Identity work performed on Twitter using @, hashtags and retweets

Abstract: The realm of computer mediated discourse (CMC) has evolved rapidly over the last three years. While literature has focused on CMC represented by online forums and chat rooms, the introduction of social networking sites such as Twitter have created an online environment which is far more reflective of real-life everyday talk. This paper outlines the workings of the social network tool of Twitter and examines the membership categorization functions of three discursive markers used by the community on Twitter. These markers are more commonly known as Hashtag, @ and Retweet (or RT). The paper outlines how Twitter users (Twitterers) strategically employ these markers to create online identities and membership categorization relationships.

Membership categorization and identity work achieved by discourse markers on Twitter

COMM6410 Discourse Analysis

Professor Karen Tracy

By Joanne White


Global  uptake of internet-based computer-mediated communication (CMC) has risen exponentially over the last five years. Not only has the internet changed the way we consume media, buy goods and services and communicate with people we already know in real life, CMC has developed to the point where people around the world are discovering an entirely new version of society online. The purpose of this paper is to demonstrate how discourse is used on Twitter to both builds towards, and be reflective of, a social structure. First I will outline literature concerning CMC and locate Twitter within that environment – to assist in this aim I have attached a glossary of terms to aid the layperson in understanding Twitter ‘language’ as they travel the journey of the paper.

Secondly, I will outline what discourse markers exist on Twitter, and provide examples of how these are used by Twitterers in performing identity work and as membership categorization devices to present themselves as particular types of people with particular memberships, supporting the premise that an online structure of society exists. In making this claim, I will be drawing on the tradition of ethnography through membership categorization analysis, as well as AIDA in considering how Twitterers strategically use these markers to create an identity online, and how these strategic uses suggest there exists more than simply a ‘pointless’ phatic network in social media.


Over the last five years, Internet-based CMC has seen a progression from chat rooms and forums to the more recently introduced social media platforms such as myspace and Facebook. These social media sites are designed to support content creation and sharing among their members. These types of sites have themselves morphed over time and as more people use the services. Where once users would create entire pages devoted to themselves and their activities, using text and images and linking with other people doing the same, these sites have evolved into more immediate conversation-tools which variously use text, images and video, and focus on immediate communication. These new social media tools have become known as ‘microblogging’ sites and are arguably a new breed of Web 2.0 social media. One such tool from this space is Twitter.

To date, literature dealing with the new developments in CMC are scant and in those that do exist, much time is focused on comparing early-stage social media sites (such as myspace and Facebook) with traditional forms of discourse, equating formal communication with writing and informal with verbal communication. From that standpoint, authors have used supposition and guesswork estimates what will be the next realization and impact of CMC. For example, in New Media, Networking and Phatic Culture (2008), Vincent Miller states that Twitter represents the epitome of how ‘connected presence’ leads to a ‘phatic culture’ where connections and database networks take precedence over content and narrative. Miller asserts that CMC has become focused on databases made up of bite-sized chunks of data which enable people to collect ‘friends’ and produce a network out of them based on what are essentially ‘pointless’ utterances rather than through connecting with people in the more traditional sense of ‘friendship’, that which is underpinned by substantive communication and dialogue. Miller concludes that Twitter, as the next step in social media, is a great example of a ‘phatic technology’ given its concentration on limited content. He asserts this technology builds relationships and sustains social interaction through pervasive (but non-informational) contact and intimacy, and he refers to Wittel’s (2001) work in describing the effects of this technology as a ‘flattening’ of communication.

In further examination of the idea of ‘phatic communication’,  in their work on small talk, Karen Tracy and Julie Naughton (2000) refer to the original definition of phatic communion (rather than Miller’s reference to ‘phatic communication’) from Malinowski as discourse that ‘serves to establish bonds of personal union between people … and does not serve any purpose of communicating ideas” (p. 70). In their work, Tracy and Naughton demonstrate that discourse turns that are phatic communion do in fact convey information and ideas about society, and that the work done by this is important, even if the observable content seems ‘pointless’ or superficial.

Additionally, Tracy and Naughton reflect on the work of Watzlawick, Beavin and Jackson (1967) in considering the relational level of a message, and how relationships between the communicators are reflected and reinforced by discourse. This reinforces the practice of identity work and reflects the psychology theories proposed by McClelland and McGuire more widely used in academic theory surrounding marketing and consumer behavior. Both these theorists are discussed at length in many marketing texts (for example, Webb 2005), however (with apologies for the lack of detail) they may be outlined very simply as follows:

In his Theory of Three Basic Needs, McClelland proposed that people are inherently driven by needs which fall into one of three categories:

  1. Need for achievement. These are known as ‘high flyers’. People who have a need to be seen as successful in whatever they have done. Typically these people have achieved a financial and/or personal status which demonstrates accomplishment, such as being the CEO of a large organization, or a successful artist.
  2. Affiliation needs. These are people who are known as ‘followers’ and those who want to ‘belong’. From a marketing context, the best representation of this need is those who will buy a brand to ‘belong’ to a group or fad.
  3. Power needs. These people have a need to be leaders. More commonly known as innovators or early adopters, these are the people who are going to be queuing for the latest mobile phone, for example.

Other work in understanding human behavior includes McGuire’s Internal and External Motives Theory, also discussed in marketing theory (and in particular at length by Webb). McGuire’s theory outlines self-concept (how people see themselves) as being the primary motivation for how people present themselves and interact with others. McGuire states that individuals have separate internal and external motives. External motives are aimed at positioning oneself in society, and internal motives are focused on the sense of self, or self-assurance. The relationship between these psychological theories and the work done within the realm of discourse, particularly in CMC, is shown by the following quote in Tracy and Naughton: ‘Relationship statements are about one or several of the following assertions: “This is how I see myself… this is how I see you… this is how I see you seeing me” and so forth ‘(p.52).

In further consideration of the role of discourse in reflecting notions of self-concept and how people are perceived, Bamberg and Georgakopoulou (2008) argue that while an overview of identity is appropriate, more detailed examination of the individual utterances can show work done to create that overall impression. While these authors were considering the conversation of young boys who were seeking to create an identity within their friends, this claim can readily be transferred to the identity work undertaken in CMC.

These theories and research provide a framework within which I am attempting to undertake a cohesive discourse analysis of the Twitter environment in order to ascertain if purposeful attempts to create and share identities and memberships are sought by people on Twitter. Furthermore, the level of satisfaction discovered by people on Twitter is directly related to how efficiently they create an identity which is honest and transparent to their purpose for being there (Pornsakulvanich, Haridakis & Rubin, 2008), and this goal-directed behavior can be analyzed on a more detailed level through the examination of the structure of individual utterances, known on Twitter as tweets.


The tie between identity work and membership categorizations is unquestionable. One relies on, and serves, the other. This is particularly evident in CMC when the goal of discourse is to build a social network and presence. In social media, as anywhere else, ‘In establishing a self, one seeks to establish a separateness. However, a clear and precise separation can only be established by identifying with something and disidentifying with something else (italics in original) (Perinbanayagam, 1991, p.13).’

On Twitter in particular, it can be shown that membership categorization is sought because it is an efficient way of presenting oneself to the online community. Membership categorization has been described as ‘inference rich’ (Tracy, 2002, p. 58). A single membership categorization is often heavily loaded with an understanding of a person’s ideology, outlook and attitude. In real life, discourse is just one factor which builds identity, and one important aspect of that discourse is accent, word choices and intonation. Other tools used to do so include physical presentation, clothing choices, etc. On Twitter, where there are limited opportunities to exhibit your identity, leveraging an opportunity to be seen as part of a particular group or category makes discourse the primary tool. In the interests of efficiency and in the face of the limited tools they can use, a Twitterer can strategically use membership categorization devices and discourse markers to perform their identity work.


The Twitter Fan Wiki (2008) describes Twitter as follows:

Twitter is an online social media application launched to the general public in 2006. There is no cost for subscription. After a quick sign-up members can search for other members using whatever identifier they desire (eg: location, interest), and select to ‘follow’ them. The members you follow then have their ‘Twitter streams’ fed through a single frame onscreen, headlined with the words, ‘What are you doing?’ Each message members produce in Twitter has a 140-character limit. This compels users to keep their messages short and immediate, and encourages conversations rather than simply the broadcasting of opinions, reflecting the social aspect of the medium. The Twitter community strongly supports the tool’s use in this context and supports new members in becoming familiar with the etiquette of the community.

Twitter is an ‘in the cloud’ application, which means it exists on the web and no part of its framework needs to actually be downloaded. It can be accessed through numerous devices, including computers and mobile phones, as well as through various third-party applications which allow Twitter’s information to be presented in various formats. When Twitter is used from the standard web-based home page, a member’s page is quite fluid, with the ‘stream’ of posts flowing from top to bottom under the heading banner. A Twitterer’s post will appear in a very structured format – first their avatar (a thumbnail image that represents the Twitterer), their handle (name they are known by on Twitter), and then their 140-character or less message. In the web-based Twitter home page these posts flow asynchronously, one after the other, based purely on time they are posted. (It is worth noting that some third-party applications exist which can help aggregate and arrange a Twitterer’s stream, such as TweetDeck, at Aspects of tweets are highlighted and colored differently to the rest of the timeline – the Twitterers’ handles (online names) and any web links the Twitterer chooses to provide as part of their message. An image of the screen presented by the home page of Twitter on a normal computer screen is shown below:

Twitter has attained an extraordinary focus in traditional media, particularly over the last six months as celebrities and mainstream media both use the application and speak about it. This particular social media platform has grown at an exponential rate, with Nielsen (2009) reporting its growth in just one year from February 2008 to February 2009 of 1,382% and describing it as ‘no longer just a platform for friends to stay connected in real time, it has evolved into an important component of brand marketing’. Twitter’s largest user base falls in the adult category, aged 35-49. Nearly 3 million unique visitors within that age group (42% of the total audience) visited Twitter in February 2009.

As for any new technology, those we might call ‘early adopters’ were the first regular users of Twitter. However, as celebrities who are known around the world such as Ashton Kutcher (, Oprah Winfrey (, Stephen Fry (, Demetria Lovato (, Miley Cyrus ( and many others create their own Twitter accounts and actively engage in public conversations with other Twitterers, the general population is following them online. As a result of these celebrities and their fans moving onto Twitter, an increase in Twitterers more reflective of general society than of just the technical wiz-kids, including all ages and interests, is rising very quickly.

It is also extremely important to note that when Twitter was first developed, it was not planned to be a conversational platform. Instead, the developers believed it would offer people an opportunity to see what other people were doing – a type of broadcasting tool that would let users tell the world a message, and not rely on any turn taking at all (Williams, 2008). That premise reflects other social media popular at the time of its launch such as myspace and Facebook, which focus on the individual using the service and visiting other peoples’ ‘spaces’ instead of regular interaction as part of a community. Over time, however, the people on Twitter have developed it to the conversational social media platform it is today. According to the research of Honeycutt & Herring (2009), nearly half the people using Twitter do not actually directly respond to the query, “What are you doing?” and instead use Twitter to communicate either as part of a direct or broader conversation, or to provide  information delivery and links to areas of interest or stories on the web. This morphing of its use over such a short term is very interesting in that it suggests that people seek community online rather than simply a presence in the space.

Twitter has a very limited range of official rules, which has meant it is very much community-directed and controlled. Unlike CMC of the past, there are no moderators to approve content. Instead, people follow others based on what they believe the value is they offer to their own interests. Ultimately, the number of followers a Twitterer has can be considered one of the criteria for value in following that Twitterer. (However, it should be noted that some third-party applications have been created that seek to create connections with as many people as possible in an attempt to make a Twitterer look more ‘popular’. These appear to be ‘frowned’ upon by Twitterers other than multi-level marketers). With the most follower-rich Twitterers having well in excess of a million ‘followers’, Twitter’s deceptively simple framework may well be representative of a completely new channel of communicating and building relationships in the 21st century.

For the purposes of this paper, I am drawing specifically on the tradition of AIDA to identify the strategic use of discourse markers and how they perform work in the formation of online identities and memberships. I believe this will provide basis for further research in addressing what I am calling the latent sphere of community that I propose exists in CMC. The data I have used have been taken from two places. I have done this to demonstrate that the discourse practices concerned are not purely focused on one type of discourse or subject area. Firstly, I am randomly taking posts from general posts in the public Twitter stream over a one-week period in April, 2009. These were easily obtained simply by browsing Twitter’s public interface at after signing up. Secondly, I wanted to use data focused on a specific topic. This topic-specific data would best show Twitterers’ focus within the topic they were discussing and allow a balanced playing field in which to perform detailed analysis of the hashtag discourse marker. I wanted to see how this might relate to identity work within a membership category. I decided to search Twitter for tweets about bushfires at the height of the Australian bushfire tragedy earlier this year, when over 200 people in rural Victoria died in devastating bushfires that attracted worldwide public attention. I collected an aggregation of posts using the hashtag discourse marker of #bushfires. Then in order to obtain a focused number of tweets at the height of the topic of bushfires, I used a trending application for Twitter called Twist ( and found there was a very high reference to the bushfires between the dates of February 8 and 10, 2009. I then used the Twitter search application to locate and list all the #bushfire tweets during that period. For the purpose of this paper, I then decided to focus on 50 tweets made on February 10 2009 between 3.46pm and 7.38pm (early evening, Australian time) as my sample. This amounted to tweets from a total of 39 Twitterers that I could draw on.


Many traits of membership categorization and social identity researched in the oral discourse area are also able to be identified in Twitter. Membership categories, as defined by Sacks and outlined by Hester and Eglin (1997) are ‘classifications of social types that may be used to describe persons’ (p.12). For example, common membership categories could be mothers, firemen, etc.  Demonstration of work done by individuals to fit a membership category include behavioral and presentation by way of dress, as well as use of language. For example, in real life a group of women may discuss topics such as bottle feeding, babysitting and husbands, standing outside a school gate at about 3pm. These aspects work together to inform an observer that these women belong to the membership category of mothers. The members of the group themselves are also able to identify these characteristics to find affiliation with other people sharing the same membership – demonstrating the fulfillment of an affiliation need. More specifically, in looking at the use of discourse Hester & Eglin describe the ethnomethodological perspective as follows:

“… a possible reading of MCDs (Membership Categorization Devices) is that they comprise pre-existing structures of category-organized knowledge which flesh out, or contextualize, componential-analytic accounts of the semantic structure of language. (p.14)”

That is to say discourse serves to both contextualize and define the membership category as well as identify the person’s membership within it.

On Twitter, with conversation in the form of written discourse being the primary method of communication, membership categorization relies on text-based tweets. Within tweets, topics which are of mutual concern to the group category help define membership. There is an interrelationship in play between identity work and membership. This is shown through an examination of the overall discourse on Twitter as well as the identity work done through a more detailed analysis of the components of individual utterances demonstrated by the discourse markers used in tweets. Discourse markers indicate strategic moves taken in discourse which frame the direction and attitude in which that discourse is delivered. These can be found on Twitter, just as they can be identified in everyday talk.


One way Twitterers are able to perform identity work and align themselves with a particular membership categorization is by only tweeting on a particular topic.  For example, some Twitterers focus on a profession, and only contribute information within that topic area. One example of this practice is @Jay_Rosen ( Jay is a university professor of journalism at New York University. Jay’s Twitter profile reveals he is not lifestreaming, but rather mind-casting, and every tweet he makes is directly linked to the topic of journalism and media. Alternatively, others will ‘lifestream’ – tweet about anything and everything that interests them throughout the day. Either way, the topics they tweet about strategically serve to both identify and align themselves with a membership category as well as perform identity work, and define their sense of self-concept, both to themselves and how others see them. If a decision is made to decide only to ever tweet about professional areas, as @Jay_Rosen does, then it reduces the possibility for Twitter to become a personal tool and for deeper friendships to be made. This decision appears to have resonance with Miller’s idea of phatic communion.

It is interesting that while Twitter’s structure includes a message length of 140 characters or less, the majority of people do not shorten their language, as has happened in short messaging or mobile phone text CMC (also known as SMS communication). It is notable that the difference between mobile phone communication and Twitter is the larger community, including a large proportion of people previously not met in person by a Twitterer. It could be suggested that this decision to use the longer form of words in this restricted environment directly corresponds to the need to construct credibility in identity and membership categorization in a public space. It may seem that shortened, SMS-style writing portrays the author as a lower class or less educated to a public audience than those who write with standard English. This choice reflects on how the author is seen by the reader, and also how the author sees themself.


It has been argued that discourse markers rarely appear in written texts (for example, Cameron, 2001), however their predominance in the written text of Twitter supports the argument that the tool has morphed into one primarily focused on conversation and interaction rather than of a traditional understanding of writing. There are myriad discourse markers on Twitter, just as there are in everyday talk. For the purposes of this paper I am considering three which are standardized and regularly used by people as part of their twittering: @, # and RT (or retweet).

Luvschweetheart@Mediamum That is fantastic!! I am going to pass that around, I honestly never had heard of that. Thanks for sharing with me.

Karoli@rjamestaylor I’m loving the flip mino too. lots of fun.

pugofwar@aamore Glad your health has returned.


Karoli@rjamestaylor I’m loving the flip mino too. lots of fun.

pugofwar@aamore Glad your health has returned.


The use of @ in CMC directs the rest of the tweet that follows it to a particular Twitterer, although it remains in the public timeline, so all followers of the authoring Twitterer can see the tweet. In the timeline, the author’s handle is first and the Twitterer to whom the tweet is specifically directed is named immediately after the @ symbol (as shown above). Twitter did not begin with this convention in place. It is one that has developed and become standard over the three years Twitter has been publicly available, and it is the practice most obviously connected to gaze and turn taking seen in everyday talk on Twitter. The origin of the @ tag has been identified by Werry as a legacy of another form of CMC, Internet Relay Chat (IRC). Werry also notes that a high level of addressivity is required in multi-participant CMC environments (such as Twitter) because the addressee’s attention must be recaptured with every utterance (Honeycutt & Herring, 2009). Whereas gaze is easily undertaken separate to the actual discourse in a real life conversation, in an online environment the function of gaze must be discursively demonstrated. Furthermore, the need to include this addressivity function within the limited range of characters available demonstrates Twitterers’ belief of the importance of its inclusion for effective communication – they appear to be willing to sacrifice up to approximately 10% of their actual content purely for addressivity purposes.

There are numerous identity implications of address which come with using @ on Twitter. Given that the environment is one which presents a single space to converse with the entire Twitter community, using @ treats every user in the same way, no matter how many followers or what their perceived status might be. There is no more formal method of address in existence. In fact, the only way to adjust the level of formality is to state someone’s real name as part of the tweet after using the @, as shown in the examples below:

marcsilber @scoble hey Robert–just put up a new Tip with footage of Annie Leibovitz

Ladywolf55 @aplusk Thanks Ashton. Is it better than Tweetdeck, IYO?

pugofwar Ef, don’t pretend every day with us is not the best day of your life.

Adding the Twitterer’s real name in addition to what has become a standardized addressivity of @ gives a greater sense of familiarity and personalization to the utterance.

The use of the @ also performs membership categorization work. In this online space, if one Twitterer @s particular Twitterers and manages to engage them in conversation built on what actually becomes turn taking represented by @s in a to and fro manner on a regular basis, then the general community will associate the Twitterers as being of similar status, credibility and knowledge – often resulting in an increase in followers by those with affiliation needs. Many Twitterers can be observed attempting to do this, particularly trying to engage conversation with well known celebrities on Twitter. Typical tweets showing this strategy directed at the actress Demi Moore (mrskutcher) are shown here:

SoCaSunshine @mrskutcher Can you tell me the cut of your engagement ring? Its unique & beautiful 🙂

AmSo78 @mrskutcher Hi Demi! I’ve been having a prob with fine lines around my eyes. Have bought lots of eye cream! What do you recommend?

In both of these tweets, placing @mrskutcher at the very start shows the twitterer’s strategy of conscious addressivity. The primary purpose of the utterance is to attract the attention of mrskutcher. The content of both utterances demonstrate familiarity. The tweet from SoCaSunshine relating to the cut of mrskutcher’s engagement ring relates to a very personal item. The addition of the ‘smiley’ emoticon at the end provides a happy invitation for a reply – just as a smile in a real life conversation would. The identity work performed here is one where SoCaSunshine wants to be viewed by mrskutcher as a caring, personable individual who is genuinely interested in something she believes mrskutcher also holds dear – her engagement ring.

In the second example, AmSo78 also uses an informal mode of address. She adds familiarity to the utterance with ‘Hi Demi!’ at the beginning of the tweet, which infers friendship. The topic of skin care is one which AmSo78 believes mrskutcher will also identify with. The strategy of including the words, ‘Ive been having a prob with fine lines around my eyes. Have bought lots of eye cream!’ positions AmSo78 as someone who is not knowledgeable on the topic, yet familiar enough with the celebrity to ask. The invitation for a response is strategically accomplished with the final question, ‘What do you recommend?’

Identity work is similarly achieved if the high profile Twitterer responds. There are numerous high profile people on Twitter. Some are publicly known as celebrities (such as those already mentioned) and others are considered celebrities purely for their niche prominence in the technological space. One such person is Robert Scoble (scobleizer at Robert Scoble is a very high profile technical commentator, and a key ‘leading edge’ source of information for the technically-savvy. Two tweets from his stream are as follows:

@Rumford I’ll try to make next week. Sounds like a great event.

@tonyweeg boring? Hah! We should debate that on the Gillmor Gang with you someday.… is far from boring.

Identity work is achieved in the above tweets by Rumford and tonyweeg, simply because they are being addressed by scobleizer. This work is publicly shown as part of the twitter stream of anyone following scobleizer (currently over 80,000 people). The identity sought by each Twitterer is demonstrated by the membership categorization sought – to enter into discourse with scobleizer is to identify yourself as a technically knowledgeable Twitterer. And it demonstrates the membership categorization stake held in either the Twitterer or the content the Twitterer is providing being credible or important enough for scobleizer to reply to. These both represent the fulfillment of a power need.

In scobleizer’s reply to Rumford, identity work is achieved through scobleizer’s endorsement of Rumford’s event. By providing a weblink within the utterance for his followers to click on and find out more, along with the recommendation that it ‘Sounds like a great event’ provides a level of approval which is extended to Rumford as well. In the second example, identity work is also achieved at the same level, even though scobleizer appears to be chastising tonyweeg over an accusation that something was boring. The membership categorization of ‘technological wiz’ is confirmed by scobleizer’s invitation for tonyweeg to join him on his program, the Gillmor Gang (which consists of a panel of technical experts discussing current information technology trends and is featured in an online format). The identity work for both scobleizer and tonyweeg that is achieved shows a synchronicity where they both reaffirm identities as experts in technology, and both fulfil power needs. The need for achievement is reflected in scobleizer’s reference to his own program – ‘We should debate that on the Gillmor Gang someday’. Here scobleizer is publicly stating his own accomplishments. Instead of simply offering to debate it one day, or discuss it further offline, scobleizer’s inclusion of his recognized program adds credibility and accomplishment to his profile as well as his invitation. As a result, both scobleizer and tonyweeg benefit from a rise in status.

These examples demonstrate the attraction of celebrity of two different forms (movie stars and technological commentators). It is apparent that a different type of membership categorization and identity is sought by the different type of Twitterers who are likely to undertake discourse with these two very different types of celebrities, and this was reflected in the type of content they tweeted.


Hashtags, symbolized by the # character, are used by Twitterers to group tweets by topic areas, and are most obviously used as membership categorization devices (MCDs). Just as the use of the @ was something developed over time on Twitter, hashtags were also a later addition to the framing of discourse. They are usually used by Twitterers who are confident with Twitter to the extent that they can see functionality beyond the 140-character limit. Hashtags were adopted by Twitterers due to a desire to classify tweets on a common topic, and to make this classification clear in the public timeline. Various Twitter aggregator programs allow searches on the global Twitter stream for particular key words and ‘trending topics’. More recently, Twitter has made an adjustment to include this function within the home page of its web-based application, so it is now easier for Twitterers to track trending topics than previously, when they needed to move away from the home page to track the trend. The inclusion of a hashtag in your tweet directs people for what key word to search for. Most Twitterers seem to understand that the inclusion of the hashtag as a discourse marker is not actually necessary for the aggregator tools to pick up on the trending topic and therefore is not needed at all in the conversation. However, the hashtag operates on a different level within the discourse. The hashtag allows membership to a trending topic with Twitterers who are not part of your own Twitter stream. Through a global network of people talking about the same trending topic area, the hashtag identifier creates an inclusion to a separate group – even if it is one in which Twitterers have never conversed before. The hashtag also identifies to a regular stream of followers that a Twitterer is currently involved in a trending topic, which may or may not be something they’d also like to partake in. For the purposes of considering hashtags and how they are used as MCDs, I have drawn on the tweets I collected specifically dealing with the Australian bushfires.

It is possible to analyze the placement of the hashtag within a tweet. The marker can be placed at the very start of the tweet such as:

Rexster: #bushfires arranged a tour of Flowerdale for some touring bikies, they loved it.

When this is done, the placement is performing the work of addressivity. We have considered the way the @ tag points the tweet that follows it to another Twitterer in the public space. In this case the hashtag performs that same role of addressivity, except it points to a topic area rather than an individual. In the above example, it is shown that Rexster is addressing the trending topic of bushfires.

The placement of the hashtag within a tweet may indicate the importance and focus a Twitterer has in its inclusion. It is the most important subject of the tweet, and this relates specifically to identity work. Placing the hashtag at the start of a tweet classifies the words following it, and also the Twitterer, as having as their focus a particular membership categorization – in this case, people concerned about the bushfires. Similar to verbally saying, ‘I’m going to talk about bushfires’ and then following it with a comment on that topic area, it presents the information which follows as having a particular attachment, one which is shared by numerous others because it is a trending topic. The hashtag is not always placed at the start of a tweet. It is interesting to look at where else the hashtag is placed and analyse what identity work this change in structure might do.

The hashtag is also regularly placed within the tweet as part of the content itself. For example:

ausil: @peace yes all the #bushfires threats are over for now. Thanks for asking

The inclusion of a hashtag within the body of a tweet offers a membership categorization which is content-focused. Rather than talking around a trending topic area, the Twitterer is making a specific comment which includes the trending topic and it demonstrates a high level of involvement with the trending topic, possibly ‘insider knowledge’. Therefore we can say that the membership categorization is inherent within the content exchanged. Additionally, the placement of the hashtag in this position indicates the Twitterer is a confident user of Twitter and its frame of communication. Being able to contain the discourse marker within the actual content of the tweet saves space but doesn’t detract from the message. This could be compared with a person in real life having a broad vocabulary, or clever turn of phrase. It builds the identity of the Twitterer as well as their membership categorization.

Rosshill: @steveouttrim thanks for the xbox360’s for flowerdale – you’re a legend! @rexster will deliver them soon #bushfires

Finally, the hashtag is also regularly placed at the end of a tweet, such as:

Placing the hashtag discourse marker at the end of the tweet performs different identity work. While it still enables membership, this time the trending topic is not the primary focus of the tweet. The trending topic is not the foremost thought in the Twitterer’s mind. In the above tweet it is demonstrated that the Twitterer’s thanks is the subject and therefore most important part of the utterance. The addition of the hashtag at the conclusion appears to be almost an afterthought, aimed to contextualize the content of the tweet.


On Twitter, it is common practice to RT (or retweet) another Twitterer’s content. The etiquette involved in doing so consists of typing the RT marker (or even sometimes the complete word, ‘Retweet’) followed by the @handle of the originating Twitterer. For example,

asgerd RT @MisfortnCookie When the disappointments in life pile up, there will still be more.

mrskutcher RT @Cryptogenetic: R daughter Cassie Eyre died April 25th at 4:35 pm from Ecstasy. Join ‘Cassie Eyre’s Remembrance’ on FB Rasise Awareness

As seen above, RTs can be used to simply broadcasting information, which can be viewed an action that fulfills a Twitterer’s power and affilitiation needs. This is where the Twitterer is able to associate themselves with a) the information being broadcast (and which they have found very worthwhile to the extent they will RT it), and b) the originating Twitterer, whom they will reference as part of the tweet.

Instead of just broadcasting generally, the RT can also be used as part of a conversation when addressivity is included. Here the tweet is directed to a particular Twitterer using the @, and then the RT as usual is added. An example is:

geniusboywonder @freever RT @amazonmp3 BTW, here’s our editors’ picks for the 100 greatest jazz albums of all time, topped by today’s Daily Deal:

This strategy personalizes the utterance as something which is especially relevant or important to a particular Twitterer, but which comes from elsewhere in a Twitterer’s stream. In the above example, you can imagine that freever had perhaps been talking about jazz in recent tweets, and that geniusboywonder is referring freever to amazonmp3’s information as a response.

Having your content retweeted on Twitter raises your profile in the community as being someone who has valuable, relevant content. Retweeting is therefore key in identity work. Retweeting someone’s tweet can be seen as a Twitter’s version of a compliment. As an aspect of identity work, compliments ‘build’ the self. On Twitter, compliments build status and social creditability of a Twitterer within the community, and also as a reassurance in their own sense of self concept and alignment with membership categorization. For example, regularly retweeting Amber Alerts and Breaking News tweets identifies someone as particularly knowledgeable and concerned about current affairs and social issues, fulfilling the affiliation need to be seen as socially responsible which ties directly to identity work and any related membership category. This contrasts someone who regularly retweets amusing links to funny pictures or videos online. A Twitterer who performs that type of identity work through the use of RT is focused more fulfilling an affiliation need in a different way, and as a result will be seen to be part of a different type of membership categorization. The more shallow or humorous retwitterer could be seen as fun-loving and ready to have a good time on Twitter rather than the more serious, apparently globally conscious retwitterer.

To consider further identity work undertaken using RT, it would appear that numerous Twitterers tweet what could be termed pithy one liners in an attempt to be retweeted. For example, @weloytty has tweeted the following:

@weloytty my parents went to mexico this morning. If they die of #swineflu, I’m going to be PISSED

If this message is retweeted by LauraMast, it would then appear in her followers’ streams as:

LauraMast RT @weloytty my parents went to mexico this morning. If they die of #swineflu, I’m going to be PISSED

The overall work done through the original tweet shows @weloytty to be a humorous person, making light of a current affairs event which has international focus. In tweeting this post, weloytty has performed a lot of identity work. He has identified himself as being aware of the location of swine flu origin (Mexico), he has used the hashtag #swineflu which is currently a trending topic (a topic lots of people are talking about), and is a savvy enough user of Twitter to include it as a key part of the content of the tweet rather than simply classifying it at the beginning or end. He then finishes the tweet with a humorous statement about his love for his parents. Therefore it can be seen that the overall humor as well as the structure of the tweet perform identity work.

LauraMast’s retweet of weloytty’s original tweet does further identity work, both for weloytty and LauraMast. Firstly, weloytty has found a new ‘audience’ if you like, by having his attributed content tweeted to LauraMast’s followers. This gives weloytty status by association with LauraMast (particularly beneficial if LauraMast has more followers, or a higher status of followers than weloytty does). Additionally, it gives LauraMast stake in the credibility of weloytty. She would appear to agree with his humour, have found it funny, and believes it is worth sharing with her own stream. An interesting more recent development is what appears to be regular thanks given by Twitterers to those who retweet their content as shown below:

skavhellen @thecowgurl @kurlicu Thanks for the RT Glad you liked it 😉

brianfeldman @mattsimantov Thanks for the #miniature RT!

This practice seems to be more and more common, and could be considered to be approaching the stage of correct etiquette on Twitter. Returning to the RT example between weloytty and LauraMast example, the thanks which follows LauraMast’s RT could appear as a post such as:

weloytty @LauraMast Thanks for the RT!

This leads us to consider if it is possible to compliment (or retweet) too often? In reflection of the frame outlined by Perinbanayagam (p.119), it is possible to overuse compliments, and to do so lessens the authenticity of the compliment. On Twitter, a Twitterer who does little else than retweeting others’ content can be viewed by the community as shallow and inauthentic rather than complimentary. In fact, the retweets, even though attributed, can be seen to be an attempt to leverage popularity based on someone else’s content, and actually an impolite action.


Just as discursive markers perform work in everyday talk, similar work can be shown through the discourse markers of @, hashtag and retweet on Twitter. This work can be aligned with accomplishing the particular identity needs of power, affiliation and leadership. There are many inviting possibilities for research of discourse and identity work on Twitter. People present themselves and define themselves in online social media, and particularly on Twitter in many ways. For example, research from the CA tradition into what profile information (the background detail of a Twitterer) to post in the limited space, the avatar a Twitter chooses to use and the way a Twitterer decorates their background page and how this speaks to the identity and membership category one seeks would be a wonderful addition to research in this dynamic sphere.

I would assert that for Twitter to be considered a ‘phatic technology’ which supports a reduction in the depth of community and a ‘flattening of social ties’ as Miller and Wittel suggest, there would be minimal ongoing identity work or membership categorization identifiable on Twitter. Just as was apparent in earlier social media such as myspace and Facebook, the users’ presence would be largely static. However, as this analysis has demonstrated, on Twitter this is not the case. The morphing of Twitter from a straightforward broadcasting tool at its inception through to the fully-functioning conversational and sharing tool it has become reflects the strength and influence of the community which uses it. This suggests there is a continuing depth of community sought by people and this new breed of social media is being used to address this need. To offer further support for this claim, Huberman et al also identify that even though Twitter does not compel users to reciprocate following status, 90% of Twitterers do in fact return following status, and there is a great predominance of conversation rather than pure information broadcasting (2009, p. 2).

This study indicate further research is necessary to explore the existence of what the authors call a ‘hidden social network’ which exists within a latent sphere beyond the tweets – the online community. The discursive markers @, hashtag and retweet are strategically used to perform both identity work and act as membership categorization devices. Further work acknowledging the myriad other practices of Twitterers in this space, their identity constructions and analysis of exactly how discourse practice has changed over time on social media will better reflect this potential revolution of standard CMC as its user base becomes even more recognized through continuing widespread public acceptance and growth of Twitter.


Bamberg, M., & Georgakopoulou, A. (2008). Small stories as a new perspective in narrative and identity analysis. Text & Talk – An Interdisciplinary Journal of Language, Discourse Communication Studies, 28, 377-396.

Cameron, D. (2001). Working with spoken discourse. London: Sage.

Hester, S., & Eglin, P. (Eds.), (1997). Culture in action: Studies in membership categorization analysis. Lanham, MD: University Press of America [chapter 1].

Honeycutt, C., and Herring, S. C. (2009). Beyond microblogging: Conversation and collaboration via Twitter. Proceedings of the Forty-Second Hawai’i International Conference on System Sciences. Los Alamitos, CA: IEEE Press.

Huberman, B., Romero, D., Wu, F. (2009). Social networks that matter: Twitter under the microscope. First Monday, Vol 14(1).

McClelland, D. (1958). Methods of Measuring Human Motivation, in John W. Atkinson, (Ed.), Motives in Fantasy, Action and Society. Princeton, N.J.: D. Van Nos-trand, pp. 12-13.

McGiboney, Michelle, Twitter’s Tweet Smell of Success. (2009, March 18). Neilsen Wire. Retrieved April 29, 2009 from

Miller, V. (2008). New Media, Networking and Phatic Culture. Convergence: The International Journal of Research into New Media Technologies Vol 14(4):387-400.

Perinbanayagam, R.S. (1991). Discursive Acts. New York: Aldine De Gruyter.

Pornsakulvanich, V., Haridakis, P., Rubin, A. (2008). The influence of dispositions and internet motivation on online communication satisfaction and relationship closeness. Computers in Human Behavior, 24, 2292-2310.

Tracy, K. (2002). Everyday talk: building and reflecting identities. New York: Guilford.

Tracy, K., and Naughton, J. (2000). Institutional identity-work: a better lens. In J. Coupland (Ed.) Small Talk. Essex, England: Pearson Education Limited.

Twitter Fan Wiki. (2008). Twitter Etiquette. Retrieved 15 November, 2008 from

Webb, K. (2005) Consumer Behaviour. Sydney, Australia: McGraw-Hill Australia Limited.

Williams, E. (2008) The origin and evolution of twitter. Speech at TED conference, accessed April 28, 2009 via the Using Twitter blog. Retrieved from

Williams, J., Sligo, F., Wallace, C. (2005) Free internet as an agent of community transformation. The Journal of Community Informatics Vol 2 (1): 53-67.

Wittel, A. (2001) Toward a Network Sociality. Theory, Culture & Society Vol 18(6): 51-76.

Twitter Glossary

@:                               When replying or ‘speaking’ directly to one or more other Twitterers, using an @<name> identifies who you are talking to, even though the stream is usually public.

#:                                 Hashtag. A marker that indicates the topic being ‘spoken’ about is one that others around Twitter are also interested in. (See also Trending Topic.)

Avatar:                        A small graphic representation of the Twitterer. This can be a photograph, illustration or other image.

DM:                             Direct message. A private way of sending tweets to people who follow you.

Favorites:                    Tweets you have ‘starred’ are aggregated in a separate window which is publicly viewable.

Followers:                   People who have connected to you so that your tweets appear in their Twitter stream.

Following:                   People you have connected to and decided to have appear in your Twitter stream.

Handle:                       The ‘name’ someone has on Twitter. When signing up to Twitter a Twitterer claims a name which they feel they would like to be called. There is only one person called each Handle on Twitter. Some decide to use their real name, others do not.

Locked Account:        There is an option to keep your twitter stream private, and only allow people you choose to view it. A padlock appears on the account names of those Twitterers.

RT (retweet):               Copying a tweet from one Twitterer and placing it within your own posts. The RT at the beginning of a tweet indicates it was posted by another Twitterer first.

Stream:                        A shortened version of Twitter Stream (See Twitter Stream.)

Tweet:                         A 140-character or less message posted to Twitter. Twitter’s version of an utterance.

Twitter:                       An online global social network which invites its members to respond to the question, “What are you doing?” in 140 characters or less. Twitter represents one tool in the realm known as microblogging – a miniaturized version of blogging which is focused on conversation.

Twitterer:                    A person or entity posting messages to Twitter.

Twitter stream:            The list of tweets coming through your Twitter page.

Tweeple:                      People on Twitter. Twitterers may refer to ‘my tweeple’ which is a way of categorizing the people who follow them.

Trending Topic:           A subject that many people are discussing on Twitter, normally identifiable by an associated hashtag (#). Trending topics usually have a heightened tweet ratio over 24-36 hours, and then fall away. Identifying a tweet as part of a trending topic indicates it belongs in context with the larger discussion.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s