Sometime during the year 2013, I became vaguely aware of the Twitter phenomenon that had apparently been raging all around me for years. Frankly, I had only had a Facebook account for a couple years, and I certainly had no awareness of what a hashtag was, how it was used, or the major effect it was having socially and linguistically across almost every demographic group. The beginning of my awareness of hashtags was this Subway Tuscan Chicken Melt television commercial (https://www.youtube.com/watch?v=abkVrXLzvPU), which depicts two young men eating the sandwiches (Unknown Director, 2013). One of them is tweeting excitedly about the sandwich, using multiple hashtags to describe it, his experience, and his intended actions, while the other quietly observes. While the tweeting man is distracted with his phone, the second man steals his sandwich and begins to eat it. The tweeting man is depicted as rather ridiculous and, certainly in this case, is outsmarted and bested by his friend, so my already negative impression of social media sites and apps was deepened even further.
It was not until recently, as I noticed several of my friends quite liberally sprinkling their Facebook posts with hashtags that I became fascinated with the idea that this punctuation mark or symbol was changing the way that people spoke about and around topics. I learned quickly that many of these hashtags were being employed well outside of the intended use of the signifier, which is to semantically tag and categorize content to make it easier to find or organize (Dwyer & Marsh, 2014). Certainly, there were examples of typical use in promoting an idea, expressing sarcasm, offering subtext, or repeating a theme, however, some examples of its use seemed so bizarrely cryptic as to be inscrutable (Daer, Hoffman, & Goodman, 2014). As evidence mounted of a subtext to these expressions, I began to also wonder why individuals used the hashtag at all; why not just make the statement or express the feeling through the text of the post or message? How did using the hashtag differentiate their language actions from plain speech or text?
This, then, is the question that this exploratory research will attempt to answer. In order to accomplish this, it was key to find a venue in which the hashtag did not really belong, since applications in which the hashtag behaves naturally would be overwhelmed with “typical” use of it and fail to highlight the nuanced information sought here. The hashtag did not originate on Facebook, nor is it the expected or “natural” place to use it, so this platform was chosen. It was also of interest to take a brief detour into the use of the hashtag in spoken language to determine if there was a difference between its use when posting on Facebook and when speaking, both of which are not, in the strictest sense, occasions when hashtags would normally be used.
Peer-reviewed literature regarding the semiotics, purpose, and meaning of hashtags is currently sparse and extremely diverse. Research that specifically examines the linguistic and/or social significance of hashtags is even more rare. The studies can also be highly specific, as in the case of Dwyer and Marsh’s work in using hashtags to interpret how users conceptualize trust (Dwyer & Marsh, 2014). Often, the research may investigate the purpose or content of the hashtag, as has been done in several studies regarding the use of sarcasm, but does not touch on the underlying function of the symbol (Gonzalez-Ibanez, Muresan, & Wacholder, 2011). Other papers, but these latter especially, approach the research from a machine learning perspective and are attempts to teach computers to identify and categorize social media posts on a big data scale (Kunneman, Liebrecht, van Mulken, & van den Bosch, 2015).
Hashtags can also be placed into type categories, in addition to the content categories, based on their content. Some identified broad categories of purpose like Critiquing and Iterating, both of which, again, concern the purpose of the hashtag’s words, rather than the function of the hashtag itself (Daer, Hoffman, & Goodman, 2014). Others point to the use of the hashtag as a type of “metadata” associated with the content of the post (Wang, Wei, Liu, Zhou, & Zhang, 2011). These categories do point to the hashtag as a functional linguistic item, but in a very roundabout way.
In virtually all of these instances, the researcher acts as a distant observer, collecting information about the hashtag use through automated systems or mass downloads. One thing on which they consistently agree, though, is the idea of the hashtag as an ad hoc tagging system that has been dubbed a “folksonomy” (Yang, Sun, Zhang, & Mei, 2012). This term is an agglutination of the words “folk” and “taxonomy,” indicating a “user-created bottom-up categorical structure development with an emergent thesaurus” (Vander Wal, 2007). This, of course, refers to the function of the hashtag to categorize content, and does not explore any deeper meaning. However, this function may contain a kernel of the social characteristic of the hashtag investigated in other studies.
This pragmatic characterization of the hashtag has a relation to more meaningful definitions, such as that of Smith and Smith, who define the hashtag as “an index, an identifier, a filter, and a promoter; more important, it can connect a virtual community of users” (Smith & Smith, 2012). These are communities which, importantly, are created by the users. Other studies mention these properties, but this particular study rigorously tests the assumption using Twitter data from the 2012 College World Series of baseball. Thousands of tweets with hashtags were collected and examined during the first two games of the series and the “Twittersphere” was dubbed the new watercooler around which sports fans gather to discuss the events of the game. The use of hashtags was also classified as an active form of social-identity theory carried into virtual spaces and, in fact, creating a new kind of space.
Research surrounding hashtag activism also proved fruitful for this study, as it tends to focus on the community dynamic surrounding the hashtags, rather than the content of the hashtags themselves. Two such studies offered intriguing ideas that will be developed further in this study, as they dovetail well with the findings from the interviews in this analysis. Here, it is sufficient to say that those themes are hashtags as a shared temporality (Bonilla & Rosa, 2015) and as a use of narrative agency (Yang G. , 2016).
In order to explore the motivations and meaning behind using hashtags on Facebook and in spoken language, qualitative and quantitative methods were used. Since this is an exploratory research paper and intended only to gather broad data about the subject to determine whether there is a basis for further research, strict scientific methods were not employed. For example, no attempts have been made to randomize samples or make the researcher unaware of the identities of the survey participants.
First, a short survey was distributed using the researcher’s own Facebook account, Slack social media account, and off-line social circle. This consisted of several demographic questions and a few basic, open-ended questions regarding the users’ behavior online and off-line with regard to hashtags. It was designed to be very short and concise in order to maximize participation. The questions from the survey are detailed below.
- What is your age?
- What is your job role?
- Do you have any children under 18?
- In a typical week, which of the following social networking websites do you use most often?
- In a typical week, about how often do you use hashtags when posting to Facebook?
- If you have posted to Facebook using a hashtag, please briefly describe the post (picture, text, subject matter), the hashtag you used, and how and why you used the hashtag. For example, was it to highlight something in the post, to express sarcasm, or to share an inside joke with certain friends?
- Have you ever used the word “hashtag” in spoken language? If so, please describe the circumstances and intent of the use and context for the conversation.
Second, multiple interviews with friends of the researcher who evidenced higher than average use of hashtags on their Facebook accounts (as observed by the researcher) were conducted to obtain a qualitative, deeper understanding of the nature of their use of hashtags. The interviews began with some broad questions about the subject of hashtags, and the conversation was permitted to range freely as topics surfaced and follow-up questions were presented. The subject of the second part of the interviews were ten of the interviewee’s posts collected by the researcher. The subjects were requested to, without restriction by the interviewer, explain the meaning and motivation of the hashtags used in each post. The interviewer would then interject or ask follow-up questions regarding the information conveyed. The interviews lasted approximately 35-45 minutes each. Some examples of the introductory questions are detailed in the list below.
- Define what you think a hashtag is and how it is used in general.
- How do you think that other people use hashtags?
- Are there any social norms surrounding the use of hashtags that you are aware of?
- How active are you on social media and which ones do you use?
- Have you ever used hashtags in spoken language? Can you give an example?
Survey Results and Discussion
Overall, most of the 55 respondents (80%) to the survey indicated that they used hashtags on Facebook “Not at all often,” with the highest percentage of use in the 25 to 34 year-old age range reporting use “Not so often” or “Somewhat often.” The only other group that showed a significant figure (20%) of “Somewhat often” use was the 45 to 54 year-old age range, which seemed surprising until it was revealed that 100% of those who responded this way have children under the age of 18. This low overall rate of hashtag usage was predictable due to the aforementioned fact that Facebook is not really the indigenous environment of the hashtag and is also in line with the hashtag percentages found by Caleffi in her investigation (Caleffi, 2015). The higher use by younger people is also an expected result, although the fact that the 18 to 24 year-old range showed lower usage may indicate a generational awareness of the norms surrounding the use of hashtags in certain forums.
A large portion of respondents stated in the open-ended questions that they do not use hashtags on Facebook at all, and approximately the same number did not respond to that question. The most used purposes of the respondents were, in descending order, sarcasm and humor (7% each), emphasis (5%), and rallying (4%). The last two categories are in line with the findings of Daer, et al., who coded “metacommunicative tagging” of Tweets, although it seems unusual that their findings did not include any mention of humor or sarcasm as a category, since this seems so prevalent (Daer, Hoffman, & Goodman, 2014).
The final point of interest is the responses to the question whether the respondents used the word “hashtag” in spoken language. As might be expected, those who reported not using hashtags on Facebook also consistently reported not using them in speech. However, almost a quarter of respondents across all Facebook hashtag use were either very adamant in their denial of use in speech or reported using it to mock the use of hashtags. This points to a certain social stigma surrounding the use of hashtags that echoes the depiction of the hashtagging Subway customer cited earlier.
Interview Results and Discussion
As stated earlier, there are some uses of hashtags that cannot be immediately categorized, and this revealed itself much more clearly in the individual interviews than in the survey or in the literature. First and most importantly, the hashtags used on Facebook by the interviewees revealed far more personal and interpersonal content than has even been hinted at in much of the literature; the studies of hashtag activism perhaps come close, but are still at a very large scale and use observational techniques. This is not to say that the uses of hashtags cited in these studies do not exist or are not practiced, but there also exists a deep well of information that has not been plumbed and must be studied from a participant observational standpoint to be appreciated. Of course, this latter category is not mutually exclusive of the former; they can be employed in concert with one another.
In response to the general questions, the interviewees consistently quoted the functional aspect of hashtags as the generally accepted use of hashtags; that is, they stated that hashtags are defined as a marker to categorize and group posts and pictures, or perhaps to make a post funny. Interestingly, they also initially characterized their own use of hashtags as falling in this category. When interviewees were speaking about their own hashtag posts, though, a much different story unfolded. For example, one interviewee used the hashtag #deydontevenknow as an inside joke to only one other person and another had a set of identity hashtags that he uses each time he posts anything, regardless of the content of the post. These uses are much more social than verbally reported by the interviewees or the survey respondents.
In fact, overall, the information gathered from the interviewees pointed toward social use as one of the primary functions of the hashtag and functioned under several overall themes, either singly or in combination. Even those posts that may, at face value, appear to be strictly functional have deeper social meaning. For example, #birchwood (name of a restaurant) was used, not just because the individual was at the restaurant or that they wanted to promote the restaurant. Although these were also reasons to use the hashtag, the person reported using it because it is one of their favorite restaurants and they go there all the time with close family and friends, which is a primarily social reason for using a seemingly unemotional hashtag. Similarly, #pitbull (the artist) was used to speak directly to the poster’s mother who had joked with him about saying hi to Pitbull when he travelled to Miami. Importantly, had I researched these hashtags by simple observation, they would have been categorized these as purely promotional hashtags and missed the entire point of why they were used.
After reviewing the surveys and interviews, it became clear that the hashtag itself has a socio-linguistic function of its own, independent of the intent or content of the post or hashtag. One way this became clear was when I asked interview respondents to “translate” their hashtags into plain language or to describe how they would have expressed the sentiments without hashtags. On most occasions, the interviewee stated that they would have dropped at least some of the hashtags altogether because there was not an emotive equivalent in plain language or the story it represented would have taken too long to explain in long form. There was also the sense that it would have changed the meaning if it had to be explained, rather similar to when one tells a joke.
There are also clues to the hashtag’s function in the specific uses of some of the interviewees. One of the more straightforward uses that was revealed was group or social identification. Hashtags like #livingthedream (reference to specific college shared experience), #team[lastname], #armyaviation, and #eraualumni are clearly direct appeals or statements of membership in a certain group with which the poster identifies. This is part of the social identity studied by Smith and Smith in their research into baseball fan interactions on Twitter and is related to the human desire to be part of an organization or group (Smith & Smith, 2012). To take this one step further, one can argue that even hashtags that seem to have a different purpose are actually displays of social identity. For example, #breastcancersupport and #sundayfunday may seem to be rallying or simply functional, but peeling back a layer of this onion exposes the underlying motivation to be part of a certain group, since an individual could certainly post in support of breast cancer without using a hashtag. Minus the hashtag, however, the tweet or post does not invoke the social request or statement of the poster to be a part of a larger group and to be known as such. Even a hashtag like #irmagoaway (Hurricane Irma), when explained by the respondent, was used because she is a teacher and her students used it heavily and she was, in that moment, identifying with them.
In many other posts, and arguably all of the ones examined, there is also a temporal character layered below the social character. Temporalities or temporalizing practices are defined by Nancy D. Munn as devices “whereby the inherent temporal character of social life is brought out” and further as a sort of “hinge” that connects and relates different subjects and time-spaces (Hodges, 2008). One interviewee almost primarily uses hashtags in this way and even maintains a special relationship with her best friend through the use of hashtags. The bulk of these hashtags are inside jokes or stories, created at the time the story occurred, that, when posted to social media, have meaning for the two of them and few others. But they are not just jokes, per se; some of them entail long stories or shared experiences that are memorialized in the hashtag and are linguistic shortcuts for the time, place, and event being evoked.
Although slightly different, this usage has all the temporal functionality of Apache place names, in that they provide a link to a place and time without need for exposition or explication (Basso, 1996). Another key element in this case is that the respondent plainly stated that these insider stories, which have clearly enriched their friendship, would not make sense and perhaps would not have been developed without the existence of the hashtag device, thereby affirming that the hashtag is a new and unique device that does not have an immediately available substitute in the English language. Also importantly, this usage has a social facet as well, in that it creates a new social group or space, containing a time-related dimension, that did not previously exist, and it also serves to renew and reinforce the connection within that group.
There are many more hashtags among those used by respondents and interviewees that exhibit these characteristics. Returning to the original research question of how the hashtag differentiates language actions, one may tentatively define the hashtag as a type of original and unique temporal-social linguistic device, currently residing only in online spaces, which allows users to strengthen, create, or identify with social groups in real-time or in a constructed temporal space. It serves the function of stating some or all of the following: What follows is an offer or request to be part of a social practice or group, or a statement that I am part of this practice or group that exists in a temporal space defined by myself or previously defined by others. Clearly, this research should be expanded in a more scientific manner and some of the survey glitches should be cleaned up, but it is evident that a subject for research of the hashtag as a social-linguistic device does exist.
Basso, K. (1996). Wisdom sits in places: Landscape and language among the Western Apache (Kindle DX version). Retrieved from Amazon.com.
Bonilla, Y., & Rosa, J. (2015). #Ferguson: Digital protest, hashtag ethnography, and the racial politics of social media in the United States. American Ethnologist, 42(1), 4-17.
Caleffi, P.-M. (2015, December). The ‘hashtag’: A new word or a new rule? SKASE Journal of Theoretical Linguistics, 12(2), 46-60.
Daer, A. R., Hoffman, R., & Goodman, S. (2014). Rhetorical Functions of Hashtag Forms Across Social Media Applications. Proceedings of the 32nd ACM International Conference on The Design of Communication CD-ROM (SIGDOC ’14). Article 16, p. 3 pages. New York, NY, USA: Association for Computing Machinery. doi:http://dx.doi.org/10.1145/2666216.2666231
Dwyer, N., & Marsh, S. (2014). What can the hashtag #trust tell us about how users conceptualise trust? 2014 Twelfth Annual International Conference on Privacy, Security and Trust, (pp. 398-402). Toronto, Ontario, CA. doi:https://doi.org/10.1109/PST.2014.6890966
Gonzalez-Ibanez, R., Muresan, S., & Wacholder, N. (2011). Identifying Sarcasm in Twitter: A Closer Look. Proceedings of the 49th Annual Meeting of the Association for Computational Linguistics:shortpapers – Volume 2 (HLT ’11). 2, pp. 581-586. Portland, OR, USA: Association for Computational Linguistics.
Hodges, M. (2008). Rethinking time’s arrow: Bergson, Deleuze and the anthropology of time. Anthropological Theory, 8, 399-430. doi:10.1177/1463499608096646
Kunneman, F., Liebrecht, C., van Mulken, M., & van den Bosch, A. (2015, July). Signaling sarcasm: From hyperbole to hashtag. Information Processing & Management, 51(4), 500-509. doi:https://doi.org/10.1016/j.ipm.2014.07.006
Smith, L. R., & Smith, K. D. (2012). Identity in Twitter’s Hashtag Culture: A Sport-Media-Consumption Case Study. International Journal of Sport Communication, 5, 539-557.
Unknown Director (Director). (2013). Subway Tuscan Chicken Melt TV Commercial, ‘Hashtag’ [Motion Picture]. Retrieved November 30, 2017, from https://www.ispot.tv/ad/72WW/subway-tuscan-chicken-melt-hashtag?autoplay=1
Vander Wal, T. (2007, February 2). Folksonomy Coinage and Definition. Retrieved November 30, 2017, from Folksonomy: http://vanderwal.net/folksonomy.html
Wang, X., Wei, F., Liu, X., Zhou, M., & Zhang, M. (2011). Topic sentiment analysis in twitter: a graph-based hashtag sentiment classification approach. In B. Berendt, A. de Vries, W. Fan, C. Macdonald, I. Ounis, & I. Ruthven (Ed.), Proceedings of the 20th ACM international conference on Information and knowledge management (CIKM ’11) (pp. 1031-1040). New York, NY, USA: Association for Computing Machinery. doi:https://doi.org/10.1145/2063576.2063726
Yang, G. (2016). Narrative Agency in Hashtag Activism: The Case of #BlackLivesMatter. Media and Communication, 4(4), 13-17.
Yang, L., Sun, T., Zhang, M., & Mei, Q. (2012). We Know What @You #Tag: Does the Dual Role Affect Hashtag Adoption? Proceedings of the 21st international conference on World Wide Web (WWW ’12) (pp. 261-270). New York, NY, USA: Association for Computing Machinery. doi:http://dx.doi.org/10.1145/2187836.2187872