d'une grande utilite aux chercheurs experimentaux. Il leur suffira tout simplement de programmer le systeme en vue de modifier les messages electroniques pendant l'intervale qui separe leur production par un sujet et leur reception par un autre sujet.

A tous ces elements qui incitent les chercheurs a explorer les possibilites du discours assiste par ordinateur vient s'ajouter le discours tel qu'il se produit sur l'Internet, et qui presente egalement un interet theorique considerable en tant que moyen d'interaction sociale croissant. En raison du fait qu'il se produit exclusivement par le langage, le systeme de lan age base sur des textes et assiste par ordinateur met plus que jamais a nu le role du langage dans la presentation du moi et dans la genese et l'organisation des us et coutumes.

Malgre les enormes possibilites qu'offre le discours assiste par ordinateur, les specialistes du langage, y inclus les praticiens de ma propre discipline, a savoir la linguistique, ne s'en sont jusqu'a present peu preoccupes. On pourrait avancer toutes sortes de raisons de cet etat de choses; il suffit de mentionner la vogue actuelle qui prevaut dans la discipline de la linguistique et qui se traduit par une analyse recherchee des phrases inventees ou prises hors de leur contexte. Cet engouement porte les chercheurs a delaisser la recherche sur l'usage propre du langage. On constate donc que la plupart des travaux de recherche faits sur le langage assiste par ordinateur et publies jusqu'a present (par exemple, Herring, Ed., 1996) ont pour auteurs des specialistes qui s'interessent peu ou prou a la linguistique dominante. Les raisons evoquees ci-dessus et le nombre restreint de chercheurs expliquent que ces specialistes de l'etude du langage assitee par ordinateur se soient retrouves isoles et livres a eux-memes dans leurs recherches. Il n'en est pas moins vrai que des travaux importants ont ete produits dans de telles circonstances. Cependant si l'on veut realiser les promesses que tiennent les etudes du langage assistees par ordinateur, il faut avant tout grouper les travaux deja realises pour que les decouvertes faites par certains chercheurs isoles puissent en faire profiter d'autres, que l'on puisse assortir leurs connaissances et evaluer plus facilement l'etat de notre acquis actuel.

Ce numero du journal electronique de la communication est le premier a ma connaissance qui soit consacre exclusivement a l'analyse du discours de la communication assistee par ordinateur. Trois des quatre articles - ceux de Paollilo, de Condon & Cech et de Miller et coll. - firent d'abord l'objet d'une communication faite a un colloque intitule: "L'analyse du discours assistee par ordinateur" et que j'avais organise en mars 1995. Ce colloque devait servir de reunion preliminaire a la table ronde sur le langage et la linguistique qui se tint a l'universite de Georgetown. Le quatrieme article qui a ete insere dans le Journal electronique de la communication fut redige par Kwang- Kuy Ko. Cet article nous parvint apres que nous eumes lance une demande d'articles, laquelle par la suite parut sur l'Internet.

Ces quatre articles font tous etat d'etudes empiriques sur le langage assiste par ordinateur. De plus, ces etudes posent des questions d'ordre pratique et mettent clairement en evidence les suites que pourraient avoir l'usage possible de la communication assistee par ordinateur dans un cadre formatif, les prises de positions dans les organismes d'Etat, la traduction par machine et la situation des langues, a l'exclusion de l'anglais, sur l'Internet. La s'arretent cependant leurs ressemblances. Chaque article constitue une analyse d'un type different de communication assistee par ordinateur - bavardage synchrone dans un milieu naturel et experimental, enregistrement asynchrone a un << newsgroup USENET>>, retransmission vocale assistee par ordinateur. Chaque article, d'une facon frappante, aborde diversement l'analyse du discours. Dans son analyse linguistique du texte, Ko analyse la repartition des reperes grammaticaux et lexicaux dans un corpus assite par ordinateur en vue de resituer la communication synchrone assistee par ordinateur par rapport au langage ecrit et parle. Condon et Chech, psycholinguistes, se preoccupent avant tout du traitement du discours axe sur les differences entre la negociation face-a-face et la negociation assistee par ordinateur des taches decisionnelles. Miller, LuperFoy, Kim et Duff font de la recherche dans le domaine de l'interface homme-ordinateur. Ils sont en train de mettre au point un module de discours pour la gestion du dialogue dans un systeme de machine a traduire. Enfin, Paollilo, dans sa propre optique sociolinguistique, nous donne une analyse d'echange de codes culturels dans un groupe Usenet bilingue. Bien que ce recueil ne soit pas volumineux et qu'il ne contienne pas toutes les methodes d'analyse du discours, il nous offre neanmoins une bonne idee des diversites theoriques et methodologiques que l'on trouve dans l'etude de l'usage du langage et de l'utilite de ces memes methodes pour l'etude du langage assistee par ordinateur.

Il vat sans dire que ce recueil n'est qu'une ebauche. Bien des questions concernant le langage assiste par ordinateur ne demandent pas mieux que d'etre posees d'autant plus que les techniques de la communication assistees par ordinateur evoluent pour permettrent la multiplication des modes de transmission des messages. D'autres methodes d'analyse du langage tirees - de l'analyse du discours, de la sociolinguistique, des corpus linguistiques et d'ailleurs-pourront a l'avenir etre appliquees aux donnees assistees par ordinateur. Elles fourniront des resultats prometteurs et des connaissances nouvelles que l'on ne peut qu'imaginer a present.

Nous tenons a remercier les personnes suivantes qui ont mis leur immense talent professionnel au service de ce numero: les auteurs de cet article, les 18 critiques anonymes et la redactrice du EJC/REC Teresa Harrison.

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

Susan Herring
Programme en linguistique
Universite du Texas
Arlington, TX 76019 EUA

Telephone. (817)272-3233
Telecopie. (817)272-2731
Courrier electronique. susan@utafll.uta.edu

Abstract. This study investigates the structural characteristics of computer-mediated language by comparing one form of synchronous computer-mediated communication (CMC), Daedalus InterChange, with analogous spoken and written corpora. The analysis focuses on 28 linguistic features claimed in previous research to charac- terize prototypical attributes of speaking and writing. Each feature was examined and quantified to determine the structural characteristics of InterChange in comparison with the two other modes.

Three distinctive feature occurrence patterns emerge as a result of the analysis. In the first pattern, the linguistic features of InterChange occur with higher frequency than in written dis- course, but lower than in spoken. This occurrence pattern reflects a general tendency for InterChange to pattern like spoken language, offset to varying degrees by written-like constraints on the medium. In the second pattern, linguistic features occur with higher frequency in InterChange than in either spoken or written discourse. This pattern is claimed to be the result of users compensating linguistically for electronically-determined communicative constraints, especially the lack of turn-taking and interactional cues. In the third pattern, linguistic features occur less often in InterChange than in either spoken or written discourse. The lower occurrence of these features, which tend to be associated with elaborateness in writing, is analyzed as a consequence of the heavy processing burden placed on InterChange users, who are under both physical demands imposed by the need to type, and cognitive demands imposed by the real- time communication situation.

These results reveal that the InterChange discourse mode is not merely intermediate between (or a "combination of") speaking and writing; rather the electronic medium uniquely fosters some behaviors and inhibits others, in support of the view that the physical mode of communication shapes language use.

Abstract. This article compares discourse management strategies in face-to-face and computer-mediated interactions involving four decision-making tasks. A schema-based approach to discourse processing is adopted which claims that participants will employ the same decision- making routine in both modalities. However, due to differing constraints in the two modalities, we predict that management strategies should differ across modality. In particular, the greater effort and attentional demands of computer-mediated communication suggest that users will attempt to find more efficient ways of communicating.

We examine these issues in qualitative and quantitative analyses of the data using an utterance-unit coding system to identify discourse functions. The analyses show that participants in computer-mediated interaction encode discourse management functions more explicitly than those in face-to-face interactions. In addition, there is a clear preference for encoding management functions as first pair-parts of adjacency pairs, which results in utterances that manage turn- taking as well. The results suggest that participants in electronic communications compensate for decreased efficiency by adopting management strategies that pack more information into fewer utterances, i.e., by relying heavily on the implicit knowledge of a shared problem-solving schema.

Abstract. The empirical study reported on here is an effort to extract design constraints for the dialogue manager for a complex genre of computer- mediated communication: real-time, spoken language dialogue between two telephone conversants speaking two different languages in a voice-only exchange. This family of applications for bilingual spoken communication is termed the interpreting telephone (IT). We sought to discover desiderata for the discourse module of such a system, abstracting away from all aspects of the problem other than those related to discourse processing. To simulate a zero computational cost for machine translation and speech processing, a "Wizard of Oz" study was designed. All input and output language was in the native language of the wizard, so that "translation" was from English to English.

Even with the cost of all other computation reduced to near zero, lengthy delays entailed by half-duplex message transmission and the absence of system status indicators caused users to become impatient and easily distracted by other activities, and to lose track of whose turn it was to speak. This article describes the data collection procedure and our analysis of the discourse results, including the discovery of several inadequacies of the standard system design for interpreting telephone systems. We conclude by proposing a design for a multi-modal system to address these inadequacies.

Abstract. This article examines factors influencing language choice on the Usenet newsgroup soc.culture.punjab, a forum chartered for discussion of the culture of the Punjab region of India and Pakistan. English and Punjabi are both used on soc.culture.punjab, although English is statistically predominant, and the majority of Punjabi words are isolated borrowings occurring in otherwise English messages. The results of a functional analysis of code-switching reveal that Punjabi is used for marked discourse functions (often negatively), and only with interlocutors who are Punjabi community members. This limited usage is accounted for in terms of four factors which disfavor Punjabi use: inter-generational language shift, cultural ambivalence among expatriates, the prestige status of English in South Asia, and the predominance of English on the Internet.