COVID-19 and socio-materially bounded experimentation in food practices: insights from seven countries

Abstract COVID-19 has caused unprecedented disruption to previously settled everyday routines, prompting a period of forced experimentation as people have adjusted to rapid changes in their private and working lives. For discussions regarding consumption, this period of experimentation has been interesting, as the apparent instability has disturbed the ongoing trajectory of consumption practices, and with it has created possibilities for a transition toward sustainability. In this article, we examine food practices (e.g., food shopping, preparation, and eating) in seven countries (France, Germany, Italy, Netherlands, Norway, UK, and Vietnam) to assess what we can learn to accelerate transitions toward sustainable consumption. Grounded in a practice theoretical approach, our empirical analysis shows how disruption of everyday routines has generated socio-materially bounded experimentation. We demonstrate commonalities across contexts in how lockdown measures have restricted the performance of previously taken-for-granted practices. We also show diversity in experimentation as food consumption is entangled in other everyday practices. Our study, on one hand, portrays how the adaptation of food practices allows disruption to be managed, demonstrating creativity in working within and around restrictions to continue to provide services for everyday life. On the other hand, we reveal that the capacity of experimentation is not evenly distributed among people and this variation helps in identifying the wider socio-material conditions that constrain and enable opportunities for readjustment. Understanding disparities that affect experimentation (e.g., integration of food practices with work and caring practices) is informative when thinking about how to stimulate sustainability transformations in food practices and provides critical reflections on strategies to enable sustainable consumption.


Introduction
COVID-19 has had far-reaching implications for everyday lifeconcomitantly disrupting everyday routines and systems of provision. Its seemingly transformative nature (Sovacool et al. 2020) provides a starting point to investigate the dynamics of everyday practices and raises interesting questions regarding consumption, as the ongoing trajectory of (presently unsustainable) consumption practices has been disturbed, and possibilities for alternative (potentially more sustainable) practices are presented. In this study, we examine how food practices have changed in the light of the pandemic, which simultaneously affected systems of provision (e.g., the closure of restaurants and hygiene measures in supermarkets) and everyday routines (e.g., a widespread shift to homeworking and schooling). We examine lived experiences of disruption in seven different countries (France, Germany, Italy, Netherlands, Norway, UK, and Vietnam), exploring what these experiences reveal about the processes that embed and sustain food practices.
This research is embedded in practice theories, which have been widely applied to the study of sustainable consumption, especially with respect to food (e.g., Castelo, Sch€ afer, and Silva 2021;Domaneschi 2019;Warde et al. 2007;Watson et al. 2020;Wertheim-Heck and Raneri 2020). Within this literature, the term food practices is used to describe activities involved in procuring, storing, preparing, eating, and managing leftovers, although food practices are closely entwined with other practices (e.g., mobility, parenting, and socializing), which blurs their boundaries. Meals, and mealtimes, are socially significant and practice-oriented research has demonstrated how personal routines, social interactions, and societal rhythms are coordinated around them (Southerton, D ıaz-M endez, and Warde 2012;Yates and Warde 2017). Consequently, food practices are intriguing units of analysis and there is growing interest in how the connections within and between practices could be remade to better serve sustainability objectives. These questions are situated in a wider discussion about food-system transformation, given the problems surrounding contemporary food systems, particularly health (e.g., nutrition and non-infectious health conditions), sustainability (e.g., emissions, energy use, and food waste), and social justice (e.g., food poverty and the "heat or eat" dilemma). Some researchers challenge the possibility of purposeful, coordinated transformation of consumption, given the distributed nature of agency and interconnectivity of practices (Shove and Walker 2014). Nevertheless, de-routinizing established practices and re-routinizing alternatives is fundamental to the transformation of food systems (Spaargaren, Oosterveer, and Loeber 2012).
This article proceeds by outlining developments in social practice research that provide foundations for our analysis, followed by an outline of the methodology underpinning the investigation. Our analysis begins by exploring the disruptive nature of COVID-19, identifying various ways in which lockdowns created friction for previously established practices, to reveal how routines are stabilized and maintained, and where there are sources of dynamism. We continue by exploring how food practices have been reassembled during this time to maintain normality. Our analysis illustrates how practices are reconstructed in everyday performance, through both adaptive reactions to disruption and purposeful experimentation, as individuals engage in new ways of doing. Finally, we reflect on how the bundling of food practices with others (particularly work, mobility, and care) has affected experimentation. This latter set of reflections links discussions of social transformation to questions related to equality and diversity, and in the discussion section, we consider the implications of these findings for an agenda on transformative social change. This is followed by concluding reflections on the implications of this study for reframing and progressing a conversation about sustainable consumption.

Disruption and food practices
The identification of practice elements and the processes involved in their making and re-making has been a core task for practice-oriented research (e.g., Halkier and Jensen 2011;Biermann and Rau 2020). More recently, the puzzling question of how practices change (and how change might be directed) has become a focal issue for those studying sustainable consumption . For Shove et al. (2012, 14-15), "practices emerge, persist, shift and disappear when connections between [social and material] elements are made, sustained or broken." Though not all theories of practice emphasize the same elements (Gram-Hanssen 2011), key features commonly include embodied understanding (knowhow and skills), collective ideas of ordinary and appropriate conduct (images and meanings), and materials (infrastructures and objects). There is growing interest from researchers and policymakers in how connections within and between practices are made, sustained, and broken as a means of understanding continuity and change in consumption .
A fundamental concept in practice-oriented research is distributed agency which means that the source of direction and change in practices is located neither in consumers nor systems of provision, but within practices themselves (Browne et al. 2014;Sahakian and Wilhite 2014). It follows that anyone interested in enabling sustainable transformations should attend to the collective social, cultural, and material developments with which practices emerge, and the diverse, multi-sited performances of practice involved in their evolution. For example, empirical research has illustrated the complex relationship between "good" parenting practices and convenience food (Daniels et al. 2012(Daniels et al. , 2015 and ideas of "proper food" and emerging food-services models (e.g., meal boxes) (Hertz and Halkier 2017). Simultaneously, these studies demonstrate how practices are involved in constructing the meanings associated with food practices (see, for example, Domaneschi (2012) on food quality and Biermann and Rau (2020) on the role of meat-eating in performing luxury). The inclusion of material agency within practice-oriented studies of food reveals how objects, technologies, and the built environment influence the direction of food practices (Shove and Southerton 2000;Truninger 2011;Wilson 2012;Twine 2018). Local infrastructures are shown to shape procurement (Darmon and Warde 2013;Middha et al. 2021), and the food itself is identified as "an active inducer-producer of salient public effects, rather than a passive resource at the disposal of consumers" (Bennett 2007, 134).
A second important insight from practice-oriented research is the significance of the links between practices. Studying interconnected practices, or bundles, uncovers various ways that practices mutually reinforce each other, with practices becoming resilient and increasingly stable as they become integrated (Castelo, Sch€ afer, and Silva 2021;Higginson, Thomson, and Bhamra 2014). With respect to sustainable food, practice-oriented research has shown how policy challenges, such as food safety (Wertheim-Heck, Vellema, and Spaargaren 2014), food waste (Evans 2014), energy and emissions (Hoolohan, McLachlan, and Mander 2018) are contingent on bundles of practices that vastly exceed the discrete practices involved in shopping, preparing, and eating. The bundling of social practices contributes to the coordination of practices across time and space, and the synchronization of daily life. For example, Nicholls and Strengers (2015) reveal how codependent family practices (e.g., mealtimes, children's entertainment, and bath times) "congeal," becoming practically and emotionally stuck together. Sometimes food practices compete with other practices, with implications for their performance. For example, grocery shopping competes with leisure practices as both are arranged around temporally fixed practices of work and childcare (Wertheim-Heck, Vellema, and Spaargaren 2014). Plessz and Wahlen (2020) (following Welch and Yates 2018) show food practices are collective activities performed in similar ways by dispersed practitioners. Coordination extends beyond those participating in practices together, and similarities in the performance of practice are constitutive of culture and collective identity. These interconnectivities demand novel policy approaches that are simultaneously more ambitious in their scope and more modest with respect to their single-handed ability to affect change (Kuijer and Bakker 2015).
Recognizing practices as distributed and interlinked units of analysis has implications for sustainability transitions, raising questions regarding how connections within and between practices are broken and reformed (Spurling and Mcmeekin 2015;Watson et al. 2020). One way of exploring these questions has been to examine disruptions. Disruption occurs when the continuity of practice evolution is compromised and the relations between elements or practice bundles become unsettled (Nicholls and Strengers 2015;Wethal 2020). Chappells and Trentmann (2018, 198) propose that disruptions provide "momentary glimpses of the fabric of 'normality' as it is fraying and reveal the patterns in which practices and infrastructures are woven together." During periods of disruption, it is possible to observe ingenuity in practice performance and the emergence of precarious proto-practices as elements are speculatively recombined. Though disruption may not give rise to an enduring change in practices, the mundane experimental activity that occurs provides valuable opportunities to reflect on the dynamics of everyday practice (Browne, Jack, and Hitchings 2019).
COVID-19 is unusual in the depth, breadth, and suddenness of disruption, as well as the sharpness of contrast between pre-pandemic and lockdown routines, and therefore presents a novel opportunity to study practice dynamics. Much of the existing body of literature is oriented toward understanding gradual societal change (Hand, Shove, and Southerton 2005;Kuijer and Watson 2017), geographically limited, short-duration events, such as blackouts (Wethal 2020) and droughts (Chappells, Medd, and Shove 2011), and disruptions in personal routines associated with life-course events (Sch€ afer, Jaeger-Erben, and Bamberg 2012; Plessz et al. 2016;Burningham and Venn 2020) or migration (Maller and Strengers 2013;Brons, Oosterveer, and Wertheim-Heck 2020). These studies valuably reveal processes by which practices are de-and re-routinized, and the prosaic activity involved in maintaining the appearance of stability (Tavory and Fine 2020;Wertheim-Heck and Raneri 2020). However, the pandemic resulted in concomitant disruptions in systems of food provision and everyday routines, with profound impacts on food practices ( Figure 1). Therefore, COVID-19 provides a unique lens through which to observe experimentation with previously taken-for-granted practices and creates a space to reflect on the differentiated ways that disruptions are experienced.
To understand how COVID-19 affected food consumption we explored three key questions: How have food practices changed in response to pandemicrelated restrictions? How and why are new ways of doing emerging? And how are everyday experiences of disruption differentiated? With this research, we aim to contribute to understandings of disruption and societal transformation. Our analysis explores these questions focusing on the implications of COVID-19 for arrangements of practices, practice performance, and bounded experimentation.

Methods
Given the exploratory nature of our study, we conducted 119 semi-structured interviews with households in seven countries (France, Germany, Italy, Netherlands, Norway, UK, and Vietnam) between May and July 2020. Purposive sampling was used to achieve a sample that reflected a variety of lifestyle conditions, with a focus on working conditions (e.g., no/reduced work, homeworking, essential workers, retired), household structures (e.g., single occupancy, couples, multi-adult households, and a variety of family types), and life stage (early adulthood, young parents, empty nesters), given that existing research emphasizes how these factors affect everyday routines (see Table 1 for an overview of the sample and Appendix 1 for participant details). An interview guide (see Appendix 2) informed by social practice research was used to elicit responses on how the pandemic had disrupted mundane aspects of everyday life (not limited to food practices), asking participants to reflect on the similarities and differences between the pre-COVID period and during the lockdown.
Interviews were undertaken by research teams in the respective countries under investigation and conducted in the native language of each case. The interviews were recorded and transcribed with relevant extracts translated into English for analysis by members of the multinational team. The analysis was carried out iteratively; commonalities and differences in the interview data were identified and collated to produce a thematic coding structure to further analyze each transcript. During this process, we were looking not only at food practices directly but also at how food practices were embedded in wider constellations of practices that were disrupted during the lockdown. For example, initial screening identified the implications of hygiene measures on grocery-shopping practices and efforts undertaken to recreate dining experiences as commonly occurring themes. Individual researchers returned to their data to further investigate emerging themes and to identify points of difference and contradiction. This process was pursued with a regular discussion between contributors until a sense of saturation was achieved.  Results: the breaking and making of food practices Government responses to COVID-19 were different in their timing, level of restriction, and enforcement throughout the duration of this study, yet the nature of restrictions was broadly similar across the countries that we analyzed. Schools were closed and homeworking was mandatory (France, Italy, and Norway) or strongly advised (UK, Germany, Netherlands, and Vietnam) except for essential workers. Grocery stores remained open (except food markets in France), while restaurants were closed (except in Norway). Leaving home was limited to essential activities (e.g., food shopping, exercising, and going to medical appointments), though the stringency of restriction and the extent of enforcement varied. Other common measures involved limiting the number of people in stores, wearing masks (except in Norway and Netherlands), and sanitizing hands. There were also restrictions on social interaction and travel. In most cases, these restrictions were intended to limit contact between households and to reduce viral transmission, nevertheless, as our analysis shows, the implications for food practices were profound.
In this section, we present our findings, which show how food practices were disrupted and reconfigured during the early stages of the pandemic. We report our findings in three sections. First, we illustrate how responses to COVID-19 created friction for established practices. With the example of grocery shopping, we describe how usual sequences of practice become unsettled, resulting in the disconnection of food practices from established practice bundles (particularly commuting). We also identify alterations in practice elements, especially new senses of risk and measures to prevent viral transmission. Our findings illustrate how food practices adapted to accommodate these uncertainties, affecting food planning, shopping, and storing. Second, our findings demonstrate that concerns about COVID-19 intersected with general understandings about the role of food in healthy, comfortable lifestyles as well as illustrating how the pandemic has prompted experimentation in food practices, involving the incorporation of new technologies, services, spaces, and sensitivities in the mundane practice performances. Finally, we reflect on the bounded nature of adaptation and flexibility. We find that capacity for experimentation with food practices has varied between households, reflecting differences in household structure, life stage, work commitments, and caring responsibilities. These differences provide insight into (constrained) innovation in practice. In the next section, we use these findings to discuss possibilities for stimulating sustainable consumption.
Unsettling established practices: breaking and remaking (links between) practices In this section, we examine personal accounts of coping with disruption in everyday life, which show how food practices became disconnected from previously established sequences of everyday practices. The implications of a shift to homeworking and closure of childcare facilities are particularly evident within the findings, with many respondents reflecting on the disconnection of grocery shopping from everyday travel. We also explore how hygiene measures introduced to reduce viral transmission, along with emerging notions of risk, have compromised previously taken-for-granted modes of shopping and have shaped the positioning of responsibility for grocery shopping. And, finally, we show how interrupted systems of food provision have destabilized established ways of planning, shopping, and storing food in the home.

Breaking connections between mundane practices
With work and school displaced into the home, grocery shopping became disconnected from these practices, with implications for timing, location, and frequency. Existing research demonstrates that many food practices are flexible practices, fit in between others that depend on institutional temporal rhythms, like work and school (Southerton 2006). Consequently, food practices are ordinarily temporally and spatially contingent on these other practices. During the lockdown, these sequences of practice were unsettled. For example, the following interview extract describes how grocery shopping became disconnected from established sequences of practice, with implications for the timing, location, and frequency of shopping: [Before] I might go to the shop most days, if I'm coming back from work … Now it's changed completely. I've changed the shops that I use, because I feel a bit more comfortable in them, you know, with the arrangements that they've put in place. Also, I'm more cautious to try and get all the items that I need in one hit. Rather than thinking, well, if I do miss something, I can always go for it tomorrow.  This extract indicates the changing relationship between work, particularly the commute, and grocery shopping. For this participant, a regular pattern of travel to and from work enabled spontaneous grocery shopping for small quantities of food. Without these other practices requiring travel, food shopping became a purposeful practice in itself, with preparation guiding its performance. Moreover, several respondents described how the timing and location of food shopping were adapted to accommodate concerns about social interaction and viral transmission: I tried to go shopping at times where not so many people go to the supermarket. And I thought that I better go somewhere where there is little going on, in small supermarkets.  This quote resonates with other extracts, which reveal how food practices that would ordinarily follow consistent patterns, such as a daily coffee at a bar (IT-05) or a weekend trip to an open market (UK-04) were similarly unsettled. These findings demonstrate that when links between practices are broken, the timing, location, and nature of food practices also change. Practice-oriented research emphasizes how, though we may observe practices like grocery shopping as discrete performances, practices are distributed and interdependent with other practices [e.g., cooking (Foden et al. 2019) and commuting (Warde 2016)]. Recognizing interdependencies between practices is vital to our understanding of how practices change (and are stabilized) (Hui, Schatzki, and Shove 2016), and imagining possibilities for reconfiguring links between practices provides a way of expanding the scope of intervention in food systems (Spurling and Mcmeekin 2015;Watson et al. 2020). Our findings add weight to this argument, demonstrating how seemingly stable practices become unsettled and are reconfigured in relation to work-life routines.
Navigating risks that compromise convenient ways of doing In addition to the uncoupling of food practices from established practice bundles, new notions of risk and hygiene measures produced friction for ordinary shopping practices. The introduction of objects (e.g., masks and gloves) and actions (e.g., queuing and sanitizing) resist previously convenient ways of doing: Shopping is not as easy. You can't just pop in. You've got to make sure you've got your mask [and] I always had to make sure I have my gloves. And really, you know, you feel quite stressed in a way you never were before. (UK-04) Interviewees described different ways of negotiating risks, changing the timing and location of shopping as well as introducing new practices in selecting and storing food. For some, bigger supermarketswith more space, visible cleaning regimes, and stricter regulationswere felt to be more secure. Others described seeking smaller stores or outdoor markets where the potential for viral transmission was lower and they could support small businesses. The variety of different sensitivities and practices involved in grocery shopping during lockdown is indicative of the range of matters of concern that COVID-19 introduced into the landscape of food; concerns around viral transmission, vulnerability of small businesses to reduced patronage, and heightened risk associated with (non-COVID-19) sickness when healthcare services were under pressure. These various sensitivities culminated in a collective aversion to shopping, with most participants describing a reduction in the frequency, and increase in the displeasure, associated with the activity. I don't feel good about extensive shopping, so I'm quickly doing my shopping and quickly getting out of there.  This aversion to shopping also prompted reflection on the distribution of responsibilities for grocery shopping. Participants described changes intended to protect other members of the household and extended family from harm. The designation of responsibility was complex and varied. In some cases, rather than a task shared between household members, shopping became a one-person practice or alternated between members of the household to limit exposure. Sometimes the person who interacted with more people (due to the nature of their work, for example) adopted responsibilities; in others, it was whoever was least medically vulnerable: It was I who went shopping. I didn't want to expose them [spouse and children] knowing perfectly all about how you could get contaminated, I thought to myself there is no way they were going to go shopping. It was me [and] I cleaned everything. It was super strict.  Practice-oriented research draws attention to the various ways that food practices are entangled with affective notions of convenience, security, responsibility, and care (Hertz and Halkier 2017;Molander and Hartmann 2018;Warde, Paddock, and Whillans 2020). As COVID-19 has resisted the performance of established practices, some seemingly obdurate affective dimensions of food practices have also been compromised. Convenience has taken on new meanings as the spatio-temporal configuration of domestic routines has become unsettled, and responsibility for domestic tasks has been redistributed in light of emergent risks. In these ways, we see how the affective qualities of food practices are fluid and contingent on matters of concern and material conditions. In this way, COVID-19 helps reveal the co-productive relationship between practices and meanings, underscoring the importance of these affective dimensions in stabilizing practices, and practices in stabilizing responsibilities and meanings (Schatzki 2010). Such observations call for further attention to how patterns of ordinary activity shape, and are shaped by, general understandings and teleoaffective systems of meaning (Warde 2016;Wertheim-Heck and Raneri 2020).
Reconfiguring taken-for-granted practices: planning, stocking, and storing food The unsettling of convenient, spontaneous shopping practices and the emergence of a collective aversion to shopping has also had implications for planning, stocking, and storing food. In some places, there were food shortages. While these were mostly shortlived, they nevertheless challenged the taken-forgranted predictability of supplies of staple ingredients (Hobbs 2020).
We cooked what we could buy in the market … there was not much to buy at the market. The street vendors were prohibited and thus everyone had to go to the market. Everyone scrambled to buy. Everything was gone very quickly. We couldn't choose what food to buy and eat.  The sense that food supplies were less secure than usual prompted intensive planning practices and experimentation in storing and cooking. Participants described multiple forms of tacit understanding involved in planning meals; understandings of food durability and embodied sense of deterioration, knowledge of food systems, past experiences of food insecurity, and advice on planning for crises. For example, in the following extract, the respondent describes drawing on crisis advice to guide their planning and grocery shopping: I have focused more on buying some of the things that are recommended for ordinary crises, ordinary crisis advice. That you should have dry goods and canned goods.  Existing research demonstrates that households often have a repertoire of "tried and tested" meals that accommodate the needs and tastes of household members and fit with normal rhythms of daily life (Evans 2011;Watson et al. 2020). During the lockdown, this repertoire was compromised by interruptions to food supplies and normal rhythms, prompting flexibility in regular cooking and eating habits, and in practice, participants described purchasing dried and canned goods, and adapting meals to match the availability and durability of products (also evident in quantitative analysis by Janssen et al. 2021).
I bought more dried food than before … I also tried to find different ways of cooking such as I mixed many kinds of vegetables to make a dish or made different kinds of cakes from the flour. (VN-08) And, for the grocery shopping, we learned to miss out some ingredients. At some point, we ran out of eggs, we ran out of flour. I told myself, never mind, I will do it another way.  These extracts illustrate that rather than the needs and tastes of the household, or the normal rhythms of daily life, grocery shopping, cooking, and eating became more affected by product availability than usual. Expectations about cooking with fresh and varied ingredients were challenged, and creative ingenuity was involved in finding "different ways" of working with available ingredients.
In time, participants described the emergence of meticulous planning practices. Several respondents related how they would plan at least a week's meals and where to shop, accounting for the availability and durability of foodstuffs as well as a need to limit the number of shopping trips, a pattern of practice that showed some evidence of stabilizing as the pandemic endured.
Considering that we could only go out once every ten days, we got to the table and started to write what we wanted to cook for the whole week. And so, we were wondering what we needed to buy, where we could go to buy it and things like that … it became just a little routine during the quarantine. (IT-05) We have planned more andand yes, maybe got a little better at using up the food, and knowing a bit better what we have and stuff.  These quotes illustrate how planning practices depend on both intuitive understandings of "what we have," resourcefulness in using up ingredients already stocked in the home, and forward-thinking about what meals will be served throughout the week. Though mundane, these reflections were contrasted with ordinary, more spontaneous food shopping.
I never used to plan shopping before. But now I have started planning shopping, because now I have wanted to shop for one week at a time. So, I have been shopping for a lot of dinners at once.  The links between infrastructures of food provision and everyday routines have been documented throughout the literaturefor example, highlighting connections between bulk-buying and freezers (Shove and Southerton 2000) and infrastructures for on-the-go eating with informal meals (Middha et al. 2021). COVID-19 challenged assumptions about the reliability of food supplies and the simplicity of food procurement, prompting experimentation with alternative ways of planning and storing as well as preparing different meals. Participants' accounts demonstrate how food practices co-evolved with accessible infrastructures of provision and expectations regarding the reliability of supplies. When these foundations are unsettled, ingenuity is required to continue managing food for the household. These experiences of alternative ways of planning, shopping, preparing, and eating food add to stocks of embodied knowledge (Wallenborn and Wilhite 2014). However, the continued performance of alternative practices is codependent on infrastructures and routines (Sheth 2020). For example, Hobbs (2020) highlights that although the pandemic increased a trend in "shopping local," the failure to alter the fundamental economics of the sector means the permanency of this trend is unlikely.
Experimentation: practice emergence during disruption As well as creating friction for established practices, disruption paved the way not only for reactive adaptation but also for purposeful experimentation, involving the recombination of objects, spaces, and sensitivities. In this section, we discuss how new configurations of practice elements, new sequences of practice, and new spatial and temporal dimensions of food contributed to experimentation in everyday practices.

Reconstructing healthy and comfortable lifestyles
The pandemic provoked reflexivity on the role of food practices in everyday life in terms of their contribution to healthy and comfortable lifestyles. Existing research demonstrates the role of food practices in caring. Mothers in particular are responsible for providing "good" food and raising educated young consumers (Cairns and Johnston 2018), as well as sharing values on ethical food, in some households (Cairns, Johnston, and Mackendrick 2013). For many of our participants, lockdown strengthened intentions to keep the self and the family healthy, contributing to the emergence of alternative food practices.
The thing that has changed the most in my routine is eating healthy. Stop eating in my work cafeteria. I learned to cook. Introducing the children to other things than pasta. Introducing them to all types of cuisine. (FR-16) Normally, we [students] just care about things outside and do not care about our health. Since COVID-19 we started to take care of ourselves better, because we are afraid of death … My parents said we have to eat nutritious food to improve our immune system and resistance so that no virus could enter our body. (VN-10) Lockdown accentuated the role of healthy eating in looking after one's self and family in the immediate term (VN-10) as well as providing children with the ability to care for themselves in the longer term (FR-16). The concerns described in these extracts were similar for other participants, and several reflected on how they had more time to enact these concerns during the lockdown as food became less contingent on other practices. I looked for food stuff with more nutrients. More vitamins etc. To stay healthy but also because one always wanted to care more about it. And now I had the time to deal with it, what one can do with certain products and how to prepare them etc.  Existing research shows that people often report feeling that they are short of time for food preparation and cooking, as practices involved in planning, preparing, and eating is infringed on by others, such as leisure and work (Daniels et al. 2012;Southerton 2020; Etil e 2019). Our study arrived at similar conclusions, with participants reflecting on how a reduction in other activities and a shift to homeworking and schooling reduced the competition between practices for time.
Existing research also highlights how ideas about healthy food exceed formal nutritional guidelines (Spaargaren, Oosterveer, and Loeber 2012), and participants reflected on how food provided comfort and consolation throughout lockdown. Frequently, the practices described were not nutritiousinvolving the consumption of high calorie "snack" foods and alcohol. Instead, food provided comfort and reward during a period where other means of achieving these ends were restricted. I still have my normal working rhythm. My kids [teenage boys] are different. They don't go to school and go to bed late at night. They eat a late sandwich in the evening and get up late. My lunch is their breakfast. … Since it is new to share lunch together, the lunch has become more special, extra luxurious … We had higher expenses in food shopping, due to more products bought in the snack category, like cookies, potato chips, ice creams, and more luxurious products such as sushi.  The extract above illustrates how lockdown undermined temporal order and household organization of eating, as well as the continued importance placed on eating together. Lockdown prevented participation in many practices, leisure as well as work and school, with the result that household members were more often at home to share a meal. Though the temporal pattern of household members' days varied, coming together to eat meals together (in this case an overlapping breakfast and lunch) was a commonly reported advantage of the lockdown. Yet for others, working from home meant that snack food was more readily available, and without others to share a meal, snacking undermined a more usual "proper meal." I have been snacking more often, and that … well, it is a shorter distance to the fridge. And maybe haven't felt that I have had time to have a proper meal, you know.  The relatively high consumption of snack foods during lockdown has been widely reported (e.g., Gerritsen et al. 2021). However, here we see various other ways that lockdown affected food practices and evidence of contradictory eating practices as people sought to fortify their well-being. These findings reveal the complex relationship between nutrition, ideas around "good food" (Spaargaren, Oosterveer, and Loeber 2012), and extended notions of caring for the family (Molander and Hartmann 2018). In participants' accounts, good food is not only that which provides nutrition, but also comfort and reward. At the same time, the sense of having eaten "properly" was compromised, showing the conflicting relationship between these different dimensions of food practices. When thinking about possibilities for purposeful social transformation, it is important that these complexities are considered. General understandings of "proper" or "good" food are neither static nor universal, nevertheless, they shape food practices and reveal tensions (and contradictions) between lived and scientific definitions of healthy or sustainable food.

Relocating
and reassembling sociable food practices Many participants described ways in which they relocated dining out into the home by recombining elements to recreate the sensation of variety, sociability, and entertainment fulfilled by a restaurant meal. Several interviewees reflected on the incorporation of elements previously associated with dining out into practices of dining in. For example, in the following extract, certain foods, an aperitif, and "dressing up" were involved.
Bread and pizza and so on, a lot of these things we didn't ever do before. The aperitif was almost never prepared at home … we mostly liked to take it outside … it's synonymous with free time, really to start the weekend after work. So you dress up, and that was a habit we had before, basically every weekend, then when it was no longer possible, we began to think that it could work in the house. (IT-04) These elements are similarly apparent in the quote below, where the respondent describes measures taken to replicate "date night." Since we could not go out, we made nice food here at home and dressed up a bit to have date night here at home and make something a little special out of it. And it worked like that. It was really nice.  For others, the novelty of a meal in a different place was stressed, and efforts to replicate the sensation of dining out made use of novel spaces in the home: [We have] a small terrace that before we hardly ever used much, just for laundry. We have discovered in this period that it is a great place to have breakfast, make some dinners or even just the aperitif. We discovered this part of the house that we did not consider before.  Finally, for others it was the experience of sharing meals with others, which was emulated using online platforms, or "clandestine" meals with neighbors.
There is a very big difference if you make all the meals yourself and you eat alone. I found out a bit by chance that if I was going to FaceTime friends that I liked to do it over a meal to get some company. (NOR-02) My best friend, she lives near here, she's single, she has no children, she works all day on her own. We had a clandestine aperitif. We started on Saturday night, with the impression that Saturday night was [special]. And after that we did it every Saturday night, we had a clandestine aperitif until the end of the lockdown.  Eating is a well-recognized sociable practice (Yates and Warde 2017) and dining out is a means by which people achieve a sense of leisure and luxury (Biermann and Rau 2020). Here, our findings illustrate the creative ingenuity involved in reassembling practices to maintain these functions during a period of highly restricted social interaction. Aperitifs, "dressing up," and a location different from the usual each add a sense of luxury and entertainment to a meal. As government restrictions prevented social interaction, the essential role of mealtimes in maintaining relationships is revealed. This resonates with other studies of domestic life during COVID-19, particularly the emergence of virtual dinner parties as a means of maintaining connections while social distancing (e.g., Kirk and Rifkin 2020). Though the benefits for health, nutrition, or sustainability are questionable, the disruption reveals the role of sociable food practices and dining out in generalized ideas of a good life and emotional well-being (Spaargaren, Oosterveer, and Loeber 2012), suggesting a need to attend to these qualities of food practices when thinking about strategic change.
Exploring novel culinary practices For many participants in this study, lockdown sparked purposeful experimentation with novel culinary practices. Respondents reported cooking different meals, purchasing equipment, trialing food services, and connecting differently with friends. Though the permanency of these experiments remains unknown, studying people's participation in provisional practices (proto-practices) and the potential for these alternative practices to normalize alternative food practices, can add to understandings of innovation in food practices more broadly. For several participants, cooking became an accessible leisure practice (and coping strategy) during a time when options were limited (see also Kirk and Rifkin 2020). Cooking is a declining practice in some countries (Plessz and Etil e 2019), and a practice that people report feeling increasingly squeezed in societies with consumer cultures that are oriented around convenience (Southerton 2020). Yet during the lockdown, participants described how cooking provided a sense of purpose, reflecting on experiences of bringing sourdough to life or cooking with homegrown rhubarb (NOR-28). For others, cooking provided a means of entertaining the household (e.g., FR-16) and providing relief from boredom and relative inactivity (UK-02). In this extract, it is not only cooking but also planning that provides entertainment, and this respondentdespite not being much involved in cooking beforehad set a task to sustain their exploration. Similarly, in the extract below, efforts to avoid being adversely affected by food shortages led to an abundance of both ordinary and unusual ingredients, prompting "culinary experiments." Let's say that the first 15-20 days we all found ourselves with an avalanche of food at home and the primary thought was "what am I cooking today?" and we all baked hundreds of cakes and cuisines of a thousand nationalities. We did a lot of culinary experiments.  For others, working from home enabled a break from the usual repertoire of "tried and tested meals" ) and the opportunity to explore forms of cooking that they previously aspired to.
I was a bit more experimental [during lockdown]. I got the opportunity because I'm at home all day and have more time and not so tired because when I go into work, driving can be quite tiring in a city. So, when you get home, you're tired and you usually want to cook something really quickly. And so, because I was at home all day, I was able to have more time to then start looking at more complicated recipes.  It is important to recognize that the relative luxury of experimenting with foods in the extracts above was not a universal experience. Though many participants reflected on having additional time to cook or otherwise enjoy food practices, for some respondents, COVID-19 introduced additional pressures on already time-scarce households. For others, restrictions on the availability of products, or dependence on others to shop, meant experimentation was essential rather than recreational, a means of working with produce that was available (e.g., VN-06 and UK-05) or would last until they (or someone else) could shop again (e.g., FR-12).
Another common theme in the findings was the acquisition of items to facilitate culinary experiments. To enable the alphabet project above, NOR-01 described purchasing small items for the kitchen, such as spatulas and scales. Others bought cookbooks (NOR-09 and NOR-23) and a Vietnamese family bought an oven (uncommon in Vietnamese households) now that all meals were prepared at home (VN-01). Similarly, other participants bought into alternative food services. Food boxes and delivery services provided additional ways of navigating risk and adding variety to meals. Food boxes in particular were found to have facilitated culinary explorations, as participants described that they introduced unusual foodstuffs and required different ways of preparing food, adding a sense of ceremony to otherwise mundane acts of preparing a meal.
I ordered a package with ingredients like that and it came with a recipe and so on. It became a little more like that, what should I say, almost like a ceremony. Yes, I put more time and thought into cooking.  In contrast, food-delivery services enabled continuation of access to a more usual routine of frequent shopping, while negotiating risks and avoiding additional trips.
We used apps. This is also what explained that the food budget has increased … we started using this app more frequently, sometimes my wife even ordered two or three times a day, a little something here and there.  In combination, these extracts reveal how unusual circumstances prompted new ways of doing; exploring different foodstuffs, ways of cooking, and food services. Though participation may be temporary, these experiments have implications for the ongoing organization of food practices in the home. The acquisition of new objects and technologies provided resources for the continuation of, or further experimentation with, how cooking and food-delivery services affect the budget for food and the timing of restocking (Hertz and Halkier 2017). Similarly, forced adaptation can lead to experimentation with alternative ways of procuring food, building familiarity with ways of doing that can contribute to solidifying alternative practices (Wallenborn and Wilhite 2014;Hobbs 2020). In this way, we see how the performance of food practices during COVID-19 contributed to the ongoing evolution of food practices adding competencies, materials, and memories to guide future performance. It is this open-ended quality of food practices that we turn to in thinking about the implications of these findings for policy intervention in a later section of this article. However, first, we reflect on the differentiated nature of experimentation.

Bounded experimentation
The previous sections have reported on how disruptions affected food practices with striking similarities across the countries included in the study. However, it is important to recognize that participants' experiences of lockdown were diverse, and innovation in practices was bounded by socio-material circumstances, embodied knowledge, affective teleologies, and established routines. Hence, household composition and work-related activities during confinement were critical to the (re-)organization of food practices. For some participants, lockdown prompted changes in the composition of households, with implications for food practices.
My son-in-law, who is a vegan, came to live with me together with my daughter and her three-year old. I visited the vegetarian butcher and started to also prepare vegan food. So, the menu changed drastically.  In other cases, changes in working patterns and reduced social commitments enabled the coordination of multiple households whose shopping, cooking, and eating would have otherwise occurred independently.
I do not often cook, I normally eat out [street food]. However, during the lockdown, I could not go out to eat. My neighbors and I, we cooked and ate together. We are all single. We contribute money, we assign who goes shopping and who cooks and then we enjoy eating together. Normally, we had different time schedules, so we couldn't do the same thing. During the lockdown, we all stayed home.  In both of these cases, changes to diet and food practices were experienced with pleasure, providing a source of enjoyment during an otherwise stressful period. For others, food practices became an additional burden, or a way in which the detrimental impact of lockdowns on otherwise contented lifestyles became apparent: We have not eaten on time and together since I lost my job [cleaning lady in hotel that closed during the lockdown]. My husband and I now go out for work whenever people need us to run some errands. My mother-in-law stays home to feed the children on time.  In this example, the participant lost her job at the beginning of the pandemic and the household income became dependent on precarious work. Food practices like shopping and cooking were constrained by these new obligations to work and the loss of a regular shared meal and became symbolic of a lifestyle currently inaccessible to them. Similarly, the extract below describes how, initially, food practices provided a sense of purpose, but over time became frustrating and undermined their sense of identity.
The question I can't stand is: "What are we going to eat?" … Cooking used to be a shared task, because sometimes I came home at 10/11pm from work and well they managed to cook, but during the lockdown I was always home. So, every day they looked at me and asked me: "What are we going to eat?" At first, I didn't care much, because after I stopped my job, I felt useful, mom, mother, wife, etc. … But after a while, I was sick of it, it is not everything I am.  The few essential workers within our sample described how lockdown increased their responsibilities at work, infringing on their desire to perform different food practices (in this case, buying local foods). I didn't understand all those people that started baking and cooking. I didn't recognize myself in those activities. I was busier than ever. We had to reinvent education for small children. I am very proud of what we have accomplished … I was working day and night.  These extracts show the continued importance of the relationships between household composition, work, care, and food practices during COVID-19, resonating with various bodies of literature (Beagan et al. 2008;Cairns and Johnston 2018;Holm et al. 2015;Power 2020). For some respondents, the responsibility of providing for a family while also working and/or parenting became an additional burden and source of stress. Here change is shaped and constrained by necessity, and latitude to explore new ways of doing things is differentiated by household conditions and various responsibilities to others. It is important that these nuances do not go unrecognized, as they provide vital insight into the sources of fluidity and stability that affect transformations in food practice. We have only scratched the surface of these issues in this article and more research would enhance our understanding of how diverse living arrangements and working situations have mediated experiences of COVID-19 and their impacts on domestic practices.

Discussion
We began this article by proposing that studying lived experiences of the pandemic could contribute to our understanding of the dynamics of social practices. In the introductory sections, we outlined how practice theories enable us to comprehend consumption as distributed and interlinked, in such a way that acts of consumption can be understood as embedded within cultural, political, and material developments, both in food systems and everyday routines. We argued that the pandemic has provided a lens through which to observe the implications of concomitant disruption in food systems of provision and everyday routines for food practices. Our study contributes to understanding what enables or hampers change as we demonstrate the various ways that COVID-19 restrictions have influenced food practices and destabilized taken-for-granted ways of doing and prompting experimentation. Here we return to the question of social transformation in light of problems surrounding contemporary food systems, reflecting on these findings to consider what insights can be gleaned to inform food policy. Few of our findings directly suggest any shift toward sustainability during the lockdown. Although there are hints of changes in food planning and procurement, as well as greater autonomy and reflexivity around consumption, these do not assure sustainability and the longevity of change remains uncertain. Instead, our findings provide insights into how ordinary (unsustainable) routines are held in place, and what happens when links within and between practices are broken and (at least temporarily) remade.
It follows that our findings support others in calling for more systemic modes of intervention if we are to reduce the problems associated with contemporary food systems (e.g., Welch and Southerton 2019;Labanca et al. 2020;Larkin, Hoolohan, and Mclachlan 2020). Fundamentally, we have demonstrated how food practices are affected by public policies that are intended to limit viral transmission. These policies disrupted food practices by affecting spatial and temporal patterns of daily activity and introduced new notions of risk and materialities. Existing research shows that food practices are flexible, fitted in among other practices that depend on institutional rhythms, such as commuting, working, and making school runs (Southerton 2020). Our findings resonate with this literature, showing how the arrangements of practices within the day affect the timing, location, frequency, duration, and nature of practices, as well as peoples' experience of practice (i.e., whether food practices are a pleasure or a burden). Existing policy interventions typically target food practices directly, focusing on planning, storing, and cooking [and more often still, the people who perform these practices (Evans 2011;Crivits and Paredis 2013)]. Our findings suggest that reconfiguration of food practices resulted from fracturing these practice bundles, adding weight to arguments for initiatives that engage in shaping the ways that practices interlock and recrafting the social and material developments that contribute to their continuation (e.g., for grocery shopping this may include urban planning, working hours, and mobility infrastructure) (Spurling and Mcmeekin 2015;Watson et al. 2020).
Our results also highlight how the experiment enables continued access to the intangible functions (e.g., social connection, comfort, and well-being) that food practices provide and are preserved, emphasizing the importance of these affective qualities in directing the performance of food practices. Achieving these outcomes during a period of unprecedented disruption involved the incorporation of new technologies, services, spaces, and sensitivities in the mundane performance of food practices. For example, we saw how the experience of dining out was recreated within the home environment, and how sharing a meal involved different logistics to organize household members or technologies to involve distant friends. Existing research has repeatedly emphasized that food practices are comforting and sociable (Yates and Warde 2017), and deeply involved in caring for oneself and others (Dubuisson-Quellier and Gojard 2016). This remains unchanged during the pandemic. However, our findings reveal some of the adaptations in food practices that enable the fulfillment of these roles under turbulent conditions. The pandemic reminds us that policy interventions must be alert to the social and cultural dimensions of food practices (Biermann and Rau 2020;Dubuisson-Quellier and Gojard 2016). Preparing meals, selecting ingredients, and sharing food (whether corporeally with household members or virtually with distant friends) continued to be comforting and caring practices, even at this time of "social distance" (and perhaps even more so). For effective policymaking for sustainability, it is important to consider how food practicesas much as providing nutritionalso sustain social connections.
Our findings also underscore the importance of recognizing diversity and peoples' differentiated capacities for experimentation, capturing how essential workers, those in precarious work, and people living alone experienced COVID-19 differently from multi-person households and homeworkers. Our findings demonstrate the bounded nature of experimentation, which we have shown to be affected by wider socioeconomic conditions and connected to everyday routines. Less well-represented in our study are the experiences of people living in poverty, with long-term health conditions and disabilities, or from minority racial and ethnic groups. As the pandemic has continued, it has become increasingly evident that people in these groups are disproportionately affected, both in terms of health and socioeconomic impacts (Bambra et al. 2020;Hawkins, Charles, and Mehaffey 2020;Hu 2020;Shakespeare, Ndagire, and Seketi 2021).
Furthermore, people in these groups are marginalized throughout spaces of production and consumption, such that sustainable and healthy food practices are inaccessible to many of them. Policy commitments, such as the United Nations Sustainable Development Goals are intended to build toward a just, inclusive, and resilient world, yet policy interventions that enact a generalized understanding of consumption risk exclude those people who most need support. Disruptions fracture and reconstruct practices (Devaney and Davies 2017). However, in thinking about how more sustainable food practices might be stimulated, we must be alert to heterogeneity, asking how to create conditions that allow all practitioners, given their different commitments, to adapt Watson et al. 2020). Finally, our findings reveal a degree of continuity between nations, linked to common responses to COVID-19 in the countries included in this study at the time of the interviews (closure of schools, extensive homeworking, limitations on "non-essential" business and travel). Though we noted some crossnational differences, the design of our study reveals more about the similarities in how the pandemic interfered with everyday life, and the differentiated impacts experienced by people within countries. Though there is heterogeneity in food systems and practices between countries, our findings show the overarching importance of other aspects of daily life that further evidence the entanglement of food practices in, for example, commuting and childcare practices. As the pandemic continues, the differences in both impacts and responses between nations are becoming more apparent. Consequently, research should seek to understand whether dissimilarities among the countries have become more pronounced. Understanding these heterogenous experiences, along with longitudinal perspectives on the pandemic, will enhance our understanding of the dynamics of everyday practice in valuable ways.

Conclusion
The purposeful transformation toward sustainability (including public health) demands reframing the question of what transformation is for. It is insufficient to ask how we achieve a sustainable healthy diet or how we ensure that people get the nutrition they need in a low-carbon way. We must also ask how practices can be re-coordinated, and how the services that food provides (e.g., social connection, wellness, and enjoyment) can be achieved in alternative ways. Through a practice lens, consumption is emergent and ever-changing (Spurling and Mcmeekin 2015). Usually, change is gradual and continuous, as cultural and material conditions of society develop. Sometimes, change is sudden and discontinuous. This is where studies of disruption, such as this one, contribute. The pandemicand responses to limit viral transmissionoffered a lens through which to view how concomitant disruption in systems of provision and everyday routines affected food practices. This disruption was systemic; socially, materially, spatially, and temporally influencing mundane ways of doing for people in all of society.
As the relations between practices and elements become unsettled (Wethal 2020), we are granted "momentary glimpses of the fabric of normality" (Chappells and Trentmann 2018, 198), and though COVID-19 may not give rise to the enduring redirection of practice evolution, nevertheless, these glances inform our understanding of continuity and change. The unsettling of routine during these times enables us to observe both adaptability and resilience of dynamics within and between social practices. This in turn enables consideration of alternative futures of consumption and what could be done to disrupt unsustainable practices, highlighting the importance of systemic interventions that engage in the wider constellation of activities and infrastructures with which food practices are entangled. This is not to advocate for disruption on the scale of COVID-19, nor to minimize the disastrous social consequences of the pandemic. However, such unusual conditions provide an important perspective on disruption and change in everyday consumption practices, as the apparent instability creates possibilities for changes in the trajectory of consumption practice, and with it the possibility of transition toward sustainability. Civil engineer in public transport