Speech Acts within SNSs Running Head: Speech Acts within SNSs
Speech Act Analysis within Social Network Sites’ Status Messages
Caleb T. Carr Michigan State University Department of Telecommunication, Information Studies and Media
David B. Schrock Ferris State University Department of Humanities
Patricia R. Dauterman
*Correspondences should be directed to the first author, Caleb Carr.
1
Speech Acts within SNSs
2
Abstract This research was conducted in an attempt to expand previous research of speech acts and online messaging, by examining how individuals use the status messages of social network sites (specifically Facebook and MySpace) to communicate socially and construct their identity. Seventy-four students' SNS status messages were captured three times daily over fourteen consecutive days. Content analysis of these data revealed that status messages are primarily constructed with expressive speech acts, followed by assertives. Additionally, humor was integrated into almost 20% of status messages. These findings demonstrate differences in how users express themselves in alternate media, and are framed not only to discuss self-presentation in social networks, but also to offer suggestions for theoretical implications for computer-mediated communication research.
Keywords: speech act, social network site, identity construction, CMC
Speech Acts within SNSs
3
Review of Literature As new media technologies emerge, it is important to look at the commonalities and differences in traits and usages of these communication tools. One emergent technology quickly becoming a societal staple is the social network site (d. m. boyd & Ellison, 2008). Facebook.com, one of the more popular of these web services, currently has over 68 million active users (Facebook, 2008). With millions of users on these social network sites (SNSs) engaging in both self-presentation and interpersonal communication, SNSs afford a new lens through which to examine human interaction. The present study is an attempt to expand upon previous research which examined the content of Instant Messenger away messages (Nastri, Pena, & Hancock, 2006) by exploring similar messages presented in social network sites. This study drew heavily from Nastri et al.’s study in an attempt to see if language is used in these asynchronous media in ways that are similar or different from the ways language is used in the asynchronous applications of a traditionally-synchronous media such as Instant Messaging applications. Especially in the field of communication technology there is a tendency for empirical research to be application-specific, even though most theories and models are developed relatively independent of individual applications being used to mediate communication. One way to begin closing the rifts that have begin to emerge in the body of literature is to expand upon previous research by replicating (or at least recreating) studies in a wider variety of online contexts and with differing virtual affordances. This study sought to understand how language (and consequently identity) is constructed in emergent SNSs. Speech Acts
Speech Acts within SNSs
4
Communication and psychology scholars have long understood that there is a connection between speech and action. The theory of reason action (Ajzen & Fishbein, 1980; Fishbein & Azjen, 1975) suggests that a person’s behavioural intention (and ultimately action) is a function of the person’s attitudes and social norms about that behaviour. Previous empirical research has illustrated that the messages a person delivers can affect that person’s attitudes about a subject, even going to far as to turn an unfavorable attitude into a favorable one (e.g., Freedman & Fraser, 1966). Stampe (1975) noted the recursive relationship of speech and action, noting that, “What one says determines what one may hope to do in so saying, and what one hopes to do determines what one may say in the effort to do it” (p. 1). From a constructionist perspective, not only does communication influence action, but communication also creates the reality and contexts in which individuals exist and interact. It therefore seems appropriate to study humans, in part, through studying the messages that individuals produce. Speech acts are one means by which such study can be conducted. Speech acts are units of dialogue that provide both meaning and reality. Searle (1970) defined a speech act as language that both describes and is the action. By using different types of speech acts—such as statements, giving commands, or asking questions—an individual begins not only to operate within the world around him, but interact with the world around him, and in doing so impacts the attitudes and actions of themselves and their interactants (Cooren, 2003). Speech act analyses haves taken many forms and have been applied to many situations. For example, Overbey and PrestonMatto (2002) looked at how speech acts were used to develop the characters, mythos and social interactions within the television show Buffy the Vampire Slayer. Another
Speech Acts within SNSs
5
application is that of Egner (2006), who looked at how cultural differences between Africans and Westerners affected meaning of the speech act of “promising” in international relations. A third example can be seen in Chirrey’s (2003) work looking at the way messages were constructed and used by homosexuals as they “outed” their sexual orientation. Although the applications of speech act research are broad, the unilateral application of speech acts is to understand how individuals construct messages to communicate and create meaning through language, and specifically the construction and purposes of messages. Searle (1969) proposed a model of speech acts which addresses categories of phrases, each with a specific communicative purpose and each classification mutually exclusive of others. Searle (1969) explicated five categories of speech acts: assertives, directives, commissives, expressives and declarations. Clark (1996) later distinguished two subsets that existed within declarations: effective and verdictive speech acts. Nastri et al. (2006), guided by Baron, Squires, Tench and Thompson’s (2005) findings, further considered quotations as a seventh type of speech act. Explanations and examples of the resultant scheme of seven mutually exclusive speech acts can be seen in Table 1. Guiding Research Nastri, Pena and Hancock (2006) were interested in several aspects of the speech acts utilized to construct away messages in an application of a synchronous chat program. Firstly, Nastri et al. (2006) wanted to understand how various specific speech acts were used in away messages. Based on the earlier suggestions of Baron et al. (2005), Nastri et al. (2006) hypothesized assertive, commissive and expressive speech acts should be most commonly observed, as they may serve informative and entertainment purposes; whereas
Speech Acts within SNSs
6
directives, verdictives, and effectives should be observed the least. In other words, based on the predicted role of away messages in developing messages, Nastri and colleagues expected to find away message users would construct messages primarily to inform or entertain the message recipients, most likely friends of the message creator. Additionally, the authors sought to understand the role of humor in away messages, specifically by examining the frequency of humor produced in speech acts. Their exploration of away messages was an attempt to establish speech acts as a useful framework for studying computer-mediated communication, as well as to determine how away messages were used to achieve social functions beyond the intended design of away messages—how asynchronous away messages were being actively and cognitively put to use in an asynchronous feature of a synchronous computer-mediated communication program. To examine speech acts in away messages, Nastri et al. (2006) looked at 483 Instant Messenger away messages posted over the course of fourteen consecutive days by 44 participants between 18 and 22 years of age. After completing questions about their experience using America Online’s Instant Messenger chat program, their screen names were collected by the researchers. Away messages posted by those screen names were collected for either seven (n=28) or fourteen (n=21) days three times daily: 10 a.m., 5 p.m., and 10 p.m. Two raters individually first parsed each of the 483 collected away messages into individual speech acts, then coded for number of speech acts in each away message (finding a total of 574) the type of speech act, use of non-standard orthography, and finally for humor content. A content analysis of speech acts found that assertives were the most common (M=.68, s.d.=.03), followed by expressives (M=.14, s.d.=.02), commissives (M=.12,
Speech Acts within SNSs
7
s.d.=.03), and directives (M=.06, s.d.=.01). These findings, supported using Wilcoxon signed-rank tests, were consistent with the hypothesis that away messages were constructed primarily through assertive, expressive and commissive speech acts. Nastri et al.’s (2006) other area of inquiry relevant to this study regarded the use of humor in away messages, and identified that approximately one-fifth (M=16%, s.d.=21%) of the messages produced by each participant were humorous. Nastri, Pena and Hancock’s (2006) study illustrated several key aspects of the way people construct asynchronous messages within a synchronous computer-mediated communication application. Firstly, their study showed that away messages were used frequently (participants each posted an average of .93 message each day), more so than the message features of other communication technologies like answering machines. Secondly, this research supported previous findings (Baron et al., 2005) that away messages explain one’s absence while simultaneously providing either informational or entertainment value to the recipient. Thirdly, this research demonstrated that users adapted their language to the constraints of the medium through the strategic use of CMC- based orthographies such as common abbreviations, emoticons, intentional misspellings and non-standard usage of punctuation. Finally, although Nastri et al.’s conclusion that humor is “often” (p. 1040) used in away messages may be overstated (as humor only appeared in about one-fifth of posted messages), the fact that humor appeared more often than all but assertive speech acts (in 68% of away messages) and quotations (in 17% of away messages) indicates that humor should be considered a significant aspect of away message construction, both in conceptualizing message construction and in future research. The authors concluded their findings with the suggestion that
Speech Acts within SNSs
8
“participants made active use of away messages for self-presentation purposes” (p. 1041), noting that the demonstrated ability to express a personal identity while maintaining ties with acquaintances was not necessarily part of the application’s original design. Based on these findings, it seems pertinent to consider how messages are constructed in other media, particularly those dealing heavily with self-presentation and relationship maintenance. One means of executing such an extension is to look at the construction of messages within a popular new telecommunicative application: social networking sites. By looking at how individuals on these web applications construct messages, Nastri et al.’s work may be continued and compared against messages in emergent media. Social Network Sites boyd and Ellison (2008) defined social network sites (SNSs) as web-based applications allowing three functions: 1) users construct a public or semi-public profile; 2) present a list of other users to whom an individual is connected; and 3) view and follow that list and the lists of others within the system. These emerging technologies, enabling users to articulate and make visible their social networks, have recently taken the forefront of both commercial and academic interest (d. m. boyd & Ellison, 2008, p. 211). The two SNS frontrunners, MySpace and Facebook, “serve exclusively as virtual social networks where an individual is able to present and maintain a personal image of themselves [sic] for others to observe and interpret” (Reese, Ziegerer-Behnken, Sundar, & Kleck, 2007, p. 3). These sites allow users to (amongst other functions) post information about themselves, “friend” and interact with other users, and build associations with others, both on an interpersonal level and through social groups.
Speech Acts within SNSs
9
Ellison, Steinfield and Lampe (2007) found Facebook users utilized the service to maintain and reinforce social bonds, primarily within preexisting relationships. However, the first function most users are exposed to is constructing their profile, not unlike creating a personal homepage, albeit with a number of specific existing information fields to begin developing an identity on the site. Providing information about yourself, affording entertainment to yourself and others (both familiar and unknown to the user) and the fulfillment of goals of self-expression are common utilities afforded by such expressions online (Papacharissi, 2002). Indeed, it has been posited that the ability to relate to others online has the potential to both provide a healthy outlet for demonstration and construction of identity (Turkle, 1995) as well as facilitate deeper and more meaningful relationships (Walther, 1992; Walther & Burgoon, 1992). The rise in the popularity of these sites not only affords an opportunity to explore interpersonal and group relationships online, but to further examine how people construct their identity (Donath & boyd, 2004) and interpersonally relate to others (d. boyd, 2008; N. B. Ellison et al., 2007) in an online environment. Because of both the technological and social affordances of SNSs, the phenomena of speech acts are especially relevant. Previous research has illustrated the affordances of new media to construct an identity online (e.g., N. Ellison, Heino, & Gibbs, 2006; Filiciak, 2003; Turkle, 1995), and SNS consequently present a new medium in which to explore how individuals construct messages to create and maintain their presence and connections with others online. Although users can both post and transmit pictures, videos and music in SNSs, the primary means of communication is still text. Messages are composed and read in text-format, group affiliations and interests are denoted
Speech Acts within SNSs
10
textually, and users interact via written messages. Consequently, although SNSs afford a tapestry of images and sounds, the cornerstone of communication is still at the messagelevel. It therefore seems appropriate and relevant to extend Nastri et al.’s (2006) work and findings into a new medium. This extension of previous empirical work has two potential benefits. First, it allows a type of replication of the previous work on speech acts in computer-mediated communication, so as to determine the validity of Nastri et al.’s (2006) findings and discussion of the role of messages in CMC. Secondly, such an extension allows the findings of speech acts to be moved out from a specific communication tool (instant messaging) and be more generalized to CMC holistically, rather than limited to specific technological or social practices within a certain medium. Current Study Nastri et al. (2006) examined the speech acts constructed in an asynchronous application of a synchronous medium—away messages in instant messaging. The nature of away messages is such that even if a user is not at his/her computer for synchronous chat, they can still receive information through posted (and more asynchronous) away messages (Baron et al., 2005; Carr, 2008). Status messages in SNSs serve a similar purpose to away messages—they afford an open text field for an individual to provide his status for others to view. For example, Facebook.com allows a user to type in “is excited for the Louis Hamilton’s Grand Prix win this weekend!” that others will see when viewing the user’s profile. What makes studying status messages in SNSs interesting is the reverse in media characteristics—SNSs are certainly asynchronous media, not facilitating direct interpersonal communication, yet status messages can be used for more immediate communication to those who view it. In the SNS world, status messages are a
Speech Acts within SNSs
11
means of getting snapshots of an individual’s thoughts, behaviors or actions. In short, while instant messaging affords highly synchronous communication but asynchronous away messaging, social network sites afford highly asynchronous communication but more synchronous status messaging. The differences in the communication facilitated by these media direct the guiding research question: RQ1: What messages do users of social network sites construct in status update? As suggested by Nastri et al.’s (2006) study, one way to resolve such research question is to look at the types of messages being developed and communicated in these social network sites’ status updates. There are differences in the way that instant messengers and social network sites are intended to be used to communicate. IMing is typically a synchronous activity that has been shown to facilitate task-related information, both in workplace (Carr, 2008; Isaacs, Walendowski, Whittaker, Schiano, & Kamm, 2002) and social contexts (Huang & Yen, 2003). Meanwhile, using a SNS is an asynchronous activity, and by definition serves primarily a social purpose, facilitating socioemotional information exchange and relationship maintenance (N. B. Ellison et al., 2007; Stutzman, 2007). As expressive speech acts are those which express feeling towards the receiver, it can be reasoned that if one’s goal is to present socioemotional information, more expressive messages would be constructed. This shift in perceived focus from task-related (facilitated by assertive speech acts) in IM to socioemotional interaction (facilitated by expressive speech acts) in SNS guides the first hypothesis:
Speech Acts within SNSs
12
H1a: Expressive speech acts will be most common in social networking site status update messages. However, as SNSs also provide a forum for self-presentation and assertion of identity characteristics, much like personal web pages (see Papacharissi, 2002), it further stands to reason that in addition to expressive comments, assertive messages may be used to present aspects of the self to others. Assertive speech acts are used to get the receiver to form a belief, and as such would be apropos for identity construction and display, developing the self for others to see and interpret. This potential guides the extension of the first hypothesis: H1b: After expressive speech acts, the greatest proportion of speech acts will be assertive, commissive and directive, respectively. Nastri et al. (2006) also found slightly less than one fifth of away messages (16%, SD=21%) used humor, while about a quarter of away messages were the messaging application’s default message. If the speech acts being constructed by participants in SNS status messages are more expressive than the messages constructed in IM away messages (as predicted by H1a), and humor is an effective means of expressing emotions (Francis, 1994), it follows that greater amounts of humor will be used to construct SNS status messages, thus governing the second hypothesis: H2: Humor will be used greater than 16% of status messages in social network sites. Due to the relative novelty of SNSs, as well as the status message function (which for Facebook and MySpace is only a new feature within the last eighteen months), many users have less experience at using social network sites’ status messages than they do
Speech Acts within SNSs
13
using instant messaging. As use of a technology increases as populations have more time to become familiar with the technology (Nowak & Rauh, 2006; Rogers, 2003), it can be expected that many users, even those that use social network sites regularly, have not yet taken to using status updates. However, while instant messengers are installed with a default away message, social network sites have no default and simply leave the status update blank until created or updated. An additional difference between away messages and SNS status messages is that status messages can be left stable over many days, even while the user is communicating on the SNS, unlike instant messengers which typically require users to turn off their away message to resume the program’s functionality. This critical difference guides the second research question: RQ2: How often do social network site users modify their status message? The study of status messages in SNSs affords an opportunity to see how users of these sites are using language to communicate, both interpersonally and to manage their identity. As more people continue to turn to social network sites as a means of communication and relationship maintenance, developing a deeper understanding of the messages used to socially construct these relationships and identities provides an application for technology-based theories. As discussed previously, expanding the scope of applications of technology-based theories offers empirical validation of the theory by increasing the theory’s reliability by expanding its applicability to technologies in a more holistic and general manner than theories tied to a specific medium, or at least not tested in alternate media. Method Participants
Speech Acts within SNSs
14
In-line with the methodology of Nastri et al.’s (2006) study, this study used a convenience sample of students recruited from a telecommunication survey course at a large public university in the Midwest United States. The 74 participants (18 females), whose ages ranged from 18 to 22 years old (M=19.9, SD=.95), were given course extra credit for their participation. Twenty-eight participants did not produce any status messages during the two weeks of investigation, resulting in an ultimate sample size of 46 participants. These results are comparable to Nastri et al.’s (2006) participants in age, class standing and number of subjects, though are skewed with male respondents. Materials Participants completed a self-administered questionnaire about their computer and social network site usage at their personal computers, so as to acquire more accurate and precise responses to items such as number of SNS friends, as respondents would have access to their SNS profile as they would be able to check their specific number of friends, thereby increasing the accuracy and specificity of responses. Survey instruments inquired about participants’ online habits, including how much time they spent online each day, time spent on social network websites, number of months of experience using social network sites and the number of “friends” in their networks—both their primary network and total friends throughout networks. Additionally, participants completed the Facebook Intensity Scale (N. B. Ellison et al., 2007), including items such as “How many Facebook friends do you have” and “Facebook is part of my daily routine,” as well as an adapted version pertaining to MySpace usage (Appendix A) to gauge how connected and reliant participants were on these SNSs as a means of communicating with friends. Procedure
Speech Acts within SNSs
15
After consenting to participate in the study, participants completed the aforementioned measures. After completing the questions on the survey participants initiated a “friend request” with a profile created on Facebook and MySpace (as applicable, based on the participant’s SNS use) for the purposes of this study, so that status messages could be collected based on the research profile’s friend list. Upon completing the survey instrument and initiating a request for affiliation with a Facebook profile established for the purposes of this research, participants returned the instrument to the researcher. Participants were told that their profiles would be looked at over a twoweek period, though they were not told how often their profiles would be checked, nor were they told what specific elements of their SNS profiles were being examined. Status updates were collected at 9:00am, 2:30pm, and 9pm daily, during fourteen consecutive days in April and May of 2008. At the end of the collection period, participants were sent a message through the SNS debriefing them, and then were “defriended.” Content Analysis Collected status updates were coded following the speech acts used by Nastri et al. (2006). The unit of analysis was the speech act, defined as punctuation or propositional units as follows. Status update messages were categorized into their component speech acts, as single status updates could contain more than one sentence, and a sentence could contain more than one speech act. For example, a status update could read, “I’ve got a headache from last night, but I’m going to class anyway.” In that status update there are two speech acts, the first referring to “a headache from last night,” the second denoting the writer as “going to class.” Speech act categories, however, were
Speech Acts within SNSs
16
mutually exclusive and exhaustive—all speech acts could only be assigned to one and only one of category. Status messages were first analyzed for the number of component speech acts that comprised the update. Second, utilizing the Speech Act Taxonomy described in Table 1, the speech acts were coded as assertive, directive, commissive, expressive, effective, or verdictive. Following Nastri et al.’s (2006) protocol, quotations within status messages were coded as a separate category and not categorized into speech acts. Two raters independently coded all messages. Initial intercoder reliability of individual speech acts (i.e., parsing a complex status message into unique speech acts) was acceptable (kappa=.54, p<.001), and a moderate intercoder reliability of humor (alpha=.65, p<.05). Coders were asked to first resolve disagreement through discussing the particular speech act, resolving the act’s categorization amongst themselves. When coders could not agree on the categorization of a speech act, a third trained coder was asked to code the speech act in question, and the agreement of the third coder’s assessment with one of the first two coders was used for the speech act. Results A total of 207 unique SNS status messages were recorded during the two week period, with a mean of .18 (SD=.25) messages produced per day by each participant. Of the 207 messages produced, only 3 were posted on MySpace; the remaining 204 were posted on Facebook profiles. Because of the lack of use of MySpace status updates, the three messages produced were subsequently excluded from analysis. Twenty-eight participants did not post a status message in either Facebook or MySpace during the 14-day period, and were excluded from further analysis.
Speech Acts within SNSs
17
Descriptive Analysis Forty-six participants were ultimately used for analysis. Participants reported a mean of 266 (S.D. = 202.79) Facebook and 231 (S.D. = 593.44) MySpace friends. Furthermore, results of the Facebook and MySpace Intensity scales, the latter adapted illustrated that participants were more ingrained into Facebook than MySpace, with a mean Facebook intensity score of 0.22 (S.D. = .95), and a mean intensity score of -.66 (S.D. = .81) for MySpace. Comparative results of this study and the results of Ellison et al.’s (2007) can be seen in Table 2. Consequently, these participants represent a slightly younger and male-dominated sample, but an increased adoption of Facebook, up to 100% from 94% from surveys conducted on a similar sample the previous year. Speech Act Analysis Status messages were analyzed according to speech act(s). Following Nardi et. al’s protocol, quotations were excluded from analysis. Consequently, all remaining speech acts were coded and analyzed as one of six exhaustive and mutually exclusive speech act categories: assertive, directive, commissive, expressive, effective or verdictive. This yielded a total of 233 speech acts, or an average of 2.54 (SD = 3.50) per message. The proportion of each speech act category produced per participant was calculated by dividing the number of speech acts in a given category by the total number of speech acts produced by the participant, so as to understand how status messages were constructed at the level of the individual. There were no effective speech acts and no verdictive speech acts produced by participants, and these types consequently were not included in the analysis. Means and standard deviations for all speech act categories are
Speech Acts within SNSs
18
presented in Figure 1. Due to the categorical nature of the data, non-parametric statistics were employed for the speech act analysis (Siegel, 1956). Pairwise comparisons among the four remaining speech act categories (expressives, assertives, commissives and directives) using on signed-rank tests demonstrated that participants constructed status messages with expressives more than any other speech act (z = -1.86, p = .063), providing weaker support than desired (α=.05) for Hypothesis 1, but illustrating expressives are most commonly used to develop status messages. Assertives were produced more frequently than commissives (z = -4.464, p < .001) or directives (z = 4.465, p < .001); however, commissives and directives were not significantly different from one another with regards to frequency of construction (z = -1.514, p > .1). This finding provides partial support for Hypothesis 1a. These analyses are described in Table 3. Taken together, the data suggest that status messages are constructed primarily with expressive and assertive speech acts, consistent with Hypothesis 1 and Hypothesis 1a. Humor Recalling the second hypothesis that humor should appear in greater than 16% of status messages, the proportion of humor per message was calculated by dividing the total number of status messages containing humor by the total number of messages produced per person. On average, 21% (SD=31%) of status messages contained some modicum of humor. A t-test illustrates the participants of this study did use humor more frequently in Facebook than reported by Nastri et al. (2006) with regard to away messages (t=1.36, p<.05, one-tailed). Consequently, H2 was supported. Complete comparisons between the results of this study and the results of Nastri et al. (2006) are provided in Table 4. The implications of the support of H1 and H2, as well as the
Speech Acts within SNSs
19
implications of H1a will be discussed in the next section, specifically with relation to the research questions. Discussion This study sought to understand how status messages are constructed in social networking sites. An analysis of speech acts illustrated that expressive and assertive acts were used most commonly, whereas humor was present in almost 20% of status messages. The findings illustrate differences between SNS status messages and IM away messages as reported by Nastri et al. (2006). These findings help us develop a greater understanding not only of the creation and display of messages within the context of SNS, but also begin to shed light on the differences of messages created in comparable media applications. Of immediate note is the disparity between the use of Instant Messaging away messages reported by Nastri et al.’s (2006) and the use of Social Networking Site status messages reported here. The comparable participant pool in this study produced approximately two-fifths the messages produced by Nastri et al.’s (2006) participants. During the observation period, only 2 of the 26 participants who had a MySpace profile posted status messages and only 46 of 86 participants who had a Facebook profile posted status updates. Compared to the mean .93 (S.D. = .63) away messages produced per day reported by Nastri et al. (2006), SNS users (posting a mean of .181, S.D.=.249 messages per day) do not appear as predisposed to regularly post status updates as do IM users to post away messages. One potential reason for this disparity between media may be the relative novelty of status messages. At the time of the study, status messages had only been recently added features to both SNS tools. As SNS are typically considered means
Speech Acts within SNSs
20
of asynchronous communication, it is possible SNS users are not yet aware of or readily adopting the new tool for constructing messages targeted at friends. However, with users posting an average of .18 messages per day after only a few months of the feature being made available, it seems that the status message feature is beginning to be adopted. Even changing one’s status messages once every five days, as the data suggests, still means that individuals are updating their Facebook status more frequently than they update voice mail or other asynchronous services (Ehrlich, 1987). Another explanation of the reduced use of status messaged compared to away messages is the role of status messages in SNSs. Away messages represent an integral and meaningful part in IMs, indicating that the target is not available for synchronous chat, and in doing so conveying message about the immediate disposition of the target. In SNSs, status messages do not play such a necessary role, and their presence may simply be a means of self expression rather than serving a structural function within the medium. Consequently, status messages may appear less frequently strictly because they play a different role within the medium. Previous research has proposed that SNSs provide a forum for social interaction and interpersonal exchange (Donath & boyd, 2004). The findings of this research support such a proposition, with almost 60% of status messages containing an expressive speech act and reflective almost 52% of the speech acts produced to convey emotion towards the receiver(s). Further, 39% of speech acts were produced to have the receiver form an impression or belief (i.e., assertives), 6% to seek to get the receiver(s) to do something (i.e., directive) and 3% to commit to future action (i.e., commissive). The proportion of expressive speech acts within status messages was higher than all other categories, and
Speech Acts within SNSs
21
expressive and assertive speech acts were significantly higher than other categories. Clark (1996) noted that effective and verdictive speech acts (statements about the status of an institution) are typically made within institutional or organizational settings, and as SNSs do not constitute such an environ the absence of such speech acts may not be surprising. The role of SNS in supporting interpersonal interactions is further supported by the presence of humor in 21% of status messages. About one-fifth of Facebook users are integrating jovial content into their status messages. Instances of humor have been conceptualized as an indicator of socioemotional communication in previous research (Hiltz, Johnson, & Turoff, 1986), and its presence in SNS status messages may be indicative of the same. The amount of humor is significantly greater in SNS status messages than in IM away messages, indicating that SNS may reflect a medium in which individuals express themselves more openly and interpersonally, especially as they are providing the status messages to individuals with whom they are acquainted and have “friended.” This increase in humor content also may reflect the parallels between SNS profiles and personal homepages. Papacharissi (2002), in looking at the content of personal websites, found that homepages tended to be developed for one of three primary purposes: to provide information about the individual, communicating with friends and family , and to entertain friends. In this way, self-presentation seems to be replicating the role of personal websites. The first two purposes (information and communication) are inherent in SNS: one constructs a profile (self-presentation) and joins the site to keep in touch (communicate). This study suggests that the third goal may, too, be sought via SNSs—individuals may seek to entertain others via humor in their status messages.
Speech Acts within SNSs
22
Previous research has thoroughly acknowledged the role and function of specific types of speech acts within various settings (e.g., Baron et al., 2005; Nastri et al., 2006). What is of importance to the focus of this research is the notable differences between the findings of this study and comparable research conducted with similar participants in a similar yet different communication application, as it illustrates the different messages constructed in different online media. Culnan and Markus (1987) made two important notes regarding researching effects of mediation communication. The first was that while face-to-face communication differs from computer-mediated communication, the fundamental communicative processes may not be as radically different as many scholars assume. However, their second note is of great significance to the findings of the present research—different media have the potential to facilitate different communication and via different processes. It is the implications of this research mapped onto their second notion which will drive the remainder of this discussion. Goal Achievement and Strategic Use. Theorists have long discussed the impacts of communicative goals on media selection. Early CMC research suggested that CMC was unable to transmit cues that facilitated interpersonal interaction, and instead focused on the role of CMC in task-related situations (e.g., Daft & Lengel, 1986; Kiesler, Siegel, & McGuire, 1984). More recent research has acknowledged the potential for CMC to foster rich and highly interpersonal interactions, sometimes more so than face-to-face counterparts (Walther, 1996). While the early theories looked at how individuals selected media based on the nature of the task-related goals they desired to achieve, subsequent research on media has instead focused on the mechanisms by which CMC facilitates interpersonal interactions. The findings of this research, and specifically that expressive
Speech Acts within SNSs
23
(i.e., socioemotional) messages are a primary component of status message development in contrast to the predominance of assertives in IM, implies that there may be some strategic cognitive processes that users are going through to decide what messages to use to construct away messages given the nature of SNS. Consequently, it may be prudent for future research to begin to once again look at the process of media selection, but to do so from a relational perspective. Walther and Bazarova (2008) offer one example of the beginnings of such research, as they look at the perceptions of propinquity amongst members of computer-mediated groups, taking specific note of information complexity and available media as dependent variables. By demonstrating differences in messages constructed in different media for similar purposes, this research further illustrates the need to look at the mechanisms and processes by which users select amongst various media to accomplish communicative goals. Modality Switching. Most CMC research (particularly experimental research) has focused on interactions that occur specifically online, frequently with zero-history groups that will not interact again outside of the research context, and often will not interact outside of computer-mediated medium. One notable exception to this is the work of Ramirez and Zhang (2007) which explored how transitioning from mediated to face-toface relationships impacted relational outcomes. What has been ignored, in comparison, is the nature of interaction as the relationship moves from a face-to-face to an online context. How do individuals’ goals and interactions alter as a result of that shift in modality? Previous SNS research has illustrated that most of an individual’s “friends” in SNS are also acquaintances in real-life; in other words, people don’t have friends on SNS that they don’t know outside of the web application (d. m. boyd & Ellison, 2008; N. B.
Speech Acts within SNSs
24
Ellison et al., 2007). Although the causality behind the preponderance of expressive speech acts is beyond the scope of this research, it would be of interest to see if the amount of expressives is a result of interactants having prior knowledge and relationships with those who are expected to see the status update. Said another way, does the number of expressives in status message posts reflect the number of expressives individuals use in their offline communication with the same individuals? Predictive factors. While the previous issues have been applied in an attempt to provide explanatory analyses of the research, it is also necessary to address the predictive role of theory, and its absence in the study. The hypotheses in this study were derived under the assumption that social networking sites were better suited to facilitating socioemotional communication specifically of the expressive nature. This assertion was supported quantitatively within the research. However, the fundamental nature of the medium may and should not be the only factor to consider when theorizing differences between media. For a theory to be useful, it must not only be able to retrospectively explain why a phenomenon occurred, it must also be able to proactively predict results based on a specified series of precedent variables (Shoemaker, Tankard, & Lasorsa, 2003). Although this study has illustrated different communication resulting from two media, it not fully predicted why such differentiations occurred in a theoreticallygrounded way. As new media begin to emerge and become integrated into society, it becomes increasingly important for theories to emerge to enable researchers to not only conceptualize how specific media are constructed and function, but to further be able to develop functional models that could detail the mechanisms and interactions that guide both interaction and use of media.
Speech Acts within SNSs
25
Conclusion This study was conceptualized and executed in an attempt to replicate previous research (Nastri et al., 2006) to explore how individuals express themselves online, and to also extend such research into a new medium. Such an extension was intended to explore how different media may individuals using different media may manifest different types of expressions. The findings of this study support previous claims that social networking sites are inherently social and emotive in nature, demonstrating the preponderance of expressive acts of communication, displaying and sharing emotion amongst other SNS users. In addition to the predominance of expressive speech acts, the increased proportion of humor used in SNSs over IMs further demonstrates the social and emotive nature of SNS and their role in maintaining interpersonal and emotional ties amongst networked individuals. The results of this study, while illuminative in their own right, become even more significant when interpreted in tandem with previous research. The findings presented here illustrate the role of media on self-presentation and demonstrate that generalizations of findings across multiple media may not be fairly made based on exploration of a phenomenon within a single context or technology. As the opportunities and means for individuals to communicate continues to proliferate, there is a need to begin to develop theories outside of specific media to begin to explore relational media holistically and how they impact the nature and processes of communication amongst their users. The development of such a theory would be advantageous to understand the nature of communication in multiple situations, especially in online contexts and in situations
Speech Acts within SNSs
26
where individuals choose (either strategically or unconsciously) which medium to use for a specific communicative goal.
Speech Acts within SNSs
27
References Ajzen, I., & Fishbein, M. (1980). Understanding attitudes and predicting social behavior. Englewood Cliffs, NJ: Prentice-Hall. Baron, N. S., Squires, L., Tench, S., & Thompson, M. (2005). Tethered or mobile? Use of away messages in instant messaging by American college students. In R. Ling & P. Pedersen (Eds.), Mobile communications: Re-negotiation of the social sphere (pp. 293-311). London: Springer-Verlag. boyd, d. (2008). Why youth (heart) social network sites: The role of networked publics in teenage social life. In D. Buckingham (Ed.), Youth, identity, and digital media (pp. 119-142). Cambridge, MA: MIT Press. boyd, d. m., & Ellison, N. B. (2008). Social network sites: Definition, history, and scholarship. Journal of Computer-Mediated Communication, 13, 210-230. Carr, C. T. (2008, May 25, 2008). Social influence on an organization’s successful adoption of Instant Messaging. Paper presented at the annual meeting of the International Communication Association, Montreal, Quebec. Chirrey, D. A. (2003). ‘I hereby come out’: What sort of speech act is coming out? Journal of Sociolinguistics, 7, 24-37. Clark, H. H. (1996). Using language. Cambridge, UK: Cambridge University Press. Cooren, F. (2003). The contribution of speech act theory to the analysis of conversation: How presequences work. Paper presented at the annual meeting of the International Communication Association, San Diego, CA. Culnan, M. J., & Markus, M. L. (1987). Information technologies. In F. M. Jablin, L. L. Putnam, K. H. Roberts & L. W. Porter (Eds.), Handbook of organizational
Speech Acts within SNSs
28
communication: An interdisciplinary perspective (pp. 421-443). Newbury Park, CA: Sage. Daft, R. L., & Lengel, R. H. (1986). Organizational information requirements, media richness and structural design. Management Science, 32, 554-571. Donath, J., & boyd, d. (2004). Public displays of connection. BT Technology Journal, 22(4), 71-82. Egner, I. (2006). Intercultural aspects of the speech act of promising: Western and African practices. Intercultural Pragmatics, 3, 443-464. Ehrlich, S. F. (1987). Social and psychological factors influencing the design of office communication systems. Paper presented at the ACM Conference on Human Factors in Computing Systems and Graphics Interface (CHI1GI ’87). Ellison, N., Heino, R., & Gibbs, J. (2006). Managing impressions online: Selfpresentation processes in online dating environments. Journal of ComputerMediated Communication, 11, 415-441. Ellison, N. B., Steinfield, C., & Lampe, C. (2007). The benefits of Facebook ‘‘friends:’’ Social capital and college students’ use of online social network sites. Journal of Computer-Mediated Communication, 12, 1143-1168. Facebook. (2008). Factsheet. Retrieved March 17, 2008, from http://www.facebook.com/press/info.php?factsheet Filiciak, M. (2003). Hyperidentities: Postmodern identity patterns in massively multiplayer online role-playing games. In M. J. P. Wolf & B. Perron (Eds.), Video game theory reader (pp. 87-102). New York: Routledge.
Speech Acts within SNSs
29
Fishbein, M., & Azjen, I. (1975). Belief, attitude, intention, and behavior: An introduction to theory and research. Reading, MA: Addison-Wesley. Francis, L. E. (1994). Laughter, the best mediation: Humor as emotion management in interaction. Symbolic interaction, 17(2), 147–163 Freedman, J. L., & Fraser, S. C. (1966). Compliance without pressure: The foot-in-thedoor technique. Journal of Personality and Social Psychology, 4, 195-202. Hiltz, S. R., Johnson, K., & Turoff, M. (1986). Experiments in group decision making: Communication processes and outcome in face-to-face versus computerized conferences. Human Communication Research, 13, 225-252. Huang, A. H., & Yen, D. C. (2003). Usefulness of instant messaging among young users: Social vs. work perspective. Human Systems Management, 22, 63-72. Isaacs, E., Walendowski, A., Whittaker, S., Schiano, D. J., & Kamm, C. (2002). The character, functions, and styles of instant messaging in the workplace. Paper presented at the the 2002 ACM conference on Computer supported cooperative work, New Orleans, LA. Kiesler, S., Siegel, J., & McGuire, T. W. (1984). Social psychological aspects of computer-mediated communication. American Psychologist, 39, 1123-1134. Nastri, J., Pena, J., & Hancock, J. T. (2006). The construction of away messages: A speech act analysis. Journal of Computer-Mediated Communication, 11, 10251045. Nowak, K. L., & Rauh, C. (2006). The influence of the avatar on online perceptions of anthromorphism, androgyny, credibility, homophily, and attraction. Journal of Computer-Mediated Communication, 11, 153-178.
Speech Acts within SNSs
30
Overbey, K. E., & Preston-Matto, L. (2002). Staking in tongues: Speech act as weapon in Buffy. In R. V. Wilcox & D. Lavery (Eds.), Fighting the forces: What's at stake in Buffy the Vampire Slayer (pp. 73-84). Lanham, MA: Rowman & Littlefield Publishers, Inc. Papacharissi, Z. (2002). The self online: The utility of personal home pages. Journal of Broadcasting & Electronic Media, 46, 346-368. Ramirez, A., Jr., & Zhang, S. (2007). When online meets offline: The effect of modality switching on relational communication. Communication Monographs, 74, 287310. Reese, C., Ziegerer-Behnken, D., Sundar, S. S., & Kleck, C. (2007, May 24-28). The company you keep and the image your project: Putting your best face forward in online social networks. Paper presented at the annual meeting of the International Communication Association, San Francisco, CA. Rogers, E. (2003). Diffusion of innovations (5th ed.). New York: Free Press. Searle, J. R. (1969). Speech acts. Cambridge, UK: Cambridge University Press. Searle, J. R. (1970). Speech acts: An essay in the philosophy of language. Cambridge, UK: Cambridge University Press. Shoemaker, P. J., Tankard, J. W., & Lasorsa, D. L. (2003). How to build social science theories. Thousand Oaks, CA: Sage. Siegel, S. (1956). Non-parametric statistics. New York: McGraw-Hill. Stampe, D. W. (1975). Meaning and truth in the theory of speech acts. In P. Cole & J. L. Morgan (Eds.), Speech acts (pp. 1-39). New York: Academic Press.
Speech Acts within SNSs
31
Stutzman, F. (2007). Impression formation and management in social network websites. Paper presented at the annual meeting of the International Communication Association, San Francisco, CA. Turkle, S. (1995). Life on the screen: Identity in the age of the internet. New York: Simon & Schuster. Walther, J. B. (1992). Interpersonal effects in computer-mediated interaction: A relational perspective. Communication Research, 19, 52-90. Walther, J. B. (1996). Computer-mediated communication: Impersonal, interpersonal, and hyperpersonal interaction. Communication Research, 23, 3-43. Walther, J. B., & Bazarova, N. N. (2008). Validation and application of electronic propinquity theory to computer-mediated communication in groups. Communication Research, 35, 622-645. Walther, J. B., & Burgoon, J. K. (1992). Relational communication in computer-mediated interaction. Human Communication Research, 19, 50-88.
Speech Acts within SNSs
32
Appendix A
Social Network Website Survey Thank you for answering our questions about your experiences using social network websites. Please answer all questions to the best of your ability. Name: ___________________ Biological gender:
Male
Year in school:
Freshman (0-30 credit hours) Junior (61-90 credits) Graduate/Professional
Sophomore (31-60 credits) Senior (91+ credits) Other
Local Residence:
On-Campus
Off-Campus
My home (permanent) residence is:
Female
In-State (MI)
Are you a member of a sorority/fraternity?
Out-of-State
Yes
No
How many on-campus extracurricular groups are you actively participating in? None 1 2 3 4 5 6 7 8 9 10 11+ In an average day, how many hours do you use the Internet for either work or pleasure? None Less than 1 1-2 2-3 3-4 4-5 5-6 6-7 7-8 8-9 9-10 10-11 11-12 12-13 13-14 14-15 15-16 16-17 17-18 18+ Do you use some form of Instant Messaging chat program (ex. AIM, ICQ, etc.)? Yes No If yes, about how long have you used an IM program? ____ years & ____months
Facebook Do you have a Facebook account? Yes If “No,” please move on to the next page.
No
About how long have you had your Facebook account? ____ years & _____ months
Speech Acts within SNSs
33
How many total Facebook friends do you have at MSU or elsewhere? _____ In the past week, on average, approximately how many minutes have you spent on Facebook each day? _____ How many total Facebook groups do you belong to? ______ Strongly Disagree Neutral Disagree
Agree
Strongly Agree
Facebook is part of my everyday activity I am proud to tell people I’m on Facebook Facebook has become part of my daily routine I feel out of touch when I haven’t logged onto Facebook for a while I feel I am part of the Facebook community I would be sorry if Facebook shut down I use Facebook to learn more about other people in my classes I use Facebook to learn more about other people living near me I use Facebook to keep in touch with my old friends When you are done with this part of the survey, please send a friend request to “Telecom Study.” It will have a TISM logo for a profile photo.
Speech Acts within SNSs
34
MySpace Do you have a MySpace account? Yes No If “No,” please move on to the end of this page (**). About how long have you had your MySpace account? ____ years & _____ months How many total MySpace friends do you have at MSU or elsewhere? _____ In the past week, on average, approximately how many minutes have you spent on MySpace each day? _____ How many total MySpace groups do you belong to? ______ Strongly Disagree Neutral Disagree
Agree
Strongly Agree
MySpace is part of my everyday activity I am proud to tell people I’m on MySpace MySpace has become part of my daily routine I feel out of touch when I haven’t logged onto MySpace for a while I feel I am part of the MySpace community I would be sorry if MySpace shut down I use MySpace to learn more about other people in my classes I use MySpace to learn more about other people living near me I use MySpace to keep in touch with my old friends
When you are done with this part of the survey, please send a friend request to the display name, “TC Research.” It will have a TISM logo for a profile photo. **Thank you for completing this survey. Please return the completed survey to the researcher.
Speech Acts within SNSs
35
Table 1 – Coding scheme for analyzing speech acts and quotations in Facebook status messages Speech Act
Properties of Speech Act
Example(s)
Assertive
Statements of fact, getting the viewer to
“GRILLING on this
form or attend a belief
beautttiful day,” “is home.”
The sender uses this to get the receiver to
“saying watch payback,”
do something (i.e., a command)
“hey do it”
The sender commits himself to do
“only 26 more days,” “80 to
something
go”
Sender expresses feeling towards (though
“hoping something amazing
not necessarily about) the receiver
will happen...,” “wishing he
Directive
Commissive
Expressive
was a weapon master.” Effective
To change an institutional state of affairs
“You’re fired,” “Play ball,” “Chapter at 7 or you’re fined”
Verdictive
Quotation
To determine what is the case in an
“I find him innocent,”
institution
“strike”
The message is not originally produced by
“crying blood and bleeding
the sender
tears,” “"You hit Me Like A Cosmic Blast”
Note. Most examples were selected from the status messages collected. Because they did not appear in this study, effective and verdictive examples were drawn from Nastri et al. (2006).
Speech Acts within SNSs
36
Table 2 – Participant demographic comparison between present study and Ellison et al. (2007) Present Study
Present
Ellison et al.
Ellison et
(2008) Mean
Study s.d.
(2007) Mean or
al. s.d.
or % (n)
% (n)
male
76% (56)
34% (98)
female
24% (18)
66% (188)
Gender:
Year in school 1
1.95
0.95
2.55
1.07
Home residence: in-state
85% (63)
91% (259)
out-of-state
15% (11)
09% (25)
on campus
73% (54)
55% (157)
off campus
27% (20)
45% (127)
08% (6)
08% (23)
Local residence:
Member of fraternity or sorority Hours of Internet use per day
Facebook members
2
3:08
100% (74)
2 hours 56 min.
1:52
94% (268)
Facebook friends
266.14
202.79
Facebook groups
27.58
30.29
Facebook Intensity
0.22
0.95
MySpace members
1
5 hours 18 min
2
28% (20)
-0.08
0.79
N/A
MySpace friends
231.71
593.44
N/A
MySpace groups
1.38
2.77
N/A
MySpace intensity
-0.66
0.81
N/A
1 = freshman, 2 = sophomore, 3 = junior, 4 = senior, 5 = graduate student converted from ordinal scale using mid-point of response category (e.g., 1-2 hours = 1 hour 30 minutes)
Speech Acts within SNSs
37
Table 3 – Pairwise Wilcoxon test comparisons among proportions of speech acts
Z
Effectives/
Effectives/
Effectives/
Assertives/
Assertives/ Commissives/
Assertives
Commissives
Directives
Commissives
Directives
-1.860
-5.095**
-4.907**
-4.907**
-4.564**
-1.514
.063
.000
.000
.000
.000
.130
Directives
Pairwise comparisons were based on the proportion of each speech act category, calculated by dividing the total of speech acts in a given category by the total number of speech acts produced by the participant. * p=.06, two tailed. **p<.01
Speech Acts within SNSs
38
Table 4 – Speech Act construction comparison between present study and Nastri et al. (2006) Present Study
Present
Nastri et al.
Nastri et
(2008) Mean
Study s.d.
(2006) Mean
al. (2006)
Speech Act/Post
s.d.
Speech Act/Post Assertive
.19 (86)
0.27
.68
0.03
Directive
.01 (6)
.08
.06
0.01
Commissive
.02 (12)
.06
.12
0.12
Expressive
.27 (110)
.34
.14
0.03
Effective
.00 (0)
.00
.00
0.02
Verdictive
.00 (0)
.00
.00
0.00
Quotation
.05 (22)
.17
.17
.22
Humor
.19 (65)
.29
.16
0.21
Total Messages
204
Messages per day per
.18
483 .25
.93
.63
participant Total Speech Acts
233
Acts per message
2.54
574 3.50
1.14
0.44
Speech Acts within SNSs Figure 1 – Frequency of speech acts within SNS status messages
*Figures presented are means of speech acts within status messages per person, while numbers in parentheses are standard deviations.
39