“Rather uncharacteristic of an Englishman to criticise plain food”: Invoking category entitlements in food assessments

Abstract

Discursive psychologists have shown how food assessments serve a variety of functions during mealtime interactions. Anthropological studies attest to the cultural significance of food and eating practices as reported through interviews and ethnographies. This study extends these literatures by studying food assessment in small talk on a social media platform. The data comprised 53 posts selected from an extended discussion on Twitter of a traditional south Indian dish which provoked a furore of media responses. Conversation and discursive psychological analysis showed how the initial food assessment was warranted and challenged by mobilising category-related knowledge and entitlement. In addition, cultural categories and humour were used to dismiss the permissibility of making an assessment. We conclude that the cultural significance of a particular food and the entitlement to assess it are practically accomplished, and that online small talk provides a useful context to study how and thereby make a novel contribution to the literature.

Introduction

This paper explores how, in online forums, making assessments of food can also address issues of identity and culture. Drawing on conversation analysis (CA) and discursive psychology (DP) of family mealtimes, previous work has shown how food assessments are collaboratively worked up and serve practical and communicative purposes in the context of eating (van der Heijden and Wiggins, 2025; Wiggins et al., 2001). Our aim is to extend this approach to food assessments by examining how they are made categorially and culturally significant in ostensibly mundane online interaction by examining assessments on Twitter of a south Indian breakfast food, idli. We explore how the assessment of this food is constructed, ratified, and challenged in interaction and how its social and cultural relevance is mobilised as a discursive resource. We propose that studying mundane interactions (or “small talk”) on social media is a novel and worthwhile context for studying food assessment not least because it is a significant avenue for issues of culture, category-membership, and community. Our work therefore draws on and sits at the intersection of contemporary discursive literature on food assessments, culture and food, and conversation analytic studies of small talk.

Food assessments in family mealtimes

Food assessments during family mealtimes have been successfully studied using conversation analytic (CA) and discursive psychological (DP) techniques (Lindström et al., 2019; van der Heijden and Wiggins, 2025; Wiggins, 2002, 2004). This work builds on seminal studies of assessments that show how they are coordinated between parties in an interaction and oriented to social action (Goodwin and Goodwin, 1992; Pomerantz, 1984; Stivers and Rossano, 2010), and on findings concerning the role of epistemic authority and entitlements in assessments (Hayano, 2011, 2016; Heritage and Raymond, 2005).

Studies also show that during a meal, food assessments can be used to accomplish social actions related to eating, such as requesting second helpings, complimenting the chef, refusing servings, justifying leaving food on the plate, and persuading children to eat a particular food (van der Heijden et al., 2022; Wiggins and Potter, 2003). They can also change the topic of conversation by reorienting people’s attention towards the food or diverting the conversation to avoid sensitive topics (Mondada, 2009). Wiggins and Potter (2003) show that the actions accomplished are facilitated through the way the food assessments are formulated. For example, food assessments can be constructed as subjective or objective assessments, and they can focus on the food item or category. Objective assessments convey a sense of the quality of the food (e.g. “The potatoes are delicious”) and are used in complimenting it (Wiggins and Potter, 2003: 522). Subjective assessments (e.g. “I don’t like red wine”), are used to distinguish the assessor’s own tastes and preferences from those of others, or to refuse a second serving of food (Wiggins and Potter, 2003: 520–521). Category assessments (e.g. “I don’t like carrots”) are used to indicate an enduring dislike of a food group which in turn allows a speaker to refuse the dish being served. Lastly, item assessments (e.g. “I love this carrot cake”) which are specific to a given interactional occasion are used to give compliments to the chef (Wiggins and Potter, 2003).

In addition, family mealtime interactions involving children reveal patterns of coordinated social activity in food assessments. For instance, van der Heijden et al. (2022) showed that in mealtime interactions, a child’s positive assessment (liking) of a food item (“These potatoes are tasty”: 7) did not often receive a response from the parent; they thus suggested that a positive food assessment does not anticipate a response. By contrast, a child’s negative assessment (disliking) of a food item regularly invited a response from the parent because it was treated as employed to refuse food. Similarly, Wiggins (2004) examined authority and entitlement to evaluate food during eating, including parents’ claims about their own and their children’s preferences. For example, she showed how challenging a negative evaluation (e.g. “I don’t like red wine”) can involve questioning the evaluator’s experience with the food. Further, a subjective assessment (e.g. “I don’t like cheese fondues”) may implicate epistemic authority or privileged knowledge over an experience with the given food. Challenging these kinds of food assessments can therefore involve questioning the culinary experience of the assessor (e.g. “When have you ever tried them?”; Wiggins, 2004: 35). These findings underscore how entitlement to make an evaluation of food can be undermined and the veracity or reasonableness of the assessment thereby rejected. However, little attention has been paid in this literature to the use of cultural knowledge and identity categories which other work suggests may be used as possible epistemic resources. The approach is therefore usefully extended to the study of food assessments online where they are not embedded in eating activities but cultural ones.

Food and culture

Anthropological research on food attests to its cultural, social, and practical significance (Jönsson, 2023; Karrebæk et al., 2018). For example, food choices relating to preparation, ingredients, and which recipes are passed through generations are found to be shaped by history and religion (Srinivas, 2011; Zhang et al., 2019); eating practices and food preferences are, in turn, taken to be key indicators of people’s cultural and social group memberships (Silva et al., 2014). In focus groups, citizens and immigrants reported the importance of eating practices for preserving identity and for managing fear of identity loss respectively (Reddy and van Dam, 2020). Talking about food, and the adoption of specific eating practices is therefore said to have evolved as a means of marking people’s membership of various social groups or identities like castes, religions, classes, and regions (Gordon and İkizoğlu, 2017; Karrebæk et, al., 2018; Srinivas, 2011; Wright and Ransom, 2005). The object of food assessments with which we are concerned here, idlis, are associated with the cuisine of south India. They are widely recognised as a staple breakfast food in the region and sold at most south Indian breakfast places across states including Karnataka, Andhra Pradesh, Telangana, Kerala, and Tamil Nadu (Swasthi, 2015).

Clearly, food preferences and practices are important in signifying the cultural identity of people. Nonetheless, research has not attended in significant detail to how issues of culture and identity may be employed as discursive resources in warranting food assessments in mundane interactions or “small talk” on social media.

Food preferences in small talk

Despite the diminutive reference to talk as “small,” studies of talk about topics such as the weather have shown it to be very important in “oiling the interpersonal wheels” of interaction (Holmes, 2000: 50) and in the accomplishment of institutional tasks (e.g. Iversen et al., 2022). CA studies have shown that small talk is used to build and maintain social relationships (Holmes, 2000), to support or enable the accomplishment of institutional tasks such as helping to start meetings in medical settings (Maynard and Hudak, 2008), to distract from an intrusive procedure (Benwell and McCreaddie, 2016), and in advice-giving (Iversen et al., 2022). Topics like the weather offer an object of shared experience which, though apparently neutral and impersonal (cf. Coupland and Ylänne-McEwen, 2000) in practice serves as “an inference-rich anchor point from which people can express their membership in shared and divergent normative communities” (Iversen et al., 2022: 1059). Our aim is to apply this approach to small talk outside an institutional setting and to a different topic, that of food. Indeed, in addition to the everyday familiarity of questions such as “what is your favourite food?,” a Google search identifies food as one of the nine most common topics of small talk.

Most studies of small talk are of verbal conversations but since its inception, social media has provided a platform for people to express views and to interact with others. Moreover, food-related topics are commonly found on social media sites like Twitter. Social media is thus arguably an appropriate and interesting context for research on food assessments in small talk. Some evidence for this comes from a paper by Jönsson (2023) which examined how cultural understanding of eating habits in Sweden led to a controversy on Twitter known as #Swedengate. The study showed how a cultural practice became a point of contention in talking about food, but importantly, it highlighted how the ensuing clashes were publicly displayed through the use of social media phenomena such as hashtags, and more generally how one post can trigger a media controversy (Jönsson, 2023). In addition, the widespread use of social media within and between different geographical regions makes this an ideal place to appreciate the kinds of category membership and normative issues observed by Iversen et al. (2022). This study will therefore examine talk about food assessment, “Idli are the most boring thing in the world” on Twitter, produced in response to a whimsical question, “What’s the one dish you could never understand why people like soo much.” It will attend to the ways interactional and categorial issues may be mobilised in this context.

Present study

The present study will extend CA/DP work on food assessments from mealtimes to a mundane non-eating context by examining small talk on Twitter (now X), in which the assessment of idli is at stake. We explore how food assessments are done, and whether and how the cultural, social, and categorial nature of food is made relevant.

Methodology

In the analysis that follows, we examine a collection of three-part sequences. The first part is a whimsical type of question posted by Zomato India (ZI; a food-delivery service based in India) which was quote-tweeted by someone we have code-named “EA” in responding to it by producing the initial assessment of idli with which we are concerned in the analysis. The assessment by EA is the second part of the interactions, and effectively the original post (OP) as it triggered the series of responses or second assessments we examine below. These responses are the third part of sequence.

The data was originally collected as part of a larger project on categories and identities which used the keywords “north Indian south Indian” as search terms in collecting data, and the timeframe October 9–16th, 2020. The dataset for this study included all 53 replies to the (initial) assessment of idlis (the OP) posted within the timeframe.

Data was collected using a script in RStudio and the official Twitter API (Twitter API Documentation, n.d.; Earth Data Science - Earth Lab, Twitter Data in R Using Rtweet, 2017). Only publicly accessible tweets were collected and, in line with the ethical guidelines outlined by the BPS, no additional consent was required from the users (BPS, Ethics Guidelines for Internet-Mediated Research, 2021). However, the anonymity of users has been protected. This project was approved by the University’s Research Ethics Committee.

There is a substantial body of work that attests to the successful application of Conversation Analysis (CA)/Discursive Psychology (DP) to a variety of social media data including Twitter (e.g. Giles, 2021; Housley et al., 2017; Meredith et al., 2021). Social media sites provide unique resources for interactions such as reply-to, or “@” mentions, or quoting a prior tweet, which perform the task of addressing recipients and this reproduces the sequential aspect of interaction (Meredith and Potter, 2014).

Preliminary analysis of the responses to EA’s assessment indicated four ways in which the assessment was replied to which we examine in more detail. The first set of responses were those that agreed with EA’s assessment and used epistemics as a resource in doing so. The second set also drew on epistemic resources, but here to warrant disagreement with the initial assessment. Responses in the third collection invoked membership categories to undermine or dismiss the assessment. Finally, in the fourth set responses similarly dismissed EA’s assessment but here by constructing it as not permitted. In the following section, we present and examine in detail a few examples from each of these sets.

Analysis

The data we examine was in response to the following question:

Extract 1 (First Assessment): Idli are the most boring things in the world

(ZI is Zomato India; EA is a respondent)

1 ZI (Oct 4): What’s the one dish you could never understand 2 why people like soo much 3 EA (Oct 6): Idli are the most boring things in the world

This question received multiple answers but the initial response we are interested in here named idlis as the dish, as requested and provided an assessment; idlis are boring. There are two notable points here: first, that as observed in the Introduction, this is a traditional South Indian dish and this in turn makes culture and identity potentially relevant in responses to the assessment. Second, we can identify several ways in which EA warrants naming idli as a candidate dish for others’ incomprehensible taste. The item is characterised using extreme case formulations (Edwards, 2000; Pomerantz, 1986) “the most boring things in the world” that reflect the extreme or exaggerated terms used in the initial request (“never understand,” 1:1; “like soo much,” 1:2) and portray a mildly negative assessment. However, EA produces an object-side assessment (Wiggins and Potter, 2003) in response to the subject-side request, in that the object (idli) rather than third-party preferences is assessed. In other words, “boring” is characterised as a property of the food rather than a matter of the assessor’s opinion of the food. Similar distinctions between the use of s-side and o-side assessments were observed by Wiggins and Potter (2003).

We noted above that this initial assessment provoked a number of responses, and The Times of India even ran a story on the ensuing debate provoked by an apparently innocent question (Hussain, 2020). In the following sections, we analyse 6 such responses, selected with the aim of presenting the range of responses observed; how the food assessments are done; and how the assessments invoke the social and cultural relevance of idlis.

Warranting agreement using category affiliation

The first response examined does agreement with the assessment of idlis as boring. There were six instances of agreement in our dataset. Although there were fewer of them than disagreements, they are interesting because of how posters extend the initial negative assessment and display the categorial relevance of idli by invoking category-bound knowledge to warrant this assessment.

Extract 2: Idlis are strictly ok.

1 EA (Oct 6): Idli are the most boring things in the world 2 ZI (Oct 4): [Quoted] What’s the one dish you could never 3 understand why people like soo much 4 NV (Oct 9): as an Indian married in a South Indian family 5 I am going to agree with you. 6 Idlis are strictly OK. I had them once a week 7 past two months and I feel I am done with 8 idlis for good. 9 KI (Oct 31): As a South Indian, I concur. I can never 10 understand the hype for idlis

The posters state and extend agreement with EA’s tweet in several ways. One is explicitly by writing “I am going to agree with you” and “I concur” (2:5 2:9). There is also an explicit statement of what the posters agree with which preserves something of the original assessment while transforming it. For example, idlis are described as “strictly OK” which modifies and delimits the mildly negative character of the initial assessment (boring). Finally, KI expands the requested naming of a food that others like by characterising it as “the hype for idlis” (2:10). The term “hyped” implies the excessive and unwarranted promotion of the associated object; here, it implies that idli do not merit the positive view others hold of them. The term “hype” also extends the original assessment (of liking so much). In each case, then, the explicit agreement is reinforced by expanding part of the initial assessment. A similar finding regarding the way agreement extends the initial assessment is reported in Pomerantz’s (1984) classic study and she shows how this demonstrates independence of view while doing agreement with a prior speaker. A similar effect is achieved in this data.

Additionally, the credibility of the posters’ agreement is established through invoking category membership and through displaying reluctance. In relation to the first of these, posters describe the categorial basis of their agreement with EA’s assessment of idlis. NV ascribes their agreement as related to being “an Indian married in a South Indian family” (2:4-5). A person who is married and a member of a South Indian family would be expected to share regular meals and therefore have knowledge of culinary traditions. Indeed, NV provides a description of the relevant and regular experience, “I had them once a week past two months” (2:6-7). Moreover, the relevance of this identity to the assessment is made clear through the construction “As an Indian . . .” (2:4) which suggests the co-occurrence of states of affairs (family membership experience and agreement). These descriptions portray and warrant NV’s assessment of idlis as an informed one due to category affiliation. Having thus described the basis for the assessment of idlis, NV states “I feel I am done with idlis for good” (2:7-8). KI describes themselves as “South Indian” and thereby achieves a similar effect, of making knowledge and experience relevant as the basis of the response. The implied absence of any alternative breakfast reinforces the claim to have regular experience of and need to eat idlis.

In summary, in these extracts we observed how posters extended the assessment and described its basis in category-related experience. The users ground their independent view of idlis in the cultural knowledge of eating practices and frequency of eating available to them owing to membership of categories. These were used to warrant the agreement with the view that idlis are boring. This in turn has several effects. One is to establish category-related epistemic knowledge of the object and therefore its assessment is shown to be informed; the second is that likes and dislikes are treated as tied to category membership; and the third is to indicate independence of view and thereby offset “blind agreement.”

Warranting disagreement

Many of the responses to EA’s first assessment we observed disagreed with EA’s negative assessment of idli. In the analysis that follows, we examine three types of sequences in which users do disagreement, and the cultural knowledge and category membership they invoked in warranting these. In the first, EA’s knowledge of how to eat idli is used to disagree with the assessment directly. In the second, posters question EA’s category-related entitlement to assess idli and thereby reject the assessment implicitly. In the third, posters invoke the cultural significance of idli to dismiss the permissibility of any assessment of idli. We will examine each of these patterns in turn.

Questioning grounds for disagreement through normative knowledge

Below, we show that disagreement with the first assessment is done through producing an object-side assessment of idlis as being intrinsically tasty (and healthy) when eaten in the right way; attributing disagreement to a lack of the right normative (culinary) experience; and questioning EA’s ability to make a reasonable assessment of idli on category-related grounds. They thereby warrant disagreement by questioning EA’s epistemic authority and entitlement to assess idlis. There were 20 instances of this kind of response in our corpus.

Extract 3: No its not, you haven’t had the right idly

1 EA (Oct 6): Idli are the most boring things in the world 2 ZI (Oct 4): [Quoted] What’s the one dish you could never 3 understand why people like soo much 4 SY (Oct 9): No it’s not, you haven’t had the right idly 5 if you think this.

Extract 4: I think you’ve not tried the right combination

1 EA (Oct 6): Idli are the most boring things in the world 2 ZI (Oct 4) : [Quoted] What’s the one dish you could never 3 understand why people like soo much 4 KR (Oct 9): I think you’ve not tried with proper 5 combination. It’s tasty and healthy 6 breakfast. I can say you’re unlucky

In extract 3, the poster disagrees directly with EA’s assessment, “No it’s not”; in extract 3, disagreement is inferable from what follows. In both, posters account for EA’s negative assessment of idli on the grounds of lack of experience. Thus, in extract 3, thinking that idli are boring is made conditional (“if you think this,” 3:4-5) on not having had the right idlis (“you haven’t had the right idli,” 3:4). Therefore, EA’s assessment of idli is attributed to the absence of relevant experience and it is implied that eaten in the right way, idlis are not boring. Similarly, in extract 4, the responder suggests “you’ve not tried with proper combination” and thereby accounts for EA’s assessment as due to his lacking the right experience rather than as an attribute of the food itself or personal taste. This is reinforced through an object-side assessment (Wiggins and Potter, 2003) of idlis as having the qualities of being tasty and healthy and, in extract 4, by characterising EA as unlucky (“I can say you’re unlucky,” 4:6). The negative assessment is thereby treated as an unfortunate outcome of circumstances rather than the intrinsic qualities of idli. It is worth noting that in portraying the grounds for not liking idlis, KR uses “I think” (4:4) which assuages the subsequent claim about lack of experience. It shows some delicacy around claims to know others’ experiences.

In summary, the initial negative assessment is dismissed using an object-side assessment which, it is implied, is based on evidence and therefore correct, and by accounting for EA’s (erroneous) assessment by claiming he does not have the right kind of culinary experience. Here, then, we see how cultural rather than personal knowledge is recruited as a resource for warranting disagreement, and how the assessment of idli is treated as fact (relating to its intrinsic qualities) rather than a matter of personal preference. Food preference here is treated as a matter of cultural practices of preparation and accompaniments rather than personal taste.

Questioning category entitlement to assess idli

In the extract below, we examine the second way in which posters rejected EA’s initial assessment of idli: by relating it to the assessor’s category membership. Specifically, we show how category-bound features of that membership category are used to question entitlement and the ability to assess idli. There were 15 instances of this strategy in our corpus.

Extract 5: Uncharacteristic of an Englishman . . .

1 EA (Oct 6): Idli are the most boring things in the world 2 ZI (Oct 4) : [Quoted] What’s the one dish you could never 3 understand why people like soo much 4 DS (Oct 9): Rather uncharcteristic of the an Englishman 5 to crticise plain food, becoz English ppl 6 like their food bland. Even their chicken is 7 boiled, urggh.., If an englishman has chicken 8 fried in desi style, with oil, onions, red 9 chilli and masala, he would literally and 10 practically choke.

In this extract, DS questions EA’s ability to assess idli on category-related grounds and thereby dismisses his negative assessment rather than disagreeing with it explicitly. Specifically, DS observes: “Rather uncharacteristic of the an Englishman to criticize plain food” (5:4-5). There are several features of this comment which are worth noting. For example, the assessment of idli as boring is reformulated as EA criticising the plain quality of food. That is, the focus is shifted from the assessment of the food to the assessor’s act of assessing it, and “plain” and “boring” are equated. Moreover, the reference to the criticism as “uncharacteristic” indicates the exceptional nature of this act and this is tied to EA’s category membership (English). In addition, the category-bound significance of making an assessment of plain food is reinforced in several ways. One is directly, by providing an account of its noteworthiness: “becoz English ppl like their food bland” (5:5-6). The implication is there is something questionable about EA’s assessment since it involves him criticising a dish for a fault that could equally apply to English food, and this in turn undermines the reasonableness of making the assessment. A second way in which the category-based unreasonableness of assessing idli as boring is called into question is by describing the preparation of “their” food: “even their chicken is boiled” (5:6-7). Preceding the description with “even” suggests that if even chicken is cooked in this way, other foods are too. The resulting dish is assessed negatively using a disgust marker, “urggh” (5:7; see Wiggins, 2013). In this way, DS makes relevant certain category-bound attributes of being English and in the following statement, this is illustrated through a hypothetical description of an Englishman eating chicken cooked in a different way, that is, fried with spices, and the consequences (“he would ‘literally and practically choke,” 5:9-10). The terms, “literally” and “practically” construct and emphasise the “real” status of the claim and therefore the likelihood of the choking. It is thereby implied that an Englishman would not be able to eat spicy food, is therefore restricted to plain (non-spicy) food, and thus not best placed to make judgements of taste. Finally, DS employs both “the” and “an” prior to the category “Englishman” (5:4). The use of both articles is interesting here, since “the” ascribes identity specifically to EA whereas “an” treats him as a member of the category. The dual terms thus function to make both referents live and thereby make claims about the act of assessing personally and categorially relevant. Taste is thus treated as a category-bound attribute of EA and “Englishmen” more widely and this in turn undermines entitlement to make the judgement of idlis as boring. Although not stated directly, in questioning the categorial ability to judge, the assessment of idli as boring is also undermined. Thus, the right to assess is called into question through attributing category membership to EA and this is used as a resource for rejecting the validity of the first assessment. In other words, the act of assessment is dismissed on the grounds of the poster’s category membership.

Dismissing the act of food assessment because of the categorial significance of the food

In the final set of extracts below, responders similarly treat the act of assessment as inadmissible or insulting, but here the grounds for doing so relate to the category-bound significance of idli as a South Indian dish which, it is claimed, make any assessment inadmissible.

Extract 6: Btw how could you?

1 EA (Oct 6): Idli are the most boring things in the world 2 ZI (Oct 4) : [Quoted] What’s the one dish you could never 3 understand why people like soo much 4 SR (Oct 9): beg for forgiveness before the idli 5 fatwa!! BTW how could you ?? Seriously!! No 6 sympathy. . . am not even a southie and am hurt

Extract 7: Oh Man, you just stirred the hornet’s nest!

1 EA (Oct 6): Idli are the most boring things in the world 2 ZI (Oct 4) : [Quoted] What’s the one dish you could never 3 understand why people like soo much 4 RM (Oct 9): Oh maaan ! You just stirred a hornet s nest ! 5 We are a passionate and emotional lot. Even 6 if we hate idlis we will fight for what it 7 represents. Btw That’s all we are permitted 8 in india these days! Lol.

In these extracts, responders portray reactions to the act of assessing idli. In extract 6, SR instructs EA to “beg for forgiveness before the idli fatwa” (6:4-5). A fatwa refers to a legal ruling or opinion on a point of Islamic law given by an authority (Fatwa, 2023); employing this concept in relation to an idli fatwa implicates an authority or authoritative view on idli which is contrary to EA’s assessment of idli as boring. That is, instructing EA to beg forgiveness constructs his assessment as wrong, even blasphemous, and potentially worthy of punishment. The improper nature of the act is reinforced in the expressions that follow. For example, the phrase, “how could you?? Seriously!!” (6:5), here presented as an aside (“Btw,” 6:5), portrays and emphasises shock or dismay at another’s actions and thereby further condemns the act of assessing idli. In extract 7, RM characterises EA’s act of assessing as “You just stirred a hornets nest!” (7:4). This is a colloquial expression which implies the poster has done something to create controversy. This in turn implies the act of assessment is provocative. Moreover, the tweeted response is prefaced with “Oh maaan!” (7:4) which is used to express reactions such as shock or disbelief, and here the affective reaction conveyed is emphasised by extending the aaa sound and adding an exclamation mark. In both cases, then, the act of producing an assessment of idli is treated as reprehensible or offensive. In extract 7, this is followed by a statement about what “we” are like, thereby indicating the category-related nature of the claimed insult. Similarly, begging forgiveness and reference to fatwa makes a social normative aspect relevant (extract 6).

Both posters also provide an account for their characterisation as problematic of the negative assessment of idli that invokes the category membership of the insulted party. For example, SR states “No sympathy . . . am not even a southie and am hurt” (6:5-6). By implying that a “southie” would be hurt by the act of assessing idli, SR indicates a relationship between the category “southie” and idli that accounts for the reaction. By describing himself as sharing the reaction (“hurt”) despite not being a category member, he emphasises the offensiveness to members of EA’s assessment (the implication being that this would be even more so for a member). Moreover, he prefaces this claim with “No sympathy,” thereby characterising EA’s actions as not worthy of understanding or mitigation in relation to their offensive character. In extract 7, RM invokes category relevance using the pronoun “we”: “We are a passionate and emotional lot. Even if we hate idlis we will fight for what it represents,” (7:5-7). This claims membership of a category (by implication south Indian) and ties attributes, “passionate and emotional” and of being willing to “fight” for what idlis represent to the category. In both cases, then, the condemnation of the act of assessing idli is accounted for in category-related terms.

The censorious formulations are, however, offset with several markers of humour. For example, there is a mismatch between the mildly unfavourable assessment of idlis as boring and the instruction to beg forgiveness for an act akin to blasphemy which conveys humour through absurdity. In addition, SR’s instruction to “beg for forgiveness before idli fatwa!!” (6:4-5) is preceded by three grinning emojis, and non-seriousness is further conveyed through the use of multiple exclamation and question marks (e.g. “how could you??” and “Seriously!!,” 6:5). RM tweets, “Btw that’s all we are permitted in India these days! Lol.” (7:7-8). This implies that fighting in defence of idli is the limit of protest or reaction in India, but the use of an exclamation mark and “Lol” (Laugh out loud) indicates that this is to be treated as ironic.

In summary, responders here reject the assessment of idli by treating it as offensive and as having category relevant consequences for the members of the category to which idli are bound, while implicitly acknowledging the trivial nature of the assessed object using humour.

Discussion

In this paper, we examined several responses to a (negative) assessment of a food item (idlis) and the discursive strategies users employed in producing their (dis)agreements. When agreeing with the assessor’s evaluation (Extract 2), they extended part of the initial assessment and portrayed mild reluctance in agreeing. Users also described their category-bound experience (as a “Indian married in south Indian family”) as the basis of the agreement. Posters thereby ratified EA’s claim regarding the quality of idlis (as boring). However, users disagreed more often than they agreed, and their disagreements included an object-side assessment of idlis as intrinsically tasty and healthy. They accounted for EA’s (wrong) negative assessment by attributing it to the absence of the right eating experience or inability to eat idli correctly (with spicy accompaniments). In the third part of the analysis, we showed how a category membership (English) and category-bound attributes (bland food) were attributed to EA and used as a resource for rejecting his entitlement to assess idli. In the final section, we showed how claims about the category-related consequences of producing a negative assessment were used to reject the very act of assessing idli, while implicitly acknowledging the trivial nature of this reaction through humour.

Several interesting themes thus emerged from the analysis which speak to and extend DP/CA work on food assessments. These include the object-side design of the second assessments; epistemic issues in (dis)agreeing with the initial assessment; and the use of categories in establishing and disputing entitlement to assess idli. Firstly, judgement of food is treated as a matter of fact rather than personal preference. Similarly, in her mealtimes conversation data, Wiggins (2004) shows that subject-side assessments are harder to challenge than assessments of food quality precisely because they treat taste as personal rather than a feature of food. In addition, the disagreements here were about more than questioning judgement, they questioned experience (“how do you know if you haven’t tried it?”). However, the treatment of experience was more nuanced, in that posters argued that the initial assessor lacked the right kind of experience (rather than experience per se). The use of experience with the food is thus particularly interesting because it takes place outside the context of eating and knowledge of assessor’s eating habits; it nevertheless works as a resource in dismissing or undermining the assessor’s epistemic entitlement to assess idlis.

Secondly, the analysis examined the assessor’s turn and entitlement, and the findings speak to work on epistemics in assessments (Hayano, 2011, 2016; Pomerantz, 1984). We showed that category attributions and mobilising category-related entitlements were used in several ways to undermine the initial assessor’s entitlement to assess. For example, attributing the category membership “English” and category-related experience was used to dismiss EA’s view of idli as boring. In another instance, not being a member of a category (south Indian) was used to work up the credibility of agreement with the assessment of idli. Establishing credibility of the assessment was different in these cases, and may be particularly important, because food assessments are still coordinated social activities even if claimants cannot appeal to immediate taste or gustatory reaction as evidence. Thus, we have shown that and how posters use category memberships flexibly to negotiate eligibility to assess and draw boundaries around who is and who is not permitted to assess the south Indian food.

Thirdly, in providing an opinion, food can be treated as a cultural item rather than something to be eaten or tasted and therefore as having cultural relevance. Perhaps this is not surprising given that food has been associated with culture and social identities through eating practices (Gordon and İkizoğlu, 2017; Reddy and van Dam, 2020; Wright and Ransom, 2005). However, this an interesting divergence from work on food assessments situated within the ongoing activity of eating by showing how interactants treat the cultural significance of a food as a practical resource and a concern to be rhetorically used to reject the relevance of who can and cannot assess idlis.

Our work thus extends previous work on food assessments and epistemics by observing the use of categories in asserting and dismissing entitlements to assess and in showing how culture was produced as a practical accomplishment in dismissing the act of assessing a food item. Moreover, in conducting this research, we were also interested in social media as a non-eating context in which food assessments are sought through opinion-seeking questions, which trigger mundane, but socially significant talk. Despite the difference in the kind of interactions to those reported in the literature, we found similar constructions of (Goodwin and Goodwin, 1992; Pomerantz, 1984) and negotiations surrounding the proffering of (dis)agreements on social media in general and to those that have been observed during eating (Edwards and Potter, 2017; Wiggins, 2002, 2004. Importantly, disagreements outnumbered the agreements and were proffered with directness and lacking palliatives or other modifiers that would demonstrate hesitation in providing a challenging or resisting response to the assessor’s stance. In some instances, the initial opinion was dismissed, which is a type of response not previously found (to our knowledge) in verbal assessment-response sequences. Therefore, by looking at one thread (with one assessment and several responses) using techniques of microanalysis of data, we were able to appreciate the diverse ways in which people draw on culture and epistemics to manage self and others’ rights to (negatively) assess food.

Conclusion

By examining one food assessment and various responses to it, we show how food assessments are coordinated social activities even outside an eating context, in “small talk” on social media. We illuminate in this as yet unexplored context, how category-memberships, related knowledge, and category-bound epistemic entitlements are important and useful resources for users to negotiate boundaries around who is and who is not permitted to assess the given food and its cultural significance. Through this micro-analysis of online interactions, we conclude that posters treat food assessments and social media as sites in which culture is made a practical concern.