#vegansofInstagram: Studying Vegan Identity Performance on Instagram

(1)

#vegansofInstagram: Studying Vegan Identity Performance on

Instagram

Nora Lauff

MA New Media and Digital Culture

(2)

Abstract

This thesis provides an insight into how highly followed vegan on Instagram perform their vegan identity to their online audience. Set within frameworks of identity performance online, Instagram as a platform enabling performativity first and foremost through visual content and veganism as a growing movement in form of lifestyle changes, this study aims to outline vegans’ posting behaviour in relation to how they publicly identify a vegan online. To do so, the most recent 100 posts of ten different vegan Instagram users were studied using the Google Vision API to analyse and categorise visual content, and Voyant Tools to make sense of written content in form of post captions. This study concludes that vegans on Instagram portray an overly positive, partly idealised version of veganism to their audience, likely in an attempt to dismantle existing negative stereotypes around vegans, but probably also related to the platform specific community norms of Instagram.

Key words:

(3)

CHAPTER 1: INTRODUCTION

1. Veganism and its rise in popularity

The Vegan Society, the oldest, still active society on all things veganism, defines veganism as “a way of living which seeks to exclude, as far as is possible and practicable, all forms of exploitation of, and cruelty to, animals for food, clothing or any other purpose” (The Vegan Society). In other words, people following a vegan way of life adapt their diet, as well as other consumption practices such as shopping, with the goal to eliminate anything connected to potential animal cruelty from their lives. Starting out with dietary practices, this means that vegans will not consume any meat or fish, as well as animal by-products such as dairy, eggs, or even honey. Fashion-wise, for instance, they will also avoid buying anything that contains animal-derived materials such as wool, leather or silk. Lastly, the lifestyle even extends to cosmetics, as vegans will avoid any products that may have been tested on animals. Veganism in itself is thus acted out through “what one does (and does not) eat, consume and purchase” (Greenebaum 132), rooted in personal choice and practice.

Even though veganism thus goes against the consumption of meat and dairy or leather goods, practices enrooted in modern society as “normal,” often making vegans appear like “the odd one out” amongst other people, the cruelty-free lifestyle has reportedly experienced a growth in popularity. In 2016, The Guardian published an article in which was stated that over the last decade alone, the UK had seen an increase in vegans by 350% (Marsh and Guardian readers). More recently, other sources even pointed out how “experts are predicting veganism [to] go mainstream” in 2018 (Molony) as more and more people develop an interest in the lifestyle. Such claims could, for one, be supported by the increasing availability of “vegan consumer goods” (Greenebaum 137), ranging from meat and dairy substitutes to plant-based leather alternatives. Since following a vegan lifestyle may thus become progressively easier to implement in every day with more information on the overall subject of animal cruelty, as well as wider choice in vegan products becoming available, veganism is slowly experiencing a shift to becoming more mainstream and socially acceptable.

(6)

2. The role of social media

In regard to the rising popularity of veganism, the movement is apparently one “driven by the young” and heavily influenced by social media (Marsh and Guardian readers). Editor of Vegan Life Magazine Maria Chiorando addresses the importance of social media in the context of veganism as well, as she comments, “I think social media is a huge driver of veganism and this may tie into why it is more prevalent among younger people” (Meager). One of the reasons for this may be the way in which traditional media, as opposed to new, often user-generated media, portrays veganism or vegans overall. More often than not, “old,” mainstream media promote “negative stereotypes of vegans” (Bosworth 20), something Matthew Cole and Karen Morgan even call “vegaphobia,” in so far that the lifestyle is discredited by making it seem unattainable or even entirely impossible, sometimes even resulting in the ridiculing of the practice or the people participating in it (qtd. in Bosworth 20). The negative and potentially wrong representation of veganism in traditional media thus represents one of the reasons why people may turn to the internet as a form of “mass self-communication” to research veganism or find like-minded people within the lifestyle. In this case, the online world allows people to be autonomous in their practices and shaping of opinions as they go against societal norms of consumption or everyday practices (Castells qtd. in Bosworth 21).

Linking the internet, and more specifically social media to my personal experiences of finding out about veganism and communities built around it, the online world impacted my change in diet and lifestyle tremendously. Social media platforms such as YouTube or Instagram, for example, played an essential role in how I perceived veganism as a “newcomer.” By watching videos, reading posts or scrolling through photo galleries, I was able to gather information related to veganism ranging from food (recipes, tips on food replacements, restaurants) to health (medical studies), the environment (documentaries) and even animal cruelty (documentaries, short clips). On top of that, seeing how others identifying as vegans implemented a cruelty-free lifestyle in their everyday practices, evoked a sense of community and belonging for me as I became aware of how many others were interested in or already following a vegan lifestyle.

(7)

3. Research questions

With social media playing an important role for people interested in veganism or those already implementing the lifestyle in their everyday life, how is the cruelty-free lifestyle represented on social media platforms? As Chiorando mentions veganism to likely be more common amongst “younger” people, how do platforms popular amid this age group allow individuals to perform veganism online? With Instagram users, for example, being made of by “young adults” for larger parts (Silva et al. 124), how do vegan Instagram users portray their vegan identity on the visual-centric platform? Differently put, this research project sets out to answer the overarching question of

how vegan identity is performed on Instagram. In an attempt to answer this

overarching question of vegan identity performance on Instagram, the following sub-questions will allow for a more detailed study: what kind of image of veganism is represented by Instagram users identifying as vegan on the platform? How is veganism represented as defining part of these users’ identity? More precisely, what kind of visual or textual content do vegan Instagram users share on the platform, and which aspects of the lifestyle are highlighted the most by them? These sub-questions will first and foremost be answered based on content analysis, while the overarching question will allow to draw conclusions on how vegans perform their identity on Instagram.

4. Previous research

Veganism as a whole, in part including the communication of it online, is not a new object of study in itself. However, more often than not, studies on veganism that were conducted within the field of humanities typically explored veganism from sociological and cultural perspectives, with the dominant research method being that of participant interviews and focusing on vegans’ offline practices for the most part (Cherry, ‘It’s Not

Just a Diet’: Identity, Commitment, and Social Networks in Vegans; Greenebaum;

Bosworth). The way in which veganism and vegan identity is communicated has typically been restricted to the “physical world,” with less of a focus on its communication online (Bosworth 21), particularly in relation to content shared by vegans on social media platforms.

(8)

Whilst taking into account sociological and cultural perspectives, as well as findings from previously mentioned studies, this research project sets out to further study the representation of veganism online. By using digital methods, I am attempting to understand the way veganism is performed as a defining part of self-identity online, and more specifically on the social media platform Instagram. Even though previous work on vegan identity on Instagram exists (Xavier et al.; van t’ Laar), research has often been restricted in breadth due to lack of time or available tools (and skills to use said tools), which is why I am hoping to further research in this area of study. By using tools such as data scrapers and visualisation APIs, I am hoping to generate a bigger picture of vegan Identity and its performance on Instagram by having the means to analyse larger sets of data than has previously been done.

5. Research structure

To answer the question of identity performance of vegans on Instagram, this research project takes into account different frameworks based on relevant theoretical concepts. Firstly, I establish how identity has been studied as an essential part of the self and one’s self-presentation in everyday life, drawing on works of Erving Goffman (1959) or Anthony Giddens (1991). In applying a new media angle to my research project, I then delve into how such concepts on identity formation and performance have been translated to social media studies and the way individuals present themselves online. In order to offer a greater understanding of veganism as a social movement, as well as an alternative lifestyle, I proceed to explain what veganism entails and what people’s motivations are to participate in it. Following this, I discuss previous studies conducted on veganism and the way vegans communicate in an attempt to frame existing debates on veganism which then allow me to draw parallels to my own research objective previously outlined through my research questions. Finally, the last part of my theoretical framework addresses Instagram as an object of study, especially in relation to the importance of visuals in online communication and identity performance.

Based on these theories, this research project sets out to explore how vegans on Instagram perform their vegan identity, first and foremost through an analysis of visual and textual content shared by them, namely photos and their respective captions. In

(9)

setting up a method to scrape and analyse said content, the section following the methodological set up and conduct offer an insight into the results and findings founded on the visualisation tools implemented. Finally, the last two chapters of this project discuss the outcome of the study and its potential meaning, as well as draw a conclusion and setup a base for potential further research.

(10)

CHAPTER 2: SETTING UP A THEORETICAL FRAMEWORK

1. Identity performance 1.1. Impression management

Generally speaking, identity makes up the part of one’s self that is shown to others and by which one becomes recognisable by them (Altheide 2). Identity is thus what sets one apart from others and a fundamental part of self-presentation. In the 1950s, sociologist Erving Goffman published a book on the topic of identity and self-presentation in 1959 titled The Presentation of Self in Everyday Life. In it, he documents his observations and analyses of people’s behaviours and actions in everyday life and how they present themselves to others. He does so by using dramaturgy as a means to draw parallels between theatrical performances and individuals’ performance of the self in everyday life (Gonzales and Hancock 169). In doing so, Goffman outlines the similarities between actors’ behaviour depending on whether they are finding themselves on the “front” stage, within the presence of an audience, or whether they are “back” stage, hidden from the sight of others watching (Goffman). By using this metaphor, he thus argues that, similarly to the case of actors in a play, for instance, individuals’ everyday behaviour and actions change depending on whether they are in public, surrounded by other people, or alone in private. Based on this analogy, Goffman states that people purposely choose to “project a given identity” in social situations (Bullingham and Vasconcelos 101), depending on the social environment they find themselves in at the time (Goffman 18).

One of Goffman’s key arguments is thus that identity and self-presentation are performative in nature, with the individual’s ultimate goal being that of influencing and guiding the perception others may have of them (Goffman 26). Goffman calls this performance of the self “impression management,” and states that it is conducted through the performance of presentational cues. Moreover, he differentiates between said presentational cues in that some expressions of the self are “given” by the individual, whereas others are “given off.” Expressions “given” typically make up direct and intentional presentational cues, commonly in the shape of verbal communication, with the goal to convey information about the self to others. Expressions “given off,” on the other hand, are cues less directly, or even entirely unintentional and

(11)

unconscious in nature (Goffman 14). Although both expressions “given” and “given off” contribute towards a person’s self-presentation, cues actively “given” tend to make up a bigger part of it (Ellison et al. 418). It is thence through the implementation of these presentational cues that the individual forms their self-presentation in front of others, eventually turning their public identity into a “fact in its own right” and thus setting it in stone (Goffman 34).

1.2. The reflexive self in post-traditional times

Drawing on Goffman’s earlier work on identity and performance of the self (Jacobsen and Kristiansen 151), sociologist Anthony Giddens argues that questions about the self and how to act in everyday life have increased in importance as they are “focal questions” in what he identifies as “late modernity” in the early 1990s (70). As Giddens contextualises this concept of identity in what he calls “late modernity” or “our present-day world” (keeping in mind that he published his book in 1991), he argues the search for the self and one’s own identity, whether portrayed to the outside world or not, has its origins in Western beliefs centred around notions of individualism (74). He supports this claim by quoting Baumeister, who underlines that pre-modern times showed less concern about individuality in people, as “individuality was not prized” (Giddens 74– 75). Using the example of appearance, for example, Giddens states that even though the way someone looks made up part of their social identity in pre-modernity, it was not so much related to personal identity (99). In modern times, however, “appearance […] becomes a central element” of personal identity (Giddens 99) as it allows for an individual to either fit in or stand out amongst other people. In this sense, the way each person decides to appear in front of others is highly individualised and can, for one, be done so through the changing of one’s appearance.

In line with Goffman’s theory on expressions “given” and “given off” by individuals, Giddens emphasises that, rather than (exclusively) being influenced by external influences, “the self is not a passive entity” but is instead rooted in people’s choices and actions actively crafted on an everyday basis (2). In other words, the self becomes a “reflexive project, for which the individual is responsible” (Giddens 75). Furthermore, Giddens points out that this “reflexive project of the self” does not only encompass the creating of one’s identity, but also the adapting and revising of the self in an attempt

(12)

to create a continuous, somewhat biographical narrative (5). Even though one’s self-identity can thus be edited and changed over time, these changes still need to be done within the previously determined framework of the self so as to create a coherent image that is projected to others.

1.3. Identity performance through lifestyle choice

Taking this one step further, Giddens proceeds to shift his attention to “lifestyle” choices, which he defines as “a more or less integrated set of practices which an individual embraces,” particularly as said practices contribute to the individual’s formation of a narrative of their self-identity (81). In his article After Identity from 2007, the academic Jonathan Rutherford also discusses identity formation and performance in relation to lifestyle choices in applying it to a neo-liberal context. He links it to economics and consumption, arguing personal identity to be increasingly defined through the consumption of goods, which can, in turn, affect someone’s societal status (Rutherford 6). Though Giddens also takes into account consumption practices as defining identity and self-presentation in relation to lifestyle, he deems it to be wrong to look at it from an economic perspective solely (81). Nevertheless, he points out that, whether through consumption or in other ways, lifestyles are not simply something individuals follow out of personal choice, but much rather because it is somewhat required of them. In his own words, “we have no choice but to choose” (Giddens 81). In this regard, self-identity has become increasingly defined by lifestyle choices (Sneijder and te Molder 621) made by individuals on an everyday basis. Such lifestyle choices shaping people’s sense of the self and their identity exist in many forms, one of which could be that of following a vegan lifestyle as introduced earlier. In adopting a vegan lifestyle, people are arguably seen to be creating a new identity for themselves in how they navigate everyday life as they differentiate themselves from non-vegans (Cherry, ‘It’s Not Just a Diet’: Identity, Commitment, and Social Networks in Vegans 20). From an identity studies perspective, veganism thus presents an interesting social phenomenon to investigate.

(13)

2. Identity online

2.1. Performing the self online

Theories on identity performance such as the work by Goffman or Giddens have been implemented in the study field of social media, ranging from studies on online dating (Ellison et al.) to Facebook (Zhao et al.) or even user-generated online reviews (Vásquez), generally within the context of Web 2.0 and the rise of user-generated content (Seargeant and Tagg, ‘Audience Design and Language Choice in the Construction and Maintenance of Translocal Communities on Social Networking Sites’). This is in part justified in the claim that self-presentation is one of the two “fundamental social dynamics” of social media (Seargeant and Tagg, ‘Introduction: The Language of Social Media’ 5), or, as danah boyd puts it, social media quite literally allows users to “write [themselves] into being” (boyd). As social media is rooted in not only promoting, but also relying on user-generated content as one of its “defining characteristics,” it gives users the freedom to decide what to post or share and who to interact with, as they can shape the way they are seen by others online (Vásquez 65). This is in particular done through visual and textual context, since, as opposed to “offline” identity performance, the online environment makes it harder for things such as speaking tone or body language to be translated unless for content in video format (Seargeant and Tagg, ‘Introduction: The Language of Social Media’ 7). Nevertheless, users online are able to perform their identity in front of an “audience” that is no longer required to be physically present as said performance can be mediated through the use of new media technologies (Meyrowitz qtd. in Giddens 84). The sheer activity of posting or sharing content revolving around oneself online thus comes with the awareness of an “audience” watching.

In an article discussing identity performance amongst bloggers in the online sphere, the authors Liam Bullingham and Ana Vasconcelos compare differing opinions on how “online” identities compare to “offline” ones. Whereas Luiz Carlos Baptista argues online identities to add layers to “non-virtual” ones, Emmanuelle Vaast recognises online identities as entirely “new selves” (Bullingham and Vasconcelos 102). Vaast’s opinion in particular could thus be argued to draw on Goffman’s “front” versus “back” stage identity as online identities take place in the “front” stage in the sense that people perform their self the way they want to be perceived, whereas their offline self is less

(14)

performative in nature, thus moving to the “back” stage (Bullingham and Vasconcelos 103). Nevertheless, Bullingham and Vasconcelos point out that both Baptista and Vaast agree that online environments make it easier for people to adopt a certain persona, whether it relates to their offline identity or not (102).

2.2. Anonymous vs nonymous identity performance

In line with Goffman’s argument that people’s behaviour changes according to the social setting they are finding themselves in, studies on identity performance on social media concluded that online users’ behaviour is similarly affected by the “nature of setting” of the online environment (Zhao et al. 1817). Studies on this subject mainly differentiated between anonymous and nonymous (non-anonymous) online environments, as users of anonymous online settings such as chat rooms, were more likely to slip into specific roles, however close to their “actual” selves, than users of nonymous sites or platforms (Zhao et al. 1817). With the rise in popularity for typically more nonymous platforms like Facebook, users are more likely to present their online identity closer to their non-virtual one, as the nonymous nature of the social media network constricts “the freedom of identity claims” (Zhao et al. 1818). Whilst both anonymous and nonymous online platforms thus allow for a certain degree of performance of the self, anonymous ones appear to allow for greater freedom to do so.

One way or another, social media, in general, allows for a highly selective process when it comes to self-identity, as it is driven by “self-mediation,” with users being in a position to decide what expression they want to “give” in curating their identity into being (Khamis et al. 196). Thus, similarly to offline identity performance outlined in Goffman’s work, people are able to highlight parts of their identity perceived as more desirable by themselves, whilst hiding potentially less desirable aspects at the same time (Zhao et al. 121). This selection and editing process ultimately marks online identity as “always under construction” and thus never fully “finished” (Helmond 22), as the very essence of social media revolves around posting and sharing new content and updates.

(15)

2.3. Microcelebrity

Whereas offline identity performance is founded on the physical presence of others making up the “audience,” online identity performance is reliant on others visiting a user’s profile or users being part of the same network in general, differing in nature based on which social media platform is being used. Whether this online audience consists of network connections such as “friends” in the case of Facebook, or asymmetric networks such as “followers” on Instagram (Hu et al. 596), online users are likely to be “keenly aware” of being seen by others as they partake in “mediated self-presentation” (Gonzales and Hancock 171). Hence, just as constructing an interesting, appealing and coherent narrative (Giddens 5) of oneself in “real life” may make a person more attractive to on-watchers, creating a “compelling narrative” online can lead to a growing audience (Khamis et al. 196) as increasingly more people become interested in others’ mediated selves online. In this sense, the self on social media is more likely performed to be consumed by others, rather than serving as a tool to look back on the self over time (Khamis et al. 196).

Especially platforms making use of previously mentioned “asymmetric networks” in relation to type and size of audience of users’ persona such as Instagram, allow individual users to gather high amounts of followers within a relatively short time span. In this sense, the audience observing a person’s identity performance online can grow in the form of follower count, turning the high number of followers or high engagement on the social media profile overall as a form of success or at least acceptance of the user’s performed identity (Khamis et al. 197). As growing a large audience online is thus conceivably easier to do in online environments than in non-virtual life, the lines between performer and audience, or “amateur” and “professional,” are easily blurred (Seargeant and Tagg, ‘Introduction: The Language of Social Media’ 3). A person’s “success” in attracting a large audience online is thus not dependent on their expertise knowledge on a subject, for instance, but rather on whether or not their self and the narrative they create around themselves is recognised as interesting to others. These types of people who perform their online persona to a large audience online, and potentially gain attention through it, are what Alice Marwick calls “microcelebrities” (Marwick, ‘You May Know Me from YouTube: (Micro-)Celebrity in Social Media’) Though such microcelebrities typically start out as “everyday users” with no intent in

(16)

commercialising their self (Abidin n. pag.), they are normally known or even famous exclusively within “small, niche networks,” as their persona is niche-specific, allowing them to keep the narrative of their mediated self coherent (Marwick, Status Update:

Celebrity, Publicity & Branding in the Social Media Age 117). In the case of some

social media platforms, such as Instagram, these microcelebrities are also known as “influencers.” As the term suggests, these highly followed or engaged with online users having attained influencer or microcelebrity status may be perceived as knowledgeable by their audience, with the potential to affect or even change the way their “audience” or followers acts or behave (Cambridge Dictionary). As previously mentioned, however, this perception of being knowledgeable does not necessarily stem from being an “expert” in what the purvey of themselves online. Rather, authenticity and consistency in the way influencers present themselves to their audience are what makes them stand out and reliable for others.

In some cases, the audience may identify with the performed persona of these online influencers as they sympathise with them based on shared opinions or cultural practices (Seargeant and Tagg, ‘Introduction: The Language of Social Media’ 9). In agreeing with what a highly followed user shares online, people making up the audience may thus relate to them, creating a sense of togetherness as they identify with the influencer and what they represent (Seargeant and Tagg, ‘Introduction: The Language of Social Media’ 9). Against this background, Instagram users identifying with veganism and performing this identity on their social media profiles not only shape the way others perceive them through the lens of veganism but might also influence their audience’s behaviours and practices. In some cases, the way vegans on Instagram present themselves to their audience may also inspire others to look into the alternative lifestyle and potentially adopt it themselves based on the image provided by certain users.

3. Veganism as a lifestyle 3.1. Motivations to go vegan

Adapting a vegan lifestyle is typically outlined by three main motivations; health, ethics or preservation of the environment. Even though, as defined by The Vegan Society,

(17)

be perceived differently based on an individual’s motivation to change their consumption practices and overall lifestyle to begin with. “Health” vegans for one, may implement an animal-product-free diet to improve their health and overall well-being, but in doing so do not necessarily protest the buying and wearing of animal-derived materials or different types of products tested on animals. Similar to “health” vegans, people adopting a vegan diet or lifestyle for environmental reasons will cut out animal products from their diet but not necessarily from their overall consumption, dependent on whether they deem materials such as leather to be more sustainable than plant-based imitations. Lastly, those choosing a vegan lifestyle from an ethics perspective, live an animal-product-free life founded on political and ethical beliefs against animal cruelty in all aspects of life, sometimes even expressing “anti-establishment” beliefs (Greenebaum 130–31). Whatever the motivations of someone publicly identifying as vegan are, doing so does always comes with some kind of “declaration of […] identity, morals and lifestyle,” making self-identifying vegans stand out amongst non-vegans (Greenebaum 129). In terms of self-representation, self-identifying vegans thus find themselves in a position of having to edit and adapt their public selves in line with their lifestyle changes as they question and ultimately go against meat-eating, for example, something that is often perceived as “cultural norm” (Cherry, ‘It’s Not Just a Diet’:

Identity, Commitment, and Social Networks in Vegans 9). Overall, no matter what kind

of external influences vegans encounter, the success of the lifestyle is primarily based on changes in everyday habits and practices conducted by each individual personally (Bosworth 9), regardless of what others may do or not do.

3.2. Different levels of veganism

In relation to differing reasons and motivations for people to identify as vegan, the way they perform it as part of their public self may vary. One of the main studies on veganism referred to often in academic work, is that of Elizabeth Cherry dating from 2003, where the author conducted a study with a large group of participants, self-identifying as vegans. One of Cherry’s main findings was that, even though all participants identified as vegan, an overall distinction became noticeable in how the participants presented themselves and performed their vegan identity in everyday life, and particularly in public. Overall, Cherry categorises the participants into “punk” vegans and “nonpunk” vegans, with the former being associated with more “militant” veganism, putting a strong emphasis on fighting animal cruelty and being publicly

(18)

outspoken about it, whereas the latter type of vegans showed to be less outspoken about their vegan practices in presence of others, and especially non-vegans (Cherry,

‘It’s Not Just a Diet’: Identity, Commitment, and Social Networks in Vegans). Relating

this to motivations behind why one may become vegan, the author also pointed out how punk vegans were largely motivated by ethical reasons, whereas nonpunk vegans leaned more towards the health reasoning, outcomes in line with studies stating health and ethics to be the main drivers in veganism (Radnitz et al. 32). Drawing on similar conclusions in her study on authenticity in relation to veganism, Jessica Greenebaum differentiates between self-identifying vegans “who ‘eat’ vegan […] and those who ‘live’ vegan” (131).

What proved to be problematic in this case, was that nonpunk vegans appeared to be less “strict” about their vegan practices, as they in part still purchased non-vegan products in the form of clothing or cosmetic products, or even ate animal products on occasion (Cherry, ‘It’s Not Just a Diet’: Identity, Commitment, and Social Networks in

Vegans). In this sense, nonpunk vegans were for the majority of the time not adhering

to the official definition of veganism, encompassing the cutting out of animal-derived products in all aspects of life, a common point of critique by punk vegans in discussions about differing approaches to veganism throughout the study (Cherry, ‘It’s

Not Just a Diet’: Identity, Commitment, and Social Networks in Vegans 41). In this

respect, the question may arise whether the category Cherry defined as nonpunk vegans actually falls within the category of practicing veganism based on The Vegan Society’s definition, as “veganism is more than a diet: it is a philosophy and ethics” (Greenebaum 129), and nonpunk vegans generally regarded veganism as more of a diet than an overall lifestyle.

3.3. Communicating veganism

Beyond differing motivations for self-identifying vegans, Cherry also investigated the way in which vegans may or may not experience a sense of community, typically in the form of social networks between themselves and other self-identifying vegans. She even goes so far to state that some vegans (usually those categorised as “punk”) created a kind of “collective identity” as being vegan, forming a network based on the same beliefs and practices (Cherry, ‘It’s Not Just a Diet’: Identity, Commitment, and

(19)

through friends, music or even punk subculture, founded on beliefs centred around animal rights, nonpunk vegans learned about the cruelty free lifestyle by means of personal research (Cherry, ‘It’s Not Just a Diet’: Identity, Commitment, and Social

Networks in Vegans 56). Punk vegans were thus shown to be surrounded by a

stronger social network founded on common beliefs regarding veganism than nonpunk vegans (Cherry, ‘It’s Not Just a Diet’: Identity, Commitment, and Social Networks in

Vegans). In relation to how “strictly” punk versus nonpunk vegans adhered to

veganism in relation to The Vegan Society’s definition, social networks arguably play an important role in how consumption behaviours are influenced (Kennedy) as Cherry notes that having other vegans “in one’s social networks” affects the practices of self-identifying vegans (Cherry, ‘Veganism as a Cultural Movement: A Relational Approach’ 165).

In an era where creating and sustaining social networks has become increasingly easier through the presence of online spaces in the form of social media platforms, vegan “offline” networks can now be easily translated to the online environment. Self-identifying vegans wishing to read, see or watch content created and shared by other vegans, are now able to do so using keywords or hashtags on social media platforms such as Facebook or Instagram. However, as outlined previously, the way in which veganism may be performed as part of the self tends to vary in non-virtual life, raising the question whether this phenomenon also takes place online and moreover, how so.

4. Studying Instagram

4.1. Self-identity in photography

Whether online or offline, self-presentation and performance of the self can take on different forms. In the non-virtual world, what Goffman described as presentational cues, can be performed through tone of voice or body language, whereas online environments require textual or visual information to be shared about the self. Photography featuring the self is one way identity performance is facilitated, with the camera becoming the tool with which self-identity is affirmed (Caldeira 142). Whereas usually, the presence of other people would make up the stage on which a person may perform their identity, in the case of photography, the photograph itself can act

(20)

as the stage “allowing individuals the performance of their preferred image of themselves (Caldeira 142), which then may be looked at by others. According to José van Dijck, this is particularly the case due to changes in photography over the last couple of decades, as the focus of photography has moved to the individual, making it one of the main outlets in which self-presentation can be performed (60). By being able to change the way a person presents themselves in a photograph, notably through changes in posture or facial expression, they are able to direct the way others perceive them. According to Brian McNely, this shaping of others’ perception of the self through imagery is both “strategic” and “self-reflective,” putting the individual in a position of control in how others see them (2), making documentation of the self through the means of photography highly performative in nature.

Similar to non-virtual self-presentation practices, communicating the self through visuals on social media also requires the presence of an audience (Schreiber 147). Even though the majority of the most popular social media platforms such as Facebook, Twitter or Instagram feature visual content to some extent, the way in which it is presented through the platform’s design still differs, resulting in each platform attracting different types of audiences as well. By equipping the smartphone with a constantly improving high-quality camera, “photography is placed [at] the centre of the everyday experience,” with apps like Instagram facilitating the sharing of visual content with others (Martin qtd. in Caldeira 135). Unsurprisingly, Instagram has established itself as “the most popular photo capturing and sharing application” since its launch in 2010 (Hu et al. 595), with a total of 800 million users at the end of 2017 ('Instagram active users 2017). With some opinions stating that the online platform purposely “exists for people’s [self-promotion],” (Sheldon and Bryant 90), it thus represents an interesting object of study based on which identity performance can be investigated.

4.2. Curating the self on Instagram

Social networking sites in general, revolve around the individual user and their interaction with the site in an attempt to share, post and update content about themselves (Ellison et al. qtd. in Bakhshi et al. 966). In the case of Instagram, this sharing of the self is primarily encouraged through the curation of photographic or the occasional video graphic content, further supported by the option to accompany the visual with a written caption so as to give contextual information to the shared photo.

(21)

In line with boyd’s statement that social media allows people to “write themselves into being” (boyd) Instagram, in particular, could be seen as a platform allowing its users to curate themselves into being primarily through the sharing of visual content. This “curation” of the self does not only take place in users’ individual posts which appear in their followers’ news feed and ultimately create a “photo gallery” on their own profiles, but already in the way they present themselves through choice of username, profile picture and profile description, commonly known as the Instagram “bio.” However, while choosing a username presents the condition under which people may sign up for the platform, the name they choose for themselves, alongside what they do or do not display in their profile picture or description, is up to them. In this sense, such affordances allowing users to customise their profiles are “understood as potentiality” (Schreiber 146), even if they are typically followed by most users and thus recognised as the normative way to engage with the platform (Stanfill 1060). Additional options in the profile section, such as including links to external sources like websites or other social media profiles, further allow users to connect their Instagram selves to other (online) parts of themselves. Following how users thus decide to present themselves in the profile section of their account, photographs users decide to share on their accounts then build the base upon which their narration of the self is constructed (Caldeira 143).

Differently put, the social media platform allows its users to “narrate everyday life and practice” (Carah and Shaul 82). With the introduction of new elements such as Instagram “stories” in 2016 (Instagram), the app even allows for “live” updates, similarly customisable due to editing and filtering options. However, as Instagram worded it themselves in one of last year’s press releases, stories were introduced to allow the app’s users to share “all of their moments – the highlights and everything in between,” potentially placing “stories” in less of a “curated” setting, encouraging the sharing of more unedited or “raw” content (Instagram), which in turn is only available to view for 24 hours, making it ephemeral in nature.

Yet, considering Instagram’s mission statement as the platform having become “the home for visual storytelling,” with an emphasis on a passion for creativity from its users (The Team), the platform targets individuals who not only want to capture and share photographic content but also edit it. This concept of editing an image to one’s liking

(22)

before posting it on the social media platform is especially supported through the presence of in-app editing. This in-app editing of images featuring basic editing options such as adjusting brightness or contrast, or more advanced options such as the use of filters, invite the user to carefully modify the image they are about to share, creating what Zachary McCune describes as “personal craft culture” (39). Just as people would change or adapt the way they act in front of other people in “real” life in order to provide a certain image of themselves, Instagram allows its users to carefully alter an image they are about to share on their profile, available to be looked at by other people. In this sense, Instagram has become one of the main apps allowing “identitary dramaturgy” of the self, often in the shape of curating an idealised image of the self online (Caldeira 155). By being selective on what to share and what not to share on the smartphone app, users thus have the possibility to present themselves in a light that is perceived as favourable to them and which may lead to positive reactions by their audience.

Even though imagery of the self in the form of “selfies” for one, has proven to be popular amongst Instagram users, both in relation to people posting photos of themselves, as well as user engagement on such photos, literally showing oneself in the visual content does not act as the only way of identity construction within the app (Caldeira 146). Instead, users may also choose to share photographs of objects or their physical environment in an attempt to create an “identitary narrative” shaped by their lifestyle choices and practices (Caldeira 155). They thus make decisions on “how to capture, edit, and circulate” imagery narrating “their lived experiences,” using the app as what has been called an “image machine” (Carah and Shaul 71) and an outlet to present themselves to others online.

Whether the audience accepts the user’s performance of the self or not through the shared imagery, can then be recognised in the “feedback mechanism” of the app. Receiving a “like” on a recently posted image can thus offer “a small form of quick validation,” alongside positive comments, whereas the lack of “likes” or presence of negative comments may indicate the audience’s disapproval of the shared content (Caldeira 153). Based on this feedback mechanism, users can thus gain an insight into how others perceive their online identity on the app specifically and adapt it as needed to guide the impression others have of them. However, even though this

(23)

feedback mechanism in the form of likes and comments is available for anyone using the app, engagement and interaction on one’s own profile and content shared is affected by whether or not the user chooses to keep their account “private” (i.e. they specifically choose who has access to their content) or “public” (i.e. anyone can access their content, both in-app and using an online browser). Even though Instagram accounts are “public by default” (Bakhshi et al. 967), users are able to change their privacy settings so as to be in control of who can access their content. Whether a user chooses to share their self to an open public or only a distinct group of people within the platform, their awareness of who can or cannot be watching them shapes the way in which they may decide to present themselves to others.

(24)

CHAPTER 3: CREATING A METHODOLOGY TO STUDY

INSTAGRAM

1. Selecting Instagram accounts

Given the nature of this research project and its focus on individual identity performance, the analysis of vegans’ identity performance on Instagram has been conducted on a pre-defined set of Instagram accounts, making up a set-in-stone, stable data corpus to evaluate. Based on my own experience and online research, it has become evident that a lot of vegan-related Instagram accounts (or online content on the topic of veganism in general) revolve around food, resulting in highly followed vegan-related Instagram accounts being food-centric, with no clear indication or representation of the user behind the account. Whilst I am aware of these accounts existing and arguably being popular amongst users interested in veganism, and the food aspects of the lifestyle specifically, this research project focuses on identity performance of individual users through the lens of veganism. Thus, only highly followed vegan-related Instagram accounts clearly owned by individual people identifying as vegan, subsequently building veganism into their personal narrative on their profiles, are taken into account.

Overall, the selection process of Instagram accounts to be studied in this research project contained four essential components. First of all, only Instagram accounts that explicitly included the terms “vegan” or “veganism” in either their username or user profile description (“bio”) were taken into consideration. In including veganism in what Zhao et al. consider “explicit identity statements,” in this case in the form of “autobiographic descriptions” put together and shared to others by the user themselves (1820), it can be assumed that veganism plays an important role in their lives and in how they present themselves to others – so much that they deem it important enough to include in a relatively small section describing themselves at the very top of their profile. In other words, the goal is to study so-called “organised vegans,” these being vegans openly identifying as vegan in the presence of others (i.e. their online audience), with the possible, though not definite, goal of starting a conversation on the topic and getting others interested in it (Cherry, ‘It’s Not Just a

(25)

Instagram accounts to be considered had to be of “public” nature, meaning that they were accessible to anyone, whether through the app by other Instagram users or through different online browsers my non-Instagram users. This assured that the users’ audience or “public” was not restricted to a certain amount of people chosen by the user, but instead open to anyone interested in the user or the content they share. Moreover, only Instagram accounts with a relatively substantial following were of interest for this study. By looking at accounts with a minimum of 100.000 followers (and, more often than not, even several hundred thousand), the users owning the chosen accounts could be regarded as microcelebrities or “influencers” based on the notion of the potential influence they might have on their comparably large audience. Finally, in relation to audience size or follower count, only Instagram accounts containing written content in English were taken into consideration. Since follower count makes up one of the selection criteria, it appears logical to focus on English language accounts due to the language’s wide reach. Stylistic choices affecting users’ content shared on the platform, such as the language in which they communicate in, can arguably be “shaped by [the user’s] attempts to accommodate to their addressees” in so far that the user makes their content accessible to a specific group of people (Seargeant and Tagg, 161 ‘Audience Design and Language Choice in the Construction and Maintenance of Translocal Communities on Social Networking Sites’). By concentrating on English language accounts only, the respective users thus may be sharing content to an audience with “global potential” (Seargeant and Tagg 7, ‘Introduction: The Language of Social Media’), regardless of their own or their followers’ geographical location.

Built on these steps making up the selection process, finding Instagram accounts fitting the previously mentioned criteria required the use of different approaches. This was done in the combining of my own knowledge of vegan Instagram users based on my own following practices on Instagram, online research consisting of looking up well-known or highly followed vegan Instagram accounts, and lastly, crowdsourcing potentially fitting accounts in veganism related groups on social media platforms such as Facebook. Overall, these research approaches resulted in approximately 50 Instagram accounts. After about half of these were eliminated from the selection process due to follower numbers below 100.000, the remaining half was then narrowed down to ten individual accounts, five of which belong to male users and the

(26)

other five belonging to female users. These final ten accounts were chosen on the basis of diverse representation of vegan-related content in relation to lifestyle choices or identity performance, so as to get a multi-dimensional insight into how vegans may differ in performing vegan identity on the platform, depending on their own values and ideals.

2. Chosen accounts

Something that quickly became noticeable during the research process based on the previously outlined approach, was that distinctly more highly followed or “popular” vegan Instagram accounts belonged to women, rather than men. Nevertheless, in order to gain a wider insight into vegan identity performance on Instagram, I chose to study five female, as well as five male accounts, potentially leading to interesting differences and contrasts between how each user performed their vegan identity, in part related to their gender, for one.

The selected accounts for this study are as follows; the chosen five women’s accounts consist of @ellenfisher, @fullyrawkristina, @naturallystefanie, @highcarbhannah and @frommybowl (see table 1). @ellenfisher is the account of a mother of three who lives on Hawaii and shares a lot of family, nature and food-related content, as well as advertising her YouTube channel and “health tips + recipe ebook” in her profile description. @fullyrawkristina is run by a self-named “author & speaker” and “badass unicorn Youtuber,” living and advertising a “fully raw” lifestyle, in which solely raw foods, such as fruits, vegetables, nuts and seeds are consumed, and likely one of the highest followed vegan Instagram accounts, holding one million followers up to date. @naturallystefanie is run by a Scottish YouTuber, athlete and entrepreneur collaborating with fitness brands like Gymshark (activewear) and Myprotein (online fitness supplements). Through her fitness-related content and her collaboration with fitness-related brands, she highlights the vegan fitness community, simultaneously advertising her own “12-week challenge” in her profile description. @highcarbhannah is an account centred around vegan weight loss, with the user broadcasting her own weight loss on a vegan lifestyle (“-70lbs”) and linking her own recipe books and weight loss guides, as well as her YouTube channel in her “bio.” Finally, @frommybowl is the

(27)

Instagram account of a “vegan recipe developer,” emphasising the simplicity and deliciousness of vegan food and linking to her website, as well as her recipe ebook. On the other hand, the five selected men’s accounts are: @therawboy, @nimai_delgado, @veganfatkid, @conscious_muscle and @theveganzombie (see table 2). @therawboy is likely (one of) the most followed vegan men’s accounts with over half a million followers and revolves mainly around “easy and flavourful” recipes. Similar to the above-mentioned women’s accounts, @therawboy also links to his YouTube account which people can watch to see his “journey” and recipes. @nimai_delgado is the account of a “born” vegetarian (“never eaten meat”) and engineer, recognised as a “pro” of the International Federation of Bodybuilding and Fitness (IFBB), advertising both his YouTube channel as well as a discount code to Vivolife, an online vegan, natural foods supplements shop he likely works in collaboration with. @veganfatkid advertises “eating thru the best of Los Angeles so u don’t have to,” focusing on (often fast) food, but also highlighting fitness with a link to the user’s own website where his collection of activewear can be purchased. @conscious_muscle is run by a self-proclaimed online vegan fitness coach who is “vegan for everything,” with a link to his own website where meal and workout programmes can be found, but the user also highlights his monthly donations to animal sanctuaries. Lastly, @theveganzombie is owned by an award-winning cookbook author and YouTuber sharing recipes, travel, lifestyle and vegan-related content on his channel, linking both his website and latest YouTube video.

(28)

username followers name in bio

@ellenfisher 445k ELLEN FISHER

@fullyrawkristina 1M YouTube: FullyRawKristina 🍉

@naturallystefanie 212k STEFANIE MOIR | Vegan💫

@highcarbhannah 191k Hannah Howlett

@frommybowl 268k Caitlin Shoemaker

Table 1: women’s Instagram accounts1_.

username followers name in bio

@therawboy 595k Neto

@nimai_delgado 207k Vegan IFBB Pro Nimai Delgado

@veganfatkid 216k VeganFatKid

@conscious_muscle 131k Online Vegan Fitness Coaching

@theveganzombie 134k The Vegan Zombie

Table 2: men’s Instagram accounts.

3. Employing digital methods for textual and visual content analyses

Whilst elements such as a user’s profile picture or their profile description function as “first impressions” on Instagram, which arguably make up an important part in how the user is perceived by others (Goffman 22), this research project goes beyond user profiles, taking into consideration both visual and textual content shared by the user in form of images and respective captions posted. Although online identity tends to be studied mainly through the analysis of user profiles (Vásquez 6), I would argue that a more thorough analysis of recently shared user content will provide a more in-depth insight into how the chosen users perform their vegan identity on Instagram. Even though visuals are a “key means” by which identity can be performed and presented in online environments (Highfield and Leaver 49), taking into account exclusively

1_{I am aware of the ethical implications that come with openly naming the Instagram users analysed} for this study. However, given the “public” nature of all ten accounts used, the individual users are

(29)

imagery shared on Instagram may not provide a full picture of a user’s identity performance, due to the potential lack of context in which the image was shared. To create and include this context into the study and analysis of the chosen vegan Instagram accounts, captions accompanying images will also be taken into account in an attempt “clarify intent and context” of the users’ published photos (Highfield and Leaver 51). This research project thus aims to study digitally native content, or differently put, content “born” on the platform (Rogers, ‘The End of the Virtual: Digital Methods’ 19), taking into consideration both visual as well as textual content created and published for the precise purpose of being consumed by the users’ audiences. In order to get a somewhat in-depth insight into how the chosen ten Instagram users perform veganism as part of their identity, I settled on studying each user’s 100 most recent Instagram posts. Since this would result in the study of a total of 1000 Instagram posts, I decided to make use of digital methods both for data collection purposes, as well as for its analysis. Thus, I employed scraping to collect my data, as scraping online content represents one of the existing “web techniques” to aid and conduct the gathering of relatively large amounts of data (Rogers, ‘Introduction: Situating Digital Methods’ 1). Furthermore, I decided to use a vision API for “broader analytical practice” (Weltevrede 30) and visualisation tools to both analyse the previously gathered data, as well as create visualisations of the results to facilitate comprehension.

3.1. Instagram scraper

In regard to scraping tools, Esther Weltevrede argues that different types of scrapers need to be implemented depending on the site or platform of interest. Given the differing formats and affordances of each platform, scrapers allowing for data collection thus need to be tailored “for the specifics of the sites,” so as to produce a useful outcome (Weltevrede 30). After my initial research into finding an appropriate scraper to collect Instagram data of interest to my research project, I decided to use the free and publicly available Instagram scraper hosted on GitHub. Created by Richard Arcega under the username rarcega, it is written in Python and executed through the computer’s own terminal (GitHub, ‘Rarcega (Richard Arcega)’). The Instagram scraper proved to be appropriate for my research purpose mainly due to its

(30)

customisable nature, which allowed me to specify which accounts I would like to scrape, as well as how much and which type of content I would like to collect.

In an attempt to create a comparable data corpus between the ten chosen Instagram accounts, I scraped all of the imagery and textual content of the 1000 relevant Instagram posts on one day and within a relatively short time frame; 10March 2018 between 6:00 and 6:30 pm. The collected data was then cleaned appropriately so it could be run through different APIs or visualisation tools. The first step in doing so consisted of making sure that any text making up the collected image captions that was not in English but in another language, serving as a translation of the English text, was removed. This was mainly necessary in the case of @therawboy’s account, as the user first posts his captions in English, followed by a Spanish translation underneath. Captions of @fullyrawkristina also contained a few Spanish translations, mostly in the form of hashtags, which in turn were removed as well. Doing so ensured that only English written content was being used for the analysis overall. Furthermore, from a textual analysis perspective, the scraped captions had to be cleaned from so-called “filler words” such as “the,” “and” or “is” so as to assure that these “meaningless” words did not drown out any more meaningful content or keywords for a later textual analysis. To enable a separate analysis of visual and textual content using different tools, separate documents were created, containing, respectively, the image links and captions, ahead of the analysis. These were also filtered, in additional documents, to isolate content specific to male and female accounts. By separating and organising visual and textual content, it then became easier to run it through APIs and visualisation tools for analytical purposes. Separating male from female data would then also allow for a more gender-specific analysis of visual and textual content, something that will be addressed further on.

3.2. Visual analysis: Google Vision API

Given the large amount of visual content scraped from the various Instagram accounts, I chose to employ the Google Vision API openly available online and run through GitHub’s Memespector programme, written by GitHub user Bernhard Rieder under the username bernorieder (GitHub, ‘Bernorieder (Bernhard Rieder)’). The Google Vision API facilitates the grasping of content present in visuals by classifying

(31)

large amounts of images into categories, resulting in an overview of the images’ content. Overall, the programme has the ability to detect anything from objects to faces, as well as the presence and the content of text on imagery (Google Cloud). Due to the programme’s potential to look at hundreds or even thousands of images within a relatively short time frame, it is able to go beyond what the human eye can observe and detect in such a large corpus, making it a potentially more effective analytical approach in contrast to manual analysis.

One of the API’s weaknesses or limitations that has been studied in academic context is the potential lack of reliability of the programme when it comes to poor quality or manipulated images that cannot be recognised correctly in return (Hosseini et al.). Even though the Vision API may be superior to the human eye when it comes to analysing large amounts of visual content, a study has shown that through the simple editing of images, namely by adding “noise” to them and thus changing the images’ clarity, the programme was no longer able to recognise the content correctly, resulting in incorrect categorisation of what could actually be seen on the images (Hosseini et al. 2). However, given the relatively high quality and clarity of the images in this context, the risk of the API wrongly recognising the content of the visuals, and thus categorising it incorrectly as result, would be relatively low. In an attempt to test this assumption, I selected a handful of images of different quality (higher or lower resolution, more or less bright) from my data set in order to conduct a manual verification. Figures 1 to 4 show the results of this testing technique2_.

(32)

Figure 1: tags associated by the Google Vision API to a photo of @therawboy.

(33)

Figure 3: tags associated by the Google Vision API to a photo of @ellenfisher.

Figure 4: tags associated by the Google Vision API to a photo of @fullyrawkristina.

In an attempt to further assure the accuracy of the API’s content categorisation, as well as to facilitate the understanding of the programme’s output, the resulting data was cleaned using Alteryx. In doing so, only API results bearing a probability of 55% or higher in relation to the tool’s accurate recognition of visual content were kept for future analysis. This means that any image tag or category recognised with a

(34)

probability of less than 55% by the programme were removed from the dataset. Furthermore, only the six highest tags associated to each image by the API were kept, meaning that the six tags with the highest percentage of probability would be used in the process of visualising the API’s output so as to determine which kind of visual content is featured the most in the data set, based on the API’s output.

This process of cleaning the programme’s output facilitated the final step in visualising the results using Gephi. In order to be able to visualise the Vision API’s output based on the 1000 images it analysed, the table containing the cleaned output data had to be run through Table2Net. Table2Net is an online tool which aids to “extract a network from a table” in CVS format (Table 2 Net) and thus makes the content of a table readable in visualisation software like Gephi. Using Table2Net, I created a network with one type of node. In order to visualise the most commonly recognised image tags by the API, the nodes would represent the tag determined by the programme based on an image’s content, whereas the edges connecting the nodes to each other, would represent individual images and how they may relate to different tags.

Finally, uploading the resulting network in Gephi would allow for a visual representation of the Google Vision API’s results. The reasoning behind choosing Gephi to visualise the results I obtained from the Google Vision API is rooted in the adaptability of graphs given the software’s various layout options. Depending on the layout, data visualised in Gephi allow for the recognition of thematic clusters in relation to the content of the analysed data, as well as the clusters’ relation to each other. In my study, this translates to a clear visualisation of not only the frequency of occurrence of certain types of visual content but also how these types of content relate to each other thematically.

Thus, in order to obtain a readable graph, the layout ForceAtlas 2 was selected in Gephi with the scaling set to 40. Furthermore, node and label size were ranked according to degree, meaning that nodes and their labels representing tags recognised by the API more frequently would be bigger in size, with the minimum size being set to 0.5 and the maximum to 70. In order to understand how the API associated tags to larger categories, the graph was coloured according to modularity, resulting in

(35)

only the eight biggest tag categories were coloured. The resulting visualisation in form of a network displaying image tags appointed by the API thus allows for an analysis of the most often recognised elements in images, as well as how different tags relate to each other, resulting in “tag clusters” consisting of frequently used tags being surrounded by related, but less frequently occurring tags, coloured accordingly. This procedure resulted in one main Gephi map, the “mother map,” visualising the Vision API’s results for all 1000 studied Instagram images after the original cleaning. Off of this, another two, gender-specific Gephi maps were created in order to allow a side-by-side analysis of women’s versus men’s visual content.

3.3. Textual analysis: Voyant Tools

In order to investigate and understand the content of the 1000 Instagram captions accompanying the respective images, I utilised Voyant Tools, a “free, versatile suite of web-based, text analysis and visuali[s]ation tool” (Welsh 96). The publicly available online tool allows for the analysis of either text that is uploaded to the website or pasted directly from its original source. Having combined all 1000, but also the respective 500 captions from both the male and female accounts in separate documents, I opted to upload the individual files containing the captions. Apart from displaying the total amount of words, as well as the amount of unique form of words used in the uploaded textual corpus, Voyant Tools also ranks the words from most to least used, highlighting the five most used words, and creating a word cloud of the 25 most frequently occurring words, which can be further adjusted to show a higher amount of “most used words.” Additionally, the tool also shows the collocate words of the most used words, as well as the distinct sentences in which they appear, facilitating the understanding of context and relevance of wording used in the captions. The tool thus covers “any scholar’s text analysis needs,” especially when it comes to “frequency and context of specific words” (Welsh 97). In determining the most frequently occurring words used in the Instagram captions, Voyant Tools thus allows for a better overarching analysis and understanding of the visual content by providing contextual information by offering a greater understanding of the images’ accompanying text. Making use of this tool hence resulted in a word cloud showing the 25 most used words throughout all 1000 captions and two separate ones representing the 25 most used words from the captions of male, as well those of female accounts. To get a better idea of the context

#vegansofInstagram: Studying Vegan Identity Performance on Instagram