Abstract

This article reports on a study of the use and form of greetings and closings in the emails of two New Zealand workplaces: an educational organization and a manufacturing plant. Using discourse analytic techniques, 515 emails were analyzed and a number of differences were identified. In the educational organization, where restructuring has resulted in low staff morale and a mistrust of management, indirect and socially distant styles of communication prevailed and greetings and closings were not widely used. In the manufacturing plant, the more extensive use of greetings and closings reflected and constructed the open and positive relationships between staff and management and the direct, friendly, and familial workplace culture. The findings suggest that workplace culture is a more important factor accounting for the frequency and form of greetings and closings than are relative status, social distance, and gender

Introduction

The closing years of the 20th century saw the introduction and widespread adoption of email as a means of workplace communication. Email is now a fact of life in many workplaces, where it has largely replaced written memos and much telephone and face-to-face interaction. In some workplaces in the corporate world, email has become the primarycommunication medium, and many of today's workplaces could no longer function without it. It plays an important role in the transmission of information and, in general, in dealing with everyday administrivia at work (Waldvogel, 2005). The main advantage of email over other modes of communication is that it enables people to communicate speedily the same information to many others in diverse locations and time zones. It is also valued because it provides an audit trail and record of the communication.

Interestingly, greetings and closings perform as important a social role in email as in other forms of interactions. As Eckert and McConnell-Ginet (2003) note: "Greetings and farewells offer formulas to ease the strain created for face by the beginnings and ends of interactions" (p. 138). The absence or presence of a greeting and the type of greeting set the tone for the email conversation that follows. The greeting is one means by which the writer constructs his or her social and professional identity and relationship with the addressee(s). A closing can help consolidate the relationship and establish a relational basis for future encounters. A study of greetings and closings can therefore provide valuable insights into people's relational practices at work and, on an organizational level, into the organizational culture of the workplace, since the aggregate tone of individual emails plays a constitutive role in constructing the organizational culture.

This study investigates greetings and closings as distinctive stylistic features of workplace email, in terms of the relationships that exist between their form and use, the workplace or organizational culture, and the sociolinguistic variables of status, social distance, and gender of interlocutors. Focusing on greetings and closings, the article explores some of the ways in which email communication in two workplaces contributed to the construction of aspects of social and professional identity and provided indications of the nature of the workplace culture and its current climate or "state of health."

The article first reviews previous research on email in the workplace. This is followed by a description of the discourse analysis methodology used in the study. In the findings section, the use of greetings and closings in the two organizations is described, along with how these linguistic features contribute to the construction of aspects of social identity. The findings are then discussed and explanations are proposed for the differing patterns found in the two organizations.

Literature Review

Initially, computer-mediated communication (CMC), including email, was viewed as being less personal than face-to-face or telephone communication because email filtered out the intonational and body language cues present in other modes of communication and lacked social presence, thus rendering its messages more impersonal. The proponents of Media Richness Theory (Daft & Lengel, 1984) described email as a "lean medium" incapable of communicating rich information and suitable only for unequivocal or single-meaning, task-based informational messages.

More recent research is showing increasingly, however, that far from being "lean," email is capable of conveying rich information (Abdullah, 2003; Huang, Watson, & Wei, 1998; Markus, 1994; Ngwenyama & Lee, 1997; Waldvogel, 2005; Williams, 1999; Zmud & Carlson, 1999). The extent to which email is able to do this is largely dependent on the relationship between the communication participants and the kind of organization to which they belong. The active construction of rich meaning is accomplished most effectively between participants who have a well-established relationship in organizations with a similar culture. Abdullah's research demonstrates that workplace emails do much affective as well as transactional work. She describes email as "a rich repository of relational communication" that allows writers the flexibility to personalize their messages (Abdullah, 2003, p. ii.). She shows that, using email technology together with their own linguistic resources, writers are able to convey relational information in task-based messages.

The hyper-personal model posited by Walther (1996) also offers an alternative to the "cues-filtered-out" perspective. According to this model, the lack of non-verbal cues is actually an advantage for the writers of email messages, because they have more control over the planning, composing, editing, and delivering of their messages than do face-to-face communicators. This may help them create more polite messages. An analysis of requests made via email and voicemail (Duthler, 2006) indicates that, overall, email requests were more polite than voicemail requests, providing support for Walther's observation that CMC technologies, particularly asynchronous text-based CMC, can facilitate socially desirable communication.

How email writers express relational aspects of communication, namely concern for and interest in others (positive politeness) together with consideration for the need of others not to be imposed upon (negative politeness), has been addressed only very incidentally in the literature to date. Murray notes that several studies found an increase in politeness markers, which she attributes to the transactional forms of many computer-mediated communications. "The absence of politeness markers could leave readers uncertain about the illocutionary force of the request or annoyed at the impoliteness and perhaps the inappropriate assumption of authority" (Murray, cited in Mulholland, 1999, p. 75). Mulholland's own data also show the presence of many politeness markers, although the forms chosen tended to be brief ones, e.g., OK.

In as much as greetings and closings pay attention to the recipient and are oriented to the addressee's face needs (see Goffman, 1967), they are politeness markers. Like other politeness markers, they serve an important function in constructing and maintaining workplace relationships. Greetings and closings enable the writer to express warmth or distance, expressions that are otherwise difficult to do in email, and they are a strategy for personalizing messages as well as a means of reinforcing status relationships and underlining positional expectations.

Kankaanranta (2005) notes that salutations (greetings), closings, and signatures frame messages as being relational and involved. She found that a high percentage of the messages in a multinational corporation, written in "lingua franca English" by Swedes and Finns, started with a salutation and first name, and she notes that the use of salutations in messages seems to be more common among non-native English speakers. She suggests two reasons for the frequent use of salutations with first names in particular. One is that because email is a descendant of the American internal memo, email writers familiar with the memo format were more likely to adopt its "no salutation" usage than those who were not familiar with it, such as the Swedish and Finnish writers whose emails she studied. The second reason is that by using salutations, the writer "constructs a relationship with the recipient, and the usage thus contributes to the maintenance of good social relations" (p. 359). The use of signatures and closings, another widespread practice in the multinational corporation studied by Kankaanranta (2005), also seemed to contribute to this and helped give the messages a positive tone.

Recent non workplace-related research also supports the importance of politeness in email messages in constructing and maintaining workplace relationships. Bunz and Campbell (2002) found that messages containing both verbal (e.g., please, thanks) and structural (greetings and closings) indicators elicited the most polite responses. They observed that email recipients detect politeness indicators and accommodate to this by including similar politeness indicators in their email responses. In another study, Jessmer and Anderson (2001) noted that message recipients viewed more positively messages that were polite and grammatically correct than messages that were impolite and ungrammatical. Polite messages were viewed as having been written by a more friendly and likeable person than impolite messages. Not surprisingly, recipients were more likely to want to work with the senders of polite messages and with the senders of grammatical messages, whom the recipients perceived as being more concerned with them.

Status has been shown to affect the use of politeness markers, including signatures, in email. Since the sender is identified at the top of the email, signatures are, strictly speaking, redundant. Bearing this in mind, Sherblom (1988) studied the email files of a large organization and found that relative social position in the organizational hierarchy influenced the use of signatures. None of the messages sent down the organizational chain were signed, whereas one-third of those sent upwards had signatures.

In the Malaysian context, Abdullah (2003) found that writers were particularly sensitive to the relative status of the recipients of their messages and to the "weight of the imposition" (Brown & Levinson, 1987) they wished to convey. Findings in a pilot sample of 50 New Zealand emails showed a different pattern, however (Waldvogel, 1999). Nearly all the women (97%) and most of the men (87%) who sent messages up the hierarchical chain used some sort of sign-off or closing, and five of the six people (83%) who sent emails down the chain also signed them with their name. The greater general use of signatures may be a reflection of a more collegial atmosphere in this workplace or of the more egalitarian New Zealand attitude to interpersonal relationships at work. Whatever the explanation, these differing patterns suggest that there is good reason to consider further the influence of status and social distance on the use of politeness markers such as greetings and closings in email communication, and that patterns of use vary across cultures and organizations.

Research supports the notion that workplaces tend to develop their own unique email style, reflecting organizational cultural differences. Gains (cited in Murray, 2000) examined 116 randomly-selected email messages exchanged within an insurance company and within and between universities. The insurance company messages used a semi-formal style, did not incorporate features from conversational discourse, tended not to include an opening greeting, and used few features of simplified register. By contrast, the university emails exhibited a range of styles. They adopted features from conversational discourse (e.g., well, you see), included some form of greeting, and often referred to the medium itself.

Research on email in the workplace has thus increasingly shown that in addition to the transactional business carried out by email, affective messages are also conveyed. These messages construct, signal, and define interpersonal relationships and organizational cultures.

Method

This study of greetings and closings in workplace email is one aspect of a larger study (Waldvogel, 2005) exploring the relationship between the organizational or workplace culture and the role, status, and style of email. The research, which is essentially qualitative and ethnographic in nature, explored these relationships in two very different organizations, an educational organization (SCT) and a manufacturing plant (Revelinu). As part of this research, which was linked to The Language in the Workplace Project,¹ staff were surveyed about their use of email, their attitudes toward it, and their email practices.

The greetings and closings in emails written by the employees in each organization were analyzed to identify the influence of the sociolinguistic variables of status, social distance, and gender on their form and use. The analysis was done on a simple count of the various types of greeting or closing cross-tabulated to the variables. The following operational definitions were used:

Greeting—the use of a person's name² and or greeting word to initiate the email.

Closing—any name sign-off, farewell formula (e.g., Cheers), or phatic comment (e.g., Have a good day) used to end the email. Thanks is counted as a closing when it comes with or without the writer's name at the end of a message. In this article the term "closing" is used interchangeably with sign-off.

Data

Each organization provided two email samples: One was a week's inward and outward messages from a senior manager (SM); the other was a set of emails related to a particular issue. Because of the greater use made of email in the educational organization (Waldvogel, 2005), the SCT corpus (394 emails) was over three times as large as that from Revelinu, the manufacturing plant (121 emails). However, as the samples collected from each organization were of a similar nature, i.e., they each contained a week's emails from a senior manager and a set of emails related to one issue, the difference in corpus size should not greatly affect the patterns shown. The focus was on emails written within the organizations, so only those written by and addressed to people within each organization were included in the study.

Tables 1a and 1b show the number of male and female message writers and the percentage of messages written by each gender. The apparent disproportionate number of emails sent by males in the issue-related sample from SCT (Table 1b) is because many of the messages were written by the key person³ who provided the emails for that sample, who happened to be male.

Apart from a few messages of a confidential or personal nature that were deleted, 273 of the 394 messages in the SCT corpus were the internal work-related messages received and sent by the key person over a one-week period. The other 121 were emails written over a two-and-a-half month period and related to one issue, the restructuring of a section. The majority of the messages in the SCT corpus, approximately 75%, were to or from members of the teaching staff. The contribution of the key people to the message writing needs to be noted. In the 273-message corpus, the key person wrote 66% of the male messages and 39% of the total number of messages. The key person in the issue-related corpus contributed half of the total number of messages and nearly three-quarters (74%) of the male messages. One of the other key people wrote 14 (17%) of the male messages, one person wrote four (5%), and the other three males wrote one message each. Seventeen (44%) of the female messages were written by one person; the other 13 women each wrote between one and three messages. The overall findings may therefore be partly influenced by the personal styles of the key people.

A total of 121 messages were collected from Revelinu. These came from two main sources. The largest group was from an acting manager who saved his inward and outward emails for five days. A smaller group came from a senior person in the organization who saved all the emails he received and sent relating to a product item. In addition, there were a few emails from two people who recorded all of their communications for one day. The male acting manager wrote 29 of the messages and the male senior person in the organization wrote nine. The other authors each wrote between one and four messages.

The Two Organizations

The two organizations were very different. The educational organization (SCT), a government funded distance education provider, had, at the time of the research, a teaching and administrative staff of over 500 who were predominantly middle-class, tertiary educated, and mono-cultural European New Zealanders.

Revelinu, the manufacturing plant, is part of a multi-national organization that produces a wide range of personal care and other products. It had a staff of some 300 people; approximately one-third worked in the administrative areas and were predominantly European New Zealanders. The remainder worked on the factory floor; a high percentage of these workers were Maori or from Pasifika and other non-English speaking backgrounds. About one-quarter of the total staff was tertiary educated.

Findings

The analysis showed that while status, social distance, and gender had some influence on the presence or absence of email greetings and the form these took, the organizational culture appears to have a greater influence.

Greetings and closings like the following were common in the two workplaces studied

Many workplace emails, however, do not start with a greeting but rather launch straight into the business of the email. Many also end baldly,⁵ as in this example from the educational organization (SCT):

As Table 2 shows, the preferred form of greeting was quite different in the two organizations investigated. At SCT, 59% of the messages began without any form of greeting, and an additional 21% started simply with the person's name. Only 20% of the messages contained any general greeting. Hi was the most popular general greeting (10% of messages), followed by Dear and Hello (about 3% each) and Good Morning (2%). Two messages used the Maori greeting Kia ora.

At Revelinu, in contrast, the figures were almost reversed. Most messages (58%) began with a greeting word, usually accompanied by the addressee's name (53%). Another one-quarter began with a name only, and in sharp contrast to SCT, only 17% began baldly. Here, too, Hi, found in half of the messages, was the most popular greeting word. The second most favoured greeting word—starting 7% of the messages—was Dear. However, with only one exception, Dear was used to start messages to a group. Eighteen percent of the messages addressed to a group began in this way. Good morning was used twice, including once to a group, and Hey, once.

While it might be thought that an initiating message, or a first communication on a topic, would be more likely to have a greeting than a follow-up or response message, this seemed to have no effect in either organization.

In the SCT corpus, 161 (59%) of the messages were first or sole communications on a topic. The number of follow up messages (4%) was even smaller than at Revelinu; consequently, the presence or absence of greetings in these messages is not of great significance. None of these response messages contained a greeting word. Approximately equal numbers of them started either baldly (six messages) or with the person's name only (five messages). Thirty-seven percent of the messages were responsive. A slightly higher percentage of these messages started baldly (66%) than did those initiating a communication (59%). Eighty-seven percent of responsive messages started either baldly or with just the person's name, compared to 82% of first messages.

The 20 messages to individuals in the corpus of the SCT senior manager that started with a greeting word and or name, with one exception, included all those that introduced a matter of a fairly delicate nature, made a major request of a higher status person, or expressed appreciation for a major request. One such example is given below. In this message the formal greeting Dear shows negative rather than positive politeness to the addressee and conveys respect and deference rather than solidarity.

In the Revelinu corpus, where two-thirds of the messages were first or only communications on a topic, the percentage of messages having no greeting (17) was the same whether the message was an initiating or sole message, part of an ongoing dialogue, or a response. However, as only 10% of the messages in this corpus were follow-up messages, the numbers are too small to be more than indicative. Sixty percent of initiating messages contained a greeting word (8%) or a greeting word and a name (52%). This compares to 45% of the response messages. In these, the greeting word was always accompanied by a name.

With two exceptions, the messages that did not have a greeting or acknowledge the sender fit into one of four categories:

Crystal (2001) has observed that between people who know each other, "Greetingless messages are usually promptly sent responses, where the responder sees the message as the second part of a two-part interaction (an adjacency pair), for which an introductory greeting is inappropriate" (p. 100). Over half of all the messages in the present study, however, contained some parting formula such as Regards, Cheers, or Thanks. The two exceptions, mentioned above, were a quick query between close colleagues and a message to a group.

The number of messages to be dealt with and lack of time are sometimes given as reasons for not using a greeting in a routine email message. However, the considerable difference between the two organizations in the preferred form of greeting suggests the need to go beyond these reasons for an explanation and to view email message style, including greeting form, in the context of the organizational culture. The following is typical of the messages found in the SCT corpus:

Compare this to a typical message from the Revelinu corpus:

Whereas both messages are brief and have a transactional purpose, the informal greeting and closing of the second create a greater sense of solidarity than the more detached tone of the first. This difference may well reflect and help create the different organizational cultures.

Closings

The use of closings in the two organizations closely mirrored that of greetings, as Table 3 shows.

At SCT, in spite of the preference expressed by those surveyed⁷ to have a closing, over one-third of the messages (34%) ended without one, and another 38% ended with the writer's first name only. Only a small percentage contained some kind of farewell formula. Thanks (7% of the messages) was the most common closing, followed by Regards (6% of the messages) and Cheers (5% of the messages). Sometimes Thanks was used genuinely to express thanks for something done or in anticipation of the meeting of a request, but there were also a number of instances where it was used more as a ritual closing formula, as in AC41a above.

At Revelinu, in contrast, three-quarters of the emails signed off with a closing word such as Cheers or Regards, a phatic comment such as Have a nice day or Thanks, and the person's name. Thanks, used either as a farewell or an expression of gratitude for something done or about to be done, followed by the person's name, was the most popular closing and ended 29% of the messages. Regards (19% of the messages) and Cheers (15% of the messages) were the next most popular closings and the only others that reached double figure usage. Regards seems to be used where greater formality is sought; of the 7% of messages signed off with both a first name and a surname, nearly half used Regards.

Through the linguistic choices they make, people construct their own social identity and that of others. Like many other linguistic tokens, greetings and closings encode social information. Three important pieces of social information encoded are status, social distance, and gender. The greetings and closings were analyzed to see how their use contributed to the construction of these aspects of social identity. All three proved to be relevant to some extent.

Status

In the workplace, some people have the right to expect others to do their bidding because of the legitimate power their institutional role gives them. In both organizations, people of higher status were more likely to be greeted or acknowledged by name than were people of lower status. This suggests that greetings or the use of a person's name are seen as a form of politeness and/or as a way of constructing the recipient as worthy of respect. Table 4 below shows the effect that status had on greeting use.

At SCT, only 17% of the messages sent to higher status people started with a greeting and the person's name. This was only marginally higher than the 15% corpus average and not markedly higher than the 10% for messages sent to lower status people. Similarly, while nearly 60% of the messages in the corpus as a whole contained no greeting or acknowledgment of the receiver, the figure for messages to higher status people was 45% and 70% for messages to lower status people. Thirty percent of the messages to equals contained greeting words.

At Revelinu, of the messages sent to individuals whose status was known, about 50% were sent to people of equal status, whereas the remaining 50% were divided almost equally between people of higher and lower status than the sender. Because of the small numbers, the findings are suggestive only. Even though greetings are used widely in this plant, there was still a higher percentage in messages going to people of higher status. Nearly three-quarters of these messages contained a greeting word plus the person's name, compared to 50% to people of lower status, 62% to equals, and just over half for the corpus as a whole. However, whereas two of the messages to higher status people started baldly, all of the messages to lower status people started with either the person's name, e.g., Neil, or a greeting word, e.g., Hi or both e.g., Hi Amy, suggesting that in this organization, people of lower status are acknowledged and treated with respect.

Signing off or adding a closing to a message is also a way of doing deference or signaling respect and thus constructing the addressee as having status. At SCT, messages sent to higher-ranking people were more likely to be signed off in some way (69%) than those sent to lower-ranking people (57%). However, status had little effect on the type of closing used, although greater use was made of Thanks, with or without the sender's name, in messages sent to higher (16%) than to lower-ranking people (9%). At Revelinu, there was no great difference in the closings in messages to higher or lower status people. The most common sign-off for messages to both was Thanks + name (53% of messages to higher status people and 63% to people of lower status). The main difference was in messages to equals. Half of these messages ended with a closing of some sort, either a word of farewell or phatic comment or the person's name. The corresponding percentages for messages to higher and lower status people were 20 and 12, respectively. This may indicate greater solidarity between people of equal status. A larger sample might, however, have produced a different pattern.

Social Distance

Whereas status refers to the role relationship between the writer and reader in terms of their professional duties, social distance refers to their degree of acquaintance on both a personal and a professional level. In a large workplace, social distance is a factor of how well people know each other and the frequency of their interactions. Table 5 below shows how social distance affected greeting use. The message writers were asked to state their social distance relationship to their reader using the following simple operational definitions:

In both organizations, greetings contributed to constructing relationships as more or less distant; distant colleagues were more likely than close colleagues to be both greeted and acknowledged by name.

At SCT it was the messages to distant colleagues that were most likely to start with a greeting and name. However, unlike at Revelinu, messages to groups were more likely than messages to close colleagues to start with a greeting word or a greeting word and name.

At SCT there were no real differences in the closings of messages to distant and close colleagues. Almost equal numbers of messages ended baldly, 35 and 38%, respectively. The most noticeable difference was in the messages to groups. These messages were more than twice as likely as messages to distant and close colleagues to end with either Thanks + name or farewell word + name. Forty percent of group messages ended this way compared to almost half that number for both close (19%) and distant (22%) colleagues.

At Revelinu 76% of distant colleagues were addressed with a greeting word and name compared to 48% of close colleagues, which was fewer than for the corpus as a whole. Compared to the corpus as a whole, a higher percentage of messages to close colleagues started with just their name or a greeting word by itself. Messages being sent to a group were the least likely to start with both a greeting and name acknowledgment. Sixty percent of group messages either started without a greeting or with just a general group name compared to the corpus average of 42%. However, in contrast to SCT, a similar pattern was evident with closings, which also appeared to construct relationships as more or less distant. Messages to groups and distant colleagues had the highest percentage of polite closings, that is, a farewell word and the writer's name or Thanks and the writer's name. The percentages were 89% for distant colleagues and 75% for groups, compared to 65% for messages sent to close colleagues. This indicates that these closing forms do important interactional work in constructing the relationship between participants, and suggests an attempt to reduce social distance and establish solidarity. However, messages written to a group were the most likely to end without any kind of sign-off (18%), while for the corpus as a whole it was 10%. A higher percentage of messages addressed to close colleagues ended with just the person's name or initial than did those written to any other addressee. Nearly one-third of these messages closed with just a name, while the average for the corpus was 18%.

The following two messages from SCT illustrate how status and social distance can affect the form of greeting. The writer of both messages is the same, as is the issue. However, Petra's use of Dear in the first message to a senior manager and a distant colleague establishes a respectful formality and shows negative politeness. This contrasts with AC31d, which is written to her immediate manager who is a close colleague. The informal Hi and Cheers in this message create solidarity and show positive politeness.

Gender

Gender identity is also constructed through the linguistic choices people make. The pattern revealed by men and women's use of greetings in the two workplaces was quite different, as Table 2 indicates. At SCT women acknowledged their addressee more frequently than did men and made greater use of greetings and closings. Two-thirds of the messages written by men contained no acknowledgment of the sender or greeting of any sort, compared to 49% of those written by women. Over one-quarter of the emails written by women started with the person's name, compared to 21% of those written by men. Salutations such as Dear, Hi, and Hello, with or without the person's name, were seldom used by either sex but were favored slightly more by women. When used, they usually were accompanied by the person's name.

Interestingly, when men did start their message with the person's name or a greeting, this was more likely to be in a message to a group or to another male than to a woman. Nearly three-quarters of the 72 messages men wrote to women started without a greeting,⁸ while the corresponding figure for messages written to another male (86 messages) or a group (31 messages) was 59%. The pattern was reversed for women writing to men. Women used the addressee's first name or a greeting somewhat more often when they were writing to men (49% of 72 instances) than when they were writing to another woman (43% of 17 instances) or to a group (40% of 27 instances).

Whereas there was very little difference between the number of men who used a closing and the number of women who did so, women were less likely to end their messages with just their name. They were nearly twice as likely as men to conclude their messages with a farewell formula or Thanks accompanied by their name.

At Revelinu, however, it was men who used more forms that expressed positive politeness. The percentage of male and female messages that began baldly, that is without a greeting word or any acknowledgment of the addressee, was 16% for both male and female. However, a higher percentage of women (34%) than men (20%) started their messages with just the person's name, while a correspondingly larger percentage of the male messages started with a greeting word. Sixty-five percent of the messages written by males started this way, compared to just under half of the messages written by females. In contrast to SCT, men writing to women were more likely to begin their message with a greeting word and name than vice-versa. Seventy-four percent of the messages that men wrote to women (n=31) began this way, compared to 48% of the messages women wrote to men (n=25) and 68% of the messages men wrote to other men (n=19). Women, however, were more likely than men to begin their messages to groups with a greeting and the group name. Fifty percent of their messages to groups (n=18) began this way, compared to 29% of the male messages to groups (n=24).

There was no real difference, however, in the way men and women closed their messages. Seventy-three percent of the male messages and 77% of the female ones ended with some sort of closing word and the person's name.

Discussion

The results of this study show considerable variation in the use made of greetings and closings in the two organizations. The use of greetings to open email messages is much more a feature of the manufacturing plant (Revelinu) than of the educational organization (SCT) and possibly reflects the more friendly and familial culture of the manufactoring plant. It may also indicate what other data (Waldvogel, 2005) about the manufacturing plant have shown, namely that people matter and that staff value each other. Symmetrical forms of address have been associated with solidarity (Brown & Gilman, 1960). In the manufacturing plant there was widespread use of the greeting form, Hi + name. The lack of greetings in many of the emails collected from the educational organization suggests that this is a business first, people second culture and that there is less solidarity among the staff here. Many of the messages in this corpus were, however, brief administrative exchanges between people who were in frequent email correspondence, so this may also have contributed to the absence of greetings in many of the messages.

In terms of the presence or absence of greetings, the findings from Revelinu are not dissimilar to those of Crystal (2001). Two-thirds of a sample of 500 emails from people who knew Crystal contained an introductory greeting. However, in his study, Dear was by far the most frequent individual greeting formula. Messages containing Dear were twice as common as messages without Dear. The most repeated greetings in order of frequency were:

• Dear + name

• Name

• Hi + name

In the present study, Dear was the greeting in only 2% of the messages in the SCT corpus and 7% of those in the Revelinu corpus. The bias towards the use of Dear in the English study may be attributable to a somewhat higher value placed on formality in the section of English society to which Crystal belongs, or to Crystal's own high status due to age and reputation in the field of linguistics.

The findings from Revelinu are also similar to those of Kankaanranta (2005). About 80% of the messages written by the Swedish and Finnish staff of the large multinational corporation she studied used a salutation (greeting), and most of these also included the first name of the recipient. Hello and Dear + first name opened over a half of the 46 messages in one person's mailbag. Signatures and closings also occurred in the great majority of the messages.

There are several possible reasons for the low use of greetings and closings at SCT. One is that less friendly and more impersonal emails are part of its culture and may reflect the social distancing strategies people use when they feel alienated from what is going on. Another is that because of the greater volume of messages they receive relative to their Revelinu counterparts, managers in particular respond to them under pressure. The key person, from whom the bulk of the emails came, averaged over sixty emails received per day. In these conditions, niceties such as greetings and closings may be the first to be sacrificed. Email may thus assist in constructing a less personalized culture. A third possibility is that greetings and closings are omitted because there is an assumption that they are unnecessary in exchanges among professionals engaged in a common purpose. However, this runs counter to what most of those surveyed felt.

The analysis also shows that, in both organizations, greetings and closings would appear to be used to construct colleagues as more or less worthy of respect and as more or less socially distant. This is done through the choice of whether or not to include a greeting or closing and, if included, the choice of greeting.

Status and social distance are acknowledged to some extent through the use of greetings and closings. In both workplaces these formulae tend to be used more to people of higher status than the writer and to distant colleagues, suggesting that these formulaic components of email are valuable resources for constructing relationships at work. In the educational organization, perhaps reflecting the greater social distance that existed in general among people in this workplace, greetings and closings were also used to construct solidarity more often in messages sent to groups. However, status differences, particularly at SCT, were not great. The fact that in this organization a relatively high percentage of the messages to high status people (45%) contained no greeting, while in the manufacturing plant all of the messages contained some acknowledgment of the recipient, does suggest that in both these organizations either people are not particularly status conscious or are status conscious but inclined not to acknowledge it. The widespread use of first names and the predominance of informal forms of greeting denoting solidarity exchanged among communication partners, regardless of status, possibly reflect the egalitarian nature of New Zealand society. The use of the informal Hi, widely used at Revelinu, constructs solidarity, whereas Dear constructs greater formality. The latter form, however, was not greatly used in either organization.

Gender differences were also observable in the use of greetings and closings in both workplaces. It has been noted that in many contexts females are more likely than males to attend to the social and affective aspects of an interaction (Holmes, 1995). This is relevant also to computer-mediated language (CMC), as Herring (1996a, 1996b, 2000) has demonstrated. Her study of male and female messages on bulletin boards indicated the existence of gender marked styles.

Data on greetings and closings from the educational organization tend to support Holmes' contention (1995). However, it is not supported by data from the manufacturing plant. The patterns involving the gendered use of greetings and closings in the two organizations were very different. At SCT, where women dominate staffing numbers, they construct female identities through their greater use of greetings and closings. At Revelinu, where males dominate numerically, male messages more than female messages are likely to contain greetings and closings. A possible explanation for these different patterns in the two organizations may be that the type of male working in the manufacturing plant is more likely than his counterpart in the white color organization to have "old world values" about masculine and feminine roles. These values would see him adopting a more protective and courteous stance towards women than his counterpart in the educational organization, where women dominate numerically and where more liberal values predominate. The pattern in the manufacturing plant, where males predominate, may indicate that men are showing deference to women and constructing them as more worthy of respect. In the more liberal environment of SCT where women predominate, the patterns could be interpreted as indications that men may view women more as equals. Whereas, in general, women used more greetings than men, variations in the gendered pattern of greeting use would also seem to have some relationship to the workplace culture.

The analysis also shows a great variation in the pattern of greetings and closings used in the emails of the two organizations. Does the absence or presence of greetings and closings in an organization's emails provide insights into its culture? The findings presented here suggest that it does. An independent study of the culture of the two organizations (Waldvogel, 2005) has shown that Revelinu has a culture where staff and management are more in harmony and supportive of each other than at SCT. This finding is substantiated by the patterns evident in the use of greetings and closings at Revelinu. The much greater use made of greetings and closings in this organization suggests that staff members are concerned to establish a friendly tone in their interactions and maintain good interpersonal relationships. The informal terms of address used indicate strong feelings of solidarity.

In the educational organization, on the other hand, greetings and closings were used in only about 20% of the messages, suggesting that more importance is placed on the message than on how it is conveyed. Even though in the workday environment cooperative collaboration existed and was valued, the changing nature of the staff and their roles, brought about by the restructuring and the physical dispersion of workgroups, meant that people in different sections did not know each other very well. This, plus the mistrust and uncertainties created by the restructuring, may have created the social distancing and low solidarity reflected in the paucity of greeting use in the emails from this organization.

Although greetings and closings are politeness markers, the presence or absence of these in email messages does not necessarily mean that a writer who omits them is being less polite than one who includes them. Politeness is relative to the culture in which it is found. Watts (2003) makes the distinction between politic and polite behavior. Politic is linguistic behavior that is appropriate to the ongoing social interaction. It is thus unmarked and generally goes unnoticed. (Im)polite linguistic behavior is marked, as it goes beyond the expected, either because it is perceived as negative or because it is perceived as overly polite. Watt's definition implies that linguistic structures are not inherently polite or impolite. The less frequent use of greetings in the educational organization, while indicating less solidarity, does not inherently mean that the writers are less polite, as in this workplace, this behavior appeared to be the norm. While most of those surveyed indicated that they preferred messages to have greetings and closings because this made the messages friendlier, they did not find the absence of these "marked" in any way.

Conclusion

The importance of greetings and closings as a linguistic resource lies in the affective role they play. The choice of greeting or closing and its presence or absence in an email message conveys not only an interpersonal message enabling the writer to negotiate his or her workplace relationships but also contributes to the creation of a friendly or less friendly workplace culture and, in turn, reflects this culture. This study demonstrated that there is a need to consider cultural factors in addition to sociolinguistic variables when accounting for the linguistic choices people make. The study also demonstrates that the role played by greetings and closings has important implications for organizational communication.

The following limitation to this research needs to be noted. The research described here is a case study of the use of greetings and closing in the emails of two workplaces in New Zealand. The emails comprise messages for the most part received and sent by two people in each organization. It is possible that these are not representative for each organization, let alone for other organizations. For this reason, no statistical tests have been carried out. The study should therefore be regarded as exploratory and the conclusions and findings as suggestive. Further research based in different workplaces and in different contexts needs to be carried out to determine the extent to which these findings hold for other workplaces and might be generalized.

Acknowledgments

I would like to acknowledge the help given by Professor Janet Holmes, who read the first draft of this article and made many valuable suggestions. I would also like to thank the two peer reviewers for their constructive comments.

Notes

1. The Language in the Workplace Project (LWP), based at Victoria University of Wellington, has been investigating spoken communication in New Zealand workplaces since 1996, with the aim of identifying characteristics of effective communication in such workplaces. Instead of basing the findings on data which have been collected using self-reporting techniques such as interviews and surveys (the traditional approach to investigating workplace communication), the team has been collecting naturally-occurring workplace interactions (see Holmes & Stubbe, 2003 and Holmes, 2000 for descriptions of the project and its aims). The result is a large corpus of workplace talk recorded on audio- and videotape (approximately 2,500 interactions) from 20 different workplaces ranging from government departments and corporate organizations to factories and hospital wards. The data have been analyzed using a variety of discourse analytic techniques, including Conversational Analysis, Politeness Theory, Critical Discourse Analysis, and Interactional Sociolinguistics, and have focused on many different aspects of organizational life, including how managers get things done, workplace teams, gender in the workplace, humor, and small talk.
2. No distinction has been made between first name only and first name and surname, as in the SCT corpus there were only 13 uses of a first name and surname, and in the factory corpus, only eight.
3. The key person is the person who provided the emails. This was either a senior manager or the person responsible for the issue-related emails.
4. Because there were only 14 issue-related emails in the Revelinu corpus, the two samples from this organization have been combined.
5. As used in this article, baldly refers to a lack of a greeting or closing.
6. All of the messages were coded and numbered. AC is the code used for messages from the corpus of the senior manager at SCT. UL is the code for messages from the Revelinu corpus. The letters a, b, c, etc. indicate that the message is part of a message thread.
7. See the Methods section.
8. To save space, the table showing these figures has not been included but is available upon request from the author.

References

Abdullah, N. A. (2003). Relational Communication in Organisational Email: A Malaysian Case Study. Unpublished doctoral dissertation, Faculty of Language and Linguistics, University of Malaya, Kuala Lumpur.

Brown. R., & Gilman, A. (1960). The pronouns of power and solidarity. In T. Seboek (Ed.), Style in Language. Cambridge MA: MIT Press.

Brown, P., & Levinson, S. (1987). Some Universals in Language Usage. Cambridge: Cambridge University Press.

Bunz, U., & Campbell, A. (2002, October). Accommodating politeness indicators in personal electronic mail messages. Paper presented at the Association of Internet Researcher's 3rd Annual Conference Maastricht, The Netherlands. Retrieved December 28, 2006 from http://bunz.comm.fsu.edu/AoIR2002politeness.pdf

Crystal, D. (2001). The language of email. In Language and the Internet (pp. 94-128). Cambridge: Cambridge University Press.

Daft, R. L., & Lengel, R. H. (1984). Information richness: A new approach to managerial behavior and organizational design. In L. L. Cummings & B. M. Staw (Eds.), Research in Organizational Behavior 6 (pp. 191-233). Homewood, IL: JAI Press.

Duthler, K. W. (2006). The politeness of requests made via email and voicemail: Support for the hyperpersonal model. Journal of Computer-Mediated Communication, 11 (2), article 6. Retrieved December 28, 2006 from  http://jcmc.indiana.edu/vol11/issue2/duthler.html

Eckert, P., & McConnell-Ginet, S. (2003). Language and Gender. Cambridge: Cambridge University Press.

Goffman, E. (1967). Interaction Ritual: Essays on Face-to-Face Behavior. New York: Garden City.

Herring, S. (1996a). Gender and democracy in computer-mediated communication. In R. Kling (Ed.), Computerization and Controversy. New York: Academic Press.

Herring, S. (1996b). Two variants of an electronic message schema. In S. Herring (Ed.), Computer-Mediated Communication: Linguistic, Social and Cross-Cultural Perspectives (pp. 81-105). Amsterdam: John Benjamins.

Herring, S. (2000). Gender differences in CMC: Findings and implications. The CPSR Newsletter, 18 (1). Retrieved April 6, 2005 from http://archive.cpsr.net/publications/newsletters/issues/2000/Winter2000/herring.html

Holmes, J. (1995). Women, Men and Politeness. London: Longman.

Holmes, J. (2000). Victoria University's Language in the Workplace Project: An overview. Language in the Workplace Occasional Papers, 1. Retrieved December 28, 2006 from http://www.vuw.ac.nz/lals/research/lwp/docs/ops/op1.htm

Holmes, J., & Stubbe, M. (2003). Power and Politeness in the Workplace. London: Longman.

Huang, W., Watson, R. T., & Wei, K. K. (1998). Can a lean e-mail medium be used for rich communication?  A psychological perspective. European Journal of Information Systems, 7 (4), 260-274.

Jessmer, S. L., & Anderson, D. E. (2001). The effect of politeness and grammar on user perceptions of electronic mail. North American Journal of Psychology, 3 (2), 331-346.

Kankaanranta, A. (2005). "Hej Seppo, Could You Pls Comment on This!" Internal Email Communication in Lingua Franca English in a Multinational Company. Ph.D. dissertation, Centre for Applied Language Studies, University of Jyväskylä. Retrieved January 7, 2007 from http://ebooks.jyu.fi/solki/9513923207.pdf

Markus, M. L. (1994). Electronic mail as the medium of managerial choice. Organization Science, 5 (4), 502-527.

Mulholland, J. (1999). E-mail: Uses, issues and problems in an institutional setting. In F. Bargiela-Chiappini & C. Nickerson (Eds.), Writing Business (pp. 57-84). Singapore: Longman.

Murray, D. (2000). Protean communication: The language of computer-mediated communication. TESOL Quarterly, 24 (3), 397-423.

Ngwenyama, O. K., & Lee, A. S. (1997). Communication richness in electronic mail: Critical social theory and the contextuality of meaning. MIS Quarterly, 21 (2), 145-167.

Sherblom, J. (1988). Direction, function, and signature in electronic mail. Journal of Business Communication, 25 (4), 39-54.

Waldvogel. J. (1999). Email: User Perceptions and Aspects of Its Role and Impact in One Workplace. Unpublished master's research project, School of Linguistics and Applied Language Studies, Victoria University of Wellington, New Zealand.

Waldvogel. J. (2005). The Role, Status and Style of Workplace Email: A Study of Two New Zealand Workplaces. Unpublished doctoral dissertation, School of Linguistics and Applied Language Studies, Victoria University of Wellington, New Zealand.

Walther, J. B. (1996). Computer-mediated communication: Impersonal, interpersonal, and hyperpersonal interaction. Communication Research, 23 (1), 3-43.

Watts, R. J. (2003). Politeness. Cambridge: Cambridge University Press.

Williams, E. (1999). E-mail and the effect of future developments. First Monday, 4 (8). Retrieved December 28, 2006 from http://www.firstmonday.org/issues/issue4_8/index.html

Zmud, R. W., & Carlson, J. R. (1999). Channel expansion theory and the experiential nature of media richness perceptions. Academy of Management Journal, 42 (2), 153-183.

About the Author

Joan Waldvogel is an analyst at the Ministry of Social Development in Wellington. Her doctoral research investigated the role and style of email in two New Zealand workplaces.
Address: Ministry of Social Development, PO Box 1556, Wellington, New Zealand