SOCIOLINGUISTICS OF ELECTRONIC MAIL

by Steve Mizrach

Introduction

While there are many people currently involved in research on the new field known as CMC (computer-mediated communications), few are taking a sociolinguistic perspective. Mostly, the questions they have been asking have been technical and pragmatic. These questions have usually been along the following. Does electronic mail increase productivity in work environments? Does electronic communication help foster political participation, community involvement, or knowledge acquisition? Does telecommunication promote media centralization and concentration? Does CMC, by masking cues for race, gender, handicap, physical features, and so forth, level social hierarchies that might exist in face-to-face contexts? Such communications research is important and interesting, but it ignores the basic fact that CMC involves the use of language in a social setting. Thus it requires some sociolinguistic attention. In this paper, I will primarily be focusing on the sociolinguistic properties of e-mail, and not other communication channels on the Internet such as newsgroups, IRC chat channels, or teleconferencing.

Doing a sociolinguistic study of electronic mail would seem inappropriate in that ordinarily sociolinguists are concerned with speech events, and not written texts (except for transcriptions of taped speech), which they often leave to literary critics, communications researchers, or content analysts. Further, generally such speech events are studied in unmediated (face-to-face) contexts, rather than through mediated contexts (such as telephone or CB radio conversations.) Electronic mail is written (or more properly, typed) and mediated. However, as one researcher has noted, email shares the attributes of spoken conversation (especially when the exchange is interoffice or in some other local network), and displays the volatility, evanescence, immediacy, familiarity, and directness of speech. On the other hand, it lacks the paralinguistic features of speech (voice intonation and body language cannot be detected), and it has the deliberative quality of letter writing, in that the person can revise repeatedly what they want to "say" before "saying" it. Whether or not email is speech or writing (or something in between), it involves the use of language, and even if it seems to erase some real-world social variables (whether it actually does so is subject to debate), it still takes place in a social context.

Certain features of email have been discussed almost to death. Since an email message must be typed using the limited 256-character ASCII set (the 'lingua franca' for all computer systems), and cannot employ changes in font, style, or color, it is a highly restricted form of communication. At this point, almost everyone is familiar with the fact that, since emotional cues are missing due to the lack of paralinguistic features, email users seek to avoid miscommunication by conveying emotion through the use of the so-called "emoticons" or "smileys," small facial expressions formed through the use of sideways juxtapositions of keyboard characters. Furthermore, almost everyone is familiar with the long-discussed problem of "flaming", in which the misinterpretation of someone's comments (due to this same lack of emotional cues, mild sarcasm can come across as bitter attack) in an email message lead to a long torrent of accusations and insults. In this paper, I wish to propose some new sociolinguistic hypotheses, and present a proposal in which these hypotheses could be tested.

First, I suggest, there is an attempt to create a standard for email communication. This process of standardization is in progress, and for these reasons provides an interesting way for sociolinguists to examine the creation of standards. Second, people who fall short of this emergent standard (since it's not really yet codified) or violate other norms of electronic communication because they are new to it are likely to receive the secondary social status of "newbie," and lose face as a result of it. Third, on email lists, those people who most closely approach the standard and also display various techniques of verbal mastery are likely to have the most social power, "speak" most frequently, and be listened to the most. In concluding, I will suggest ways in which these hypotheses can be tested in a long-term research project.

The emergence of an e-mail standard

Although the Internet is perhaps less than a decade old, electronic mail, bulletin boards, and conferencing systems have been in use for three decades. However, since most of these systems had been proprietary and not interconnected, each was essentially self-policing and contained internal rules of use. Because of the global, interconnected character of the Internet, various parties have been searching for appropriate standards of conduct for use - a universal "netiquette" that would apply to all of the Internet, and perhaps, by extension, other modes of CMC also. As of now, there have been certain netiquette principles set forth, but no agreement in what to do about violations or enforcement. Mostly, the principles are tacit, assumed to be known to experienced users, and enforced informally through reprimand. Much of this netiquette is outside the focus of this paper - regarding the inappropriateness of software piracy, advertising, invasion of privacy, use of encryption, "spoofing" other peoples' address, computer trespass, and so forth. But the central 'conduct' on the Internet is textual communication, and thus a large portion of netiquette concerns what language is appropriate for such communication.

Thus, a large portion of netiquette concerns the search for a standard for electronic mail. Electronic mail is already 'standardized' to a certain extent because all systems use the same character set, physical-electronic protocols (TCP/IP), and encoding methods - a necessity for a global network. But these standards are merely for the software and hardware that makes the communication possible, not how users employ it. The de facto language of the Internet is English, for the simple fact that the numerical majority of Internet-linked computers are in English-speaking countries, especially the United States, and English is still considered a lingua franca for international business and trade and so forth. However, there is no statement of use on the Internet which specifies which variety , register, or style of English is to be used, and this results in the "Babel" of communication which users so frequently decry in their rush to create a standard. The standard that Internet users are trying to create is loosely based on ordinary Standard English, but it is a modified code with additional features in order to reflect the realities of e-mail. Here I wish to suggest what these linguistic rules may be and who the agencies are that are pushing for their implementation as standards.

The search for standards

It should be noted that within the evolving systems of "netiquette," these standards receive differing levels of attention. Stylistic faux pases are forgiveable and often not commented upon ("You know, your last message lacked impact, expression, and rhythm control!") However, violation of many these other norms may generate a range of responses, ranging from polite requests for the person to stop posting inappropriate messages to retaliation and expulsion from the list or newsgroup. Most of these norms emerge from what seems to be common-sense consideration for other e-mail users, but some of them seem extreme, and there is sharp disagreement over their application (few would agree that sarcasm and humor should be purged from e-mail.) The level of enforcement for these various norms throughout the Internet is low to null; what is more important is that they are tacitly enforced, in that persons who violate the norms are likely to be tagged with the label of "newbie" and lose face. People intrinsically correlate the level of mistakes a person makes with their ignorance of the Internet and its conventions.

At this point, these should be seen as informal expectations of people. They are in no way officially codified. Many of them are contextual to the context in which electronic mail is being used (professional or academic setting.) Indeed, there remains a great deal of disagreement over some of these items even among experienced e-mail users. People are expected to acquire knowledge of these norms through observation, imitation, and informal socialization. It is because of the informality and irregularity of this process that there have been calls for codification by the Internet Society or other organizations - something that should be familiar with sociolinguists who have researched the processes of standardization. In the meantime, new Internet users are continually being exposed to these norms in sudden and painful ways, when they are sharply attacked by old-timers for their mistakes and violations.

The perils of 'newbiehood'

If one is judged to be a 'newbie' on an electronic mail list, one can expect many responses. Some users may offer assistance or guidance, but more often those who consider themselves 'old-timers' will instead offer derision and insults. This judgement may come from many sources - based on the system the person uses to access the Internet (is it America Online[AOL] or a UNIX computer), their ignorance of expected tacit knowledge (embodied in the Frequently Asked Questions [FAQ] file for the list), or the fact that people have not "seen" them before on the list. Most often, it comes from the violation of expected tacit rules of communication. Experienced users say one frequent sign of "newbieness" is "spamming", or the placement of messages in several lists, many of which are in no way topically related to the content of the message. The lack of use of abbreviations (such as IMHO, FAQ, etc.), the lack of emoticons and other emotive textual devices, and poor spelling or grammar are often taken as signs of "newbiehood." The social distance between old-timers and newbies is real, since the old-timers consider the newbies "barbarian" invaders on the carefully civilized turf they have struggled to create.

Old-timers have social power because supposedly they were around when Internet was harder to use (everything was UNIX line prompts instead of point-and-click) and thus they are thought to have greater knowledge and mastery. Because they have 'been around' for such a long time, they are better expected to have received the informal socialization into 'netiquette.' The old-timers feel they played an active role in shaping the Internet into the open, rational, consistent system that it is today. "Newbies" who have just arrived through AOL or other commercial systems are viewed like people who watch commercial TV are by PBS watchers - as "the great unwashed," uncultured, crass, seeking the lowest common denominator. They are assumed to be ignorant of the technology and also the informal rules that are governing its use. Old-timers see newbies as parasites - people preying on the system they created, turning it into something they never wanted it to be.

Social power is linguistic power in virtual communities

While people known to be 'old-timers' often command social power on email lists, it is not necessarily people who have been the longest on Internet in general or on the particular list or channel in specific who command the most attention and respect. Rather, this is often conferred to people who display the most linguistic competence, in terms of their adherence to the standard and their ability to go "above and beyond" it through their artful, clever, and captivating juxtaposition of words. How this verbal mastery is assessed depends on a range of functions - use of general vocabulary as well as specialized jargons or registers, command of rhetorical persuasion and what is considered to be reasoned argument, and avoidance of making comments personal, vicious, or petty, among other things. What is interesting is that newbies who can display this competence may find their social status on the mailing list suddenly elevated.

Social power on mailing lists mostly means one receives more attention to their postings, more perfunctory adulations and 'right ons' for what they post, and more people willing to provide advice and information when they ask for it. Those with social power on mailing lists tend to dominate the discourse, and people tend to go along with their judgements of other users and what they post. They tend to set the tone for the entire list based on their preferences as to what needs to be discussed and what doesn't. Their minor mistakes tend to be easily overlooked and forgiven. In contrast, the social powerlessness of the newbie is often experienced through a lack of attention from other people on the list, a marginalization of their opinions and comments, and derision and ridicule. Personally, I have witnessed social power on mailing lists take on all kinds of amazing manifestations, including people leaving the list nominating their appointed successors for 'list guru!'

Whereas peoples' speech is often evaluated based on its phonology, their written communication (and by inference, their social status, etc.) is usually judged by its morphology (proper use of conjugation, contraction, etc.) and lexical fluidity. However, because a computer screen is a more versatile way of arranging text than a typewriter, and allows for text manipulations that are more precise than those in handwritten letters, other criteria come into evaluating e-mail messages. E-mail messages thus are judged also based on the textual devices the person employs, and even things such as spacing, ASCII embellishments (such as text outline boxes), and alignment. This is interesting since, while such tricks are available to people in conventional computer word processors, they are more often used in e-mail than in print documents. It is also problematic, since the versatility a person may have in creating e-mail may be dependent on the system they are using (Emacs, PINE, RiceMail, Eudora, Pegasus Mail, Tin, etc.)

A person's lack of resident software for generating the textual manipulations that give e-mail messages 'impact' often cannot be assessed by other users, and thus their non-use of such items is assumed to be based on choice or ignorance. People are routinely flamed for things that appear in their messages which are artefacts of their mail gateway, and not things they put in originally (such as the =20 ending that some mail routers insert as a linefeed marker at the end of each line.) Thus, what ends of up happening is this; those who have the most versatile mail software can generate the most versatile e-mail messages. These people often command the most social power on mailing lists. Since the best software is available to those with the money to buy it (of course), what is happening implicitly is another connection is forged between economic and social power - this time through net linguistic competence.

Research proposal

There have been many CMC studies that have narrowly focused on one 'virtual community.' In essence, the researcher has chosen one IRC channel, Usenet newsgroup, mailing list, or BBS conference for micro-analysis of its communication. However, I am proposing that this emergence of a linguistic standard for e-mail is a global phenomenon for all of Internet (and perhaps all forms of CMC), and not just occurring in specific contexts. Thus, in order to assess it, a random sample of material would be best appropriate. Since newsgroups and other text media on the Internet are slightly different from 'classical' electronic mail, the best research survey would be of about 30 Internet e-mail lists, chosen at random, so as to get a broad range of lists based on type (professional, hobbyist, etc.), origin (university or private site), and topical focus. Such lists would be monitored and all public messages archived for a period of six months.

Testing of my hypotheses would mean that the data would be analyzed in the following way. Messages would be coded (using content analysis techniques) as follows: plus one point for positively demonstrating or exceeding a norm (as outline above), minus one point for violating a norm. Each message would have a score, which would add or detract from the cumulative score of their originator. Message posters on the list would receive a cumulative score based on their repeated offenses (or demonstrations of their mastery of the standard). Thus, repeated "spammers" or "flamers" and so on would likely receive overall negative scores, whereas people who flamboyantly display adherence to these norms repeatedly would likely receive high overall positive scores. It would then be interesting to determine how long people had been on the Internet and how this correlated with their scores.

Furthermore, it could then be examined whether people with high positive scores (high obedience to the norms of the standard) commanded high social power on the list (based on side interactions with other users), and whether people were judged to be "newbies" when they had high negative scores, even if they were relatively experienced users. It could be tested whether actual "newbies" who displayed mastery of the standard (perhaps having read a book on e-mail etiquette before getting online) were judged to be experienced users, or at least allocated greater social power on the list. This research gives sociolinguists the opportunity to observe how social status is allocated according to perceived linguistic competence - in an ongoing social process. Unlike "offline" language standards, the standard for email is still coalescing. We can observe who is setting the standards, how they are getting people to comply, and most importantly, why they are doing this.

Bibliography

Return to CyberAnthropology