The Communicative Functions of Emoticons in Workplace E-Mails: :-)†
- University of Southern Denmark, Denmark
- Aalto University School of Business, Finland
- †Accepted by previous editor Maria Bakardjieva
Correspondence To: Karianne Skovholt, firstname.lastname@example.org
CMC research presents emoticons as visual representations of writers' emotions. We argue that the emoticons in authentic workplace e-mails do not primarily indicate writers' emotions. Rather, they provide information about how an utterance is supposed to be interpreted. We show that emoticons function as contextualization cues, which serve to organize interpersonal relations in written interaction. They serve 3 communicative functions. First, when following signatures, emoticons function as markers of a positive attitude. Second, when following utterances that are intended to be interpreted as humorous, they are joke/irony markers. Third, they are hedges: when following expressive speech acts (such as thanks, greetings, etc.) they function as strengtheners and when following directives (such as requests, corrections, etc.) they function as softeners.
In this article, we examine emoticons as they appear in authentic workplace e-mails. More specifically, we identify the communicative functions of emoticons in three different Nordic companies where employees use email as their primary tool for communication. We ask in which situations emoticons are used, what type of utterances are followed by emoticons, and what type of communicative functions emoticons serve. The study is carried out within the framework of discourse analysis, applying terms from speech act and politeness theory.
The word “emoticon,” a construction of the words “emotion” and “icon,” refers to graphic representations of facial expressions, which often follow utterances in written computer-mediated communication (CMC). Emoticons may be produced by ASCII symbols (:-)) or by “pictograms,” which are graphic symbols (). According to Krohn (2004), the emoticon was first used in written text in 1982 by computer scientist Scott E. Fahlman at Carnegie Mellon University in the United States. Fahlman suggested that the keyboard-based “smiley” face :-) and the “frowny” face :-(could be used to identify jokes in a computer scientist discussion forum. The overall aim was to economize computer-mediated interaction.
Within the last 30 years, emoticons have developed different forms and meanings, and a growing number of forms accompany different types of chat software. Still, using emoticons in CMC has traditionally been viewed as a typically teenage phenomenon and has been associated with young people's chat style on the Internet (Johansen, 2008). Emoticons have also been considered superfluous and a waste of bandwidth (Andrews, 1994). Not surprisingly formal guidelines for computer-mediated communication or “netiquettes” advise writers to limit their use of emoticons in workplace communication, mostly because their excessive use may signal emotional instability and a lack of control over one's feelings (Wolf, 2000). Furthermore, such guidelines (see e.g. Boone, Kurtz, & Block, 1997; Extejt, 1998; Munter, Rogers, & Rymer, 2003; Krohn, 2004) tend to be normative and colored by the author's personal values rather than reflecting the actual use and communicative functions of emoticons. In the popular press and media, emoticons are banned by some authors (e.g. Andrews, 1994; Rossavik, 2011), and praised by others (Stensland, 2011).
Since different authors provide conflicting advice for the use of emoticons, we argue that emoticons have an ambiguous status in written discourse, and therefore more in-depth knowledge is needed about the authentic use and communicative functions of emoticons in workplace communication. Our purpose is to show that the emoticon as a semiotic resource in e-mail communication is used systematically to modify speech acts, and thus has developed new and more specific functions compared with those proposed by Fahlman in 1984. In this study, we identify and categorize the communicative functions of emoticons in a way that has not been done in previous research. By using discourse-analytical methods, this study contributes to the research on the use and function of emoticons in computer-mediated communication in general and e-mail communication in particular. As we will show, emoticons represent a multifunctional semiotic resource available to e-mail writers, who can use them both to contextualize discourse and to organize social relationships.
Use of emoticons in e-mail
Research on the use of emoticons in e-mail over the years has addressed various issues such as their function as indicators of emotional state (e.g. Rezabek & Cochenour, 1995) and the differences in usage between men and women (e.g. Wolf, 2000). Although we address such investigations to provide background for the current study, our focus here is on previous research related to the impact of emoticons on message interpretation and the pragmatic function of emoticons, as that research forms the basis for our investigation of the communicative functions of emoticons.
Scholars seem to agree that emoticons are graphic signs which are used to indicate an emotional state (Rezabek & Cochenour, 1995; Raymond, 1996; Wolf, 2000; Derks, Bos, & Grumbkow, 2007; Derks, Bos, & Grumbkow, 2008). Most of these studies assume that emoticons are used to compensate for the lack of nonverbal communication cues, such as facial expressions, intonation, gestures, and other bodily indicators, in CMC (cf. Kiesler, Siegel, & McGuire, 1984; Sproull & Kielser, 1986; Shao-Kang, 2008; Krohn, 2004). In other words, emoticons are perceived as providing support to written communication, in the same way that visual and body language support face-to-face communication.
In the field of linguistics, emoticons are primarily viewed as emotion markers (see e.g. Baron, 2000, p. 242). Renowned linguist David Crystal (2001, p. 36) defines emoticons as a “combination of keyboard characters designed to show an emotional facial expression.” Furthermore, he proposes that emoticons seem to have a “purely pragmatic force – acting as a warning to the recipient(s) that the sender is worried about the effect a sentence might have” (Crystal, 2001, p. 38).
The first study (known to the authors) on emoticons was conducted in 1995 by Rezabeck and Cochenour (1995); it examined the frequency, variety, and usage patterns of emoticons in four listservs. The frequency of emoticons varied within the four listservs, and in the listserv with the highest score, every fourth message contained an emoticon. The most frequently occurring emoticon was the traditional smiley, :-), followed by the smiley without a “nose,”:). Although the listservs contained a variety of emoticons, Rezabeck and Cochenour (1995) concluded that there was most likely a shared understanding of the meaning conveyed by the emoticons occurring most frequently. The use of the smiley depended on individual preferences and varied according to the context (Rezabek & Cochenour, 1995).
Previous research has also been interested in differences between genders in the use of emoticons. In a study of emoticon use in online newsgroups, Wolf (2000) found that women used emoticons more often than men did. This finding correlates with Herring (2003) who found that women in the public Internet Relay Chat produced three times more emoticons of smiling and laughter than men. Furthermore, Baron (2004) found that the majority of emoticons in her corpus of synchronous Instant Messaging were produced by women. In contrast, Huffacker and Calvert (2005) found that male teenage bloggers used more emoticons than females.
Within the field of social psychology, scholars have sought to examine the impact of emoticons on message interpretation and have received somewhat conflicting results (Walther & D'Addario, 2001; Provine, Spencer, & Mandell, 2007; Derks et al., 2008; Thompson & Foulger, 1996). For example, through different experiments and simulated studies, Walther and D'Addario (2001) found that emoticons had little impact on how the e-mail messages in their data were interpreted: The messages containing emoticons did not generate different interpretations than those without emoticons. Thus, the emoticons served to complement the utterances, but did not serve the function of contradicting or enhancing them. In other words, the verbal message content seemed to be primary, and prevailed over the impact of the emoticon. Walther and D'Addario (2001, p. 344) concluded that the emoticon's “actual communicative effects are minimal in the context of the language cues they may accompany.” Similarly to Walther and D'Addario (2001), Provine, Spencer, and Mandell (2007) found that emotional expression was subordinate to the production of spoken phrases.
In contrast to Walther and D'Addario (2001), other experimental studies have found that emoticons do have an influence on the perception of utterances. For instance, the psychological tests conducted by Derks, Bos, and Grumbkow (2008) showed that readers perceived messages with emoticons as more positive than messages without them. Derks et al. (2008) concluded that emoticons can serve some of the same functions as nonverbal behavior. In particular, emoticons can complement and enhance the verbal message, but they are not able to contradict it. In yet another experimental study, Thompson and Foulger (1996) examined the effect of emoticons on the perception of flaming. Their results suggest that emoticons modify the perception of flaming. Emoticons alerted readers that the message could be taken less seriously; in other words, they represented a useful strategy for preventing unintentional outbreaks of flaming.
The pragmatic function of emoticons was examined by Dresner and Herring (2010), whose data was collected from Herring's archives over the last 10 years, and included private e-mail, private chat (Instant Messaging), public chat (AOL chat; Internet Relay Chat), and postings on public discussion forums. Drawing on Austin's (1962) theory of speech acts, they argue that the primary function of emoticons is not to convey emotion but rather to indicate an illocutionary force, i. e. the intended effect of the utterance.1 Moreover, they suggest that emoticons serve two additional functions: They function as indicators of emotion, and they function as indicators of what Dresner and Herring (2010, p. 263) call “nonemotional meaning, mapped conventionally onto facial expression.” They explain this use of emoticons as indicating a joke by using a wink: ;-). The wink is not signaling emotion; rather, it is conventionally indicating a joking intent (Dresner & Herring, 2010, p. 263).
In sum, the results of the psychological studies (e.g. Walther & D'Addario, 2001; Derks et al., 2008) are valuable in CMC research but to gain more specific knowledge of how emoticons are actually used, we will analyze them as graphic signs which, in interplay with verbal utterances, serve different communicative functions. Rather than investigating the interpretations some informants make of the utterances invented for the purposes of a particular study, we argue that it is more productive to analyze patterns of emoticon use as displayed in naturally occurring instances of CMC, such as we may see, for example, in Dresner & Herring (2010). Similarly to our research, they used naturally occurring data and identified communicative functions of emoticons. However, while Dresner & Herring (2010) identified emoticons in general terms as illocutionary force indicators, we argue that the communicative functions of emoticons need to be more systematically and specifically differentiated and discussed.
Speech act and politeness theory
In the present study, we use analytical concepts from two prominent theories of linguistics: speech act theory and politeness theory. The speech act theory springs from the basic idea that in saying something a speaker is performing a social action, a “speech act.” The theory has its philosophical origin in Austin (1962), who examined the difference between two types of utterances, “constatives” and “performatives.” According to Austin (1962), constatives are descriptive statements which can be either true or false (as in “It is hot in here”), whereas performatives are utterances which realize a social action (as in “I hereby pronounce you husband and wife”). Austin (1962) elaborated on this distinction by differentiating between three distinct acts of an utterance: “locutionary,” “illocutionary,” and “perlocutionary.” The locutionary act refers to the determinate sense and reference of an utterance (e.g. the hotness in “It is hot in here”), whereas the illocutionary act refers to the making of a request, offer, joke, promise, etc. in uttering a sentence, i.e. what is directly achieved by the conventional force associated with it (e.g. a request to open the window in “It is hot in here”). The perlocutionary act refers to the effect that is brought to the audience (e.g. the audience opens the window).
Following Austin (1962), several taxonomies have been proposed, firstly by Searle (1969), who distinguished between five main categories of speech acts: “directives” (e.g. requesting), “expressives” (e.g. thanking), “representatives” (e.g. asserting), “commissives” (e.g. promising) and “declarations” (e.g. appointing). Subsequent researchers have elaborated Searle's categories and proposed their own taxonomies of speech acts. However, it has become clear that it is impossible to stipulate an exhaustive taxonomy of speech acts including felicity conditions (Jaworski & Coupland, 1999, s.16).2
In the current study, we will identify what type of speech acts are accompanied by emoticons to see if they have any common characteristics. Our speech act taxonomy is developed inductively from the data and includes, for example, speech acts such as thanks, greetings, questions and requests. When we address these speech acts on a general level we use Searle's terminology of expressives, directives, commissives and representatives.
Within the frame of politeness theory (Brown & Levinson, 1987), analyzing language usage coincides with discovering how social relations are constructed. In order to build a universal model of how social relations are constructed linguistically, Brown and Levinson (1987) relied on Goffman's (1967) notion of “face,” our public self-image when participating in interaction. According to Brown and Levinson (1987), all rational agents have two particular “wants”: the want to be unimpeded, a negative face, and the want to be approved of, a positive face. In social interaction, the participants share a mutual interest in maintaining each other's face. Some speech acts, however, inherently threaten the other party's face, for example, performing a request. They are called “face-threatening acts” (FTA). When a speaker needs to perform an FTA, they may use one of five different politeness strategies: (1) perform the FTA directly, perform the FTA with compensation by using (2) positive or (3) negative politeness strategies, (4) perform the FTA indirectly or (5) simply avoid doing the FTA (Brown & Levinson, 1987, p. 69–70).3
A case in point of our speech act taxonomy is the request, which is a speech act expecting a response in the form of a verbal or a physical act (Searle, 1969). According to Brown and Levinson (1987), speakers assess the overall seriousness (“weightiness”) of this FTA by assessing the social distance between the participants, their relative power and the ranking of the imposition in the particular culture (Brown & Levinson, 1987, p. 74f). The greater the imposition of a request, the more the individual needs to compensate for the FTA.
Since the present study examines authentic workplace interaction, the participants' institutional roles and the organizations' needs are in the foreground, which means that complying with requests is more or less part of employees' job descriptions. Depending on the employees' hierarchical positions and their institutional roles, they have a legitimate right to make certain requests and/or an obligation to comply with others. Hence, making some requests is part of the regular work routine and is treated as such (e.g. Kankaanranta, 2005:111). In spite of this, the linguistic realization of the request may provide information about the social distance between the participants. As we will argue in the analysis, a smiley which accompanies a request will contribute to providing information about, or even constituting, the social relationship between the writer of the email and their addressee.
In our analysis we also make use of the concept “hedge.” Hedges are expressions with a “pragmatic value that modifies the speech act, not the logical form of the utterance” (Fretheim, 1981, p. 86 [our translation]). Moreover, they are referred to as a “direct modification of propositional content” (Verschueren, 1999, p. 193), and according to Brown and Levinson (1987), a hedge is “a particle, word, or phrase that modifies the degree of membership of a predicate or noun phrase in a set (…)” (Brown & Levinson, 1987, p. 145 [original italics]). Accordingly, hedges may be divided into “strengtheners” emphasizing the propositional content and “softeners” which soften the propositional content. They “indicate something about the speaker's commitment toward what he/she is saying, and in so doing modify the illocutionary force” (Brown & Levinson, 1987, p. 147). For example, in the utterance “A swing is definitely a toy,” the adverb “definitely” functions as a strengthener, i.e. strengthening the speaker's commitment toward the utterance. In contrast, in the utterance “A swing is sort of a toy,” the adverb “sort of” signals that the swing is only partly an object in the category of “toy.” The adverb “tentativizes” the propositional content and functions as a softener (Brown & Levinson, 1987, p. 145).
Data and methodology
The data in the present study consist of a total of 1606 e-mail messages which were gathered from three different Nordic companies.4 First, 491 internal e-mails were gathered from a distributed work group in a Norwegian telecom company (“Telecom”). Second, 504 emails were collected from the correspondence between employees of a Danish insurance company (“Codan”) and their clients. Finally, 611 internal e-mails were collected from a Finland-based paper company (“Paper Group”). Most of the e-mails from the three companies were originally written in Norwegian, Danish, Swedish, and Finnish (N = 1324), and were translated into English for the purposes of the present study; only 282 messages were originally written in English. As inhabitants of the Nordic countries and users of Scandinavian − a fluid combination of Norwegian, Danish, Swedish and Icelandic − the first two authors could understand texts in each other's native tongues (Norwegian and Danish respectively) and the Finnish author gained access to them through Swedish, the other official language of Finland. In other words, we were able to evaluate, discuss and clarify each other's translations of the original Norwegian, Danish and Swedish texts. Although the first two authors did not understand the emails written in Finnish (which does not feature in Scandinavian), the third author was able to provide accurate English translations of the texts because of her educational and professional background.5 An overview of the data is presented in Table 1.
|Language||Norwegian||Danish||Finnish, Swedish, English|
|No of messages||491||504||611|
The Telecom corpus contains e-mail messages sent between 11 members of a distributed work group, “Agenda,” in a Norwegian telecom company. The e-mail exchanges were gathered from August to December 2004 and the corpus contains 491 e-mails. The senders and recipients of the messages were primarily Agenda members. Most of them were physically located in different departments and even in different parts of Norway. They had occasional meetings at the main office, but most of their communication was carried out by e-mail and telephone. Agenda's main task was to edit all letters that were delivered to customers in the private market.
The Codan corpus contains e-mail messages from a Danish subsidiary of a British insurance company. In 2003, when the corpus was collected, the Danish subsidiary was the third-largest insurance company in Scandinavia with more than 2000 employees. The e-mail exchanges between the company's employees and clients were gathered during 3 weeks in March and contain 504 e-mails representing external communication. The clients initiated all email interactions and the e-mailing parties were unknown to each other. Indeed, the company invited their clients to e-mail them by having the following text on the corporate website: “Send us an email. If you have questions or comments, feel free to send us an email. We answer 95% of all emails within one day. Write to us.”
The Paper Group corpus consists of 611 internal e-mail messages collected from eight employees' mailboxes in 2000-2001; five of the employees were Finns and three Swedes. They came from different parts of the newly-merged Finnish-Swedish corporation with nearly 44,000 employees globally and held diverse jobs such as Vice President, Technical Manager, Project Manager, Communication Specialist and Secretary/Assistant. Around one third of the messages were written by the eight employees, and the rest represented their (mainly) Finnish or Swedish communication partners or other Paper Group employees, whose messages were found in the message chains. Approximately half of the messages were written in lingua franca English (N = 282) and the rest in either Finnish or Swedish.
In order to identify the communicative functions of the emoticons, we examined the speech acts that were followed by an emoticon. Our analytical procedure involved two different levels: descriptive and explanatory. On the descriptive level, we picked out all the messages that contained emoticons and identified the speech act preceding them. As we show below, the speech acts fell into four groups: directives, expressives, commissives, and representatives. On the explanatory level, then, our analysis related the performance of the speech acts to theories of language use and interpersonal relationships, i.e. politeness theory. For example, directives impose a “burden” on the addressee – and threaten their negative face – and are thus classified as face-threatening acts in politeness theory. Expressives, which signal the sender's emotions such as gratitude (e.g. thanking), support the addressee's positive face. Likewise, speech acts categorized as commissives and representatives were discussed within the frame of politeness theory.
The identification of speech acts was the result of pragmatic interpretation and all three authors participated in the coding process. The identification of speech acts is always a matter of close reading and interpretation of utterances in relation to their context; there is not necessarily a direct relationship between grammatical form and pragmatic function. In addition, one utterance may be interpreted and treated as more than one speech act. In order to make the coding as reliable as possible, each author proposed a list of the speech acts from their respective corpora. This identification work was carried out inductively and was based on the original e-mail messages in Norwegian (Telecom corpus), Danish (Codan corpus) and English/Swedish/Finnish (Paper Group corpus).
To ensure interpersonal reliability (McMillan, 2000, p. 4), we compared the three lists and discussed any ambiguous and problematic cases. On the joint list of 110 speech acts, there were five cases that emerged as particularly challenging and that the authors had to discuss in detail to identify the speech act. An example which illustrates one of the five ambiguous instances is the following message from the team leader in the Telecom corpus to her group: “(…) MEN fokuset er fortsatt sterkt fra ledelsen, så her er det bare å stå på :-) (…)”/“BUT management is still paying close attention to us, so we just have to keep up working :-) (…)”. One author coded the utterance as a request, while another coded it as a joke. After some discussion, we decided to code it as a request. More specifically, this is an indirect request; i. e. a request mitigated with politeness strategy 4.
Although our approach is qualitative, some quantitative results will also be presented (e.g. the total number of emoticons) to show some distributions of the analyzed phenomenon in the three different corpora. Our motivation is only to provide a richer description rather than claim any statistical significance.
Identification of speech acts
To obtain an overview of the use of emoticons in the three companies based in Norway, Denmark and Finland respectively, we counted the frequency of emoticons. Overall, the traditional smiley, :-) and :), dominated, but the emails also included ten instances of winks ;-) and one single smiley face with the ‘tongue’ sticking out :-p. There were no instances of a frowning face :-( in the corpus.
As Table 2 shows, the largest number of emoticons appeared in the Telecom corpus (N = 88); the Codan corpus contained twenty (N = 20) emoticons and the Paper Group corpus contained only two (N = 2) emoticons. In the Telecom corpus, 17.7% of the messages contained emoticons amounting to 81 of the total of 491 emails. Seven (7) of the messages had two (2) emoticons. In the Codan corpus, 3.4% of the messages contained emoticons corresponding to 17 emails of the total of 504. Three (3) of the messages had two (2) emoticons. The Paper Group corpus of 611 emails only contained two emoticons, equaling 0.003%.
Table 3 summarizes our coding results and shows the individual speech acts followed by emoticons in the three corpora. To depict the categories of the speech acts, we use the capital letters D(irectives), E(xpressives), C(ommissives) and R(epresentatives) after the respective acts. As Table 3 shows, emoticons were used after 12 individual speech acts, of which the last category (signatures) cannot be found in Searle's (1969) categorization. For the purpose of this study, we have categorized signatures as expressive speech acts, which − in our view − best describes their function as a way to sign off messages. The 12 individual speech acts could be placed into four different categories of speech acts: directives (N = 32), expressives (N = 73); commissives (N = 4) and representatives (admissions) (N = 1), of which the first two clearly dominate. The directives comprised four individual speech acts: requests, corrections, rejections and complaints. They are characterized as face-threatening acts in Brown and Levinson's (1987) politeness theory. The expressives comprised six individual speech acts: thanks, greetings, wishes, appraisals, jokes and signatures (N = 73). They contribute to supporting the addressee's “positive face” (Brown & Levinson, 1987).
|1 Requests D||20||6||0||26|
|2 Corrections D||2||1||0||3|
|3 Rejections D||2||0||0||2|
|4 Complaints D||0||1||0||1|
|5 Thanks E||4||1||0||5|
|6 Greetings E||3||0||0||3|
|7 Wishes E||1||1||0||2|
|8 Appraisals E||4||0||0||4|
|9 Promises C||3||1||0||4|
|10 Admissions R||1||1|
|11 Jokes/irony E||16||7||0||23|
|12 Signatures E||32||2||2||36|
|Total No of emoticons||88||20||2||110|
Most of the emoticons were used in the Telecom company emails; indeed, the leader and the text designer of the Agenda work group most frequently produced emoticons. 76 of the emoticons were smileys: :-) and :). 11 of the emoticons were winks: ;-) and 1 emoticon was a “teasing” smiley: :-p. Most of the emoticons were found in the signature field of the message (N = 32) and as many as 26 of them were produced by the text designer. Most of the emoticons in the leader's messages appeared after requests (N = 10). In the Codan corpus, all the emoticons were smileys and were produced by the company's clients and none produced by the employees. The largest number of emoticons were produced to accompany requests (N = 6) and jokes (N = 7); only two (N = 2) emoticons accompanied signatures. In the Paper corpus, the only two emoticons – smileys – followed signatures.
After identifying the speech acts followed by emoticons, we now show how the emoticons in our data serve three communicative functions: they mark a positive attitude, they mark jokes/irony and they function as hedges by softening FTAs in directives and by strengthening expressives.
Emoticons as markers of positive attitude
Our findings show that the emoticons following signatures were used to show a positive attitude towards the recipient, as example (1) shows:
|Ha' en fortsat god dag||Have a nice day|
|Mvh. Ditte :-)||Rgds. Ditte :-)|
The emoticon in example (1) follows a signature. It is attributed to the sender's name thus identifying her. In this context, the emoticon is a marker of the facial expression the sender brings to the written conversation. As such, the emoticon has an iconic function, signaling aspects of the sender's identity and her positive attitude towards the recipient. Thus, when used after signatures, the emoticons seem to function as markers of the sender's facial expressions.
Emoticons as markers of a joke or irony
Some emoticons were used to place utterances that could be interpreted as jokes or irony in the appropriate humorous context. Example (2) is an excerpt from a message in the Codan corpus. A young man applies for insurance for his first car. He provides the following information to the insurance company:
Jeg er 36 år. Har aldrig før haft bil. Har dog kørt i andres biler siden mit 18. år uden at lave så meget som en skramme, men det tæller jo nok
|I am 36. I have never had a car. However, I have driven others' cars since I was 18, without making a scratch, but that doesn't count, I guess. :-)|
In example (2), the client explains that he has had driving experience although he has never owned a car of his own before. This information seems irrelevant and redundant within the context. However, it becomes relevant when we know that some insurance companies provide cheaper insurance to drivers who can document their driving experience. The young man knows that the type of “undocumented” information he provides does not count as valid experience, and he makes a humoristic comment about it. The smiley thus seems to function as a contextualization cue which places his utterance within a humoristic frame.
Emoticons used as softeners served to modify four types of speech acts that can be defined as FTAs: requests, rejections, corrections, and complaints. They are directed towards the addressee's negative face and seem to mitigate the face threat. Example (3) contains a request:
|Hei Maria,||Hi Maria,|
|(…)Vi skal ha en HELHETLIG OG TYDELIG kommunikasjon. Fint med en liten oppklaring her.... :-)||(…) We are supposed to have a CONSISTENT AND CLEAR communication style. Nice with a little clarification here.... :-)|
|mvh Line||rgds Line|
In example (3), the leader of the Agenda work group (Line) asks the text designer (Maria) to give her some information (“Nice with a little clarification here”). The request, which is formulated in the indicative mood, is articulated as a personal wish, rather than an institutional request. Within this context it must be interpreted as an indirect speech act (Brown & Levinson, 1987, p. 70), which functions as a request for information. The request is accompanied by a smiley, which serves to soften the illocutionary force of the request. In other words, the emoticon serves to modify the propositional content of the utterance and functions as a softening hedge. The request is presented with less authority and the emoticon serves to downgrade the authoritative position of the leader. As pointed out earlier, the highest frequency of emoticons appears in messages containing requests from the leader of Agenda (Line) in the Telecom corpus.
The second type of FTA that emoticons were used to modify are rejections, as the exchange between Siri and Maria in example (4) illustrates:
|burde vi ikke heller ta hensyn til hva kunden ønsker og trenger av informasjon. (…)||Shouldn't we take into account customer needs for information. (…)|
|Nei, sier jeg :-p […]||No, say I :-p […]|
In example (4), text designer Maria rejects Siri's proposal with a categorical “no.” In a face-to-face conversation, a rejection is a potentially face-threatening act. It is typically constructed in a format which involves an account or an excuse, for instance: “I don't mean to be rude, but I don't think that's a good idea” (cf. Sacks, 1987; Pomerantz, 1984). The emoticon with the tongue sticking out signals that Maria's categorical rejection departs from the norms of how a rejection is supposed to be formulated. It also functions as a humoristic teasing signal. In other words, the emoticon contributes to making the rejection less categorical and functions as a hedge by softening the potentially face-threatening act.
The third type of face-threatening act that emoticons were used to modify are corrections. In example (5), one of Codan's clients has been addressed with an incorrect title. In his email message to the company, the client informs them about his correct title as follows:
|Samtidig skal jeg gøre opmærksom på at min titel er som i underskriften, og ikke forsikringsmægler som står i brevet. (Det er 16 år siden jeg var mægler :o))||At the same time I have to make clear that my title is identical with the one in the signature, and not insurance broker as written in the letter (It's been 16 years since I was a broker :o))|
The final utterance “It's been 16 years since I was a broker :o)” has the grammatical form of an informative statement, but it has also pragmatic implications. The utterance may be perceived as a correction or even as an accusation. The emoticon functions as a softener, which downgrades the illocutionary force in the client's correction. It contributes to softening the correction, and can even be interpreted as a humoristic comment. The emoticon serves to limit the negative implications of the utterance and to foreground the positive implications.
Finally, emoticons were used to modify complaints. In example (6) a client complains to Codan that he has not received the amount of money which he has outstanding:
Det er trods alt næsten 3 måneder siden jeg betalte de ca. 499 kr for meget. Var det mig der betalte for lidt i præmie skulle der jo nok komme
et rykkergebyr fra jeres side, så det er vel rimeligt at kunden bliver behandlet efter samme vilkår.
|It is actually almost 3 months since I overpaid by DKK 499. If I had paid too little, I would most likely have received a reminder from you, so it is reasonable that the client is treated on equal terms.|
|En lille fredagsbrok herfra :-)||A little Friday moaning from here :-)|
The client builds up his complaint in an indirect way. He starts out by informing how long he has been waiting for the money. He continues by presupposing that the company would have sent him a reminder if he had owed them money. Indirectly, he accuses the company of being unwilling to pay back his money. There is no doubt that he treats the insurance company's lack of adequate processes as unreasonable. Strengtheners such as “trods alt”/“actually” and modal markers such as “jo nok”/“most likely” contribute to emphasizing his message. In addition, the client's meta comment “En lille fredagsbrok herfra”/“A little Friday moaning from here” frames his speech act as a complaint. The emoticon which accompanies the utterance serves to downgrade his critique of the missing processes.
In sum, we argue that emoticons were used as softening hedges in the data when they accompanied FTAs such as requests, rejections, corrections and complaints. When emoticons follow these types of directive speech acts, they seem to compensate for the threat towards the e-mail recipient's negative face, and their communicative function is to soften the illocutionary force and render the directive less authoritative. When used as softening hedges, emoticons express respect to the addressee's integrity and right to be “unimpinged on” (Brown & Levinson, 1987, p. 131).
Some emoticons in our data served to intensify expressive speech acts such as greetings, thanks, compliments and appraisals, which are oriented towards the addressee's positive face, as the following four examples (7-10) show:
|Et annet spm, kan du være møteleder den 4.nov da jeg er bortreist den dagen.||Another question, could you chair the meeting Nov. 4. I am away that day.|
|Det skal nok gå greit!||That's okay!|
|flott og takk :-)||great and thanks :-)|
|Hei Anna, lenge siden sist:-) Håper alt står bra til i Steinkjer!!||Hi Anna, long time no see:-) Hope everything is okay in Steinkjer!!|
|Håper HVøvelsen har gått greit :-)||I hope the home guard training was fine :-)|
|Ellers så syntes de at AGP gjør en viktig og god jobb!! :-)||They think that Agenda does an important and good job!! :-)|
Finally, we found that emoticons followed promises and admissions, which may be classified as commissives and representatives according to Searle's speech act taxonomy.
|Tak for hjælpen - og nu lover jeg, at det er sidste gang, jeg henvender mig i denne sag:-)||Thanks for your help – and now I promise that I have contacted you for the last time concerning this case:-)|
|Jeg har sikkert blingset vedr signatur, så der har du sikkert rett :-)||I have probably read the signature incorrectly, so there you're probably right :-)|
Examples 7-12 illustrate how emoticons contribute to the intensification of the illocutionary force in expressive, commissive, and representative speech acts: They seem to strengthen the sender's commitment to the speech act, signal that the sender is expressing themself with sincerity, and that they are supportive and collaborative. When used to accompany speech acts which are directed to the addressee's positive face, the emoticons express interest in and approval of the addressee's “positive self image” (Brown & Levinson, 1987, p. 70).
Discussion and Conclusion
The current study has made visible the practices of using emoticons in email correspondence in three Nordic companies. While linguists have traditionally presented emoticons as emotion markers (Krohn, 2004; Crystal, 2001; Baron, 2000), the current study has shown that new communicative functions and conventions have emerged among e-mail users. First, our findings show that emoticons serve to signal positive attitude when used with signatures. They function as markers of the sender's facial expressions. This function corresponds with one of the three communicative functions that Dresner and Herring (2010, p. 263) identified in their study.
Second, emoticons are used as joke/irony markers. This finding supports much of the previous research on emoticons (e.g. Wolf, 2000; Thompson & Foulger, 1996) and confirms that emoticons are conventionally used as Fahlman (see Krohn, 2004) and also Dresner and Herring (2010) suggested, i.e. to identify jokes.
Finally, and most importantly in our opinion, the present findings show that emoticons have assumed a function which has been hitherto unidentified, namely that of hedges. On the one hand, emoticons soften speech acts which are threatening to the recipient's negative face (or face-threatening acts). In this role, they thus serve to soften directive speech acts such as requests, corrections, rejections and complaints. On the other hand, emoticons hedge speech acts which are directed to the recipient's positive face. In this role of strengtheners, they therefore strengthen such speech acts as thanks, greetings, wishes and appraisals. The communicative functions of emoticons are summarized in Table 4.
|1) MARKER OF POSITIVE ATTITUDE|
|used with signatures|
|2) MARKER OF JOKES/IRONY|
|used with humoristic utterances|
|Softener used with||Strengthener used with|
|(a) requests||(a) thanks|
|(b) corrections||(b) greetings|
|(c) rejections||(c) wishes|
|(d) complaints||(d) appraisals|
On the basis of the communicative functions we have identified in the present study and presented in Table 4, we argue that emoticons serve two more general communicative functions. First, the emoticons used as joke/irony markers and as hedges (as softeners and strengtheners) do not necessarily express the sender's inner state or emotion, as argued by previous research (e.g. Baron, 2000; Crystal, 2001), but rather seem to contextualize and modify preceding utterances. In other words, on a general level they function as contextualization cues (Gumperz, 1982) providing information to aid in the interpretation of the utterance. Thus, they can also be called pragmatic modifiers.
Second, all three of the communicative functions of emoticons, i.e. markers of positive attitude, markers of irony/jokes and hedges, seem to organize social relations. Employees can use emoticons to downplay potentially face-threatening directives. Rather than presenting requests, corrections and rejections directly, employees may modify these speech acts so that they appear less imposing, impolite and authoritative. For example, by accompanying requests with emoticons, employees may align with peers and subordinates and reduce the risk of being too demanding. Emoticons thus function as a positive politeness strategy, presupposing that the addressee is willing to comply with the request. Indeed, when superiors downplay a request with a smiley, they presuppose common ground and familiarity with their subordinates. By foregrounding the social relationship with group members, their institutional role as a superior is put in the background. Hence, superiors may use emoticons to appear less authoritative. In this way, emoticons function as solidarity markers. By creating a personal and informal style in e-mails, emoticons may contribute to reducing the social distance between colleagues. They may thus contribute to establishing and strengthening interpersonal relations in the work context where internal communication is increasingly carried out by e-mail.
Table 5 presents a systematic overview of the specific and general communicative functions of emoticons. The table shows that emoticons are multifunctional. On a specific level they function as softeners, strengtheners, markers of jokes/irony and markers of positive attitude. On a general level emoticons function as contextualization cues, which serve to organize social relations. As such they also function as solidarity markers.
|Organizing social relations|
|Hedge: softener, strengthener|
|Marker of jokes/irony|
|Marker of positive attitude|
As we have shown, emoticons contribute to modifying the propositional content and the illucutionary force of the speech act. This finding corresponds with Dresner and Herring's results (2010). However, while Dresner and Herring (2010) identified the communicative functions of emoticons in general terms as “illocutionary force indicators,” the present study has identified specific types of speech acts (e.g. requests, thanks) which are followed by emoticons in workplace e-mails and, moreover, specified the communicative functions of emoticons, i.e. markers of positive attitude, markers of jokes/irony, and hedges such as softeners and strengtheners.
Infinity Ward / Activision — Woodland Hills, California, United States
Infinity Ward / Activision — Woodland Hills, California, United States
Skydance Interactive — Marina Del Rey, California, United States
[View All Jobs]
[Next News Story] [View All]