After taste: Culture, consumption and theories of practice

Multi-disciplinary studies of consumption have proliferated in the last two decades. Heavily influenced by notions of ‘the consumer’ and tenets of ‘the cultural turn’, explanations have relied preponderantly upon models of voluntary action contextualised by webs of cultural meanings which constitute symbolic resources for individual choice. Arguably, the cultural turn has run its course and is beginning to unwind, a consequence of internal inconsistencies, misplaced emphases and the cycle of generational succession in theory development in the social sciences. Theories of practice provide a competing alternative approach which contests the colonisation of consumption by models of individual choice and cultural expressivism. To that end, this article explores the use of theories of practice as a lens to magnify aspects of common social processes which generate observable patterns of consumption. It is suggested that theories of practice might provide a general analytic framework for understanding consumption, one whose particular emphases capture important and relevant aspects overlooked by previously dominant approaches to consumption as culture. This article reviews reasons for the emergence of theories of practice and isolates some of their distinctive emphases. Strengths and weaknesses of the theory of practice as an approach to consumption are discussed.

Consumption, cultural turn, sociology, taste, theories of practice

There is widespread agreement about the importance of consumption in the contemporary world. Some argue that it is a principal driving force behind social and economic development, others that it is the core preoccupation of populations across much of the world. The topic has come to be studied extensively by most

Alan Warde, The University of Manchester, Manchester M13 9PL, UK.

Email: alan.warde@manchester.ac.uk

disciplines in the social sciences and the humanities and excellent studies have transpired. The multi-disciplinary origins of such works have produced a very diverse topic area. Despite much contestation and controversy, most explicit general formulations and accounts are based in cultural theories of various sorts. Research programmes developed after the 1970s were strongly influenced by the cultural turn, the broad public intellectual movement whose most radical manifestation was postmodernism, which altered understandings about the nature of culture and the appropriate ways to give accounts and explanations of social and cultural phenomena. Arguably, however, the cultural turn has run its course, such that fresh theoretical approaches to consumption may be anticipated.

Social scientists do not agree about what theories should be expected to do. The view with which I operate in this article is that theories are instruments of selective attention. A theory is a set of propositions (discursive or algebraic) which, when seeking to explain why or how situations (processes, events or states of affairs) come to be the way they are, identifies what entities to look out for (relevant and important entities, whose properties and dispositions will normally be described) and in what relationship those entities stand vis-a-vis one another. Such relationships may be classificatory, associational or causal. Theories necessarily bracket off most parts of complex reality to give a parsimonious account of how particular phenomena operate, with some disciplines typically seeking more parsimonious or reductive theories than others. Consequently, a principal effect of any theory is that it emphasises some features of the world and not others. Within a discipline, theories differ by virtue of their emphases.

One such view of theory is supplied by Andrew Abbott (2001a) who characterises the development of sociological theory as revolving around fundamental, and ultimately irreconcilable, analytic oppositions. He argues that the same fundamental disputes recur because their basic puzzles revolve around inerasable oppositions. In the process of disputation, some positions become discredited or fall from fashion, but only temporarily. For, to embrace one side of a core opposition makes it impossible to give sufficiently comprehensive or balanced accounts. Victory for one generic view at any one point in time will later be redefined in a more accommodating fashion or reversed. Theory is thus cyclical rather than progressive. Nevertheless, despite reverting periodically to common conceptual starting points, we become better informed, partly through having mapped more of the empirical terrain of the social world. This perspective can be used to chart the unwinding of the cultural turn and the emergence of theories of practice.

In the next section, I review schematically the emergence of sociological approaches to consumption as they developed in the later 20th century, pointing to the influence of cultural theories and identifying some significant limitations. The section 'From Culture to Practice' introduces theories of practice, which have gathered some momentum as a potential alternative to the previously dominant cultural approaches. Brief comment on their emergence is followed by a summary of their principal distinctive emphases. The section 'Theories of Practice and the Sociology of Consumption' reviews some recent empirical applications to various

fields of consumption. Subsequently, the section 'Some Problems and Limits of the Sociology of Practice' evaluates some critiques of theories of practice with a view to enhancing their value for investigating consumption.

Social scientists turned their attention to consumption in response to unprecedented material abundance during the Long Boom. Before then, interest was shown in consumption primarily in the context of the study of poverty or the normative critique of leisure and luxury. Since the late 1960s, the social science of consumption has had three broad, partly overlapping, phases of development, each of which has had a distinctive focus. Schematically, emphasis shifted between the three fundamental dimensions of consumption – acquisition, appreciation and appropriation (see Warde, 2010).

In the first phase, the focus was the economic system and its reproduction in an age of mass production and mass consumption. Sociological accounts, revolving around themes from macroeconomics and critical political economy, were 'economic' in the sense that consumption was subordinate to, and to be explained in terms of, production; the stereotypical instance was the Marxist base and superstructure theorem. Cultural phenomena, like taste, were, if not determined by, at least heavily steered by the apparatus of industry, for example advertising, and were by-products of the unequal, usually class-based, distribution of property and income in capitalist societies. By implication, welfare provision was a consumption issue. Debates centred on the relationship between needs and wants, and upon the patterns and the justice of the existing distribution of access to goods and services among the population. In this phase, the default model of the consumer was no different to that of economic theory; a utilitarian model of the sovereign consumer was sufficient to account for what individuals and households purchased. Consumption was a process involving personal deliberation, albeit heavily influenced by commercial pressures, leading to independent decisions in light of preferences.

Radical new departures coincided with 'the cultural turn' in the humanities and social sciences in the 1970s. Critical of 'economic', attention increasingly shifted from the instrumental aspects of consumption to its symbolic dimensions, and especially its capacity for communication. The new cultural studies was one manifestation providing a major stimulus to the sociology of consumption in Europe. It contested not only economic explanation but also the earlier moral condemnation of consumer behaviour. Not only did mass-produced goods and services provide comfort and entertainment, they also expanded cultural experience for many people, supplied materials to be used in personal self-development and self-expression, and, as with the example of gifts, established and consolidated social relationships (Warde, 2002). Consumption was thus re-habilitated, a cause for celebration rather than dismay. Based in processes of globalisation, aestheticisation and commodification, people's aspirations, activities and possessions were

interpreted in terms of the spread of 'consumer culture'. Research, much of it rich in semiotic and experiential detail, focused on style and taste, on sub-cultural expression, on popular culture, on uses of mass media, and on the playful and non-instrumental aspects of life. Increasingly, consumption came to be seen as a means by which individuals and groups expressed their identities through symbolic representation in taste and lifestyle, with their desires focused on symbolic rather than material reward. The key emergent figure was what might be termed 'the expressive individual', whose activities, possessions, meanings and judgements were directed towards symbolic communication of identity by means of lifestyle.¹

The cultural turn did not entirely obliterate earlier concerns. Not everyone took the cultural turn, and some who did took it without alacrity. Earlier impetuses that sustained more sophisticated and nuanced accounts of the importance of the material and economic phenomena were developed, for example, in the work of Pierre Bourdieu on the sociology of taste and research on the culture industries, drawing upon the Frankfurt School. Nevertheless, production-led understandings receded. The imperatives of the cultural turn had become effectively hegemonic at the point of the birth of a systematic empirical sociology of consumption in the 1990s. Consequently, the cultural turn and its associated research programmes indelibly marked the sociology of consumption and provided it with the bulk of its theoretical understanding and empirical findings to date. The dominance of the cultural turn might be evidenced by its trans-disciplinary impact, including the creation of unlikely genres like cultural political economy and cultural psychology, by the rapid expansion of the sociology of culture and the attention paid to explanations of taste, by flourishing research programmes in consumer behaviour like Consumer Culture Theory (CCT), and of course by the establishment of cultural studies per se (e.g. Arnould and Thompson, 2005; Hall et al., 2010; Santoro, 2011). Cultural theory came in stronger and weaker versions, depending often on the degree of embrace of postmodernist themes. Firat and Venkatesh (1995), for example, outline the radical postmodernist programme for research and analysis in the area of consumption. More moderate programmes included CCT and much work in cultural studies. Nevertheless, as Reckwitz (2002b) pointed out, cultural explanation in different versions was rampant and, as Kaufman (2004) observed, the tendency was to give exclusively cultural explanations of cultural phenomena.

As Abbott might anticipate, elements side-lined in the period of the flourishing of cultural analysis became due for re-cycling. The elaboration of the cultural turn gradually produced a battery of objections by critics who accused it of neglect of practical and routine activity, embodied procedures, the material and instrumental aspects of life and mechanisms for the transmission of culture into action. The emphases of the cultural turn diverted attention away from some empirical phenomena relevant to the analysis of consumption. Because much of the work on the culture of consumption focused on the display for others of symbols of identity the many aspects of consumption that are routine, ordinary or inconspicuous were obscured (Gronow and Warde, 2001). Also, investigation of class and status became less common, and there were fewer studies of resource distribution and

the influence of material inequalities. Moreover, the cultural turn had found little place for objects and technologies as material forces. As Reckwitz (2002a) charged, in cultural theory '[t]he material world exists only insofar as it becomes an object of interpretation within collective meaning structures' (p. 202). He argued that the main theoretical problem of the cultural turn was that material entities, the copious materiality of mass manufacture and consumer culture, are treated as objects of knowledge and not as material sui generis.

Arguably, in addition, cultural analysis of consumption contained a further and deeper set of theoretical weaknesses embedded in its general theory of action. Despite its internal diversity, and thus important exceptions, primary recourse was increasingly had to a voluntaristic theory of action, upholding models of an active, expressive, choosing consumer motivated by concerns for personal identity and a fashioned lifestyle. The model of an active and reflexive agent predominated, implying that conscious and intentional decisions steer consumption behaviour and explain its sense and direction. In key respects, its model was little different from the sovereign consumer of neo-classical Economics. The dominant template of consumption in all disciplines remains modelling the individual engaged in many discrete acts wherein personal deliberation precedes personal, independent decisions made with a view to the satisfaction of preferences. The ever greater prominence of neo-liberal political and economic doctrine has given further impetus to this tendency to postulate the autonomy of the individual and freedom of individual choice (Holmwood, 2010). Of course, all disciplinary approaches admit in addition to contextual influences on decisions (income, prices, subjective norms, socio-demographic characteristics, lifestyle group membership), but individual choice remains the core presupposition. This persists despite increasing criticism of individualistic explanations. For example, Warde and Southerton (2012: 5–6) point to the failure of standard models to capture the practical, collective, sequential, repetitive and automatic aspects of consumption. New elements of such a critique can even be found in embryo in the disciplinary heartlands of explanations based on individual intention, economics and psychology, where recent developments in behavioural economics, cognitive psychology and neuro-science have indicted the dominant models of rational action for their failure to accommodate the automatic, reactive and habitual aspects of most normal human conduct (e.g. Haidt, 2012; Kahneman, 2011; Thaler and Sunstein, 2009).

One defect of the model of the expressive individual is that it mis-defines or mis-specifies consumption – a notoriously chaotic concept of which the active agent model uses only one of the possible definitions, arguably a poor one, precisely because it pays insufficient attention to appropriation. Sociological and socio-cultural studies of consumption adopted the notion of appropriation from anthropologists who in the mid-1980s applied their discipline's insights about non-market exchange and material culture to modern consumption (e.g. Appadurai, 1986; Kopytkoff, 1986; McCracken, 1990; Miller, 1987) in order to capture the importance of people 'domesticating' mass-produced and alien products, endowing them with particular personal meanings and converting them into items to be made

use of and enjoyed for their own practical purposes. The idea of appropriation emphasises use, referring to the incorporation, adaptation and using up of items to serve practical purposes. Consumption serves the practical activities of everyday life.²

Despite the perceived inadequacies of individualistic models, there is no well-established alternative. One strong contestant arises in theories of practice. Since the turn of the century, the concept of practice (and various derivatives) has circulated widely. While Lizardo's (2009) claim that 'practices now play as central a role in sociological thinking as values and normative patterns did during the functionalist period' may be an overstatement, the ideas of practice theories have had appreciable impact (p. 714). Scholars from different disciplines and sub-disciplines discovered, identified and sought to promote the use of practice-theoretic tools. The approach has been commended in terms of a 'new paradigm' for media studies (Couldry, 2004), a 'Practice Lens' or 'Practice-Based Studies' in management learning and organisational behaviour (Gherardi, 2009), a 'practice-oriented shift' in economic geography (Jones and Murphy, 2011), practice-oriented design (Scott et al., 2012), and 'practice theory' in consumption studies (Warde, 2005) and ecological economics (Ropke, 2009). These contributions sometimes have taken the form of a manifesto for practice-theoretic approaches, although usually they have done little more than rehearse earlier established concepts. Reckwitz's (2002b) pleasing and elegant formulations have been widely quoted, particularly by those who were engaged with (for or against) the cultural turn. Others influenced by 'the linguistic turn' were more likely to fix on concepts from science and technology studies (STS) (e.g. Gherardi, 2009). Yet others owed debts to the very influential work of Lave and Wenger (Lave, 1988; Lave and Wenger, 1991; Wenger, 1998) whose concept of 'communities of practice', conceived as vehicles and products of successful social and collective learning with particular relevance for business organisations, has been disseminated and redeployed on a very broad front.

The most common explanation of the emergence of theories of practice is that they were a response to a number of fundamental problems of social theory at the point of the passing of economism and Marxism in the 1970s (e.g. Ortner, 1984). Influential, though diverse, theoretical formulations were articulated by, among others, Pierre Bourdieu, Anthony Giddens, Marshall Sahlins and Michel Foucault. Described by Postill (2010) as the first of three 'phases' in contemporary practice theory, the impetus was primarily European, social theoretical, post-marxist, and macro-sociological and was especially concerned to reconcile the opposition between agency and structure.

Rouse (2007) contends that this and other core problems remained unsolved, as indicated by a continuing struggle with such matters when renewed attention, a couple of decades later, was signalled by the rhetorical announcement of

'the practice turn' in contemporary social theory (Schatzki et al., 2001). This second phase was more concerned with the philosophy of action. Filtering the cultural turn, inspired by STS and sociology of scientific knowledge (SSK), and taking agency for granted, it concentrated on performances. Schatzki (1996), the central protagonist, added Lyotard, Garfinkel, Charles Taylor and Judith Butler to the list of founding theorists. The extent of its internal differences was hinted at by Schatzki (2001: 2–3) when he noted that three diverse currents of thought, post-functional, post-structuralist and post-humanist, all found the approach attractive.

Divisions between and within the first two generations pose dilemmas for scholars of the third and current phase, who are trying to apply the theory to substantive explanation in diverse empirical settings. One key feature of the reformulation in the 1990s, most prominent in the work of Theodore Schatzki (2001), was to proclaim practices to be core to social scientific analysis of social order and personal conduct, practices being presented as the primary entities of the social world and society itself 'a field of practices' (p. 2). Locating practices at a meso-level permitted a solution to the methodological controversy which contrasted holistic with individualistic explanations.

While theories of practice still remain very diverse (Nicolini, 2012; Rouse, 2007; Schatzki, 2011; Stern, 2003), the absence of robust and definitive theoretical resolution has not prevented the emergence of distinctive and defensible empirical analyses; new emphases drawn often eclectically from the works of the celebrated theorists have shed fresh light on social processes. Typically, promotion is directed less to developing theory, more to considering how various themes arising from the heterogeneous sources of the first two phases of theoretical development might be employed to address problems of description, interpretation and explanation of social processes and behaviour in a particular domain.

Given that it is not easy to specify what exponents, proponents and implementers of practice theories hold in common, it is somewhat hazardous to try to capture their distinctive features. Perhaps what is most definitive is that which they oppose and which they seek to minimise when offering explanations. The rather crude contrasts of Figure 1 give a schematic map of differences from dominant accounts of action based upon models of the sovereign or expressive individual.

Some, particularly the strong, versions of practice theory suggest that some of the items on the left are antecedent and prior to, and explain, features on the right. So, for example, Pragmatists may claim that doing precedes and steers thinking, that habit and routine are the fundamental basis of all action, that activity is a matter of flow and not discrete acts, and that all consciousness is effectively practical consciousness (Joas, 1996; Kilpinen, 2009, 2012; Whitford, 2002). Bourdieu (1977 [1972], 1990 [1980]) would see decisions as the corollary of dispositions, embodied sense as the foundation of deliberative capacity, and individual purposes as a function of a shared habitus attached to a position in a field. A slightly weaker set of claims would be that the items to the left are much more prevalent and important in social conduct than those to the right, in order to insist that explanations should give due credit to routine, know-how, shared understanding,

the embodied and the material (e.g. Reckwitz, 2002a, 2002b; Warde, 2005). Thus, early work in the sociology of consumption emphasised habit and routine, practical competence and regulation (e.g. Gronow and Warde, 2001), while later work has variously picked on embodiment (Wilhite, 2012), the material (Shove et al., 2012) and routine and sequence (Southerton, 2013). A third and less distinctive view would be that the items in the left-hand column have simply received insufficient attention. Hence, some versions of practice theories give emphasis to features in the right-hand column, and also, of course, some exponents of cultural analysis are perfectly content to admit the pertinence of some items to the left.

Practice theories are, then, in part an attempt to redress the 'biases' of cultural analysis which was hegemonic during the second period of consumption studies. Theories of practice seem appealing for the study of consumption because they promise to make a double correction to previous work; first, by providing an alternative framing to models of individual choice, whether based upon the sovereign or the expressive individual, and, second, by uncovering and exploring phenomena normally concealed in cultural analysis. This is best considered a matter of the emphasis put on different aspects of conduct. Against the model of the sovereign consumer, practice theories emphasise routine over actions, flow and sequence over discrete acts, dispositions over decisions, and practical consciousness over deliberation. In reaction to the cultural turn, emphasis is placed upon doing over thinking, the material over the symbolic, and embodied practical competence over expressive virtuosity in the fashioned presentation of self.

Practice theories offer a promising and vibrant basis for a programme of research. While its major impacts have occurred in organisation studies and science studies, its influence on studies of consumption is not insignificant. Leaving aside those accounts which use the term 'practice' as a mere synonym for activity, and looking instead to studies which announce themselves to be applying a practice-theoretical approach, we can find a growing corpus of empirical work. The Journal of Consumer Research and the Journal of Consumer Culture have each carried several articles since 2005. Such work sometimes seeks explicitly and simply to apply a version of practice theory as a framework for the organisation and presentation of an empirical account. Sometimes, this is done for demonstration purposes – by Schatzki (2002), for instance, when he illustrates the application of his ontological theory of the social site by way of herbal medicine manufacture and watching horse-racing. On other occasions, depictions of activity find it optimal to employ the theoretical concepts, their connections and juxtapositions, to indicate how certain activities are accomplished.

The gradual filtering of theories of practice into empirical studies of consumption has come through particular research areas, probably the most important of which has been environmental degradation, climate change and sustainability. Theoretical positions have been fashioned in relation to the very substantial role of household consumption in climate change, because the use of natural resources to fashion objects and operate machines constitutes a major political problem (Shove et al., 2012; Spaargaren, 2011; Warde and Southerton, 2012; Wilhite, 2012). A particular feature has been the attention to the mundane activities of everyday life. The use of water and electricity, for example, is incurred in practices of cleaning, washing and keeping warm or cool, about which topics the paradigms concerned with symbolic display, communication and presentation of self have little purchase.

Bringing the material and functional properties of things back into the account was signalled theoretically clearly by Reckwitz (2002a) and has been exploited extensively. Thus, Shove et al.'s (2012) three key concepts for practice theory – meanings, competences and materials – give as much weight to the things as to meanings because of their affordances for many everyday practices which surreptitiously use up natural resources. The intellectual tools for exploring the material component of practices derive partly from STS, wherein the human use of machines and objects is fundamental. The resulting research deals with activities like the washing of bodies and clothes (Shove, 2003), gardening (Hitchings, 2007), heating and cooling (Shove et al., 2013), using electronic devices (Christensen and Ropke, 2010) and waste disposal (Evans, 2011). These are practices which are dependent upon inconspicuous use of energy, water and scarce raw materials and where changing of patterns of consumption might mitigate environmental effects.

A second area which has attracted much attention is the topic of eating, which involves both mundane and socially symbolic features. As a prime instance of consumption (Wilk, 2004), eating recommends itself as a mundane and routinised

activity, which is founded in bodily habits and learned taste, of both sensual and social type. Bourdieu (1984 [1979]) was unusual among social theorists in making a contribution to studies of eating, which he does nicely in Distinction: A Critique of the Judgment of Taste. Eating is a propitious area for investigation because it can incontrovertibly be characterised in terms of the material, the corporeal and the mundane, and by repetition, routine and convention. Explicit and deliberate applications of theories of practice have dealt with mundane matters like kitchen hygiene (Martens, 2012), the operation of new kitchen machinery (Hand and Shove, 2007; Shove and Southerton, 2000; Truninger, 2011), cooking (Halkier, 2009), dealing with excess food (Evans, 2012), temporal routines surrounding meals (Southerton et al., 2012), as well as more publicly visible and distinguishing activities like weight management (Jauho et al., in press) and eating out (Warde, 2004). The study of eating necessarily addresses both physiological and aesthetic aspects of taste. It also points up some important differences between cultural analysis and practice theory, especially in relation to preferences conceived as articulated liking for particular items. It is hard to say that one likes a food that one has not eaten; the practical experience (participating in an eating event) of its consumption is a precondition for passing judgement, and new foods usually provoke discussion.

A third area where practice theories have proved attractive is in relation to enthusiast groups around recreational pastimes and aesthetic movements, as for example ones which engage brands and styled objects. The 'brand community' is one intriguing instance of the way that people establish social relationships around commercial products; ordinary people not only actively engage with their favourite products but come to feel a sense of attachment to networks of devotees who admire the same brand, as, for example, owners of MG or Saab automobiles (Muniz and O'Guinn, 2001). Schau et al. (2009) demonstrate through an examination of nine different studies that theories of practice provide a framework for a general analysis of brand communities. Arsel and Bean (2013) offer a powerful analysis of collective commitment to an aesthetic style in interior design – 'soft modernism' – by means of a study of an online exchange of opinion and advice via the Apartment Therapy website. Groups of people, self-organised around activities which have both practical and aesthetic dimension, involving manual work and social organisation as well as appreciation, can be framed sympathetically by practice theories by virtue of their emphases on the way that things are used, possibly particularly how they are valued, for the purpose of the practice, on doing things together as a group rather than isolated individual engagement, and where shared standards of performance, where what is good, is a primary raison d'être for membership of the group and, indeed, for the very existence of the group. These cases, where a good deal of self-conscious styling and organising occurs, suggest not so much distracted ordinary consumption as project-based mobilisation of attention.

Overall, then, a varied and expanding set of empirical studies which have reported within the framework of theories of practice have displayed some of its potential. They are just now beginning to contribute to the task of revising theory

in the light of the messy business of empirical investigation and analysis. Common difficulties have been identified and acknowledged, with some resolved rather better than others. The recent work of Shove et al. (2012) is a landmark, offering preliminary solutions across a broad spectrum for pragmatic research purposes. It suggests ways to represent practices, offering plausible concepts for comprehending key processes like change and recruitment and defection, and also reviews modes of intervention to alter behaviour. It is already serving as a defining introductory handbook, lending useful tools for the reporting of research (e.g. Maller and Strengers, 2013). It is therefore timely to reflect on the problems and the limitations associated with the application of theories of practice to consumption.

Critics of practice theories have identified a variety of defects. Among them are several with which I will not deal here, including its theoretical imprecision, its methodological eclecticism, its potential political conservatism and the difficulties with its application to policy. Here, I focus on a few (five) key issues especially germane to applications of the theory to consumption. I begin with disputes about the ontological and epistemological status of the theory, and specifically whether practices should be examined as entities with powers. This leads to another important question, when analysing performances, regarding the relative weight to be attributed to agency and deliberation, as opposed to habit, social environment and practical sense. This, in turn, is related to a third set of extensive controversial issues about how to conceptualise the relationship between mind, body, things, social context and action. Ultimately, that is a matter of a general theory of action. Fourth, I discuss the misguided criticism that theories of practice cannot handle social change. I suggest that the predominantly narrative form of explanation employed in empirical studies of practice is entirely suitable to accounting for change. To that end, a wide range of data, methods and techniques of analysis may be employed. Whether that potential has yet been convincingly realised in substantive studies is questionable, and that may have led to a perception of weakness, and consequently scepticism about the theory's value. Finally, one genuine associated problem concerns the scope of theories of practice, specification of what can and cannot be explained would be most helpful. I pursue these five issues in turn, showing how these often abstract issues impinge upon application of the theory in empirical explanation.

A first key question concerns the status of practices and the viability of the programmatic injunction to treat practices as the most fundamental unit of sociological analysis. In addressing this question, second-generation practice theories distinguish between practices and performances, but remain divided about the

ontological and methodological status of practice. Schatzki (1996) attributes a social existence to practices, sharing being permitted by virtue of their having 'teleo-affective structures' which are not properties of the individual. Reckwitz's notion of the individual as a Trajer of practices, makes the distinction strongly but without explicitly clarifying the locus and substance of that which is carried or borne. Some more recent contributions constitute practices as entities, as Practices, accentuating their institutional and perhaps organised form (Shove et al., 2012). Thus, theories of practice divide over how to conceptualise emergent characteristics which go beyond the sum of the relevant performances, and whether practices should be seen as entities.

Because of Turner's (1994) critique of the hypostatizing effects of sociological use of collective concepts, many exponents of practice theory have been very reluctant to speak of Practices, preferring instead to restrict themselves to the analysis of Praxis or performances. Turner was critical, inter alia, of existing conceptualisations of collective entities and of the ways in which knowledge, especially tacit knowledge supporting the repeated and similar performances which recursively sustain common practice, could be shown to be shared and transmitted. This was barely envisaged as a problem in first-generation practice theory, for in the 1970s the use of concepts of collective action or institutions was unobjectionable and unexceptionable. Neither Giddens (1984) when presenting the relationship between routinisation and institutionalisation, nor Berger and Luckmann (1971 [1966])³ in their description of a dialectical process of habituation, objectivation and internalisation to explain the basis of social order, paused over such an objection. Subsequently, however, in a theoretical climate increasingly characterised by methodological individualism and postmodern suspicion of grand narratives, the explanatory role of concepts of collective action attracted suspicion.

There have been a number of effective ripostes to Turner (Barnes, 2001; Rouse, 2007; see also Lizardo, 2007 and Turner, 2007a) and Turner (2007b) himself seems to have accepted reformulations as adequate, though certainly not all the problems have been resolved. The case for examining practices as habitual and routine collective activity is especially well made by Barnes (2001). His argument against Turner is that habits are not just individual competences but are implicated in people's mutual orientation to one another as necessary to achieve coordination:

Shared practices are the accomplishments of competent members of collectives. These are accomplishments readily achieved by, and routinely to be expected of members acting together, but they nonetheless have to be generated on every occasion, by agents concerned all the time to retain coordination and alignment with each other in order to bring them about. (Barnes, 2001: 24–25)

Shared practices requiring mutually adjusted actions are many, including, Barnes says, singing, dancing, hunting, sailing, and doing science. Rouse comes

to similar conclusions in defending a normative model of practice against the more common view that observed regularity and repetition is sufficient to identify a practice. To the contrary, observable regularities are the consequence of the mutual accountability of performances:

A practice is not a regularity underlying its constituent performances, but a pattern of interaction among them that expresses their mutual normative accountability. On this 'normative' conception of practices, a performance belongs to a practice if it is appropriate to hold it accountable as a correct or incorrect performance of that practice. (Rouse, 2007: 529–530)

This would suggest a role for empirical inquiry into the social processes underlying accountability, although it does not necessarily justify viewing the practice as an emergent entity.

These arguments do not indicate which activities shall be considered as practices, nor how the boundaries to a practice might be constructed, commonly recognised, and sociologically analysed.⁴ If a practice is to be considered an entity, over and beyond an aggregated sum of many similar normatively mutually adjusted performances, it requires a robust means of identification. The difficulty is to reconcile the fact that performances are very varied with claims that they are instances of a common and recognised Practice; how can boundaries of a Practice be drawn in order to justify treating it as more than random personal activity and instead subject to collective formulation and regulation? Philosophical conundrums aside, for sociological purposes, the existence of an integrative Practice might be recognised in at least four ways. One might be that an instruction manual, outlining how to do a body of activity, could be written, and which a significant number of people might read. Another is its potential for inclusion as an activity in a time-use survey: enough people allocating a significant amount of time to an activity, and knowing that they are doing so when giving a report, is good evidence. A third criterion is that there are, or could be, disputes with fellow participants about the standards of the performances, in light of standards of excellence associated with the Practice. A fourth avenue might be to identify suites of specialised equipment devoted to an activity; objects typically appearing together, including being sold together implicitly or explicitly confirming their mutual association with a recognised activity, like pen and paper, automobiles and gas stations, microwaves and freezers. All these are clues to the existence of recognisable and discrete Practices. While ambiguity will never be eliminated, these constitute a reasonable set of indicators of the existence of a Practice which is collective in its prescription and evaluation. Note, however, that as with the majority of successful studies in the field of consumption, these are most readily identified in integrative practices with singular and unambiguous instrumental objectives.⁵ More complicated cases exist where criteria of success are multiple, because several related integrative practices bear upon the orchestration of competent performances, as arguably with parenting, playing and even eating (Warde, 2013).

Common to many hostile and sympathetic criticisms of theories of practice is the contention that it cannot explain change. This is demonstrably mistaken.⁹ A range of abstract mechanisms underlying the processes of change in practices have been identified and described (Schatzki, 2013; Shove et al., 2012; Warde, 2005). Moreover, change has been convincingly described in empirical studies of the

emergence, transformation, differentiation and decline of specific practices, including, for example, Nordic Walking (Pantzar and Shove, 2010), stylish veiling (Sandkici and Ger, 2010) and the use of the freezer (Shove and Southerton, 2000). Collective practices almost inevitably must be analysed as conduct evolving unevenly over time and in space. Practices lend themselves to narrative forms of explanation of change, often in association with configurational (or institutional) analysis of specific performances. To adopt methods familiar in both historical and conjunctural analysis seems straightforward. However, behind a misleading allegation lurk two problematic issues. One concerns assumptions about the agents involved in change, the other relates to the characterisation of social structure.

Recall that theories of practice were promoted initially in order to transcend the structure – agency dualism, which had typically resulted respectively in either holistic or individualistic accounts. As applied subsequently to consumption, accounts have steered closer to the pole of agency, to the neglect of matters of structure or institutions. Frequently, this commits the humanist fallacy that individuals are the sole source of change. Too often, individual reflexivity and impetus to change are conflated. Most people much of the time do not have control over the circumstances in which they find themselves, nor do they consider as sensible alternative courses of action. Actually, change in behaviour is likely to occur as often as a result of endogenous change in social circumstances; the situations in which people find themselves are neither constant nor recurrent. As Abbott (2001b: 254) postulates, fluctuation, disorder and constant change are more likely features of social co-existence than orderliness and reproduction. Exercise of individual agency as a source of social change should be considered a rare occurrence, a privilege of the powerful and a distant horizon even in the context of collective mobilisation.

Currently, practice-theoretic accounts most often analyse individual behaviour, albeit as performances rather than voluntary, deliberative personal choices. Yet, they pay little attention to the creation of norms, standards and institutions which produce shared understandings and common procedures. Sometimes, machines and artefacts dictate common procedures. Sometimes organised enthusiasts or groups involved in collective mutual informal regulation generate and then police norms and standards. Sometimes, commercial corporations and the state assume responsibility for the establishment of the rules, standards and institutions which are the structural elements of society, in most practice theorists' accounts – most explicitly in the work of Giddens (1984), but also variously present in other first- and second-generation formulations. The relative neglect of the processes lying behind the normalisation of practice leaves theories of practice in the same doldrums that have becalmed macro-sociology over the last couple of decades. Although more explicit attention to Practices as entities (and their inter-relationships) might help, a lack of explicit methodological protocol, not to mention persuasive documented cases, is debilitating.

A strong theory of practices will insist that structural characteristics are nothing other than the effects of the intermingling of many practices and that this is the object of macro-level analysis. Yet, while it is possible to conceive of the organised

and meaningful world of individuals and collectivities as constituted by inter-related, overlapping and competing practices, writing an account of the total social field in such terms is a very daunting prospect. At present, there is no consensus about how such a fiendishly difficult task might be approached. Swidler (2001) pondered whether some practices were more significant than others, such that the most important could be targeted for analysis. Intuitively, practices might be imagined to be nested, as with genres and sub-genres, or species and sub-species in other domains of knowledge. Alternatively, field theory might offer a lead. Bourdieu (2005), when contemplating how fields were related to one another, presented diagrams suggesting overlap, claiming that some fields encompassed others (e.g. the field of power contained all others (Bourdieu, 1996 [1992])), but his account was less convincing than the one offered for the internal workings of fields. Perhaps, instead, it is possible to extrapolate from conceptualisations of how performances draw upon multiple practices. Multiple competences underpin the successful performance of almost any focal practice, whether determining the focus is a matter of definition by the actor or the social scientist. Notions like practice bundles (Schatzki, 2013) practice complexes (Shove et al., 2012), meta-practices (Molander, 2011), assembled practices (Jauho et al., in press) and compound practices (Warde, 2013) all attempt to capture the way in which competent performances of some activities draw simultaneously upon several otherwise autonomous integrative or dispersed practices. However, it is far from obvious how to ramp up such concepts to the level of social systems.

Thus, it might be prudent to agree with Barnes (2001) that there are many things that a theory of practice cannot do, or cannot do as efficiently as other approaches, and to re-engineer connections with other complementary accounts. Whether there is a general solution to the question of which theories are most complementary remains to be seen. In relation to consumption, however, several options already exist, most of which locate consumption behaviour in the context of wider economic, and to a lesser extent social, processes. McMeekin and Southerton (2012) suggest the Multi-Level Perspective of Geels (2002) is a suitable partner to a practice-theoretic account of consumption because its account of the evolution of socio-technical systems shows how technologies bind producer to consumer and technology to performance (see also Watson, 2012). Fine and Leopold's (1993) concept of systems of provision suggests how the arrangements for the delivery of different products to the market frame and steer consumption. This might be seen as a specific version of supply chain approaches, but one which pays greater attention than others to the consumption moment. Harvey's (2007) instituted economic process framework, which emphasises the mutual inter-dependence of consumption with processes of production, distribution and exchange, advances an institutional and configurational mode of explanation. All of these have some affinity with more general theories of political economy from which further inspiration might be sought. However, an interim solution might be to render accounts firmly and explicitly in terms of institutionalisation. This could exploit the affinity between the concept of institution and the notions of shared understandings and

shared standards of performance which epitomise theories of practice. A disciplined commitment to provide a deep and thorough account of institutional context, with special emphasis on the impact of practice entities other than the focal one, might be sufficient for most explanatory purposes.

Theories of practice are varied in nature, but in the field of consumption studies, they have challenged both individualistic explanations and cultural excess. Even if few take much notice, they have generated a powerful critique of methodological individualism. The relationship to culture is more ambivalent, for there is no wish to deny the role of culture or belittle the achievements of the cultural turn. Following Abbott's (2001a, 2004) logic, the mode of succession to the cultural turn is likely to be less obliteration than ingestion and accommodation. One reason is that most of the reliable empirical work to date was conducted in that mould. Another is that practice theory, if Reckwitz is correct, is itself an offspring of approaches to cultural analysis. Surviving elements include some methods of investigation, a bank of empirical findings open to reinterpretation, acknowledgement of the multiple positive functions of consumption, and recognition of the role of consumption in the formation of personal and collective identity.

Nevertheless, 'the practice turn' has unsettled the study of consumption by providing a rallying call for investigations which question the previously predominant emphasis on culture. The re-emergence of concern with materiality and the affordances of objects has had significant consequences, particularly because of concern about the environmental effects of current modes of consumption. A start has also been made in exploiting the idea that consumption is primarily a process of appropriation for multifarious and often mundane use. Cogent critique of dominant individualistic models of action anticipates a general sociological framework beyond methodological individualism. Promising leads arise from general meta-theoretical considerations, as, for example, with the emphasis on habit and routine, the recognition of the importance of the local setting or environment in the steering of behaviour, and the shared and social nature of practices. These intimations have framed a set of empirical studies where the use of concepts derived from the theory permit description and analysis of the way in which social practices subend consumption.

Outstanding problems remain, some identified above in the section 'Some problems and limits of the theory of practice'. Practice theories may need supplementing with other frameworks, particularly to capture macro-level or structural aspects of consumption. This does not mean a return to the old economism, but probably entails some recovery of political economy and re-articulation of the link between consumption and economic production (e.g. Smart, 2011). While observing and exploiting the analytical distinction between production and consumption in order to study consumption as a phenomenon in its own right was an essential step in the 1990s, theoretical reconciliation is due. That is to say, theories of practice have

been much better at re-describing and analysing in a distinctive way details of the use of commodities in the performances of everyday practices than they have in elucidating the institutional or systemic conditions of existence of those practices.

I would like to thank Isabelle Darmon, Frank Geels, Jukka Gronow, Sue Scott, Dale Southerton and Luke Yates for comments on a draft of this article.

Alan Warde is Professor of Sociology in the School of Social Sciences, University of Manchester and Guest Professor, Department of Food and Nutrition, University of Uppsala, Sweden. Research interests include the sociology of consumption and the sociology of food and eating. Current projects concern developing theories of practice, strategies for changing eating habits, and changes in the organization of eating out in the UK.