‘Good food’ in an Instagram age: Rethinking hierarchies of culture,criticism and taste

A brief history of food criticism

The formal practice of critiquing what we eat began in France. ‘French writers Alexandre Grimod de la Reynière (1758–1837) and Jean Brillat-Savarin (1755–1826) are generally recognised as the founders of food, wine, and restaurant criticism, of which the connoisseurial review is a central element’ (Kobez, 2018: 264). Their work positioned elite expertise as key to legitimising culinary judgement. Here, knowledge production is unambiguously the preserve of experts, and linked to language, literacy and the printing press. This fits neatly with early restaurant criticism’s context of production: the Enlightenment and its valourisation of the scientific method, progress, objectivity and Truth’s singularity (Porter, 2001) on one hand, and its classed and racialised Eurocentrism on the other hand (Amin, 2009: 178 passim).

These modernist values and knowledge claims infuse the Michelin Guide – arguably the world’s most esteemed institution of restaurant criticism (Vincent, 2018). Printed annually since 1900 (Johnson et al., 2005: 171), ⁴ the Guide aligns itself with expertise, professionalism, impartiality and with substantial cultural and economic capital, Frenchness and a Eurocentric reading of culinary worth (Vincent, 2018: 2). The Guide consists of short, written restaurant reviews produced by rigorously trained, anonymous inspectors (Vincent, 2018). These inspectors decide which establishments receive Michelin’s coveted stars and how many stars they get. Not only are its inspectors anonymous, Michelin also refuses to disclose how many inspectors it employs (c.f. Rémy, 2004). Neither does the Guide publish guidelines used by its inspectors, explaining that doing so would create a formula by which restaurants could earn and retain stars (Johnson et al., 2005: 172). Michelin claims this lack of transparency is essential to producing an objective guidebook (Bouty et al., 2013; Lane, 2013). Cynical analysts might suggest this institutional opacity works to preserve, if not enhance, Michelin’s power by endowing its guidebooks with unchallengeable omniscience and authority.

The Michelin Guide is also associated with wealth and high-end, Eurocentric gastronomy (Vincent, 2018). It consistently focuses on expensive meals, and the high-cost ingredients, labour and material infrastructures required to produce them. The hefty price tag of eating at a Michelin-starred restaurant necessarily limits who can dine there. Michelin’s economic exclusivity also converges with its cultural claims of what counts as ‘good food’. The Guide’s reviews and award hierarchy demonstrate a stylistic preference for tasting menus and aesthetically considered, often austere plates of meticulously arranged food. The enjoyment of a Michelin-approved meal thus seems predicated on having the sort of cultural capital that appreciates a deconstructed spear of asparagus served on a largely empty plate. Michelin’s cultural bias also extends to its privileging of French cuisine. As this research finds, restaurants awarded three Michelin stars – the Guide’s top honour – are disproportionately French, serving French food and/or employing French culinary techniques.

This cultural bias reflects the Guide’s historically significant bond with France. Today, the Guide purports to identify the world’s best restaurants but it was originally ‘[devised] . . . for rating the “best” restaurants in France’ (Harp, 2001: 225) and its early editions focused solely on establishments within French borders. The Guide was also, and continues to be, published by a French company: Michelin, a now-global tyre manufacturer whose revenues topped €24 billion in 2019 (Michelin, 2020). When the company was founded in 1889, the tyre market was niche, reflecting the novelty and considerable expense of automotive travel at that time. ⁵ The first version of Michelin Guide appeared when ‘there were fewer than 3,000 vehicles [in France]’ (Meeten, 2018). To help popularise automotive travel, Michelin became France’s unofficial cartographer, mapping and numbering the nation’s roads and equipping towns with signage (Harp, 2001: 74–85). It published the Michelin Guide, which featured restaurant and hotel listings, as a way of further growing its tyre business.

In seeking to attract a financially elite clientele, the tyre maker also drew on the classed mythology of French artisanal production, ‘playing on the French bourgeoisie’s association of early French industrial goods with artisanal articles de luxe’ (Harp, 2001: 27, emphasis original). For the Michelin Guide, the company introduced an anthropomorphic mascot, Bibendum, who embodied the ideals of upper-class ‘Frenchness’. Harp (2001) suggests that

Bibendum revealed and . . . reinforced gender, racial, and class hierarchies of early-twentieth-century France . . . [He] was a white, upper-class French man, often a veritable man about town (mondain), who would advise or dominate fellow men, ‘conquer’ women, and control racial inferiors [sic]. He embodied . . . strong assumptions in prewar France about what a well-off man should be. (p. 16)

Bibendum and the Michelin Guide thus fused culinary taste with whiteness and wealth. Today, the Guide continues to exercise considerable power over popular notions of ‘good food’. The Guide’s annual releases are much anticipated by gastronomes and news media alike. The creative industries routinely invoke Michelin’s prestige – for instance, in the films A Thousand-Foot Journey (2014), Burnt (2015) and Chef (2014) and in Netflix’s glossy documentary series Chef’s Table (2015–2019.) Foodie bloggers often describe their dining activities with reference to a restaurant’s Michelin stars (e.g. Hayler, n.d.). Chefs, too, contribute to Michelin’s symbolic power. In 2003, Bernard Loiseau committed suicide after allegedly fearing that his restaurant would be downgraded from three to two stars (Boucher, 2003). Another three-star chef asked that his restaurant be removed from the Guide altogether because he could not cope with the pressures of three-star status (Street and Ganguly, 2017).

Despite the Michelin Guide’s enduring authority, its primacy as chief arbiter of culinary quality has been challenged by widespread adoption of the participatory web. It is therefore important to consider whether and how restaurant criticism has changed in the digital era of user-generated content. Contemporary food media has purchase over the voices and ideas that are respected and reproduced, and those that are not. As Victoria Jean Zeamer (2018) argues,

While the underlying purpose of the construction and consumption of food texts remains the same from analog to digital form, the authority of food culture and its complimentary narrative control has shifted as a result of the convergence of food texts and digital media affordances. (p. 3)

As the production of food texts moves increasingly online, restaurant criticism is no longer the exclusive domain of experts and professionals. Online food media allows (just about) anyone to have their say. This democratisation of voice has fuelled OCR platforms like TripAdvisor and Yelp, where non-experts rate and review restaurants. Drawing on Benkler’s (2006) ‘seminal work on the democratizing effect of the Internet’, Mellet et al. (2014) suggest OCR sites democratise culinary criticism in two ways: by allowing more eating establishments to get reviewed and by enabling more people to do the reviewing (Mellet et al., 2014: 7). These digital platforms can thus be read as anti-elitist, because they purportedly undermine the power of traditional media gatekeepers; accessible, because they are typically free and easy to use; and empowering, because they allow anyone to share an opinion.

Others have similarly suggested that social media democratise culinary expertise and judgement, and inform consumer decisions about where to eat (Beuscart et al., 2016; Mellet et al., 2014; Onorati and Giardullo, 2020). From OCR platforms and online foodie communities like Chowhound (Watson et al., 2008) to personal food blogs (Cox and Blake, 2011), YouTube channels (Lupton, 2020) and food-themed Instagram posts (McKeown and Miller, 2020; Walsh and Baker, 2020), social media sites, apps and networks are now influential and mainstream venues of culinary judgement and representation (Lupton and Feldman, 2020). The emergence of social media food spaces and foodie ‘produsers’ (Bruns, 2007) coincides with new claims (and claimants) to culinary knowledge and authority. Restaurant criticism is no longer limited to institutions like the Michelin Guide, nor is the practice necessarily linked to claims of objectivity and professionalism.

This expansion of participation is not without its detractors. As Richard Schnickel (in Kobez, 2016) laments, ‘Criticism – and its humble cousin – reviewing – is not a democratic activity. It is, or should be, an elite enterprise [. . .] a purely “democratic literary landscape” is truly a wasteland, without standards, without maps, without oases of intelligence or delight’. Schnickel is surely pleased by the Michelin Guide’s continued existence. Nevertheless, the social media-based amateur restaurant critic is unlikely to disappear anytime soon. Instead, amateurs’ culinary knowledge production offers insight into the endurance of traditional institutions of culinary authority and their distinct claims. It is significant, for instance, that OCR sites like TripAdvisor are full of negative reviews of Michelin-starred establishments (Vásquez, 2021; Vásquez and Chik, 2015), with diners complaining about small portions, poor value for money, staff snobbery and leaving hungry despite having spent a fortune.

It is also significant that today, foodie attention seems to have shifted from blogs and OCR sites to image-sharing platforms like Instagram. This shift has important consequences for how, what and whose culinary knowledge gets codified. Different platforms have different affordances, political economies, cultures and norms, and this impacts what qualifies as a restaurant review on a given site. It affects what reads as legitimate, where not all social media platforms are created equal; neither are all social media-based restaurant reviews. It is noteworthy, then, that where blogs and OCR sites prioritise narrative accounts of eating out, Instagram prioritises visual representations thereof.

Instagram was born in Silicon Valley in 2010 and acquired by Facebook for US$1 billion in 2012 (Leaver et al., 2020). As indicated earlier, food is one of the platform’s most popular content categories. The most recently available data show that the platform had 1 billion monthly active users in June 2018, ⁶ (Constine, 2018) and that year, it ‘generated an estimated 6.84 billion U.S. dollars in revenue. The . . . app’s revenue [was] projected to more than double to 14 billion U.S. dollars in 2019’ (Tankovska, 2021). However you measure it, Instagram is big and it is influential. And despite its global reach, Instagram is an American company, reflecting the acquisitive, ‘move fast and break things’ ideology of Silicon Valley capitalism (Little and Winch, 2021).

While Instagram offers space for individual, non-commercial expression and (re)invention, it is simultaneously a platform of corporate advertising, sponsored content, ‘influencers’, and content bans (Leaver et al., 2020). Instagram thus reflects the tensions of contemporary digital culture: tensions between the Internet’s democratic rhetoric and its Wild West mythology where ‘anything goes’, and the Internet’s (and social media’s) increasingly commercialised, professionalised register. And unlike earlier forms of user-generated restaurant review platforms, Instagram prioritises the visual over the textual. What does all this mean for restaurant criticism on Instagram? What qualifies as ‘good food’ within this ecosystem? The rest of this article attempts to find out.

Building two datasets

My research design began with place. After all, eating is a territorialised and material practice. Place is also a prominent feature of Michelin’s and Instagram’s topologies. The Michelin Guide is location-specific and by 2016, 28 different editions of the Guide were being published, covering dining destinations in 25 countries across the Americas, Europe and Asia (Michelin, n.d.). Place is also an important organisational device on Instagram, with ‘foodstagrammers’ (Vila et al., 2021) routinely using the platform’s geolocation feature to tag the restaurants in their posts. This automatically hyperlinks a post to a restaurant’s location on a map.

Against this technical background, I selected London as my geographic focus: a ‘global city’ of over 9 million people and 300 languages (Serrant, 2018; Greater London Authority, 2019; Sassen, 1991), and a renowned foodie destination (Ellwood, 2016). London is also a site of extreme wealth concentration and inequality (Derbyshire, 2017). The city’s complex, multi-variegated cultural and economic profile therefore provides an ideal territory through which to observe culinary hierarchies of representation. Which of the city’s restaurants gain recognition in Michelin and Instagram, and on what terms?

I constructed two datasets to answer these questions. The Michelin dataset contains London’s award-bearing restaurants, as specified in the Guide’s (Michelin, 2019) Great Britain and Ireland edition. This constitutes 113 restaurants awarded with a Bib Gourmand, one-star, two-star or three-star ranking, out of a total 402 London restaurants listed in the 2019 Guide. My Instagram dataset contains all London restaurants featured on 10 London-based, foodie Instagram accounts, and posted during October 2018 – the same month the Guide’s 2019 Great Britain and Ireland edition was published. The Instagram dataset contains 126 unique restaurants, culled from 175 qualifying posts (i.e. 49 restaurants were posted more than once).

Whereas the Michelin Guide furnished a ready-made dataset, identifying relevant Instagram accounts was a murkier affair. I made a list of prospective accounts by using Instagram’s search feature to find accounts containing the phrase ‘London food’ and by conducting Google searches for the phrases ‘London Instagram food bloggers’ and ‘London foodies on Instagram’. I then used five criteria to evaluate the suitability of the accounts I found, requiring them to (1) be London-based, (2) primarily post about restaurants, (3) not be aligned with a restrictive diet, (4) demonstrate substantial reach ⁷ and (5) be run by individuals not working in the food sector. To further harmonise and make broadly comparable the Michelin and Instagram datasets, I coded all restaurants according to three elements: culinary nationality, Michelin award ranking and cost.

I coded each restaurant by the nationality of its cuisine in order to locate the geopolitical contours of Michelin’s and Instagrammers’ culinary representations. Nationality became a proxy for cultural identity. Through this coding, I wanted to understand which culinary traditions were represented on both platforms, in what volume and at what levels of esteem. As previously noted, Michelin is routinely – and credibly – accused of privileging French gastronomy over other cuisines and geographies (Gill, 2012; Johnson et al., 2005). Given the Guide’s history, its high regard for French culinary techniques and traditions is in and of itself not necessarily problematic. Trouble arises when considering the Guide’s global reach and its evaluation of other food cultures through a French lens. A twin aim, then, of my restaurant nationality coding was to understand the extent to which the Guide continues to reproduce a French bias and whether this bias is reproduced by Instagram reviewers.

Certainly, nationality coding is a conceptually problematic instrument of division and distinction (see ‘Limitations’ section). But it nonetheless persists in the food world and needs to be treated as a meaningful epistemic and ontological category. Nationality informs the way restaurants market themselves, and it is a frequent means by which diners decide what and where they want to eat. In today’s algorithmic age, it also structures the way search engines and social media platforms organise culinary knowledge. In other words, national cuisine labels matter. To identify a restaurant’s culinary nationality, I reviewed how each described its cuisine on its website, examined menu content, and considered how each establishment was culturally-cum-geopolitically coded by Google ⁸ and by the OCR platform TripAdvisor. I also consulted the geopolitical labels contained in the Michelin Guide, but given the inconsistency of that labelling system, ⁹ I deferred to restaurants’ online self-representations, and to Google’s and TripAdvisor’s classifications.

Next, I coded each restaurant by its Michelin award ranking – or lack thereof. Which restaurants earned one, two or three stars, a Bib Gourmand or none of these? This provided insight into the inequalities of Michelin’s award distribution. This technique also helped me ascertain how restaurants in the Instagram dataset mapped onto Michelin’s hierarchal typology. To what extent did Instagrammers engage with restaurants outside Michelin’s recognition? To what extent did they reproduce Michelin’s understanding of culinary quality?

Third, I coded each restaurant by the average cost of a meal there. This offered a concrete way of locating the economic privilege of eating out. Using cost indicators from Google search’s Knowledge Panel (Schaer, 2019), I assigned restaurants to one of four price tiers: £ (inexpensive), ££ (moderate), £££ (expensive), ££££ (very expensive). ¹⁰ These tiers, though broad and imprecise, effectively gesture to the economic dimension of Instagrammers’ and Michelin’s representational practices.

To be clear, this research does not consider price to be a proxy for quality. But price is important. London is a city where 39 percent of children live in poverty (Leeser, 2020). There are many people here who cannot afford to eat out – much less eat out at a Michelin-starred restaurant. It is therefore important to draw attention to the cost of meals at establishments awarded Michelin honours and Instagrammer attention. Cost data also offers insight into how two leading platforms of food criticism – one notably exclusive, the other allegedly ‘accessible’ – contribute to classed understandings of ‘good food’. In other words, by comparing economic representation in Michelin and Instagram, we can test scholarly claims around social media’s democratising qualities against Michelin’s presumed elitism.

Finally, I cross-referenced the nationality, award ranking and cost data. This highlighted the specificities of culinary representation in both platforms, and how these operate across cultural and economic terrains. Moreover, it allowed me to identify interconnections between cultural and economic representation and exclusion, within and between the Michelin and Instagram datasets.

Limitations

This research design is not without its limitations. Foremost is how I went about choosing Instagram accounts. Despite using carefully considered criteria, it is impossible to remove all arbitrariness from what is fundamentally a subjective process of selection. In short, the 10 Instagram accounts I selected may not have been the most appropriate. My method of account selection also did not – and could not – address questions around accountholders’ ‘offline’ positionality and power. Nor did it consider how pre-existing privilege might contribute to an Instagrammer’s online success and visibility. These blindspots prop up the amateur–professional distinction this article draws between Instagram and the Michelin Guide. But as my findings below make clear, this distinction itself proved problematic.

Another limitation of this research is its utilisation of cuisine ‘nationality’ categories. Such labels ignore regional specificities, and obscure the mobility of influence and the fluidity of cultural identity. Nationality markers advance monolithic readings of food culture, giving rise to myths of culinary ‘authenticity’ and ‘purity’ (Mannur, 2007; Oum, 2005). But whether one classifies pizza as Italian, American or Turkish depends on who (and where) one asks. Equally, seemingly national foods can conceal distinctly ‘foreign’ origins. For example, the quintessentially British dish of fish and chips derives from Spain’s expulsion of its Jewish population in the late-15th century (Roden, 1996: 185). Does that make battered fried fish Spanish, Jewish or British? My point is that uncritical reliance on nationality labels conceals the complexities of cultural formation. Yet as mentioned, the ‘national’ frame continues to matter to culinary practice and knowledge production.

Inclusive

Culinary representation in my Instagram dataset is far more inclusive than it is in the Michelin dataset, in cultural and economic terms. There are 32 national cuisines in the Instagram posts analysed, while the Michelin Guide’s award-bearing restaurants represent only 17 national cuisines. Most (88%) of the Michelin cuisines are featured in the Instagram dataset. Conversely, only 53 percent of the cuisines in the Instagram dataset appear in the Michelin data. These data clearly show that cuisine nationality representation on Instagram is more diverse than in the Michelin case. However, the fact that most of the Michelin cuisines were also featured in the Instagram data suggests that the IG incorporates and reproduces some of Michelin’s claims about which culinary cultures matter.

Cuisines unique to Instagram include those of Asia (e.g. Vietnamese, Korean), Central and South America (e.g. Peruvian, Mexican), the Middle East (e.g. Lebanese, Greek), and the geographically non-specific (e.g. vegan). Thus, the majority of Instagram-only cuisines represent non-western or non-Global North cultural geographies. Only one European entry – German – features in the list of Instagram-only cuisines. By contrast, the Michelin dataset reveals the persistence of the Guide’s long-standing European and specifically French bias: French establishments account for nearly 19 percent of the Michelin dataset and two-thirds of its three-star restaurants. Meanwhile, French restaurants constituted less than 5 percent of the Instagram dataset.

Notable differences also emerge between Instagram and Michelin cultural representation when one zooms out from individual countries to global regions. That analysis shows the substantive degree to which Michelin values European culinary identity over and above any other: more than 75 percent of Michelin award-bearing restaurants serve European cuisine. Instagrammers also preferred European cooking over other cuisines. However, European restaurants represented only 44 percent of the Instagram dataset. Alongside this, Instagrammers frequently posted about New World restaurants, whereas the Guide gave scant attention to North American food cultures and did not recognise any restaurants specialising in South Pacific or Central or South American cuisine.

Economic representation in Instagram and Michelin also demonstrates how each platform defines ‘good food’. Instagram-featured restaurants, on the whole, were considerably less expensive than their Michelin counterparts. As discussed earlier, I assigned each restaurant to one of four price tiers. In the Michelin dataset, more than two-thirds of restaurants fell into the two most expensive tiers and the highest concentration of Michelin restaurants (43.4%) was in the top (££££) price tier. By contrast, only 11 percent of Instagram restaurants were situated there. The highest concentration of Instagram restaurants (60.3%) was in the moderate (££) tier. These cost data can be read to indicate that Michelin and Instagram are economic foils of each other; where Michelin leans expensive, Instagram leans more affordable. However, the data also show that the least expensive tier (£) in both datasets has the fewest entries. The Michelin dataset contains only one restaurant (0.88%) in this price category. The Instagram dataset has eight restaurants (6.4%) in the bottom tier. This makes clear that neither Michelin nor Instagram food critics favour inexpensive dining.

Cross-referencing cultural and economic representation across Instagram and Michelin reveals additional insights into how cultural identity and economic capital jointly contribute to the production of culinary knowledge and distinction. In the Michelin data, nearly 70 percent of restaurants serving European cuisine are situated in the top two price tiers. But in the Instagram dataset, almost 60 percent of its European restaurants serve food in the bottom two tiers. The data also show that there is little to no correlation between a Michelin restaurant’s price tier and its level of esteem within the Michelin hierarchy of stars. Over 90 percent of all Michelin award-bearing restaurants reside in the Guide’s two lowest tiers of esteem: the one-star and Bib Gourmand categories. Yet, in my cost coding, only one-third of Michelin’s restaurants resides in the two lowest price tiers.

Instagrammers also failed to reproduce Michelin award hierarchies. Of the 126 restaurants in the Instagram dataset, only 10 featured a Michelin award-bearing restaurant. None of these were three-star establishments. But locating cultural and economic hierarchies on Instagram proved a far trickier exercise than in the Michelin case. This is largely due to Instagram’s tierless nomenclature. The platform’s reviews do not articulate quantitative ratings or a Michelin-like hierarchy of stars. Instead, it appears that a restaurant is positioned as ‘good’ simply by being the subject of an Instagram post. This mirrors Zeamer’s (2018) findings about food blogging, wherein a participant explained that on their blog ‘We don’t rate the dish, and we don’t critique it. If a restaurant is not good, we just won’t post it’ (p. 60).

Professionalised

The second feature of the Instagram gaze is that it favours professionalism over amateurism, in both content style and content creator. The images in the Instagram dataset reflect an aesthetic professionalism, and consistently reproduce the editorial qualities of magazine spreads and advertising imagery. Image composition is also strikingly consistent across the 10 accounts analysed, deploying conventions typically found in catalogue photography – for instance, flatlays (i.e. overhead shots), tablescapes and tightly cropped object closeups (Leaver et al., 2020; Manovich, 2016).

The IG, then, can be understood as a culinary vernacular that rewards uniformity and professionalism in image-making. Restaurants in the Instagram dataset all look remarkably alike. Their aesthetic representation seems to bear no relation to the cuisine served or the price of the meal. Rather, the images reflect a cultural standardisation and pseudoindividualism (n.b. Adorno, 1941) – a set of genre conventions and representational orthodoxies bolstered by the platform’s affordances (n.b. Gibbs et al., 2015) and what I have elsewhere called ‘templatisation’ (Feldman, 2012). This is a sharp departure from the liberatory, freedom of expression rhetoric associated with amateur social media production. Indeed, this is not a world of ‘anything goes’ but a case of a shared, easily recognisable aesthetic by which Instagrammers represent eating out.

While there is, of course, restaurant imagery on Instagram that does not conform to this culinary vernacular, none of those images appeared in my dataset. This suggests a correlation between deploying specific representational techniques and having ‘success’ as a foodstagrammer. Not everyone or every representational practice will find a broad audience online. Indeed, the standardisation of content I found needs to be understood in relation to the recent emergence of the social media ‘influencer’ and the influencer economy (n.b. Goodman and Jaworska, 2020). The very existence of this kind of content creator complicates and challenges the amateur–professional binary I set out at the start of this project. My findings suggest that this amateur–professional binary obscures the increasingly blurred distinction between these two user types. Despite the filtering criteria I employed to weed out professional Instagram accounts from my project, analysis of the resulting dataset made clear that it is often impossible to tell whether an accountholder is a food industry professional or not. This is linked, in part, to contemporary public relations practices in the restaurant trade. As Kobez (2018) finds, ‘it has become commonplace for public relations practitioners in the restaurant space to employ a “hosting model,” which provides free meals, with the aim of achieving positive reviews by amateurs’ (p. 270). A free meal thus acts as payment for a social media post. This transactional model troubles the very ontology of amateurism.

It also troubles the epistemic link between professionalism, payment and authority. As Kobez (2018) further notes, the hosting model ‘has created a landscape in which audiences are oblivious to the differences between independent reviews and those that arise from PR campaigns’ (p. 271). Social media audiences do not necessarily know if they are looking at a paid-for review, which muddies distinctions between professional and amateur content. All of the Instagram accounts in this study had posts bearing #promotion or #ad disclaimers. These were primarily advertisements for food and beverage products, and occasional promotions for food industry events (e.g. Seafood Week). Such posts were easy to identify and remove from my dataset. But few restaurant posts bore these disclaimers. This made it impossible to distinguish ‘genuine’ amateur review posts from those subsidised by the PR hosting model. This is the tyranny of the hosting model. It exploits culinary curiosity. And by using ‘free’ meals as enticements, it confuses the boundary between professionalism and amateurism. Ultimately, my methodological interest in finding influential amateur foodie accounts on Instagram proved to be existentially flawed precisely because I relied on an outdated amateur–professional binary. I misunderstood the contours of contemporary culinary visibility in the Instagram ecosystem.

Detextualised

The third Instagram gaze characteristic is that it deprioritises text and narrative description. Instead, restaurant reviews on Instagram emphasise visuality and aesthetics. This marks a radical departure from traditional techniques of restaurant critique. Reviews in the Michelin Guide, for example, are often all-text. Likewise, newspaper restaurant columns are primarily narrative accounts of eating out. So too are digital reviews on pre-Instagram social media like OCR sites and blogs.

It is no surprise that Instagram values image over text. After all, it is first and foremost an image-sharing platform. However, text is essential to Instagram’s content infrastructure, ‘serving a structural role in terms of being searchable and annotative’ (Leaver et al., 2020: 74). Searching for hashtags and keywords is often how Instagram’s users find content to look at. Nonetheless, text is often regarded as aesthetically undesirable by content creators. It interferes with the clean, glossy aesthetic favoured by the Instagrammers in my study. As Leaver et al. (2020) explain, content producers thus find less conspicuous ways of incorporating annotative text within posts:

Instagram users will then hide the hashtags by either featuring them below the rest of the caption, several lines down, so that they will be cut off when seen in a user’s feed, or by putting them in a separate comment. [. . .] The workaround of hiding hashtags is a vernacular practice that then allows the user to maintain a cleaner profile, in keeping with the platform’s own promoted aesthetic.

Here, text is regarded as an ugly necessity in the Instagram ecosystem. This is also the case in my foodstagrammer dataset, where written language was routinely relegated to posts’ less visible areas.

Text in the Instagram dataset was also sparse and consisted primarily of hashtags. There were no substantive narrative accounts of eating out, and no sustained attention to a meal’s experiential qualities. As indicated earlier, this marks a profound shift from how culinary criticism has traditionally been done. There, the critic used language to communicate the multisensory realities of a restaurant visit and to tell a story. But in the Instagram case, there were no stories in the written narrative sense of the term. There were only images. A food story or restaurant review on Instagram can thus be regarded as an ocular concept, not a linguistic one.

Desocialised

The final dimension of the Instagram gaze is that it promotes a desocialised way of seeing food. It presents culinary production and consumption as wholly de-peopled, and reduces cooking and eating to mere objects. Indeed, objectification is a well-established and much-analysed effect of the gaze (Mulvey, 1975; Zulli, 2018). Despite knowing this, I was nonetheless shocked by the degree of objectification in the Instagram dataset.

None of the photos show people. There is no imagery of waitstaff bringing dishes to the table or clearing them away. There are no shots of kitchen staff prepping ingredients or loading dishwashers. There are no photos of people tucking into the food they ordered. There were no images of cooking or eating, nor any representations of restaurants as fundamentally social places populated by people and conviviality. There were only pretty pictures of food.

The desocialised quality of culinary representation on Instagram matters because it conceals food’s inherent sociality. It masks the knowledge claims, power struggles and inequalities that are inherent to the culinary field. By focusing only on the objects of culinary production and consumption, the IG works to delegitimise a people-centred approach to understanding food. This risks drawing attention away from critically important considerations around where one’s food comes from, who makes it and under what conditions, who eats it, who profits from it and so on.

Unlike Instagrammers, professional restaurant critics routinely acknowledge food’s sociality. Newspaper columnists and Michelin inspectors alike may refer to a restaurant’s supply chain, its hiring practices, its hospitality philosophy or its chef’s ecological ethics. They may also mention interactions with waitstaff or with their dining companions. Despite the out-of-touch elitism that adheres to the Michelin Guide, its critics seem surprisingly cognisant of the people and social processes that make eating out possible. The representational repertoire of Instagram food critics, by contrast, suggests that social media restaurant criticism does not result in socially contextualised or reflexive reviews. Rather, foodstagrammers’ understanding of ‘good food’ appears fundamentally estranged from sociality.