The Order of Service: the Practical Management of Customer Interaction

by Barry Brown
University of Glasgow

Sociological Research Online, Volume 9, Issue 4, <>.

Received: 12 Jan 2004     Accepted: 9 Nov 2004    Published: 30 Nov 2004


This paper discusses a pervasive yet neglected form of social interaction, that between service staff and customers. Observational and video data from two different shop settings are used to explore three aspects of service interactions. First, queues are discussed, a mundane yet massively prevalent device for managing when and how customers are served. Queues depend on customers ability to 'work the queue', to be able to see who is queuing and their place in the queue. This rests not only on the recognition of queuing behaviour, but also its production by those queuing. Second, artefacts in shop settings have not only a material role, but are resources used in interaction. The shop counter is both a surface to place goods, and a shared interactional space between customer and staff where the placement of goods has meaning for the interaction. Third, staff and customers manage their interactions using rhetorical devices, such as using three part list display can be used to show the validity of advice being given. From these observations we draw two conclusions: Behaviour in service settings has a strong moral component in that divergences from correct behaviour (such as queue skipping) are quickly sanctioned. This morality is from those in the setting, rather than an analyst's judgement, suggesting that the morality of economic markets can be studied as an endogenous feature of those markets. Second, customer service relies upon a prevalent yet powerful 'ordinary vision' - the skills of seeing, but also producing, the predictable actions that make up the order of service.

Consumption, Ethnomethodology, Shopping, Tourism, Video Methods


1.1 This paper studies a ubiquitous form of human interaction, interactions that although brief and seemingly unimportant, take place millions of times every day with little thought. Going into a shop and making a purchase is a fleeting encounter between ourselves and a stranger. While the exchange of goods in this encounter has long been a concern of economics, it is only recently that sociology has taken seriously the site and acts involved in this exchange. Indeed, although mostly ignored, much economic distribution depends upon these ordinary interactions. As Fred Davis comments in his classic study of the relationship between taxi drivers and their customers: "numerous interests beside that of simply effecting an economic exchange are customarily attended to and dealt with" (Davis, 1950).

1.2 A vast range of activities depend upon service encounters. The distribution of goods, at its final stage, requires that goods smoothly move from shops to the sites of consumption. Many service interactions (in particular, financial transactions) take place face to face, with complex interactions and talk around documents. The focus of this paper is the methods which both staff and customers need to use and display so these settings can work. Even in this fleeting encounter we see social organisation and interaction of some complexity. As with others who have looked at seemingly 'asocial' settings (Heath et al., 1999), we uncover if not a rich interactional milieu, at least a social encounter which repays further study.

1.3 This paper focuses on two settings: customer service in a covered market, and service in a tourist information centre. Studying these settings ethnographically we draw out the 'order of service': the ordinary orderly details of how customers get served, and how staff manage these interactions.

Consumerism and Customer Interaction

2.1 While the market as a topic has always been of central importance to the social sciences, it is only recently that consumers, and the activities of shopping and customer service, have gained serious attention. In part through the work those who have examined the "pleasures of consumption" or cultural appropriation (Bourdieu, 1984; Willis, 1990), consumption has become a topic. Recent work has turned to the sites of exchange themselves - supermarkets, shops and markets - to argue for the importance that the activities of purchase and shopping have in our consumption (Butler, 1991; Lancaster, 1995; Falk and Campbell, 1997; Crewe and Gregson, 1998).

2.2 While this has work at times 'exoticised' shopping, sometimes out of recognition (e.g. (Baudrillard, 1988; Chaney, 1990)), it has encouraged a move towards taking seriously the practices of shopping (Miller, 1998). This work shows that consumption is much more than ownership (Belk, 1995; Miller, 1995), but includes the acts of shopping and buying as well. For example, Miller has developed an understanding of supermarket shopping as an activity which is embedded in family settings (Miller, 1998).

2.3 While there is much to admire in accounts such as Miller's, these moves open up consumption from a simplistic exchange of goods (often portrayed as a necessarily exploitative relationship) to one with some creativity, an activity with an interplay of social forces and activity. However one clear absence in this work is at the level of the practical activities which constitute shopping. Rather straightforwardly, what is involved in going shopping, choosing items, running a shop, or selling goods and services? While Miller's work, and others in this tradition, contain descriptions of shopping they can be seen as missing out a number of the practical aspects of shopping and its management. That is to say many of the actual practices of consumerism are 'glossed over', with a discussion which instead tends towards discussion of consumerism in terms of its meanings, or broader significance.

2.4 For example, there is the question of how its that, when we are involved in customer service, we are rarely confused by the actions of others. What people are doing in these settings, be it pausing, making a purchase, or queuing, is seldom a mystery to us, despite the lack of spoken interaction. Indeed, much of what we do when we shop relies upon this - to queue one must recognise a queue when it is presented.

2.5 A move towards studying these 'details' has been made by the small body of work, motivated from an ethnomethodological standpoint (Lee and Watson, 1993; Randall and Hughes, 1995; Bowers and Martin, 2000). For example, in Randall and Hughes' studies of banking they show how staff manage their interactions with customers, and manage to "keep the queue moving" by changing the pace by which they work with customers to prevent a long queue building up. Staff used delays in the interface of their computer system to converse with the customer, managing the conversation to take advantage of these pauses. Alternatively, in Lee and Watson's (Lee et al., 1993) work on open markets they describe the work that customers do to differentiate themselves from those who are browsing the market. In particular, in this work we can see close attention to the bodily movements which customers use when managing their interactions with both staff and other shoppers. Clark and Pinch's (1995) remarkable study of market traders also demonstrates the crucial importance of interactional features such as crowd management and eye contact in producing not just orderly, but successful, selling[1].


3.1 The approach taken in this paper is ethnographic, in that it seeks to focus on the details of how shopping is done as an ordinary practice. The analysis draws on work from both an ethnomethodological and conversation analytic tradition[2] (Garfinkel, 1967; Garfinkel and Sacks, 1970; Sacks, 1995). From both these traditions this paper take a focus on the "seen but unnoticed" aspects of everyday life. That is to say not just the minutia or trivia of life, but rather the massively prevalent, yet intricately varied structures in how the world is arranged, structures which are normally taken for granted. It is this which ethnomethodology and broadly takes as its subject matter: how we interact (Sacks, 1995), talk (Atkinson and Heritage, 1984), walk (Ryave and Schenkein, 1974), work (Button, 1993), travel (McHugh et al., 1974), break the law (Sacks, 1972), argue (Goodwin and Goodwin, 2000), play with pets (Goode, 2002), or cry (Beach and LeBaron, 2002).

3.2 Conversation analysis (CA) as an approach has also made many inroads into studying the publicly available reasoning by which ordinary activities are achieved (Silverman, 1998). The sequentiality of action has been a central concern in CA, how it is in particular interactions are dependent on their place in a sequence and what has happened before and what will happen after. That is to say, gestures and the like are not context free, but develop as part of an ongoing interaction.

3.3 This paper draws on these observations in the domain of consumer interaction. In the sociology of shopping, discussed above, much work had contributed to a view of the "cultural consumer", a consumer who uses the objects they purchase to create and in response to the culture they are in. This is a step forward from the description of the consumer as the 'homo economicus', the rational optimising economic actor. Yet to an ethnomethodological approach, this work still produces a model of the consumer. We still seldom hear or see real consumers; instead the consumer is sketched out through deductive argument: because of this and this, consumers are like that. This model of the consumer which is produced explains why things are as they are since the reasons behind certain activities and behaviours can be read off from the model. Yet in this description there is little reference to what the consumer themselves actually does in detail.


4.1 To focus on these neglected details, ethnographic and video data was collected from two very different customer service settings. In the first setting a conventional ethnographic study looked at the operation of a Scottish Tourist Information Centre. Originally this data was used as part of a study into tourism (Brown and Chalmers, 2003) focusing on what advice staff gave and their interactions with tourists. However, the central role of customer service in this setting provoked an interest in how customer service worked. In particular, this setting gave access to the 'back stage' work of customer service, how staff managed their interactions so as to produce regular, unremarkable, smooth interactions. The access to the Scottish TIC allowed us to study staff and customers over a two month period, using a mix of observation and participant observation. Forty five hours of observations were collected from the TIC. Although video could not be collected in this setting, conventional ethnographic methods allowed us to study how staff managed their interactions over a prolonged period of time. While the TIC was not a commercial organisation, its commercial importance should not be ignored - the main function of the TIC was mediating between tourists and local cultural and commercial organisations.

4.2 The work of the TIC inspired us to investigate customer service in adifferent setting, to contrast with the TIC. The second setting investigated for this paper was a covered market with fixed stalls, mainly selling foodstuffs, in Göteborg, Sweden. Shoppers behaviour in this market was observed, videotaped, and analysed in group data sessions. The video data collected from this Swedish market allowed analysis of aspects of shopping which could be missed from purely observational methods, in that the videos could be watched repeated times. Using video in this way follows researchers who have developed Sacks, Schegloff and others' work on conversation analysis into an approach to studying interaction using video data (Heath, 1997; Goodwin, 2000; Heath and Luff, 2000). As opposed to a more traditionally ethnographic approach we did not interview or talk to the shoppers we were studying. The aim was not to uncover the 'meanings of shopping' which could be extracted from interview data, or even shoppers own accounts of their shopping behaviour. Instead the aim was to start with what 'anyone could see' in this setting. That is, what is it that anyone could see from the actions of shoppers which allows them to successfully shop in the same environment.

4.3 These two settings provide clear contrasts with each other in their forms of customer interaction. In one - the covered market - we have the more traditional sense of customer interaction, with the sale of goods. In the other we have the provision of information, rather than physical goods. The covered market setting was strictly commercial, again which providing contrast with the TIC's publicly funded role. Yet in both settings the interactions were constrained in time and a counter was used as a physical way of structuring the bodily interaction of customer and staff. The two setting provide a revealing contrast in the differences and similarities in customer service situations.

4.4 The discussion of the data from these two settings is structured into three sections. The paper starts by discussing the importance of the queue - as a central ordering device in service settings. In particular the focus is on the effects the queue had on how staff arranged their activities with customers - and the desire to "keep the queue moving", along with how customers managed waiting in the queue and showing themselves to be "next in line". Second, the paper discusses the objects which are involved in customer interaction - e.g. how the counter works as both an symbolic and practical prop in customer interaction. In the third section we move on to the customer interaction proper, and look at the talk involved in serving customers in a visibly competent manner, without being either too slow or too fast. In particular, staff in the TIC used a number of devices to organise their interaction: designing responses, producing 'reasonable' advice, and overlapping their interactions. This cut through data is in some ways opportunistic, in that our data allowed us to focus most closely on these three features. Yet each of these sections address an important yet ordinary building block of service interaction - queues, objects and talk. Each of these three aspects of customer interaction work together to make service possible, in how they enable the orderly organisation of service interactions on which the distribution of goods, and much more besides, rests upon.


5.1 One of the most pervasive aspects of shop settings is the queue. Shops have a strong temporal order, in that a set of customers is being served at any point in time, with a potentially larger set of customers waiting to be served. One of the key problems in any shop environment is thus managing the order and way in which customers present themselves to be served. One of the most familiar ways this is arranged is through that most ubiquitous of social inventions: the queue. "The queue" appears, at first glance, to be a simple structure which allows customers to put themselves into an order in which they will be served (Livingston, 1987). Yet, there is some considerable subtlety in how queues worked in both the setting we studied.

5.2 Queuing has been a topic of some interest to marketing researchers (Maister, 1985), in particular methods of changing the perceived wait time for customers, such as by giving information about estimated weight time (Hornik, 1984). The prevalent method here has been one which reduces queues to generalised features of a setting, with a set of factors (such as information provided, length of queue, weight time) which determine the experience for those queuing (Durrande-Moreau, 1999).

5.3 In a different way queues have also had a long standing interest for ethnomethodology, in that queues are something of a perspicuous example of social order. Garfinkel argues that queues demonstrate how social order is observable 'as is' without resort to "generic representational theories" (Garfinkel, 2002; Garfinkel and Livingston, 2003). As he puts it: "The order of service exhibits a phenomenon of order in ordinary society immediately and directly." (p5). This is an order that can only be seen in queues in vivo - asking queue-ers what they are doing will not get at the queue. An interview about queuing would not uncover the phenomena, as they quote from a hypothesised queue member: "you marvel at the way we stand in line? How do we stand in line? We stand in line, like anyone stands in line. What do you want of us?" (p25). While the order of queuing is an incessantly concern of those in the queue, to social science this achievement has been almost completely ignored. Yet even in ethnomethodological settings the abstract notion of the queue has been discussed much more than any specific examples of queues.

5.4 In the two settings we studied, queues were a method by which a small set of individuals could be served out of a larger, group. One 'challenge' to queuing is that there is little or no verbal interaction in a queue. The co-ordination is done with bodily interaction alone. Behaviour is not directly mandated or controlled by staff, or by the use of physical artefacts. As we will explore artefacts play a role, but they do not determine the organisation of the queue. Moreover, queuing must be done without undue hesitation - if one dithers when trying to queue then this can become problematic.

5.5 In the TIC how to 'work' the queue was a basic competency of both tourists, staff and management. Tourists needed to know where to queue, to watch for the next free member of staff and then approach, yet without making the faux pas of going up to a member of staff who was not free. Considerable effort had been put into organising the shop so as to 'funnel' customers into the one queue which then opened up onto the different members of staff. In an arrangement which will be familiar to most bank customers, the shop floor had a physical layout which led customers to form one queue, in front of the counters (a 'snake' arrangement (Schwartz, 1975)). A taped barrier cordoned off a channel from the counter for a meter or two into the shop. Customers would normally see the queue which was formed and join the end of the queue in the channel formed by the two taped barriers. However, at times, customers would walk around the partitions and attempt to get served from the side, especially if the TIC was quiet and there was nobody else queuing. This displays one of the key aspects of queues: they are not constituted by physical artefacts or the physical arrangement of the shop, but by people queuing. So with nobody else in the queue, there is no queue, and this provides the opportunity for individuals to make a mistake and attempt service in the 'wrong' way (e.g. by walking around the barriers). The arrangement of other customers, and the artefacts in the setting, provide resources which customers can read as instructions for proper queuing. When a queue is in progress the line of people is visible as an instruction in how to queue here and now. The partitions then physically manage that queue in formation, since if a customer joins the queue at the back they are physically forced to queue in a certain direction by the tape barriers.

5.6 For the TIC staff an important part of queue management was eye contact with customers. Managing eye contact is an essential tool for managing a waiting queue. As Goffman remarks:

"A waitress, for example, may prevent a waiting customer from "catching her eye" to prevent his initiating an order. [...] As these various examples suggest, mutual glances ordinarily must be withheld if an encounter is to be avoided, for eye contact opens one up for face engagement." (Goffman, 1963, p90).

5.7 In the TIC staff would gaze below the horizontal when busy to avoid making eye contact with customers. As one of the computer support staff explained to me, when working on the computers at the desk he would never gaze above the ground lest he catch a customers eye, have them approach and have to explain that he could not help them. In this way mutual eye gaze is an tool through which both staff and customers can request service, and answer that request.

5.8 Unlike much shop customer service work, at the front desk there were a number of task (such as filling out report forms, or booking accommodation) which needed to be done without a customer waiting. This meant that staff needed to hold off customers without a customer physically at the counter. This caused confusion for customers who would sometimes approached staff when they were busy. In this case staff would say "I'm sorry, my colleague will help you over there", even if their colleague was busy, redirecting the customer to wait for service without a direct rebuke. When staff were ready to serve customers they would look up and look directly at the start of the queue, or if a customer did not notice their inviting gaze they called out a request such as "can I help you over here?"

5.9 The queue was not without its stresses for TIC staff. There was the constant "push" of customers that needed to be seen to and helped, yet without rushing the current customer. When working the desk one could have ten or fifteen customers waiting in queue for service. Harper, Randall and Rouncefield (2000) refer to this as the pressure of "keeping the queue moving". Experienced staff in particular would resist the temptation to rush a customer because a queue built up behind them; they realised that the quality of the assistance given was more important than the relatively short delays (at the most around ten or so minutes) that a long queue would produce. If anything staff slowed when the queue grew, as not to rush customers any more than the queue itself.

5.10 There were a similar, yet different, range of issues which concerned customers in the covered market. In the covered market a ticket machine was used which would allocate service to customers in a strict first come, first served order. Customers needed to pull a ticket from a dispenser, and a display showed who was due to be served next. However, this was used in combination with a conventional queue, particularly when there were few people waiting. This could lead to some confusion when some customers were queueing in a conventional manner beside the counter, and some had tickets. The counter staff could solve this to an extent by their own use of the button which advanced the display of who was being served. When there was a mix of queuing and ticketed customers (For example, as happened when the market was getting busy) they would serve the queuing customers and leave the counter on the last previously served customer. When they had finished serving the queuing customers, or when their decision that "who was next" had a ticket, they would push the button and advance the counter. This advance also made a loud buzzing noise - signalling to customers that the ticket machine was now in use. When this happened customers without tickets could be seen then going to get a ticket, loosing their place in the queue, and getting a new place with the ticket.

5.11 Yet even with the use of tickets, there was still significant richness in how a customer could initiate a purchase. Again, this might seem like a simple activity, but if we look at video data from the market we can see how this was interactionally managed. In clip one (frames one to eight) a customer goes up to make a purchase. Her companion also approaches the counter, but does some work so as to be seen, at a glance (Sudnow, 1972), as visibly not a customer. The shopper, through her actions, shows herself to be 'a shopper', and the women's companion in turn shows herself to be 'a companion'. First, the shopper marches towards the counter, and puts her large shopping bag down on the ground with an audible "bump". She then places here small handbag on the counter which runs the length of the stall. This small shelf along the length of the counter is used by the shopper to place her handbag, allowing quick access to her purse (for payment), but also as a way of indicating her intention to make a purchase.

5.12 The shopper's friend here has a slight dilemma; she does not want to be seen as 'next in line', she wants to be seen as with the shopper. While she at first browses a second stall, she then positioning herself beside her companion without putting her bags down. As Ryave and Schenkein (1974) study of pedestrians shows, being able to "see groups" is an important part of navigating in a public space, in that we have a preference for not splitting up groups by walking between members. Accordingly, friend of shopper wants to be seen as "friend of shopper" by her actions - chatting to her friend, mutual gaze, close positioning to friend - yet not as "next shopper" - she doesn't put her bags down, she positions herself beside her friend and not in a line with other shoppers. Through the whole transaction the shopper's friend positions herself slightly further away from the counter than the shopper, but close enough to be seen as 'with' the shopper. Indeed, later in the clip (frame 6 and frame 13) when another shopper positions themselves 'next in line', she visibly shuffles behind her friend to show she is not in the queue. She is "shoppers friend", not "next shopper in the queue", and the new shopper arranges their position accordingly. The shoppers gaze (frame 6) is also directed towards the shop assistant, and although the friends at times shares this gaze she also browses by looking below the horizontal at the food which is on display.

5.13 In frames 9 to 11 these behaviours can be compared to that of a shopper who is 'browsing' and can be seen as 'browsing' by their behaviour. This shopper walks along the counter, then walks closer to the counter to stare at some goods, then walks backwards slightly, along the counter again, and then back in again to look at some goods. The shopper shows herself to be busy considering a purchase - by coming in close to the counter she can see the goods, yet when doing so she does not look up to be served, but only down at the counter. The browser needs to avoid the potential embarrassment of being asked for service, and by backing off she can show herself to be "not making a purchase", since she only stays near the counter for a few seconds.

Figure 1. Frames 1-4
Click here for video - You will need to have downloaded QuickTime at to play the video

Figure 2. Frames 5-8
Click here for video

Figure 3. Frames 9-11
Click here for video

Figure 4. Frames 12-15
Click here for video

5.14 In the last clip here (frames 12-15) the attempts of a third shopper to get served can be seen. Here we notice that there is actually a ticket machine which orders who is next in line. However, although the first customer in frames 1-4 does not grab a ticket, this shopper does. As mentioned above, if there is no one else in the queue then getting a ticket would be unnecessary. Managing getting served involves checking when to use a ticket and when not.

5.15 This customer positions herself directly behind the friend of shopper, who again moves out of the way, so as to show herself not to be 'next-in-line'. However, once the customer has been served the shop keeper is distracted with a telephone call. The shopper then has to stand waiting looking above the shop counter so as to catch the shopkeepers eye should he move to serve someone. This shopper then has a particular problem- how to constitute a queue when she is the only person in the queue. To do this she needs to visibly be waiting for service, so that anyone who comes to be served can see that she is waiting and thus before them in the queue. Although the shopper is 'doing nothing' it can be seen that she is standing about waiting to get served; she is not browsing, answering her phone, or passing through - she is 'waiting'. One would not mistake the shopper who is waiting with those passing by, even though at any moment in time they might be in similar physical positions. As Sacks (Sacks, 1984) observes, doing nothing is doing something - not just standing in place but we must be seen to be actively doing nothing, waiting with a purpose, rather than just standing in the market.

5.16 These observations on queuing may seem almost straightforward and ordinary: yet consider if it was difficult to tell the difference between those shopping or browsing, or a individuals waiting for a friend or waiting in the queue. Confusion would be rife - the simple actions of getting served would present the sort of challenge that shopping in a different culture can present to a visiting tourist. Moreover, this sort of everyday disambiguating between what people are doing goes on with no or little verbal communication between the parties involved. Unless an error occurs, people pass through this setting day-by-day with seldom a shared word between strangers. Even the artefacts used in this setting are often ignored - although the ticket machine does have a role in queuing, much of the queuing in the market took place without resort to this artefact. Indeed, the artefact itself is yet another aspect of the queue that is managed without resort to spoken communication.

5.17 Customers manage their behaviour in such a way that it is visible to others at a glance. This is what Garfinkel calls the production and recognition of activity (Garfinkel, 2002). We can recognise activities yet we also produce these actions in a recognisable way. As Livingston puts it, discussing an interaction between a father and son:

Both father and son were watching each other to see how the other was doing (producing) the thing that he was doing. Even while responding by witnessably producing a next sequence of actions, both were watching the other to see what effect his responding production had on the actions that the other would do in response to it. Not only were they both using the local methods of producing an accountable social event to produce it, both were examining those witnessable methods as means of finding what to do next. (Livingston, 1987, p8)

5.18 The queuing studied here consists in a large part of arranging ones activities so that what one is doing is visible to others. That visibility needed to be visibility at a glance without resort to further enquiry. This allows the queue to be managed straightforwardly - one can see who is 'next in queue', or who is 'last in queue' and use this to arrange your behaviour. So when we are in a queue we make a limited commitment to others -our intentions are visible to others to prevent confusion. Queuing then, is as much about being able to show ones' actions as it is being able to see the actions of others.

The Role of Artefacts

6.1 It is hardly surprising that physical objects in these settings have a prominent role - much of the job of service is passing objects (foodstuffs, tourists gifts) onto customers. Yet as well as this basic role physical objects have an importance in how they support the interaction proper. Objects, their placement and configuration, are used as resources for managing the trajectory of service interaction, placing constraints on what is easily possible for both staff and customer.

6.2 Perhaps the most important, and prevalent, object in service interaction is the counter. In the TIC the counter took the form of a wooden desk which split the shop into a front and a back area. In the front area there were displays with information for the public, a shop to buy tourist trinkets, and a partitioned area to queue for service. The counter ran the width of the centre and allowed staff to serve tourists and help them with their enquiries (see figures 5 and 6). Behind the desk there was a wall, behind which there was roughly the same amount of space as out front, where staff stored literature, 'worked' the phones (figure 7), answered emails and dealt with accommodation requests. In this way the TIC was divided into three: the shop, behind the counter and the back office.

Figures 5 and 6. The front area and the desk

Figure 7. The back office

6.3 In the Swedish covered market, there was a similar divide between the customers and the staff, with counters along the length of each stall (figure 8), acting as a physical divide between the shop and the common area of the market. Each market stall had a limited area for storage and preparation at the back of each stall, or between stalls.

Figure 8. The counter and a butcher's stall

6.4 This divide between 'front' and 'back' region is a ubiquitous feature of many settings, as most memorably discussed by Goffman (1959), and more recently (Hindmarsh and Pilnick, 2002). In both these settings the counter acted to create a 'stage' behind the counter - a region that was not open to customers in normal practice, yet one that was still on display to customers. In this way the counter took on a dividing role between the semi-public space of the shop, and the semi-private space behind the counter. There was nothing unique or special about the counters in these settings that allowed them to act as symbolic markers. Rather, 'everybody knows' how counters divide space, seeing them used as we have many, many times. Yet their sheer physical size and design acts as a constant reminder, and bodily organiser of the relationship[3].

6.5 The counter also acts as a resource for managing the interaction around the objects being purchased. When objects were given to customers (such as maps, or foodstuffs) they would be passed over the counter, passing between shop and customer. Putting objects onto the counter showed an in-between status, goods would collect there until the transaction was completed (and the customer paid). We observed customers in the market 'collecting' goods on the counter, waiting until they had paid until moving them into their bags. The possibility of theft, or of purchasing the wrong goods are ordinary concerns addressed by this arrangement. Goods are held in an in-between state before purchase - under consideration for purchase. The counter then is not merely a divider of space but an resource in the interaction. The interaction around objects can be managed by their placement on the counter, as opposed to on the shelves behind or in the shopper's bag.

6.6 In the TIC the counter had a similar role as a shared space where documents could be distributed and shared between staff and customer. Much of the interaction in the TIC involved discussion of different aspects of Scotland, using documents (such as maps) to explain and discuss different options. The counter provided a common surface, visible to both customer and staff, onto which documents could be placed and discussed. However, one unusual feature of the counter concerned the way in which documents placed on it would be upside down for one of the parties. To overcome this staff had become expert at reading maps and documents upside down. Indeed, one member of staff confessed that she could not read a map the right way up. Customers would mistakenly often turn the map around for her benefit actually making the task harder for her.

6.7 Passing objects over the counter also has an important role in the TIC. Much of the information which the TIC stored (for example, timetables) was printed on paper and duplicated multiple times so that it could be handed out to customers. This ability of paper - that it could be a "take away" - made it fit particularly well in the job of the TIC. Rather than needing to print out or write down some information, paper could simply be handed over to the customer. In Harper's work from the IMF (Harper, 1998) he talks about how handing over paper has a symbolic significance in that 'handing someone a report' symbolically stood for handing responsibility for a report onto someone else. In the TIC, the 'take away' had a similar symbolic significance, "giving" a piece of paper provided a tangible symbol of the help given, and passing the paper to a customer could be used to initiate the end of a transaction.

6.8 The paper pad which was used in the TIC (figure 9) was important for similar reasons. Information could be gathered together and written down on the pad and then given to the visitor. This meant that customers did not need to remember long lists of times or details, they could all be written down and simply handed to the visitor. As one member of staff commented to me, the pad was more important than the computer in serving customers. It was a simple mechanism which allowed information to be transferred between staff and visitor in a semi-permanent form. Nearly every exchange observed in the TIC where there was a need to impart some complex piece of information involved writing onto a page of the pad and handing a page over. The 'pad' acted in a similar way to the bag in the market, it supported the collection of information into one place and it's transportation by the customer away from the shop.

Figure 9. The paper pad

6.9 It is hardly surprising that the objects involved in these settings have a role in dividing up the space. Yet what is key is the objects which are used in the exchange (such as counters, and paper pads) also have a role in the interaction proper. So the counter is not only a space for putting objects, it has an interactional function in how passing object over the counter can indicate that a purchase is complete. Similarly, for the staff in the TIC passing a map over to the customer as a 'take away' could be used to initiate the closing of the encounter. Giving the page to a customer is not only a way of simply giving information in a portable form, the paper works as a tangible representative of the help given in the exchange.

6.10 Considering the constrained nature of this interaction and the environments they are in, it is perhaps not surprising that the interaction makes use of objects which have another purpose. In a otherwise sparse environment these objects have come to be essential resources. Although the objects in both these settings are not designed for these purpose they have been appropriated to help manage the interaction by those involved. Objects are not simply distributed to customers in these settings but their position and movement is seen, recognised and understood as interactional events which are used to control and manage the interaction.

The Service Interaction

7.1 As a temporally bound activity, the service interaction itself has a number of special features. Again, staff needed to balance the need not to rush particular customers, yet still keep the queue moving. This created a "push" on the dynamics of the customer interaction. In the TIC in particular, customers could not be relied upon to design their questions effectively, or to give any idea from their first question how long a particular enquiry could take. Since the length of the interaction was essentially managed by the customer, speed could therefore not be easily controlled. Yet there were a number of techniques which were used by staff to 'work the interaction', to both provide a good service without spending undue time helping a particular visitor:

Designing Responses

7.2 While there were many aspects of serving customers in the TIC, such as giving information, presenting the city or selling services, a key aim of the interaction was to work through what was likely to assist the customers as visitors. This focus on helping entailed that a reply would never be a simply negative; "I don't know" was not a valid response, although it was often true. Staff would work out something - anything - which would appear to assist the customer, such as giving a phone number of someone else who might know, or offering to contact them later. This 'helping' impacted on the arrangement of the interaction between staff and customer. Staff would often not directly answer questions asked by customers but would instead answer the question in a less direct but possibly more helpful manner. For example:

Older man who is with a young woman comes to the desk

Woman: My friend is going to Belfast - have you got a schedule of when they're going

Staff: There's two places boats go from, there's Troon or Stranrar which is further south

Man: I'm going from Stanrar- I've got my ticket

Staff explains how to get the bus to Stranrar, helps man with map, draws note on photocopied map with highlighter, gives him directions to the bus station and points out where to go from the door using hand and on the map

7.3 Notice that in the opening question there is no explicit mention of the mode of transport the schedule is being asked for. There are only two ways to Belfast, however, by plane and by boat. Since planes are generally booked in advance (and as such the TIC rarely answered questions on plane times), the member of staff guesses that they mean boat. Rather than answering the question simplistically the staff member points out that there's two places to go to Belfast from, also mentioning that he is assuming they are travelling by boat. The interaction continues with the staff member explaining how to get the bus to Stranrar, and giving directions to the bus station although these were not explicitly asked for. As Sacks puts it, utterances are 'designed' (p7 vol II, Sacks, 1995) to take into account the circumstances of the recipient. How we phrase what we say, and what we say, are designed in such a way as to be understandable from the perspective of the listener. In this case this general feature of conversation takes on a specific role. The staff's responses are designed to be helpful, rather than simply giving an answer to the question.

7.4 A second aspect of the customer interaction concerned not strictly helping customers, but balancing the needs of the TIC funders. Since the TIC was funded by different members of the Glasgow tourism industry, such as hotels and restaurants, there was a need for the advice given to not unfairly favour one member over another. Since many of the tourist board members were in competition with each other this could have caused problems for staff. Staff had to give advice to customers which would be seen to be fair and avoiding any charges of personal bias. There were two simple conversational devices which staff used to allow their advice to appear 'reasonable and fair': respecifying the question, and three part recommendations.

7.5 When asked to give information on amenities in the city, such as restaurants, staff would answer with a question which would put the onus on the visitor to further specify what they were looking for. So a request to recommend a restaurant would be met by another question asking what particular type of food they were looking for, where in the city, at what cost, and so on. These questions would restrict the list of potential establishments to a manageable level from which staff could recommend. In giving recommendations staff struggled with the possibility of a 'personal' rather than a 'professional' view. Any query about why a particular establishment had been recommended could then be justified by referring to the type of establishment requested: "she asked for an Japanese restaurant in town which wasn't too expensive".

7.6 When a visitor was not able to specify more exactly what they were looking for, staff would use a second device - the three part list (Jefferson, 1990). By listing three establishments staff gave the visitor a reasonable choice, not seeming to favour any particular establishment, but at the same time not overwhelming the visitor. For some types of establishment - such as accommodation - staff would have a carefully prepared set of three recommendations. So, for example, when asked to recommend a youth hostel in Glasgow staff would describe three - nearly always the euro hostel, the YHA and an independent hostel in the west end. These three covered the variety of the most popular hostels in Glasgow; a large city centre hostel, the official 'YHA' hostel, and an independent hostel in the west end of the city, which was also close to other hostels. This list was visibly fair, both to a visitor and to anyone who might dispute the advice given. Indeed, staff found giving specific recommendations problematic. In one case a staff member was asked (in a telephone enquiry) whether one hotel was better than another. The staff member working the phones resorted to asking her colleagues to see what they thought. Eventually she replied to the person on the phone that one was more "traditional" and the other was more "modern and young", and explained that since they were all fairly young, they would themselves, personally go to the more modern hotel[4]. However, this reply was given with much hesitation and hedging.

7.7 Along with issues covering the nature of the advice given, staff also endeavoured to keep interactions with customers 'manageably' quick. This was out of a concern to not keep a visitor waiting, and also to 'work the queue' and keep it moving. One method used was overlapping different activities. For example, accommodation bookings could take considerably time since there was a need to contact a hotel, negotiate suitable accommodation, fill in some paperwork, take payment and print a receipt. To speed this transaction staff would arrange their activates so as to have two actions in progress at once. They would fill in a form while discussing with the customer what sort of accommodation they were looking for (something which was also entered onto the form at the appropriate point). Staff could also overlap activities when they needed to get a customers home address. They would hand over a pad of paper and ask the customer to write down their address while they then phoned the hotel. Not only did this speed the transaction but it meant the customer was not waiting - they were busy themselves writing down their address while the staff member spoke to the hotel. Another overlapping was discussing with the hotel the nature of the customers booking, while filling in the details on the computer. Again, this was a way of interweaving different actions so as to speed the exchange and not keep the customer waiting.

7.8 Staff made a point of "working aloud" so as to make their actions visible to each other and customers. This is similar to the "filling talk" referred to by Martin et al in that staff would describe what they were doing as they did it, so as to be accountable to customers. This talk served a number of roles. Firstly, it gave the customer a sense of what the staff member was doing. That is, that their query was being actively worked upon, rather than then being 'kept waiting' while staff attended to other jobs. Working aloud also allowed for some error correction, if staff had misunderstood a customer request, this could be corrected by the customer at that point in the interaction. Some staff would often talk through a form as they filled it in; mistakes could be corrected as they filled in the form. If, while doing this, staff also needed some extra information or a decision from a customer, then the customer would understand what they meant from the context of the enquiry which had built up through their narration. In this way interaction allowed for customers to see what staff members were doing, providing for correction by the customer should a mistake be made.

7.9 Accordingly, the actions which the staff took (such as looking up a telephone number) also had an interactional import with the customer, in that the customer was aware of how their inquiry was proceeding. Staff would explain if they had to leave the desk to get some information, and would explain if it took longer than expected (such as that they had to go downstairs to the stock room, or that the photocopier had broken). However, these interactions are more than just "filling the blanks" as Randall puts it, they are a way of including the customer in the work which the member of staff is doing. This is not just simply out of politeness but has a practical role in involving the customer in the enquiry, giving customers an understanding of its development, an understanding which may be drawn on later in the interaction.

Being an effective customer

7.10 Moving back to the covered market, the data from here also lets us look at some aspects of the service interaction. In particular, we can see something more of the work which customers themselves do to manage the shop transaction. One practice of shop conduct which the videos show is that activities are arranged so as to not needlessly waste shop assistants' time. In the case of the market, shoppers appeared to have a general idea of what they wanted to buy before they were actually served, minimising the time spent being served, Certainly, if there is a queue behind a shopper then "the push of the queue", and other shoppers annoyance contributed a subtle social pressure on those who take too long to complete their transaction. In this way "the push" is felt both by the customer, as well as the shop assistant. Exceptions to this behaviour are noticed and marked as such[5].

7.11 A second implication of this rule concerns payment and ending transactions; shoppers would have their money easily accessible so that they could pay when they had received the goods (figure 9). In some cases this involved taking a purse out from a bag while the transaction was ongoing so that the money could be quickly produced when asked for. Some shoppers seemed to time this perfectly; they would take their money out just as the shopkeeper completed preparing the goods and handed them to the shopper. This meant that the shopper did not have to risk having the money out (or their purse open) for too long a period of time. The goods are prepared, the money is produced, and the goods are swapped over, in perfect timing.

Figure 10. Customers with their money ready before the end of a transaction, purses in hand.

7.12 Indeed many shops had shelves that ran along the length of the shop where a handbag could be placed. This positioned the handbag safely between the shopper and the counter (to prevent theft), as well as allowing quick access to a purse. Shoppers also waited until they were away from the counter before they would spend time arranging their bags. Shoppers complete their transactions then sort their collection of bags later, so as another shopper could be served. Again, these small activities might seem trivial but together they make purchasing and customer interaction a smooth operation - something which benefits both shops and customers, in that it minimises the time they are kept waiting.

7.13 Shoppers managed their conversations with shop assistants in a similar fashion. In one incident observed, a shopper was chatting to a member of staff after their transaction has finished. As a second customer comes up to be served, the conversation comes to a close, almost exactly at the point where the new customer gets to the counter and looks up to be served. In these ways both the customers and staff would manage their conversations so as to take into account the needs of other customers. In situations where there are no other customers waiting, staff and customer can take their time, take time over what to buy, chat, rearrange bags at the counter, or simply take longer to complete the transaction. Yet with other customers waiting one manages the transaction to be timely and efficient.

7.14 These behaviours show something of the sequential nature of these interaction. For customers and staff are orientated, in their interactions, to a sense of what is going to happen next. Although staff and customers can never be sure what is going to happen next there are certain likely events and routines. These temporal structures and routines are used to organise the activities as they unfold. So, rather simply, as a customer asks for goods it is likely at a certain point they will need to pay, so the bag is placed so that access to a purse is possible. When they have selected their last purchase the customer can lift their purse and start to pull money from the purse. More generally: this is an prospective orientation to the likely temporal organisation of the interaction. In a small way customers predict what is going to happen next and use that to organise what they do right now - they hold a prospective orientation to action.

7.15 The customer service in the TIC also shows something of this prospective orientation. Customers deciding whether to wait for service would use the length of the queue as a simple indicator of how long one might have to wait (although the number of staff serving was probably a better indicator). In the service interaction itself the TIC staff would make assumptions about extra information to provide to a tourist- if a tourist asked if an attraction was open they would pass on information about how to get to that attraction. Yet in the TIC the relatively unpredictable nature of inquiries meant that these 'predictions' were more ad hoc. Certainly, the interactions had less of the straightforward queue-purchase-pay nature of the market. This made managing the unfolding interactions less a case of orientating to a predicted sequence as opportunistically arranging information when and where it was likely to be needed. Yet even here, for some actions (such as booking a hotel) they could choreography the sequence of action with an expectation for the structure of ongoing activity.


8.1 The reader has hopefully been able to recognise their own shopping behaviour in the descriptions given. These descriptions show something of the social nature of shopping and customer service, of the work that both customer and staff need to do if this environment is to work.

8.2 Both the institutions studied here were embedded in wider economic markets and processes. These wider markets have of course been the subject of extensive enquiry. Yet at the level of customer interaction this paper shows the possibility of studying both the ordinary morality, and rule governed nature, of market in how shoppers arrange their activities they enter a moral commitment to other customers and shop staff. By this we mean that the infringement of those commitments is morally sanctionable, it is seen as 'obviously bad'. One of the strongest examples of this comes from queuing. Queue jumping, squeezing into a queue in-front of people who have been waiting, is a strongly morally sanctionable behaviour. If we are caught doing it deliberately then we are likely to be treated to considerable anger and disdain. As Livingstone comments: "we sense true evil when a person calmly walks up in front of our place in line and, without any apparent trace of emotion on her face, looks us in the eye, inviting and challenging us to make a point of her aggression. A disturbing, but order-productive phenomenon" (Livingston, 1987, p15).

8.3 Queues exist then not only as structures in our activity, and in how we shop, but as moral structures. There are practical benefit from following these commitments and commitments as well as their enforcement by sanctions and stigma. If we break these commitments we can be 'held to account' by others. This is not to make a judgement that shopping is either moral or immoral, or even the economic system overall. Rather, there is an ordinary morality at work, helping the management of interactions.

8.4 This morality is not an analysts judgement, but than of shoppers and staff themselves. This suggests the study of morality as an activity of individuals themselves, of their own judgements and perspectives (such as in Lynch's work on the Iran-contra case (Lynch et al., 1996)). While this might not be all we want to say about morality, studying this 'moral order' in markets seems promising, in that one could explore how economic morality is used by economic actors themselves, rather that as a judgement of sociologists. Indeed, Knorr Certina (Knorr-Certina and Bruegger, 2003) and Mackenzie's (Mackenzie, 2003) recent research on the interactional workings of economic markets move in this direction[6].

8.5 Goodwin introduced the notion of 'professional vision' in studying the work settings of professionals, such as scientists. The term professional vision refers to how professionals socially organise their practice so as to be able to 'see' objects at work as the proper elements of their practice (Goodwin, 1994; Goodwin, 1995; Goodwin, 2000). So, for example, in his study of archaeologists he showed how archaeologists can see colour differences in a sample of soil as suggestive of a particular historical artefact. Professional vision is the skill to see particular signs, objects, or readings as relevant features. While an uneducated observer may only see numbers on a computer screen, a professional oceanographer may see these as a 'nice' feature of a particular flow of seawater (Goodwin, 1995).

8.6 While professional vision is part of a professional community of practice, the competences described here are part of an ordinary vision - they exist as part of the background expectations of a competent member of a western economy. Customers need to see queues and staff need to see customers for shopping to work. This depends upon a long training from childhood in the ways of shopping and shop environments. We learn to recognize in shops when customers are queuing, waiting, being served and so on, the mundane constitutive features of these environments. It is only when we travel to other countries and are confronted with unfamiliar service environments that our skills of seeing are foregrounded. When in more familiar territory these skills of recognition are taken for granted and unremarkable.

8.7 As with professional vision, this ordinary vision does not exist as a lone cognitive event. This ability to see features of customer interaction and understand their relevance also depends upon the very production of those features by ourselves as shoppers. We need to display in our behaviour our actions. When we queue we need to queue in such a way that what we are doing is visible as queuing, and not waiting for a friend, or whatever. The ordinary vision in these environments is thus produced as well as recognised. The "astonishing collective achievement of predictability" in service settings (Garfinkel, 2002), comes from how we present our interactions such that our behaviour is predictable and understandable by others. We need to act as shoppers for our actions to be easily seen and understandable at a glance by others.

8.8 The 'ordinary vision' of shoppers thus depends on both the production and recognition of behaviour described here. It is these very ordinary competences that make the practice of customer service such a smooth and unremarkable event.


The author would like to thank Oskar Juhlin and Lina Larsson for their insights and help with the fieldwork in Sweden. Also thanks to Jon Hindmarsh, for many insightful comments on the paper. We thank the staff and customers of the TIC for being generous with their time.


1 The work of commercial shop designers who have studied shopping shows a clear correspondence with the observations of this work, suggesting its value beyond the academic (Underhill, 1987). While it is not a theme of this paper, shop designers often disrupt the very interactions they seek to support, and new technologies such as self-service checkouts have had only limited success (Austen, 2003).

2 Ethnomethodology differs from conversation (or interaction) analysis in that it takes an avowed 'radical' stance on the practices of research, an approach which includes hybridisation of ethnomethodology with the subjects of study (e.g. (Livingstone, 1986)). Conversation analysis has focused much more on interaction as its topic, building a substantial corpus of studies.

3 Indeed, some financial organisations have moved from using counters to using desk for some of their interactions, exploiting the different physical configuration of the surface to change the nature of the interaction (Harper et al., 2000).

4 Indeed, in an earlier interview with a youth hostel manager I had been puzzled by her explanation that she never gave personal recommendations since "if you have a bad meal in a restaurant that might just be to you, you don't know what it will be like to anyone else". It seems that this is an awareness of the variety of tastes which exist and the problems of defending a "personal view" (Lynch and Bogen, 1996).

5 Cringely comments on how Bill Gates apparently kept a supermarket queue waiting while looking for a discount coupon. A waiting customer gave Gates a dollar when he couldn't find the coupon and asked him to "pay him back when he made his first million" (Cringely, 1996).

6 And in turn it is interesting to compare this to work in socio-economics (Sciulli and Etzioni, 1995)


ATKINSON, J. M. and J. Heritage, Eds. (1984): Structures of Social Action: Studies in Conversation Analysis, Cambridge University Press.

AUSTEN, I. (2003): In Germany, Customers Scan As They Shop. New York Times: C7.

BAUDRILLARD, J. (1988): 'Consumer Society', in M. Poster (Ed.) Jean Baudrillard Selected Writings. Cambridge, Polity Press, pp. 26-43.

BEACH, W. and C. LeBaron (2002): 'Body disclosures: Attending to personal problems and reported sexual abuse during a medical encounter', Journal of Communication, Vol. 52, no. 3.

BELK, R. (1995): Collecting in a consumer society, Routledge.

BOURDIEU, P. (1984): Distinction: A social critique of the judgement of taste, Routledge, London.

BOWERS, J. and D. Martin (2000): 'Machinery in the new factories: interaction and technology in a bank's telephone call centre', in Proceeding of CSCW 2000, Philadelphia, Pennsylvania. New York, USA., ACM Press, pp. 49-58.

BROWN, B. and M. Chalmers (2003): 'Tourism and mobile technology', in K. Kuuttiet al (Eds.): Proceedings of ECSCW 2003, Helsinki, Finland. Dordrecht, Klewer Academic Press, pp. 335-355.

BUTLER, R. (1991): 'West Edmonton mall as a tourist attraction', Canadian Geographer, Vol. 35, pp. 287-295.

BUTTON, G. (1993): Technology in working order, Routledge, London.

CHANEY, D. (1990): 'Dystopia in Gateshead: the metrocenter as cultural form', Theory, culture and society, Vol. 7, pp. 49-68.

CLARK, C. and T. Pinch (1995): The Hard Sell: The Language and Lessons of Street-Wise Marketing, Harper Collins, London.

CREWE, L. and N. Gregson (1998): 'Tales of the unexpected: exploring car boot sales as marginal spaces of contemporary consumption', Transactions of the insititute of British geographers, Vol. 23, pp. 39-53.

CRINGELY, R. X. (1996): Accidental Empires: How the Boys of Silicon Valley Make Their Millions, Battle Foreign Competition and Still Can't Get a Date, Penguin, London.

DAVIS, F. (1950): 'The Cabdriver and His Fare: Facets of a Fleeting Relationship', American Journal of Sociology, Vol. 45, no. 158-165.

DURRANDE-MOREAU, A. (1999): 'Waiting for service: ten years of empirical research', International Journal of service management, Vol. 10, no. 2, pp. 171-189.

FALK, P. and C. Campbell, Eds. (1997): The Shopping Experience, Sage.

GARFINKEL, H. (1967): Studies in Ethnomethodology, Prentice Hall.

GARFINKEL, H. (2002): Ethnomethodology's Program, Rowman & Littlefield.

GARFINKEL, H. and E. Livingston (2003): 'Phenomenal field properties of the order of service in formated queues and their neglected standing in the current situation of inquiry', Visual Studies, 'Image work' special editon edited by Mike Ball, Vol. 18, no. 1, pp. 21-28.

GARFINKEL, H. and H. Sacks (1970): 'On formal structures of practical actions', in J. McKinney and E. Tiryakian (Eds.): Theoretical Sociology. New York, Appleton-Century-Crofts.

GOFFMAN, E. (1959): The Presentation of Self in Everyday Life, Penguin Books, Middlesex, England.

GOFFMAN, E. (1963): Behaviour in public places: notes on the social organization of gatherings, Free Press.

GOODE, D. (2002): Playing with Katie : An account of human-dog play utilizing autoethnography and videography, CUNY Press, New York.

GOODWIN, Charles (1994): Professional Vision. American Anthropologist, Vol. 96, no. 3, pp. 606-633.

GOODWIN, C. (1995): 'Seeing in depth', Social Studies of Science, Vol. 25, pp. 237-74.

GOODWIN, C. (2000): 'Practices of Seeing: Visual Analysis - An Ethnomethodological Approach', in T. v. Leeuwen and C. Jewitt (Eds.): Handbook of Visual Analysis. London, Sage, pp. 157-182.

GOODWIN, M. H. and C. Goodwin (2000): 'Emotion within Situated Activity', in A. Duranti and M. Malden (Eds.): Linguistic Anthropology: A Reader. Oxford, Blackwell, pp. 239-257.

HARPER, R. (1998): Inside the IMF, Academic Press, London.

HARPER, R., D. Randall and M. Rouncefield (2000): Organizational Change in Retail Finance : An Ethnographic Perspective, Routledge, London.

HEATH, C. (1997): 'Analysing work activities in face to face interaction using video', in D. Silverman (Ed.) Qualitative Methods. London, Sage.

HEATH, C., J. Hindmarsh and P. Luff (1999): 'Interaction in Isolation: The Dislocated World of the London Underground Train Driver', Sociology, Vol. 33, no. 3, pp. 555-575.

HEATH, C. and P. Luff (2000): Technology in action, Cambridge university press, Cambridge.

HINDMARSH, J. and A. Pilnick (2002): 'The Tacit Order of Teamwork: Collaboration and Embodied Conduct in Anaesthesia', The Sociological Quarterly, Vol. 43, no. 2.

HORNIK, J. (1984): 'Subjective vs. objective time measures: a note on the perception of time in consumer behaviour', Journal of Consumer Research, Vol. 11, pp. 615-18.

JEFFERSON, G. (1990): 'List construction as a task and resource', in G. Psathas (Ed.) Interaction competence. Lanham, MD, University Press of America.

KNORR-CERTINA, K. and U. Bruegger (2003): 'Global microstructures: the interaction practices of financial markets', American journal of sociology, Vol. 107, no. 4, pp. 905-950.

LANCASTER, W. (1995): The department store: a social history, Pinter, London.

LEE, J. R. E. and D. R. Watson (1993): Final Report to the Plan Urbain: Public Space as an Interactional Order. Manchester, Department of Sociology, University of Manchester.

LIVINGSTON, E. (1987): Making Sense of Ethnomethodology, Routledge and Kegan Paul, London.

LIVINGSTONE, E. (1986): The Ethnomethodological Foundations of Mathematics, Routledge & Kegan Paul, London.

LYNCH, M. and D. Bogen (1996): The spectacle of history: Speech, Text and Memory of the Iran-Contra Hearings, Duke University Press, Durham, NC.

MACKENZIE, D. (2003): ' An Equation and its Worlds: Bricolage, Exemplars, Disunity and Performativity in Financial Economics', Social Studies of Science, Vol. Forthcoming.

MAISTER, D. H. (1985): 'The psychology of waiting lines', in Czepiel (Ed.) The Service Encounter. Lexington, MA, Lexington Books, pp. 113-23.

MCHUGH, P., S. Raffel, D. C. Foss and A. F. Blum (1974): 'Tourism', in On the beginning of sociological inquiry. London, UK, Routledge.

MILLER, D., Ed. (1995): Acknowledging Consumption : A Review of New Studies, Routledge.

MILLER, D. (1998): A theory of shopping, Polity, Cambridge.

RANDALL, D. and J. A. Hughes (1995): 'Sociology, CSCW and Working with Customers', in P. Thomas (Ed.) The social and interactional dimensions of human-computer interaction, Cambridge University Press, pp. 142-160.

RYAVE, A. L. and J. N. Schenkein (1974): 'Notes on the art of walking', in R. Turner (Ed.) Ethnomethodology. Harmondsworth, Penguin, pp. 265-74.

SACKS, H. (1972): 'Notes on police assessment of moral character', in D. Sudnow (Ed.) Studies in social interaction. New York, Free Press, pp. 280-93.

SACKS, H. (1984): 'On doing 'being ordinary'', in J. M. Atkinson and J. Heritage (Eds.): Structures of social action. New York, Cambridge university press, pp. 413-429.

SACKS, H. (1995): Lectures on conversation: vol 1 & 2, Basil Blackwell, Oxford.

SCHWARTZ, B. (1975): Queuing and waiting, University of Chicago Press, Chicago.

SCIULLI, D. and A. Etzioni, Eds. (1995): Macro Socio-Economics: From Theory to Activism, Sharpe.

SILVERMAN, D. (1998): Harvey Sacks: social science and conversation analysis, Polity Press, Cambridge.

Suchman, L. (1987): Plans and situated actions: The problem of human-machine communication, Cambridge University Press, Cambridge.

SUDNOW, D. (1972): 'Temporal parameters of interpersonal observation'', in D. Sudnow (Ed.) Studies in social interaction. New York, Free Press, pp. 259-79.

UNDERHILL, P. (1987): Why we buy: the science of shopping, Texere, London.

WILLIS, P. (1990): Common culture: symbolic work at play in the everyday cultures of the young, Open University Press.