All the Web’s a Stage: The Effects of Design and Modality on Youth Performances of Identity

Abstract 
Purpose 
Online Social Networking Sites (SNSs) such as Facebook and Twitter have become increasingly popular in the last decade. Each SNS varies somewhat, with different forms of expression, communication and customization. Different sites may have different priorities, methods of interacting, social features and definitions of what it means to be ‘social’ on their sites. 
 
 
Methodology/approach 
This paper reports on 2 months of exploratory observations and interviews with participants using two of the most popular SNSs; Facebook.com and Twitter.com. Paying attention to the modal nuances of the sites and their effect on social interaction and identity portrayal, the focus of analysis is upon how these two sites are interacted with as ‘stages’ for identity performances, and how the varying aspects of design and modality on these interactive sites can result in different multimodal identity performances and social interactions. 
 
 
Findings 
Data revealed that youth are adeptly able to negotiate the different modal options presented to them online, yet the temporal aspects presented by the design of the site, the differing definitions and priorities in the framing of identity presented by the SNSs, and the modal choices present across the two sites resulted in markedly differing presentations of identity to markedly differing audiences. 
 
 
Originality/value 
This research demonstrates the impact of modality and design on how we act and interact, and highlights that as Digital Sociologists and Researchers, we should be careful not to treat all Online SNSs the same, but pay attention to the plethora of nuances these sites offer as stages for identity performances.

Findings À Data revealed that youth are adeptly able to negotiate the different modal options presented to them online, yet the temporal aspects presented by the design of the site, the differing definitions and priorities in the framing of identity presented by the SNSs, and the modal choices present across the two sites resulted in markedly differing presentations of identity to markedly differing audiences.
Originality/value À This research demonstrates the impact of modality and design on how we act and interact, and highlights that as Digital Sociologists and Researchers, we should be careful not to treat all Online SNSs the same, but pay attention to the plethora of nuances these sites offer as stages for identity performances.
Much has been made and written in the last decade of youth engagement with social networking sites, and rightly so. They are increasingly a part of daily social life; they are accessed through a growing range of technologies; they are accessible at home, on the bus, at work, and at school; and they not only reflecting current social practices, but are creating and encouraging new and novel social methods of action and interaction (Miller, 2011;Wilson, Gosling, & Graham, 2012). Facebook claims as of 2014 to have 1.35 billion active users, 864 million of whom log on daily, a 19% year over year rise (Facebook.com Newsroom, 2014). Similarly, Twitter claims as of 2014 that 500 million tweets are sent daily, with 284 million active users on the site (Twitter.com About, 2014). The increasing popularity, ubiquity, and variety of online Social Networking Sites (SNSs) means that an understanding of their uses and functions has become increasingly vital for many diverse fields of research, and that an understanding of the importance of design and modal choices upon social interaction is crucial. This paper adds to the growing body of knowledge surrounding youth engagement with SNSs to consider the importance of the many modal choices offered across the increasingly broad array of SNSs for youth to express and portray identity. Whilst research is progressively understanding these sites and spaces as increasingly important locations of social interaction for youth (Bennett & Robards, 2014;boyd, 2014, 2007, Wilson, Gosling and Graham in their review of current Facebook research in the Social Sciences (Wilson et al., 2012, p. 204) point out that '… there are dangers in treating (SNSs) as a single general category without drawing distinctions among them'. Despite the increasing variety and diversity of SNSs, little consideration is currently being given to the important role that the arrangement of these social arenas plays in shaping how youth interact, and how they construct and portray identity. As such, this research pays close attention to the design aspects and modal nuances of two prominent SNSs to understand how the layout, design and modal affordances can affect participation, identity construction, social interaction and social navigation online. This paper begins by discussing the importance of SNSs in daily social life for youth, both online and offline, before moving on to defining identity performance and the role of multimodality online. The findings are then presented, with the research finding that the differences in design across the two sites affected the roles the participants found themselves fulfilling, the audience they felt they were performing for, and the way and modes through which the participants acted, interacted, and described themselves.

WHY YOUTH SOCIAL NETWORKING SITES
Much of the research conducted on SNSs has focused on the actions and interactions of youth. There are several reasons for this, the most often cited being that they make up a large portion of the active users of SNS. The proliferation of youth in SNSs has been noted by a number of researchers, for example Lenhart et al. (2011) found that Facebook, a site founded specifically for college students, was being utilised as a site for social networking by 93% of SNS users aged between 12 and 17. Livingstone, Mascheroni, and Murru (2011) further showed that this trend was not specific to Facebook, recording the large proportion of youth across a wide range of SNSs, confirming the findings of Baker and White (2010, p. 2), who proclaimed that 'adolescents represent the largest and fastest growing demographic sector using the Internet'. However, it is not just the popularity of these sites that is important to the youth population, but rather the increasingly important role that this technology plays in modern social life. Researchers such as boyd and Ellison (2007), Lenhart and Madden (2007), Baker and White (2010), and Livingstone, Mascheroni, et al. (2011) have all noted that SNSs are increasingly important locations of social interactions for large numbers of youth, and that these sites play an integral role in how young people socialise, interact and form identities. Quinn and Oldmeadow (2013) attempted to confirm the increasingly important role of technologies and SNSs in young people's social lives by conducting a statistical analysis into how socially connected young people felt, noticing a strong positive correlation between mobile phone and SNS use, and the sense of social connectedness young people felt. Again, this is not specific to one site, and instead is a trend that has be reported across a number of sites. For example, in their study of gender differences on Twitter, Pujazon-Zazik and Park (2010, p. 78) noted that 'an important developmental task of adolescence is identity formation', and that this development was increasingly happening online. boyd (2014, p. 26) sums up the important and ubiquitous nature of SNSs for youth, stating that 'along with planes, running water, electricity, and motorized transportation, the internet is now a fundamental fact of modern life'. For the current youth generation, SNSs are increasingly important and useful tools for navigating the social landscape. So much so that in their report on a questionnaire of nearly 100 students, Mazzoni and Iannone (2014, p. 303) define SNSs as 'part of the functional organs that support emerging adults in their ability to connect and to be connected to a social network and to develop and maintain it over time'. SNSs have become another critical tool for peer-to-peer social interaction, and are increasingly playing a larger role as technology develops and becomes more common place in the lives of many young people.
The importance of SNSs for youth can in part be shown in the amount of time they spend on these sites. Agosto and Abbas (2013, p. 137) note that it is important 'to study youth as a separate sub-group, as there appear to be differences in the ways young people and adults use SNSs'. They noted that while there was little difference in the general subject matters discussed by this generation as compared to previous generations, the amount and the frequency of these interaction and activities differs greatly, and is likely to continue to grow given the popularity of these sites, and the growing ubiquity of technologies. This matches findings by other researchers who noted the increased amount of online engagement by younger generations (Christofides, Muise, & Desmarais, 2012b;Livingstone, Kirwil, Ponte, & Staksrud, 2014;O'Keeffe, Clarke-Pearson, & Council on Communications and Media., 2011;Palfrey & Gasser, 2013;Ralph, Berglas, Schwartz, & Brindis, 2011;Rideout, Foehr, & Roberts, 2010). Indeed, the amount of time the youth spend engaging with this technology lead Clark (2005, p. 203) to define the emerging youth generation as the 'constant contact generation'.
An important aspect of SNSs for youth social identity development is the potential they offer for adolescents to create a space of their own. boyd (2007) suggests that the availability of youth-controlled spaces to explore aspects of social identity away from adults is of essential importance to today's youth, and is a main aspect of why adolescents use SNSs. She points out that 'what is unique about the Internet is that it allows teens to participate in unregulated publics while located in adult-regulated physical spaces such as homes and schools ' (boyd, 2007, p. 21). Though admittedly much has changed in the SNS landscape since 2007, including a noted growth in adult users (Madden, 2010), Agosto and Abbas (2013, p. 117) make similar observations, stating that 'SNS provide youth-shared spaces for interaction and communication'. The implication here is that SNSs create a space for youths free from adults in which to explore and socially interact, allowing them to establish their own social rules and ideals. This suggestion however is called into question given the recent and increasing influx of adults onto popular sites such as Facebook and Twitter (IStrategyLabs, 2014;Christofides, Muise, & Desmarais, 2012a;Madden, 2010). Some researchers have suggested that this has led to an exodus of teenagers from Facebook as they seek other site to explore and claim as their own (Miller, 2013). This trend has been spotted as early as the beginning of this decade (Van Grove, 2010). While youth are still numerically high on these sites, the increasing adult presence may have some effect on how SNSs are used as sites of identity development and exploration for youth in the future, a topic worthy of further research and exploration.
It is worth noting that the use of SNSs for social interactions does not happen independently of existing social ideals, actions and interactions. In recent research there has been a shift towards conceptualising and grounding internet interactions by establishing their increasingly important role in offline social life, and by challenging the idea of an offline-online dichotomy, and 'digital dualism' (Bulfin & Koutsogiannis, 2012;Jurgenson, 2012;Fields & Kafai, 2009;Miller, 2011;Wilson et al., 2012;Young, 2011). This serves as an important move away from early research which optimistically suggested the potential freedoms of the new social environment offered online, highlighting the potentially freedoms and dangers a new 'blank' medium offered away from established social constructs, procedures and rubrics (Bruckman, 1996;Danet, 1998;Harasim, 1993;Rheingold, 1993;Turkle, 1995). This ontological shift towards an understanding of the interwoven nature of online and offline lives is much needed and a fascinating area of research whose discussion is active and still ongoing, and suffice it to say is worthy of much greater consideration and lengthier treatment than can be provided in this paper.
For the purposes of this paper however, it is worth noting that Facebook has been seen as strengthening ties to pre-established offline contacts in a manner that presents a marked move away from previous earlier SNSs such as MySpace and Habbo, where potentially an online presence could be maintained with some degree of separation from offline, and where a certain degree of anonymity was present and expected (Lampe, Ellison, & Steinfield, 2006, Ross et al., 2009, Wilson et al., 2012. PEW research confirmed this emphasis on pre-established offline contacts, noting that 95% of American teens use Facebook to connect to known friends (Lenhart & Madden, 2007). This connection to known offline contacts can be seen as grounding interactions, and importantly, identities within preestablished social parameters (Young, 2009b(Young, , 2011Haythornthwaite, 2005), meaning the purpose of interacting in online SNSs in often no longer grounded in anonymity but in extending and fostering existing social connections, and extending established offline identities (Baker & White, 2010;Quinn & Oldmeadow, 2013).
What is clear is that SNSs have quickly become key locations of social interaction for today's youth, who are spending increasing amounts of time using the tools provided to them by the SNSs to interact and shape social identities. Given the increased importance and the growing role of this technology for the development of social identities, and the increased reliance upon the tools presented by SNSs for identity development and portrayal, further research should be conducted into how the specifics arrangement and choices of social tools affect the manner in which social identity is shaped and formed.

PERFORMING IDENTITIES ONLINE: WHY THE INTERNET IS A STAGE FOR IDENTITY PERFORMANCES AND HOW THIS STAGE CAN AFFECT OUR PERFORMANCE
This paper's understanding of identity is informed by Goffman's work on performativity. Goffman's (1959) dramaturgical perspective is a widely influential method that considers identity to be a malleable, non-constant performance, and suggests that identity can be seen as the sum of the social actions and interactions undertaken by a person within a Social Space. It was Goffman's assertion that identity is formed in and through social interactions; rather than considering a fixed and innate core 'self', we should instead see identity as a malleable and plural socially constructed performance crafted for a specific social situation (Branaman & Lemert, 1997;Elliott, 2014). Informed by Goffman's (1959) influential work, this paper will employ a dramaturgical perspective to frame the social actions and interactions observed online as a performance of social identity. This performative notion of identity has successfully been explored in an online context. Uski and Lampinen (2014) for example studied what they defined as 'profile work'; strategic self-presentation in SNSs. Observing Facebook and the music-sharing platform Last.fm, they focused upon how users maintained a feeling of authenticity in their presentations of identity. They noted that despite the users presenting differing identities across the two sites, the presentations were still seen as an 'authentic' presentations of self.
Goffman's notion of performativity is particularly useful when we carry on this metaphor of acting, and consider how the props and staging available to the user can shape interactions. Goffman noted that social situation will shape and affect many aspects of how we act and interact; we may use different language, mannerisms, and even accents when we find ourselves at the doctor compared to when we are down the pub. He suggested that identity can be viewed as 'a dramatic effect arising diffusely from a scene that is presented' (Goffman, 1959, p. 252). Goffman's attention to the importance of environment, and his understanding of how differing social spaces can result in differing social actions and interactions (see Van den Berg, 2008) has led to him being called 'the quintessential sociologist of the "situation"' (Jensen, 2006, p. 151).
This concept holds particular interest when we consider that each online site can contain potentially different methods and modes of interaction, and different stages and props for our identity performances, which may produce different performances, and potentially different identities. When we add this to the suggestion that online social identity performances are not separate entities, but are affected and shaped by social norms and cultural conventions present offline (Uski & Lampinen, 2014;Reid & Boyer, 2013), then the identity performances taking place online can be seen as extensions of offline social constructions, moved to and performed upon a new stage, with new props and new potential methods of performance.
Much like offline performances, identity presentation and performance online can be seen as a malleable concept, performed with and through a varying array of semiotic modes, whilst remaining grounded in offline social norms and expectations. Van den Berg (2008) noted the importance of 'staging' online in regards to identity performance. Focusing upon what she defined as online 'scripts' À contextual clues that help performers understand the appropriate identity performances based upon the specific staging available to the performer À she found that alterations in situation and 'scripts' leads to changes in how the users understand their situation and their expected role within it. For Van den Berg (2008), identity becomes a concept that is dependent upon audience and situation, with any changes in situation leading to a change in the performances observed.
Following Goffman (1959) and Van den Berg (2008), identity can be understood as a malleable concept that is the result of social interactions with others within a specific environment. It is this fluid and flexible understanding of identity that underpins this paper, as we consider what role the varying environment of online SNSs play in the formation and presentation of an identity.

MULTIMODALITY À UNDERSTAND THE IMPORTANCE OF 'THINGS'
For Goffman (1959, p. 15) an identity performance was 'all the activity of a given participant on a given occasion which serves to influence in any way any of the other participants'. In order to understand all the activity of a given participant online, a myriad of potential choices and activities must be considered. SNSs offer a wide variety of activities and customisable options-from photos, videos, and textual interactions, to font size, shape, and colour. Given the focus of this paper on the impact of the specific online social environment on the performance and enactment of social identities and actions, a theory and methodology is needed to unpack these dense environments and to take into account the myriad choices and specificities made available to users online. The different range of actions available to online SNSs users, and the arrangement of the actions, can be explored using the concept of multimodality (Kress, 2004). Multimodality concerns itself with the understanding how a myriad of 'modes' affect our actions and interactions. Kress (2009, p. 79) defines a mode as 'a socially and culturally shaped resource for meaning making'. Through this definition we not only see meaning conveyed and actualised through text, but also through mediums such as images and the arrangement of content (Luff, Heath, Kuzuoka, Yamazaki, & Yamashita, 2006). As such, a specific arrangement of modes of communication can create different meanings, and lead to different manors of social engagement, different social interaction, and different identity performances. (Kress, 2004;Kress & Van Leeuwen, 2001;Norris, 2004;Van den Berg, 2008).
This understanding of the importance of design, combined with a consideration of the available modes of communication, is particularly pertinent online where a wide mix of modes of communication are present, as Young (2009a) noted of the SNS Myspace, and boyd and Heer (2006) and boyd (2004) found on Friendster. When considering multimodality online, Young (2009a) highlighted that the choices made by the website designers may affect how we are able to and choose to act and interact online. This has been the focus of a growing number of studies for a number of disciplines, who are increasingly recognising the effect and the importance of the multimodal nature of online interactions. For example, Karimov, Brengman, and Van Hove (2011) looked from a business studies perspective at how design affected aspects of consumer trust. Similarly, Lafkioui (2013) found that the resources available on SNSs affected how identity was constructed on multilingual French-based sites, and Skog (2005) found that the features present on the site 'LunarStorm' had a large influence upon social actions and behaviour, and upon aspects of identity performance, such as how much data the users chose to share on the site.
The effects of design has been found to be relevant even within websites that share similar modes, as the specific arrangements and prioritising of modes of communication can shape our interactions. Van Dijck's (2013) study highlights this, using design to focus upon to the concept of identity presentation in two online SNSs, Facebook and LinkedIn. By comparing certain features of what she defines as a 'general' site in Facebook, and a 'specific' business-based SNS in LinkedIn, Van Dijck (2013) theorises how different social aims of these two sites may alter how we can present identity using the same modes. This study also highlights the importance of not only considering the modal options available to a user, but also the consideration of these modal options in situ. Van Dijck (2013) found the combination, arrangement, and accessibility of the modes to be an important factor in how they were utilised. As such, this research seeks to show that, despite their many similarities, we should not assume all SNSs function in a similar way, but that a detailed investigation of each site is useful, and in many cases, necessary.

STYLE AND DESIGN À TAKING CHOICE AND SELECTION INTO ACCOUNT
Van Leeuwen (2005, p. 281) defined multimodality as the 'combination of different semiotic modes À for example, Language and music À in a communicative artefact or event'. This highlights an important point, namely that multimodal analysis should not only focus upon the products available to the user, but also upon the process of producing and interacting with multimodal events. In order to fully explore the participants' interactions with the available multimodal options, Meier and Pentzold (2011) suggest using the concept of 'style' as proposed by Van Leeuwen (2005), who defined style as the manner in which 'a semiotic event is performed' (2005, p. 287). Meier and Pentzold (2011) suggest that we should be interested in not only the resulting communicative artefacts, but the multimodal event itself, as this combination and choice of modes plays an integral part in the creation and performance of an individual communicative artefact. In particular, they were interested in exploring the process of production, suggesting that researchers need 'to engage with the manifold practical procedures of enacting and accomplishing such communicative products' (Meier & Pentzold, 2011, p. 1). Meier and Pentzold also noted that 'style also marks individual and social identity ' (2011, p. 6). In other words, our modal choices and specific multimodal combinations are an integral marker of our social identity performances. Seen this way, the notion of style is useful for this study as it provides a combination of a focus upon identity, the various modes of interaction and communication available online, and the multimodal nature of identity performance in order to offer real insights into how we perform identity within the complex multimodal landscape of SNSs. By utilising style, we remove the assumption that these multimodal landscape will produce certain forms of interaction, and instead take a less determinist stance to consider how the user interact with the available modes to shape and produce their ever-changing social landscape.
However, there are forces other than personal agency that may affect how the users select and combine different modes in a stylised manner to form and perform social identity performance. We should also take the time to acknowledge the choices that are made for the user, and that a user is not presented with a blank space to produce a social identity, but a space in which many choices have already been made for them. By acknowledging the importance of the staging of social identify performances, and the choices made in the design of this staging, we can acknowledge identity performances online not only as the result of stylised combinations and choices of many multimodal artefacts, but also as a result of the availability of these modes and choices made for the performers. These choices may affect how the users are able to make messages and social actions communicable, and may shape and encourage certain combinations of modes. Hutchby (2001, p. 9) touched upon this point, noting that 'assemblages afford, that is, they enable and constraint certain style-full ways of doing and saying'. Put another way, users of SNSs are not presented with a limitless multimodal landscape from which to present their identity, but are bound and affected by the design and layout of the sites, which can affect how their identities are enacted. This is acknowledged by Meier and Pentzold (2011, p. 10), who point out 'signs are realised in and through media, that is, modes of communication come into being with regard to the affordances of particular media'.
Given this understanding the role of stylised uses of multimodal options, this paper will report on two months of exploratory observations and interviews focusing upon the stylised modal expressions of identity across two SNSs, Facebook and Twitter.

METHODOLOGY
Interviews and observations with the participants were carried out over a two month period starting May 2012, with nine participants aged between sixteen and twenty-three. The participants were made up of 6 females and 3 males and were selected via the 'snowballing' method (Goodman, 1961), with the initial participants, selected from known contacts, suggesting other appropriate participants. This method of sampling allowed for observations not only of public social actions, but also of online interactions between the participants. Participants knew at least one of the other participants, and this allowed for the observation and consideration not only of the participants' public posts, but of the any public participant-toparticipant interactions.
Each participant, by their own account, accessed the sites for at least 75 minutes daily, often more than 5 times a day. For ethical reasons, no private messages were observed during the research period, and no public interactions with non-participants were recorded. The selected participants were 'followed' on each site for the duration of the study, with their public actions and inter-participant interactions observed. All data has been anonymised, and no images or traceable information will be presented in this paper to protect the participants' identities.
I observed public social actions made by the participants across the two sites for a two month period, as well as public interactions between the participants on the two sites. This observation was conducted once a day for the duration of the two month period during which notes were taken on the types of interaction and the content of the interactions. Data collection took two forms. Where appropriate, screen-captured images were collected and anonymised to record the actions and interactions directly. These screen-captures are not included in this paper in order to preserve anonymity, and to avoid copyright issues. Daily reflective field notes were also collected in order to track my observations. Notes were collected about the mode and content of any public actions and interactions made that day.
Two interviews were conducted with each participant during the two month period, during which we discussed some of the actions and interactions I had observed, how they felt they were constructing their identities, and what they felt were the important aspects of their performances. We also discussed my early analysis of the social actions and interactions, allowing them to explain their reasons for taking the observed and recorded actions. This two-way interaction helped triangulate my observations and allowed the participants to explain their actions and interactions in their own words, helping to reduce possible ethnocentrism (Babbie, 1986;Gold, 1958). Interviews were conducted in a semi-structured manner, during which we discussed and reviewed the data I had captured. The data was coded and analysed qualitatively via the thematic content of the posts, and the modes through which the participants interacted. The main topics that arose from these observations and presented and discussed below.
Though there may be questions about the authenticity of the identities observed and about the accuracy of the observed identities, this research holds that the performances, or more specifically in this case, the behaviour and social acts, are the unit of analysis, and not the individual. The identity observed, as mentioned in the literature review, is identity at it broadest point; a malleable and changing construct that it constantly considered and worked on by the performer. Rather than attempting to observe a fixed construct of a continuing 'whole' of identity, this research is interested in the moment-to-moment presentations of aspects of identity, and how the modes and content of these performances are stylistically selected by the participant and guided by the sites and the audience. As such, it does not hinder the research that only public interactions were observed, as these were the unit of analysis used to perform identity on the SNSs, and are viewed as public expressions of identity within this environment. 'Front stage' (Goffman, 1959) publicly-presented performances and interactions were observed, with justification and discussion of 'back stage' considerations of the performers (1959) unpacked during the interview process.

RESULTS: PERFORMING IDENTITY IN SNS
I will begin by presenting a brief overview of the data, foregrounding the key aspects that arose, before discussing their implications with regards to stylised multimodal identity performances. The names of the participants have been anonymised, and all images and screen-captures are described in order to avoid anonymity and copyright issues.

Facebook's Use of 'Friendship' and of a Shared Social Past Grounded in Offline Locations and Interactions
The interviews and observations suggested that Facebook's design is framed around the concept of engaging and maintaining social contact with pre-established contacts from the participants' pasts and presents. Facebook frames itself as a place 'to keep up with friends and family' (facebook.com, 01/2012). This message is further strengthened through the use of the 'friendship' motif throughout the design of the site. Users with whom the participants interact with are referred to as 'friends', and across locations on the site users are encouraged to 'find', 'search for', or 'invite' friends. Facebook also suggests list of 'People you may know'. This finding of friends on the site is tied to the offline world further by Facebook encouraging users to pick friends based upon their 'hometown, school, employer, and more'. In their interviews, the adolescent participants commented upon these ideas, suggesting that Facebook was used to maintain relationships with known contacts. When asked to describe the audience on Facebook they described them as 'school friends' or 'old mates'. Even when in some cases their privacy filters meant the information was visible to anyone, they still suggested that 'just because anyone can see it doesn't mean everyone will, right? It's only for people who know me'. Oftentimes they noted that Facebook served as a useful tool when they otherwise couldn't interact with these contacts offline. They also commented that 'it would be weird, you know? To Facebook people that didn't know you. That's just … creepy'. This notion of pre-established contacts was a commonly discussed theme of the use of Facebook. Though I attempted to discourage direct comparison of Facebook and Twitter, the topic often fell back to how the sites were fundamentally different. One participant suggested 'friends on Facebook are probably more people you actually … know … or like went to school with, but on Twitter they tend to be more random people, you know, kinda more open, so you're sharing interests'. Others discussed how they felt the audience of Facebook was 'friends … and unfortunately, family', and knew that on these sites they could discuss certain events and topics that that audience would relate to.
This focus upon a 'friendship' motif was highlighted in the design of the page by the two separate locations on every user's page showing the number of friends they had, as well as a space located on the static bar at the top of the page enabling users are able to quickly 'Find friends'.
In the users 'about' sections (which lists information about them), there were further links to family, with a section listing the family members of the user who are on Facebook. This page also facilitated the continuation of the ties to real-world social actions and existing social connections, with list provided of work and educational background, and even locations in which the user has lived.
By making the participants define themselves by a list of real-world locations and achievements, and encouraging participants to add 'People you may know' based upon 'hometown, school, employer, and more', Facebook defines and shapes what they believe to be the ideal social interaction on their site, and establishes what they believe to be the core facts and important aspects of a social connection and a social identity. Facebook's definition of the social is one built around interactions with known offline contacts, as confirmed by the participants in their interviews. Facebook's 'social' is one that builds upon already established social interactions and friendships. Facebook highlights the importance of family and pre-established friends, and encourages social interactions based upon established real-world connections. One participant noted that it was 'hard to interact with and find new people on Facebook. There's no point really'.
Facebook's 'definition of social' has an interesting temporal element too. It is a definition that focuses on the present and the past; on established relationships, locations, and events, not on fostering new connections or creating future social relationships. Indeed, the participants frequently called their pages 'timelines', viewing them as diaries of their online social lives. This idea can further be shown to be present in the site design, with the inclusion of a digital time-line which allows a user to move to a specific moment in their Facebook history and observe the interactions at that period of time. This sets the site up as an ongoing historical biography. One participant remarked that Facebook was 'a diary of their life'. Though Facebook does not actively discourage the finding of new social connections, the participants noted that they found the idea of connecting with unknown users of Facebook strange.
Facebook's design shaped the acceptable boundaries of expected social interaction on these sites; with the amplification of the ease of certain types of social interaction (the interaction with known contacts based upon historical pre-established relationships) and the reduction of ease for other types of social actions and interactions (the fostering of new relationships). Through the use of a diary-like biography based upon the users' histories, the use of the time-line, and the establishment of the notion of friendship as being built from pre-established relationships, Facebook hindered the finding of unknown people, and encouraged and defined social interaction as being between known contacts.

Customisability of Design on Facebook
Users were given few options to customise the design of Facebook, though they were presented with one space for a potentially more open-ended description of themselves, with the option to fill out a section titled 'about you'. In this section, the users were free to write whatever information they want about themselves, rather than selecting from a list of predefined options, as was prevalent for many other aspects of the site. However, tellingly, only 3 of the participants made use of this section of their profile, with the majority choosing to leave this section empty.
The participants claimed this was due to the prevalence of preestablished social relationships taking place on the site. As one participant, named Chris, put it, his friends knew 'more about me than a bunch of choices on Facebook'. The established social connections present on Facebook led to Chris potentially flouting expectations and toying with many of the descriptions in his profile. He provided false information to many of the predefined options, with his languages and religious views listed incorrectly in what he described as a 'joke'. Chris confirmed that these languages are religions were chosen because he thought they were the 'craziest option I could find!' He also tellingly wrote a message in his 'about you' section confirming that he only expected to communicate with users who already knew him. Writing that 'if ur lookin at this, the chances are u already know me. So no point in telling u is there'.
When we discussed this during one of his interviews, Chris was rather hesitant to over-explain the joke, but did say that he just did this 'as a joke, kinda random I guess … I really didn't think about it much, I just thought it was funny. Everyone who knows me knows it's a joke. All my mates on Facebook know it's me from my photo, not from this stuff'.
Chris' playfully stylised use of this customisable section again serves to highlight the notion that Facebook's definition of 'Social' is one that is built upon pre-existing relationship, and the familiarity of the relationships allows for these spaces to be less 'truthful' than they were shown to be in similar modal affordances on Twitter, as discussed below.

The Discourse of 'Followers' and the Use of Topics on Twitter
Rather than defining other users as 'friends', as is seen on Facebook, Twitter used the term 'followers' and 'following' as a signifier for the relationship between the users. This term is potentially more open-ended than 'friend', as it implies that there can be many reasons that people choose to socially interact with you and to view you updates. Indeed, When asked why they felt other users followed them, the participants suggested that not only were their established offline friends following their updates, but that other users were following them due to the content of their updates, or 'tweets'. The participants suggested this was due to the type of friendships Twitter revolved around. A number of participants mentioned how Twitter interactions and followers were chosen due to the context. One participant claimed 'Twitter is all about the stuff I like, the TV shows and celebrities I like'. Another summed this up, saying 'Twitters more about interests, so yer, they're different (from Facebook)'.
Twitter's design facilitated this feeling, enhancing and encouraging the ability of the user to talk about topics of interest, and to find other users who were talking about these topics through the use of hashtags (#). Hashtags provide users with a way to trace who is talking about a topic at any given time. When a hashtag is placed before a phrase, it becomes searchable to other users who use the same phrase. Nearly all of the participants made use of the hashtag facility, 4 participants told me they check regularly to see what is trending, and often took part in producing tweets about highly trending topics.
The use of traceable topics presents another interesting aspect of Twitter's design, namely the temporal aspects of social interaction on the site. The trending topics of discussion are constantly changing as users discuss new topics. Participants claimed they felt Twitter was 'faster moving', as it encourages discussion and awareness of events that are happening and topics that are currently being discussed.
On Twitter, in order to interact with other users and respond to posts, users had to create a new post every time. This was markedly different from Facebook, which allowed users to comment upon other peoples' updates in situ, rather than creating a new post in response every time. Given that the participants all followed at least 100 other users, with one following 347 other users, this created a large amount of updates for the user to process, and resulted in a fast moving page that discussed topics in real time. Twitter also created a temporal feel by only allowing interactions in short 'tweets'; textual interactions that are restricted to 140 characters. The participants claimed this often made them consider the content of their messages carefully in order to convey the information they wanted to in the short amount of space afforded to them, and oftentimes lead to several tweets being sent to convey longer pieces of information, creating and adding to the fast temporal feel of the present on Twitter.
Twitter presented an alternate definition of Social interaction, one that was built around topics of interest rather than pre-established offline social connections. This can be shown by revisiting Chris, whose rather playful Facebook 'about you' section was discussed previously. On Twitter, Chris was noticeably more careful in his descriptions of himself, choosing to describe himself as a 'Big cricket and football fan and Player;' and later in the research period as a 'Part Time LAD, Full time Lover, Love a bit of Banter, HUGE CHELSEA FC FAN, BLUE IS THE COLOUR!!' Twitter presented the user with only one area in which to describe themselves to other users. They were not presented with a list of option to choose from, but instead were given a section in which they could describe themselves in up to 160 characters. Many of the participants defined themselves in these sections via topics and areas of interest to them, and carefully maintained this section of their profile during the research period.
Here we see Chris defining himself in a more 'truthful' manner, based upon topics of interest to him. There is less of an assumption that he is interacting with a pre-established audience, and we see Chris presenting a more considered description of himself. He claimed in the interview that 'I guess I care a bit more on Twitter about what I say about myself, not everyone … knows me'. He later expanded that 'I don't want wierdos talking to me on Twitter, you gotta be careful what you say cos anyone can read it'.
Twitter's motif of 'followers', and the deliberate choice to facilitate the discussion of topics that are happening in real time created for the participants a prioritising of interactions that are based upon interests, not necessarily upon pre-established offline relationships. The description sections seen in the examples above are more truthful, and present a list of topics that Chris feels best describes him. When this was discussed during the interview, Chris claimed that people wouldn't necessarily know who he was on Twitter, so he chooses to define himself via 'stuff that I like'. Much like Facebook, Twitter's design encourages and aids certain ways of socially interacting and potentially discourages others. In this case, due to the potential interactions with unknown users based upon mutual topics of interest, we see the encouragement of a truthful, succinct definition of identity based upon topics of interest, and a reduction in the participant's reliance upon already established relationships.

Changing Identities on Twitter and Static Identities on Facebook
When asked about the selection criteria for relevant information on their Twitter self-descriptions, one participant answered that it was an 'ongoing process, the information I want on Twitter changes with my interests'. This highlights another aspect of identities displayed on Twitter, namely, the care taken to maintain an up-to-date description.
The information provided on Facebook's about page were not changed during the research period by any participants, and revolved around the concept of past achievements, of an established social history, with users listing places they had worked and studied, places they had lived, their family, etc. On Facebook, these identities was presented as a constant; a fixed entity that would be added to over time. This suggests that the identity portrayed on Facebook had some aspect of permanence to them, or, alternatively, as two participants claimed, that information 'just isn't important on Facebook.' On Twitter however, the identities were works in progress, evolving and changing, with a temporal pace matching the overall design of the website. Users took great care over how they described themselves, and all mentioned that they considered the context on their descriptions carefully and regularly.

The Use of Pictures
Throughout the observation period the participants all used a number of modes to communicate and interact. One of the most utilised modes was the use of photography and imagery. The participants tended to use Facebook to share their photos, as its design facilitated the easy sharing of images. One participant commented that they tended to share pictures on Facebook as they could 'tag' their friends, which would link their photos to their friends' pages, and in turn encourage social interaction and discussion of the image. The option to comment on an image in situ was not present on Twitter, and the participants also noted the temporal pace of Twitter meant photos could often get lost in the plethora of updates produced on the site.
Facebook's facilitation of photos, and the ability to comment and interact through them, lead to some interesting stylised modal choice made by the participants,. One participant in particular highlighted this during the research period. The participant, named Hannah, made two posts within an hour, one on Twitter and one on Facebook. Neither of the post can be shown to protect anonymity, therefore they will be described below. The post on twitter cannot be shown due to Twitter's searchable index of all public tweets made, but the twitter post was textual, and insinuated that there was so little food present in her house that there were mice in her parents' cupboards. Hannah's Facebook post, discussed a similar subject matters, with a picture of a near-empty fridge, with only milk, butter, eggs, cream, and Coca-Cola inside, and the words 'what the hell, parents?!!' above the image. Hannah chose to convey a similar message on both sites, a few minutes apart, but chose to highlight and express the point differently, using different modes on the two different sites.
When asked about this choice, Hannah said 'my family and everybody know what it's like at my house.' She explained that she was aware that her family members, and people who knew her and her family offline, used Facebook, and that she felt that the use of the image conveyed the message without having to explain or contextualise what was happening.
On Twitter, she felt she could be potentially more critical of her parents, and added textual details to explain the situation. The stylised choices made on Facebook suggest an awareness of appropriate content for the audience, and the ability to not have to explain or expand on the image provided. Hannah could use the modes of image and text to quickly relay the appropriate message to her perceived audience. On Twitter, her modal choices were potentially limited, and so we see the use of text to convey a message relating to her parents and food. These two different choices show an understanding of stylised modal selection to socially interact, as well as highlighting how these modal choices are framed by the site design.

Style À Timbre and Content of Performances
The previous example of modal differences highlights not only the stylised use of modal choices shown by the participants, as they select and use appropriate modes for each environment to convey their messages, but also the active consideration of appropriate content and appropriate identity for each site and each perceived audience.
The participants often had very different styles when approaching the two sites, and presented different aspects of themselves on each site. One participant, named Sue, claimed that she used Twitter as an 'escape from my boring real life'. This was reflected in her use of the two sites. On Facebook, she discussed health and family issues, and talked about issues in her daily life, often in a serious manner that elicited support and help from her friends. On Twitter however, her identity was optimistic, and often child-like. Sue chose a picture of a teddy-bear as her profile image on Twitter, and had images of the children's book 'the very hungry Caterpillar' on her page. She claimed that she felt Twitter was liberating for her, and that she could 'take a break from my real life to talk about fun stuff'.
Another participant, Mary, presented very different personalities on Facebook compare to Twitter. Whilst on Facebook she discussed family issues, and interacted with her established offline friends, on Twitter she discussed her interests, such as the works of Chaucer. She felt that her Twitter profile was more 'whimsical', and allowed her to maintain and attend to a part of her personality that she otherwise wouldn't have be able to adequately explore and express. She claimed that Twitter presented her with a receptive audience for her interest, which may have been lost on her current friends. As such, the identities presented on the two sites were, she felt, vastly different. An active separation of different identities and the thoughtful choice of content was reflected by many of the participants. For example, one participant noted that he separated events in his life into 'things just for Facebook, and things for Twitter'. This active consideration shows a careful selection and an awareness of the identities being performed on each site.

DISCUSSION
The findings show two complex sites of multimodal activity where users make various kinds of messages and meanings in and due to various situational contexts, design contexts, and contexts of culturally meaningful activity. The actions and interactions observed were influenced by a number of factors including: the allowances and modal choices available on the sites; the layout and design aspects of the sites; the discursive choices of the SNSs; and the communities and audiences the participants were performing to and for.
It is clear that the varying self-presentation strategies and social interactions were affected by several factors. The design aspects of the spaces of social interaction reflected and highlighted the use of specific Discourses of 'social' by each site, with the sites prioritising and encouraging certain aspects of social interaction and downplaying others. Facebook defined social interaction in terms of an extension of offline established contacts and events, an extension of already occurring past relationships. Twitter, on the other hand, based their understanding of social around the 'now', focusing on current events and shared topics of interest. This shaped the way the sites were approached and utilised by the participants, and affected the goals and aims of the participants on each site, the ways they used the sites, and the information they put on these sites. It also shaped the way the participants felt about and understood the maintenance of their identity. On Facebook, less care and attention was given to the identities as they were presented as static objects, based upon established past facts. On Twitter, identities were seen as ongoing objects, changeable and of increased importance in establishing and building new social relationship, rather than maintaining existing relationships as seen on Facebook.
The participants showed an awareness of these aims and displayed an adept ability to unpack these goals and 'play the game', navigating the social landscape effectively. They were aware of successful and unsuccessful strategies for dealing with each site, and showed an awareness of appropriate material and modal choices based on the audience they were interacting with and for, the platform and staging on which they were performing, and the modes available to them. The participants showed that the expectations and preferred uses of the sites could also be flouted, and that they had the ability not only to understand the implicit modal and social rules, but also, that they could unpack and test the limits of these rules, using them to their advantage. The participants displayed stylised interactions on the sites, producing individual identities that were shaped in part by the design and modal options of the sites, in part by their own agency and social aims, and in part by their social environment and offline social expectations.
Facebook presented an environment based upon family and friendship, and encouraged interactions based upon shared events and real-world issues. This echoes similar findings from other Facebook based research (Miller, 2011;Young, 2011). We can see that through the design of the site, Facebook presents a definition of 'Social Networking' that revolves around shared experiences, a temporal history, and known offline contacts and social relationships. Twitter on the other hand presented and framed 'Social Networking' as the exchange of information, social events and the discussion of current events. Given this, and the stylised manner in which each participant interacted with the sites, we should once again be careful to generalise all social networking online as similar. Though there are similarities on these sites, we should strive to look past these to fully recognise how users are approaching and using these sites, and how the performances are shaped by the SNSs and their understanding of 'Social Networking'. The modal nuances and the various definitions of 'social' present two sites that produced noticeably different performances, with each participant interacting through and with the available modes to produce a stylised performance of identity.
The intended uses, the design, and the modal choices affected many aspects of the participants' choices, actions, and interactions, as well as shaping their perceived audiences and expected social aims. Users omitted topics, tapered interactions, maintained some aspects of their identity, alter other aspects and expressed different identities based on their understanding of the sites. Topics were also presented differently on each site based upon the audiences, the participants' understanding of the offline context and the modal availabilities. This highlights that not only does the offline/ online dichotomy break down on SNSs, but that it is a malleable state that the user can use to emphasise or support other aspects of identity performance. This is highlighted through the complex, site and user specific presentation of the same topic, as highlighted in Hannah's decision to use images of Facebook and text on Twitter to discuss the same topic.
The differences in the design and audience fostered by the two sites also affected the way the participants defined and described themselves. On both sites users were afforded several indices of information for performance of identity such as personal information, historical and location based information and preference information. These were utilised differently based upon the perceived audiences and the messages the participants were trying to impart. Participants suggested that on Facebook not all this information was pertinent to the environment, the types of interaction taking place or to the performance of identity to an audience who already possessed much of this information. However, on Twitter the participants filled and maintained this information carefully. The participants suggested that the main reason for this disparity was the types of audiences that the staging and modal choice that the two sites fostered. Though both sites provided a space in which to 'write' identity, the space on Twitter was more important to the participants and was considered more integral to social interaction. Despite the similarities in certain aspects of design, when considered within the context of the sites of social interaction, we can see that the modes were used, engaged, and interacted with differently.
The participants' social interactions were also affected by the modes available on each site. Facebook allowed and favoured the posting of images, videos and longer sections of text that could be re-visited and continually added too. This variety of modes meant the participants had more flexibility and options, allowing for stylised manner to emerge when displaying messages and perform identity. The participants suggested they preferred using Facebook to share photos as they felt they were relevant to the audience they were interacting with, and liked the interactive elements present on Facebook that allowed continued engagement and commentary with the photos. The temporal pace of Facebook also aided the use of a variety of modes as the participants felt that the slower pace made the pictures and videos more permanent and easily located in comparison to Twitter.
The participants felt they all managed to successfully perform their identities and successfully interact despite the modal, temporal, design and audience differences between the sites. However, it was clear that the differences between the sites affected the manner in which they performed, and the choices they made when considering modal choices and appropriate content for the site and audience. Despite the character restrictions on Twitter, and the heavy favouring of text-based interactions, the participants still managed to effectively navigate the site and perform stylised identities, using, adapting and understanding the potential restrictions of the sites. Goffman (1959) suggested that social interaction involved the continuous strategic expression of identity and self. When the self is presented, certain aspects are emphasised and others are concealed for the audiences and contexts we find ourselves in. The same social norms can be seen extending online (Uski & Lampinen, 2014), manifesting themselves in this new modal environment. This conscious expression of self can be more considered online, given the delayed temporal nature of the internet (boyd & Ellison, 2007;Uski & Lampinen, 2014). The data presented in this paper suggests that the context and audience affect what is emphasised and what is concealed. This is guided by the modes available and the social norms implied by the site design and ethos. Stylised performances take place online, tapered not only by the users themselves, but by who they are interacting with and the modes and environments they are interacting through. The design of SNSs did not hamper the ability of the user to successfully enact and perform identity, but it did change how they approached the performance, which subjects they presented, how they understood and prioritised the tools at their disposal, how they used the tools at their disposal, and how they understood their role as a performer.

CONCLUSION
The findings here match those of previous research in noting that design and situation effect social action (De Saint-Georges, 2004;Karimov et al., 2011;Lafkioui, 2013). Of particular note is Van Dijck's (2013) work, which noted that though SNSs can share similar modes, the arrangement and presentation of these modes will affect how they are utilised and how identity is presented. Van den Berg (2008) provides perhaps the most pertinent explanation as to why this happens through the notion of online 'scripts' À the contextual clues that help users understand their role within the social landscape. In the case of this data, Facebook and Twitter both presented differing scripts and differing definitions of 'social', meaning that despite some modal similarities, the user were presented with different roles for them to fulfil, and interacted with these modes accordingly, producing differing performances that were sensitive to the perceived audience and situation.
The findings presented here suggests that Goffman's (1959) performative theories still hold relevance online, and are strengthen by the inclusion of a focus upon modality. The adolescent participants showed an ability to navigate the social landscape and produce flexible identity performances tapered to meet the requirements of the sites, the expected social interactions upon the sites and the audience to whom they were performing. The participants showed an awareness of the flexibility of social identities and an ability to unpack to 'scripts' (Van den Berg, 2008) of a social landscape and perform their identities accordingly. Rather than displaying one unified self, the participants navigated the social landscape, ably and competently picking and choosing aspects of identity to highlight, and modes through which to best express themselves. This data highlights that identity increasingly must be seen as a performance realised through many modes, shaped by a number of aspects including the availability and prevalence of certain modes, the audience, and the social aims and ideals presented through site design.
The use of 'style' (Meier & Pentzold, 2011) as a concept proved particularly useful when considering multimodal activity. By not only considering and observing the multimodal outputs produced by the users but also by discussing the way the participants considered and combined the available modes, an understanding can be reached of the how the multimodal landscape is approached and navigated. The users selected and combined the available modes based on their own needs and upon the suggested social expectations of the sites. The choice of modes both made for and by users are important and relevant to understanding a social situation (Kress, 2004;Young, 2011). The research highlights that choice of modes can be a manner of self-expression, with different modes used to send different messages. In the potentially diverse modal landscapes online, users can express themselves through the embodiment and use of a number of modes. Previous studies have noted users showing inventive uses of sites and modes, and the formation of un-anticipated modes of self-expression (Miller, 2011;Uski & Lampinen, 2014;Young, 2011). Similarly, this research observes the flouting of expected uses of the available modes, and presentation of stylised manners in which to use the available modes and perform identities. As websites continue to grow and change, the manner in which we use and interact with them will continue to adapt, pushing and creating novel, innovative identity performances.
Though there may be commonalities in these sites, the interactions vastly differed. Even when there were modal similarities the temporal choices, the arrangement of these modes, the favouring and discouraging of different social activities, and the goals of social networking on these sites differed greatly. This echoes previous findings such as Van Dijck's (2013) observations of modal similarities between Facebook and LinkedIn, two sites displaying vastly differing aims and attitudes towards social interaction. As such, the nuances and arrangements of websites should always be considered as they can play an important role in shaping the data collected from interactions on the sites. Social Networking can act as a useful general term of study, yet as shown in this research, the meaning and aims of Social Networking can be interpreted differently from one site to the next, which in turn affects the users and uses of the site.
As our everyday social interactions become more multimodal (Jewitt, 2008;Kress, 2004Kress, , 2009, so much our understanding of the social. A consideration of modality, and how it is utilised, can wield useful and in-depth findings, helping to understand how and why people are socialising in a number of diverse ways across a number of modes. Similarly, a consideration of style in modality can help unpack how similar modes can be enacted in different manners, across different stages of performance. Though the focus presented here is narrow, the application of multimodality and of stylised identity portrayal could be very broad. Other affecting factors should also be studied to add to an in-depth understanding of multimodal identity performance, such as the routes of entry to these sites À as many users are accessing these sites through a number of devices which may present differing functions, and may affect how they are accessing and using SNSs.
This paper highlights the importance of design and the effects it can have upon the interactions and identity performances taking place on these sites. As this technology becomes increasingly ubiquitous globally, the impact of design upon our social interactions becomes more and more pertinent. The data collected displays two sites that differ vastly in their design, their use and in how they 'script' and present the concept of social interaction. This not only affects how the user interacts, and the topics and modes the participants discuss on the sites, but also how the users frame and shape their identities, and how they understand their social roles and the needs of the audience. The arrangement and availability of these modes, and the design ethos of the SNSs, can be seen as affecting many aspects of how the users view and interact with and on the site, which in turn affects the tone, type and form of the messages the users are sending. Or, to butcher the words of the bard; all the web's a stage, and all the men and women merely players. How we play and perform, and how we understand our roles is, importantly, shaped by this staging, as we perform on, in and through it.