Abstract
This paper aims to describe marketization processes in terms of ‘market agencing’. The agencing framework is presented through the study of the Progressive Grocer, a trade magazine that presented new ideas about the grocery business and proposed novel ‘agencements’ for American grocery stores in the early 1920s. The case shows that agencing consists of combining the agency of grocers and market devices to shape a different retail environment; that is, a new ‘agencement’ as a situated and flexible combination of market equipment and managerial logics. The paper shows how the introduction of ‘open display’ – that is, providing a better visual access to the goods while preserving service – introduced new furniture and ideas, and thus eased the transition from counter service to self-service. The first section presents the agencing framework, the empirical source, The Progressive Grocer, and the method used to analyse it. This method is labelled ‘pico-geography’. The idea is to conduct analysis on an even smaller scale than micro-geography by focusing on spatial reconfigurations that occur at the indoor and store level. The second section presents the concept of the open display and describes how it contributed to ‘re-agence’ the grocery equipment, workforce and even consumers. The third section puts this journey into perspective, revealing that, though promoted by The Progressive Grocer, it was also part of a larger reconfiguration involving several other actors. The conclusion stresses the empirical, methodological and theoretical contribution of the paper.
Over the last 20 years, economic geographers have devoted a considerable amount of attention to the performativity programme launched by actor–network theorists at the end of the 1990s. According to this programme, instead of looking at economics as a descriptive discipline and instead of criticizing the questionable realism of its descriptions of the economy, social scientists should better consider economics as a performative knowledge, and thus study how it shapes the world it pretends to describe (Callon, 1998; MacKenzie and Millo, 2003). For instance, economic geographers took the performativity framework as a way to (a) complement the Polanyian large view of markets with a consideration for their internal functioning (Mullerleile, 2013); (b) explore the various spatial enactments of economic notions and theories (Barnes, 2008; Christophers, 2014a, 2017; Gibson-Graham, 2008; Peck, 2012); (c) describe how economic ideas circulate in the economy (Barnes, 2002); (d) bridge the performativity framework and political economy (Christophers, 2014b); or even (e) redefine the political commitment of geography itself, through its active participation in local and diverse experiments, far removed from the classic and remote critical perspective (Gibson-Graham, 2008).
Because the performativity programme is now very well-known and well developed in geography, I would like to suggest moving a step forward, by exploring the potential of one of its most recent outcomes: the notions of ‘market agencement’ (Callon, 2016, 2017) and ‘market agencing’ (Cochoy et al., 2016). As we will see, by focusing on agents and agency, these notions help studying how performativity really works; that is, how various human concepts and material objects are mobilized and articulated (i.e., ‘agencés’, in French) to produce transformative states of the world (i.e., ‘agencements’). The idea meets geographers’ efforts aimed at studying the various concrete and spatial forms of economic facts, rather than market or capitalism as abstract ungrounded notions.
I propose to study how agencing processes unfold through the study of the re-agencing of American grocery stores in the 1920s and early 1930s. I will focus on a crucial yet overlooked turning point: the development of the so-called ‘open display’ approach to grocery retailing. Open display supported the transition from counter service to self-service, reshaping the layout of grocery stores to give consumers better access to goods while simultaneously preserving service. As we will see, this approach was strongly promoted from 1927 onward by The Progressive Grocer, a trade magazine that supported independent grocery stores in an era of fierce competition with chain stores and, later, supermarkets. However, more than an idea uttered by the magazine and performed by its readers, open display was rather promoted through a continuous and subtle reconfiguration – re-agencing – of various market ideas, devices and people. As such, the re-agencing of the grocery store shows why it is useful to move beyond the initial framework of performativity.
In the first section of this paper, I provide a brief overview of the theoretical framework of ‘agencing’ processes as efforts made to shape new market ‘agencements’, or webs of material and human resources (Callon, 2016). In the same section, I present in greater detail my empirical source, The Progressive Grocer, and the method used to analyse it. I label this method ‘pico-geography’: by moving at the ‘pico’ scale, below the micro one, I wish to focus on tiny spatial transformations at the store and indoor level, well below the local outdoor level usually addressed by micro-geography. In the second section, I present the concept of the open display and discuss what it did to the geography of grocery stores through a subtle and profound re-agencing of all of their elements (i.e. their equipment, furniture, workforce and consumers). In the third section, I put this journey into perspective, showing that, though the open display concept was promoted by impressive performative rhetoric of The Progressive Grocer, it was also part of a larger reconfiguration involving several other agencies like canners, furniture designers and Cellophane.
Theory, field and method: The pico-geography of the re-agencing of grocery stores
The performativity framework has been extremely powerful to show how economic theories and business ideas happen to be adapted and enacted in various contexts. But it also raises two major difficulties. First, it implicitly acknowledges a sharp but embarrassing distinction between ideas and matters; second, it assumes a clear but debatable view where the former precedes the latter. It is somewhat surprising that the literature failed to observe that these two features are completely at odds with the classic approach of actor–network theory (ANT) which generally refuses one-way processes and prefers to see ideas and matters as the outcomes rather than the starting points of transformative processes. This is probably why Callon, who initiated the performativity programme, preferred talking in terms of ‘performation’ rather than performativity, and even moved away from these notions to focus on agencing, rather than performative processes.
On agencement and agencing
Recently, Michel Callon (2016) opposed two views of markets. Economists and many social scientists consider markets to be mere ‘interfaces’ designed to unite the two blocks of supply and demand. According to this view, goods and services are stable, unambiguous and generic entities. In such settings, the only function of the market system is to price and circulate goods. However, Callon proposed considering markets as ‘agencements’; that is, complex combinations of actors, goods and devices designed to frame bilateral relationships. According to this approach, goods are neither generic, transparent, nor stable, and the marketization process cannot be restricted to simply pricing and distributing mechanisms. Rather, markets are all about shaping market goods and market actors to ensure proper matching. This shaping activity addresses both prices and qualities. The notion of ‘agencement’ seeks to better capture such processes by describing all of the actors, actants and practices involved in each given market in the shaping of bilateral market relationships. After presenting the philosophical, sociological and mundane meanings of this notion, I will explain why it is particularly appropriate to look at the aforementioned change in grocery stores.
The agencement concept is an outcome of the Foucaldian–Deleuzian tradition. Foucault introduced the notion of a ‘device’ (in French, dispositif), which he defined as ‘a truly heterogeneous set, made of discourses, institutions, architectural layouts, regulative decisions, laws, administrative rules, scientific statements, philosophical, moral and philanthropic decisions. (…) The device itself is the network one establishes between these elements’ (Foucault, 2001:299, author’s translation). For Foucault, a device defined in this way is a strategic tool for solving urgent problems and imposing power relationships. Deleuze (1988) proposed complementing the Foucaldian device with the notion of agencement. For Deleuze, this term, by linking the ‘agent’ with the other parts of a device (or agencement), emphasized the close interconnections between subjects and devices, gathering together both a thing and the idea of the thing. For this reason, a Deleuzian agencement is always part of a transforming process through which subjects dissolve previous associations between them and objects, and then move from one place to another (de-territorialization), creating new relationships with the novel elements and places to which they relate (territorialization). This spatial dimension of agencements was later refined by DeLanda (2006), who defined the territory as a stabilized agencement, or ‘assemblage’ (Delanda’s translation of Deleuze’s ‘agencement’). Because of its emphasis on territorial, material and social dynamics, the concept of assemblage received significant attention from geographers (Castilhos et al., 2014; Kamalipour and Peimani, 2015; Müller, 2015).
Another reason for thinking in terms of agencement is sociological. Even if the term was introduced quite early in ANT-inspired market studies (Callon et al., 2007), it is only recently that Callon (2016) positioned it as the cornerstone of his sociology of marketization processes. The term fits superbly with the author’s previous ‘actor–network theory’ and its claim to account for the joint contributions of human beings and material entities (living creatures, objects, artefacts…) to social action (Callon, 1986). The term is superior to actor–network for two reasons. First, it links the agent (i.e. the human actor) to the concept of agency, or the capacity to act. This capacity is not specific to humans; rather, it is an attribute shared by all elements that populate the world, be they living or not. For example, a virus kills, a tool enhances our force, and even an immobile and mute wall lasts and closes space, thereby framing our action (Latour, 1988). Second, as both a noun and a gerund, the term ‘agencement’ points towards both a particular web of things (noun) and the action aimed at building it (gerund). In order to better stress this dual dimension, Cochoy et al. (2016) proposed shifting from ‘agencement’ to ‘agencing’. This latter notion emphasizes the dynamic character of agencements, which are better taken as processes than as material entities.
Finally, I would like to stress that, in French, the term ‘agencement’, far from being part of the sophisticated philosophical or sociological ‘French theory’ sometimes praised in Anglo-Saxon scholarly circles, is used on an everyday basis by laypeople as a very simple and ordinary word depicting highly mundane practices. One of the most frequent contexts in which French-speaking people speak in terms of agencement involves, of all things, kitchen design! For example, households dream of ‘agencing their kitchen’ (see ‘agencer ma cuisine’ 1 ), and kitchen designers try to fulfil these dreams by providing ‘ideas for kitchen agencement’ (see ‘idées d’agencement pour la cuisine’ 2 ). Agencing a kitchen means designing, planning, organizing, staging and setting the spatial distribution of the kitchen elements. Kitchens and grocery stores are, in this sense, very similar – they both handle food; they are complex bundles of furniture, tools and goods. In fact, the mundane use of the verb ‘agencer’ and the noun ‘agencement’ help us understand that re-agencing grocery stores does not differ from remodeling one’s kitchen. We should, thus, realize that the transformation of commerce, far from relying on remote economic mechanisms, complex political frameworks or abstract managerial notions, rested largely on mundane governance practices (Woolgar and Neyland, 2013) and on the tiny projects, things and gestures that reshaped both people and the geography of the space in which they resided (Cochoy, 2010b).
In the following pages, I will show that bringing together the three meanings I have just reviewed may be the best way to describe the re-agencing of US grocery stores in the early 20th century. Indeed, as in the Foucaldian tradition, this re-agencing involved reshaping a new type of market device comprising both discursive and material elements. The purpose of this device was not only to face the challenges raised by the competition of chain stores, but also to develop the market for new store equipment, as well as the circulation and advertising business of The Progressive Grocer. According to the Deleuzian–Delandian view, we will see that such re-agencing operations largely amounted to redesigning the territorial assemblage that linked the store space, its fixtures and the professionals and customers who interacted within its boundaries. Moreover, in keeping with the Callonian focus on market agencement as a combination of human and non-human agencies, I will pay equal attention to the social and material dimensions of agencing processes. As far as the material elements are concerned, I will unveil the contributions of various pieces of store equipment and devices to the reshaping of commercial boundaries, territories and actions. As far as the social components are involved, I will show that the reshaping of external environments also redefined consumers’ ‘inner geography’ by facilitating new social behaviours and dispositions (Cochoy, 2007). The re-agencing of grocery stores re-agenced the subjective ‘self’ behind self-service (Du Gay, 2004) as well as various kinds of attachments and emotions (McFall, 2014).
An historical, archaeological and pico-geographical approach for the study of agencing processes
I will describe the re-agencing of grocery stores as a pico-geographic process based on the collection of The Progressive Grocer, a trade magazine launched in 1922 to inform independent grocers and help them to improve their business. As such, just like Barnes’s textbooks (Barnes, 2002), the magazine works as an obligatory passage point that contributes to enact and circulate novel ideas and devices among the targeted public. I will study the magazine over the period from 1922 to 1932, focusing particularly on a campaign conducted in a series of articles launched in 1927. This campaign promoted ‘better grocery stores’ through the adoption of the so-called ‘open display’ concept. The Progressive Grocer operated as a two-sided market. On the one hand, in the 1920s, it was circulated on a free basis among a readership of approximately 50,000 professionals – a considerable public, since it represented one-third of all American grocers and nearly 80% of those earning more than US$1000 a year at that time (Cochoy, 2015). On the other hand, The Progressive Grocer sold its advertising space to all kinds of advertisers, including consumer goods companies and store equipment manufacturers. Therefore, the magazine faced a double concern: it had to please its public in order to be read, but it also had to serve its advertisers in order to be profitable. For this reason, every effort by the magazine to promote new store arrangements should be seen as a dual effort designed to both help grocers solve their problems and help equipment manufacturers sell their equipment. As we will see, the rhetoric of the magazine constitutes a strange and performative language. Instead of simply doing things with words, as proposed in the original linguistic theory of performativity (Austin, 1961), this language ‘re-agences’ the world; that is, it displaces, creates and combines things and words to set up new meanings and spatial contexts for commercial actions (Cochoy, 2010a).
In the next section, I will present and analyse The Progressive Grocer’s campaign for better grocery stores based on the magazine’s published articles, plans and pictures. The chosen method is threefold. It relies on classic historiography, on what I have previously called the ‘archaeology of present time’ (Cochoy, 2009), and on what I propose to label ‘pico-geography’. Like a classic historian, I will read the magazine’s articles and testimonies in order to account for the idea and rhetoric of open display. However, like an archaeologist, I will also base my analysis on the mute materials observable in my documents, including drawings, blueprints and pictures of new store arrangements, furniture, equipment, and so on. This focus on the visual and material aspects of what can be seen in pictures will provide insights that are often lacking in texts and testimonies. These two methods are coherent with ANT, an approach that claims to focus equally on the agency of both human and non-human entities. Last, but not least, I will combine these two approaches in order to develop a pico-geography of grocery stores. By pico-geography, I mean the study of the spatial transformations induced by the ‘open display’ logic and corresponding store plans at the store level. Pico-geography moves space analysis from outdoor to indoor, at a narrower level than classic micro-geography (Rammer et al., 2016). It thus complements the study of the macro or micro ‘performations’ of market ideas (Barnes, 2008) with a focus on their local enactment; it also extends the study of how mobile entities produce borders at the classic state level (Berndt and Boeckler, 2009) to the tiny level of store windows, packaging or sales counters. As we will see, the idea of open display focused on re-agencing store equipment, store operations and store layouts in order to move the grocery business away from old-fashioned retail outlets without abandoning the service focus.
Re-agencing grocery stores through open display
In its November 1927 issue, The Progressive Grocer published an article simply entitled ‘Better Grocery Stores’. Carl Dipman, the article’s author and a key figure within the magazine, presented this paper as the first outcome of the ‘most important as well as the most costly series of articles’ the magazine had ever published. This series focused on improving grocery stores through ‘better floor plans’ and ‘better store arrangements’ (The Progressive Grocer, 1927: 11, 10), and it was presented as the result of a two-year research effort aimed at ‘gathering the principles that underlie good grocery store operation’ (1927: 11, 13). The team of The Progressive Grocer, the magazine said, travelled from ‘coast to coast’ in order to consult ‘the best merchandise authorities in the land’ and to learn from ‘other fields such as the variety chains, department stores, hardware stores and chain grocery stores’ (The Progressive Grocer, 1928: 12, 20). Interestingly, the lessons of this large-scale survey were converted into a single small-scale store said to summarize all the knowledge gained. In a nutshell, the national geography of grocery methods was encapsulated in a design for a pico-geographic grocery store, per the following: ‘We then took the experience of all and combined them in a Model Store Plan’ (The Progressive Grocer, 1928: 12, 20).
The Progressive Grocer’s ‘Model Store Plan’ was not a bi-dimensional and univocal ordinary blueprint; rather, it was a full and three-dimensional ‘model grocery store’ comprising elements that could be endlessly replaced, removed, complemented and displaced: In order to make this information available to grocers the country over The Progressive Grocer has gone so far as to build a model grocery store in which the fixtures and merchandise can be moved about to meet various conditions. (The Progressive Grocer, 1927: 11, 13) [T]his model store was constructed to serve as a laboratory in which the principles of good grocery store arrangement could definitely be applied and demonstrated to retailers through actual photographs. (The Progressive Grocer, 1928: 3, 1–2)
The magazine proposed a collection of 28 store plans (The Progressive Grocer, 1928: 12, 21) to account for possible ‘various conditions’ (The Progressive Grocer, 1927: 11, 13). These multiple plans invented a kind of ‘living blueprint’, or dynamic geography; however, they were also all built around the same approach: ‘We are developing an entirely new scheme of store arrangement—a scheme of open display that lets people handle the merchandise in so far as sanitation permits’ (The Progressive Grocer, 1927: 11, 66).
More precisely, the model referred to an idea of a store that was as ‘fixed’ as its organization was flexible. ‘Let’s build what might be called America’s typical grocery store’, proposed the campaign, suggesting, ‘It will be a service store with a good volume of delivery business—yet with considerable counter trade’ (The Progressive Grocer, 1928: 1, 10). It is important to stress that the category of ‘service store’ was a starting point that was taken as a given and not negotiable. In the magazine’s context, service referred to the identity of independent American grocers who operated their stores with the help of their families and/or a few clerks. These grocers worked according to an implicit but strong model based on service, credit and delivery, rather than self-service or cash-and-carry approaches (characteristic of emerging businesses). However, the chosen model should not be considered old-fashioned based on contemporary standards. Rather, compared to the self-service stores, stores following the service model featured the most modern elements of today’s ‘drive’ stores, loyalty schemes, and so on. In addition to their service feature, such stores – with their four walls and prominent, usually main street, locations – were considered a given in every city, even as the sizes and proportions of different stores varied. These variances were addressed in ad hoc articles outlining special solutions, ranging from ‘Some Plans for the Narrow Grocery Store’ (The Progressive Grocer, 1928: 4, 10 sq.) to ‘Four Grocery Layouts for the General Store’ (1928: 5, 16 sq.).
The key idea behind The Progressive Grocer’s model and its whole campaign for better grocery stores was that of ‘open display’. By open display, the magazine referred to providing consumers with maximum visual and physical access to goods, without reducing service. The approach was premised on the logic that easy access to goods would give consumers additional ideas and, thus, encourage more sales (in addition to, as we will see, easing store operations): We want open display and open shelves, we want our stock accessible to our customers—every item displayed—so that the individual sale will be increased. (The Progressive Grocer, 1927: 1, 11) Every item sold should be so displayed that it can be seen and handled, if possible. (The Progressive Grocer, 1927: 12, 22) If we keep the store open, if we keep the aisles clear, women will circulate about a store of this type freely, wait on themselves, buy more merchandise, and have a good time doing it. (The Progressive Grocer, 1928: 2, 17) The old-fashioned store separates customers from merchandise, places barriers between them, while the new need is to bring them more closely together—so that the housewife can see and handle what she wishes to buy. (The Progressive Grocer, 1928: 12, 18–19)
Before examining the solutions proposed to implement the open display concept, it is interesting to review the reasons offered for adopting it. The Progressive Grocer provided as many such reasons as it could, drawing on economic, technical, societal and managerial rationales. In other words, it proposed that re-agencing the grocery store was first about re-agencing its operations. The main idea behind this rhetoric was to show that the context was changing and that ‘the old methods—the old system—the out-of-date store of 10 years ago—will not do today’ (The Progressive Grocer, 1928: 6, 10).
The first reason put forward to justify the remodelling of grocery stores drew on changing economic conditions. ‘Fifteen years ago’, The Progressive Grocer observed, ‘a grocer, could run his business inefficiently and still make money. Rents were low. So were wages’ (1927: 11, 1). The magazine claimed that these conditions were no longer true due to higher real estate and labour costs. It unsurprisingly concluded that the new conditions called for better efficiency, particularly since the growing ubiquity of automobiles and the competition of retail outlets with better store arrangements threatened older stores that resisted change (1927: 11, 12). The magazine warned grocers, ‘In delaying don’t forget the chain is not delaying its campaign’ (1928: 6, 14).
A second reason was technical. The magazine observed that the progressive (i.e. incremental and modern) replacement of bulk goods by packaged ones – and, thus, the growing importance of labels that had to be seen and read – required greater product display: We are now in a package age. (…) And package goods sell best when brought in contact with customers—so we no longer need rows of counters, show cases, and various contrivances for keeping the customers from the merchandise. What we need instead are ways and means of bringing them together—of letting our customers handle the merchandise—of waiting on themselves. We need opens stocks and open display. (The Progressive Grocer, 1927: 12, 22, emphasis in original).
This supposed tendency of women to engage in physical contact with merchandise was presented as ‘naturally’ calling for a more open form of display: [W]hen a woman comes into the store she is interested in merchandise. She wants contact with it. She wants to read the labels and the directions for its use. This is why the grocery store should be arranged with as much open display as possible. Give the merchandise prominence—and not the counters, show cases and antiquated fixtures. (The Progressive Grocer, 1927: 12, 23) The housewife today is more or less independent, both in how she spends her money and where. She has been taught to shop around. The automobile is playing its part. She is doing more case buying than ever before (…) This change in the buying habits of the housewife makes necessary a change in the arrangement and the floor plan of the grocery store. (The Progressive Grocer, 1927: 11, 14)
In the age of triumphant Taylorism, The Progressive Grocer’s core argument was that one could cut operations costs by optimizing labour efforts (Cochoy, 2015). Specifically, one should avoid unnecessary movement on the part of workers by better localizing supply and harnessing customers’ contributions: If you want to bring down your cost of operation, you must eliminate all lost motion. (The Progressive Grocer, 1927: 12, 22) Stocks must be so arranged that an order can be filled and customers waited on with the fewest possible steps and in the shortest possible time. If the customer can be made to some extent to help himself, so much the better. (The Progressive Grocer, 1927: 11, 58)
Why did The Progressive Grocer avoid a radical move towards self-service? Is it because this feature did not exist yet? Yes and no. Retail historians have traced the origins of self-service from a chronological standpoint, and they often mention the patent held by the Piggly Wiggly chain. This patent was filed in 1917 and published in 1921, one year before the launching of The Progressive Grocer and 15 years before the campaign for better grocery stores. The document is interesting because, like The Progressive Grocer, the Piggly Wiggly patent proposed a special store plan with an Ikea-like layout, two turnstiles at entry and exit, and a mandatory journey along a serpentine-like corridor with shelves on both sides.
However, as can be seen in Figure 2, this solution was likely too radical to be easily adopted. Indeed, the existence of a pioneering solution says nothing about its actual implementation. If The Progressive Grocer was aware of the emergence of self-service, 4 it knew that few people knew about it and that most grocery businesses remained old-fashioned counter-service stores. Furthermore, the magazine was well-informed about its readers’ strong attachment to service agencement. Therefore, it chose to propose ‘open display’ as a way to compromise between – or, rather combine – service and self-service. On one hand, the magazine acknowledged that its survey showed that customers were ready to handle goods: ‘We found that women like to wait on themselves’ (1928: 12, 20). However, on the other, it proclaimed that ‘The efficient grocery store is a skilful combination of the self-serve and the service store. It adopts the best features of each’ (1929: 1, 23).
The model grocery store. Left: The Progressive Grocer, 1928: 2, 13; right: The Progressive Grocer 1928: 2, 15.
The Piggly Wiggly patent.
This choice took into account the state of commerce. As previously mentioned, to advocate in favour of open display, the magazine suggested that 90% of goods were now packaged. Elsewhere, however, at the risk of contradiction, it also stated that 90% of consumers ended up at the wrapping counter (The Progressive Grocer, 1928: 1, 12). Given these conflicts, it is clear that what mattered most to The Progressive Grocer was the proper ‘agencement’: that is, the combination that would ensure the preservation of service and the introduction of profitable innovations. Indeed, independent grocers saw service as both a strong advantage for consumers and a key business asset, given its power to build trust and secure consumers’ loyalty. What mattered and what was seen as progress, in this respect, was service improvement, not its replacement with alternative solutions. Though self-service could reduce costs, it could also endanger the professional identity of the service grocer, as well as the consumer relationship he worked so hard to build.
But how could the concept of open display be implemented? What had to be done in order to combine self-service and counter service in the hope of increasing sales and reducing operational costs? The answers to these questions rested on a subtle re-agencing of store elements, which is best illustrated with the three plans reproduced in Figure 3.
Re-agencing moves with open display.
Better Grocery Stores, the movie.
Before and after remodeling.
As can be seen by observing from left to right in Figure 3, the most obvious changes from one plan to the next concerned the reconfiguration of the movement of shopping inside the store and the identity of the person(s) performing this movement. On the left and middle, the person behind the counter – that is, the grocer – engaged in several radiant travels in every direction inside the shop to fulfil the requests of an immobile customer. On the right, by contrast, the customer waited on herself, while the grocer remained immobile. In the middle case, the customer conducted a single short loop near the store entrance, while, on the right, she performed a long and complex journey, going all the way to the very back of the store. The difference between these first two journeys and the one on the right was strongly connected to the place where the grocer stood: in the left and middle case, the professional was near the shop entrance, while on the right, he was at the bottom of the store.
Eventually, one realizes that the evolution of moves and the repositioning of the grocer were both linked to the remodeling of the store. In the left and middle drawings, the merchandise placed on the wall shelves was separated from the customer by long counters that extended from the entrance to the back of the store, giving the customer no opportunity to be exposed to the merchandise and get new ideas. She would get only what she ordered: a lot if she asked for a lot, but only a single item if a single item was all she was looking – or, rather, asking – for: ‘The grocer and his clerks must bring to the woman standing at the counter nearly everything she wants. There is little browsing about—little opportunity for the woman to exercise her shopping instinct’ (The Progressive Grocer, 1931: 5, 46).
On the right, all the store equipment was changed: the long counters were replaced by a single one at the back of the shop, wall shelves were now accessible to the consumer and several display fixtures were spread throughout the middle of the shop. In order to pay and have her selection wrapped, the consumer now had to travel the entire breadth of the shop. Along the way, she could interact with the merchandise, be tempted and possibly pick up goods she had not considered buying upon first entering the shop: ‘Now the housewife can to some extent wait on herself. Now she can exercise that shopping instinct, browse about the merchandise, examine, handle, read and scrutinize to her heart’s content’ (The Progressive Grocer, 1931: 5, 46).
As The Progressive Grocer summarized, the supposed effect of the new store arrangement was twofold. On one hand, it was designed to increase sales: ‘Store circulation will throw [customers] in contact with a large number of items and expose them to additional sales along the way’ (1931: 5, 45). On the other hand, it was devised to reduce operation costs. From the left to the right, the grocer’s moves and physical expenses were minimized: ‘Insofar as they buy more, and especially if they to some extent wait on themselves, the grocer’s operating expense is materially reduced’ (1931: 5, 45).
All these changes were attributable to a new type of store agencement, which was rooted in three major changes. First, the major feature of open display was the setting of what the The Progressive Grocer called the ‘center of activity’. ‘The whole store must be built around our center of activity’, it said, explaining that ‘The center of activity is determined by where we place the counters. (…) That is where the activity is going to be’ (1928: 1, 12). Then, it recommended that grocers ‘locate the center of activity at the rear of the store’ (1928: 1, 13). This localization had several advantages. Because the counter was the place where ‘90% of our customers’ (1928: 1, 14) went, placing it at the rear created ‘the opportunity of exposing them to further sales on profitable items, and increase the average sale’ (1928: 1, 13). Locating the counter at the rear centred this area in both the geographical and the literal sense: the grocer could now access, with minimal movement, not only most of the goods that still required service, but also the back room, from which he could refill his stock and prepare deliveries for distant customers (1928: 1, 16). All in all, from this activity centre, the grocer could ‘sell the largest amount with the least effort—fill orders, if possible, standing in [his] tracks’ (1928: 1, 14).
This re-agencing was a matter of re(-)placing different elements, not only in the sense of relocating them, as we just saw, but also in the sense of renewing them, as we will see shortly. Indeed, moving furniture inside the store also meant removing some fixtures and acquiring new ones.
In this context, The Progressive Grocer relied on a double rhetoric. The first focused on ‘out-fashioning’ existing equipment by illustrating its inaccuracy and inefficiency in relation to alleged new external conditions: Because this is a package age and less than 10% of the average grocer’s sales are in bulk merchandise, we could find no reason for the long lines of out-of-date show cases, counters, bins and drawers that only served to obstruct sales. (The Progressive Grocer, 1928: 12, 20) A display under glass cuts off the possibility of touch. That’s why in many stores the old-fashioned show cases are being thrown out. They retard sales. (The Progressive Grocer, 1927: 11, 66) [D]o not include show cases in your plan if they are not absolutely necessary.
5
In most stores the old-fashioned show cases hide more than they show. (The Progressive Grocer, 1928: 2, 13–14)
Introducing open display: A national campaign and preceding innovations
The dynamic of the campaign for better grocery stores
The campaign for better grocery stores was designed to reshape (i.e. re-agence) both grocery stores and the grocery community. The model store was a key feature in this respect, not only because it supported the testing of various configurations through the moving of store elements, but also because it could be moved itself as a whole, as if one could displace an entire grocery store across the USA. By offering nearly unlimited options for reconfiguration, The Progressive Grocer multiplied the uses of its model store throughout the country.
We have already seen that the model illustrated various store agencements through photographs published by The Progressive Grocer. This first use was rapidly duplicated through the publication of a book called Better Grocery Stores, which compiled the magazine’s collection of papers on the topic. According to The Progressive Grocer, this book was published ‘At the request of manufacturers, jobbers and associations’ (1928: 3, 2). It was sold at cost (1928: 3, 2) and even refunded in case of dissatisfaction (1928: 12, 2–3). Not only did the magazine repeatedly advertise the book in its pages, but it also continuously reported its sales: both as proof of the campaign’s success and as a call for its further implementation. For example, we learn that 15,000 copies of Better Grocery Stores were sold in March 1928 (The Progressive Grocer, 1928: 3, 2), 25,000 were sold a month later and a third edition was planned (1928: 4, 2–3). The book was supplemented with the circulation of no less than 300,000 leaflets designed to promote the campaign among professionals (1928: 5, 2). The book was also supplemented with a special ‘portfolio’ comprising large reproductions of the model store plans, which each grocer was encouraged to use to ‘work in making [his] store along modern lines’ (1928: 12, 3). This portfolio was either proposed as a complement to the book (1928: 6, 18-–9) or circulated by a ‘thousand traveling salesmen’ working for jobbers and manufacturers (1928: 4, 2–3; 1928: 6, 10–11).
The model store was also featured as a store that could be transported and exhibited at various events and conventions. It was first used in this way during meetings of the National Wholesale Grocers Association in Chicago and Del Monte in January 1928 (The Progressive Grocer, 1928: 3, 2) and was later repeatedly displayed in similar events. Moreover, the model store feature was imitated by other private and public actors: on the private side, wholesalers built similar stores in their warehouses (The Progressive Grocer, 1928: 3, 2–3), while on the public side, the Department of Commerce invited Carl Dipman to design a model store that was visited by thousands of people (The Progressive Grocer, 1929: 4, 18).
However, the articles, books, leaflets, portfolios and model stores were not the only means The Progressive Grocer used to promote its campaign. In September 1929, the magazine’s journalists reported that they had presented their open display idea and solutions at more than 160 meetings and suggested that they would not be able to face half of the incoming demand for similar talks (The Progressive Grocer, 1929: 9, 1). To overcome this challenge, the magazine decided to produce a movie. This one-hour motion picture, entitled Better Grocery Stores, was created for meetings, conventions and other professional gatherings. It was proudly presented as the first such media for the grocery business. The film featured the adventures of Smith, a traditional grocer whose old-fashioned store struggled until a providential and fantastical re-agencing of the store – in the moving picture, ‘we see counters, show cases and shelves moving magically about the grocery store’ (The Progressive Grocer, 1929: 4, 40–41) – solved all his problems (see above, Figure 4). As it had previously done with its book, The Progressive Grocer reported the success of its new visual production. In April 1929, six prints of the movie were in continuous use (1929: 5, 2–3), and three months later, 49,046 ‘grocers, their wives and clerks’ had seen the film in no less than 156 different towns across the country (1929: 8, 50).
However, articles, books, model stores, talks and movies, no matter how numerous, didactic or spectacular they might have been, were not sufficient. Indeed, they were also simply words, talks and images. Of course, the systematic accounts produced about their circulation helped to create the impression that they publicized the campaign; however, disseminated knowledge did not necessarily imply implemented knowledge. To overcome this problem and complete its demonstration of the open display concept, The Progressive Grocer soon began to supplement exposure to its argument with the exhibition of its supposed effects. It did so by mentioning the ‘hundreds’ (1928: 6, 10 sq.), ‘thousands’ (1928: 11, 26) and, soon, ‘tens of thousands’ (1930: 10, 30–31) of testimonies the magazine received from grocers who said that they had remodelled their stores and increased their sales by following its recommendations and by posting calls asking others to share similar experiences (1928: 6, 10–11). In 1932, The Progressive Grocer stated that, according to testimonies it had received, 45,000 stores had completely modernized, 25,000 had made changes and ‘At least 10,000 of these modern food stores have bloomed during the depression year of 1931’ (1932: 1, 14). Moreover, it suggested that the shift toward open display had placed independent grocers at the forefront of the modernization of the grocery business, paradoxically forcing chain stores to follow rather than lead the movement: ‘even the chain stores are now modernizing their stores to hold their own with the rising army of efficient independent grocers’ (1932: 1, 16).
The Progressive Grocer began illustrating such reports with spectacular ‘before and after’ photographs designed to testify the changes made (see above, Figure 5). It also began to publish case studies on store transformations, with titles such as ‘What Remodeling Did for Oleson’ (1928: 9, 26 sq.), ‘Now Bailey Has a Better Store, Too’ (1929: 2, 24 sq.), ‘Miss Grimm Modernized’ (1929: 5, 16 sq.), and ‘Chapman Remodels Often’ (1929: 7, 16 sq.). Both of these tools proved particularly efficient ways to promote the new approach, leaving grocers to think that the possible future of open display was already the present, and that the present state of their stores was the past. In other words, grocers were led to believe that, if they wanted to survive, they had to remodel (Cochoy, 2015).
Based on all of these efforts, the magazine proudly concluded that ‘Open display [was] no longer an experiment,’ but a principle of merchandising, now ‘universal in application’. Furthermore, like any modernization process, re-agencement was presented as irreversible: ‘Not a single grocer who modernized to open display has been known to change back to the old way of doing business’ (The Progressive Grocer, 1930: 10, 30–31).
The distributed origins of open display: The contribution of cans, shelves and Cellophane
It would be wrong, however, to reduce open display to a performed theory, a new discourse designed to be implemented or an original concept fully controlled and enacted by The Progressive Grocer. In fact, it is now time to reveal that The Progressive Grocer simply attempted to organize, label, promote and give a name to several innovations that preceded its intervention. These innovations originated neither from The Progressive Grocer nor from grocers themselves, but from two other key actors: the manufacturers of the goods sold in grocery stores and the providers of the equipment made to sell them.
Let us start with the goods. Some of the most discreet yet most important contributors to the advent of open display were canned and packaged goods. However, it should be noted that the promoters of these goods did not fully frame and anticipate their products’ ultimate effect. The example of canned foods illustrates this phenomenon well. On the one hand, one might imagine that such foods were tightly designed and managed by canners: was it not these manufacturers who fabricated, filled, sealed and sold each can? Yet, though cans may have largely served as intended, they were also unable to accomplish some of the things manufacturers would have liked them to, and, conversely, they proved capable of other things that manufacturers did not expect. Cans were made to preserve foods and they were good at it, but they were also highly consumer un-friendly. They were opaque, hermetic and terribly difficult to open. As such, they deprived shoppers from experiencing their contents through their senses. To overcome this problem, canners were forced to replace sensorial information with scriptural substitutes: they stuck labels on their cans displaying the product’s name, brand and composition. Unexpectedly, this labelling process played a key role in shifting markets from generic, local and anonymous goods to differentiated, national and branded products. They re-agenced consumer cognition by replacing the sensorial qualification of goods with a packaged one. With this new regime of product evaluation, knowledge about a given product was no longer acquired through a single subjective experience (‘tasting’), but through the reading of a detailed, standardized and objective list of product characteristics (‘testing’) (Cochoy, 2017). However, cans did more than this. Because they were hermetic, they could be handled without any sanitary risk; because they were robust, they could overcome grocers’ fears of letting consumers handle merchandise; because they had a flat bottom, they could be piled and, thus, arranged in ‘mass displays’ to better show their content. These properties were soon identified and promoted by can manufacturers.
As it can be seen in the advertisement for Pet Milk Company (Figure 6), cans were clearly presented as the pivotal point in the shift from classic ‘old-time’ store agencements to ‘modern merchandising’. On the one hand, they were still attached to counter sales, since they still required a grocer and his antiquated furniture, through whom the manufacturer had to go to reach consumers: thus, the counter and the ‘store keeper’ implicitly operated as a gate and gatekeeper. On the other hand, cans facilitated the movement of goods from the shelves to consumers: in the ad, they are piled up just in between the consumer and the grocer, on the counter, thereby evidencing their ability to be equally accessible to both parties. Moreover, cans spoke for themselves, repeating brand names sold through advertisements. In other words, cans introduced open display without telling it: they worked as a true transition between the counter service of yesteryear and the self-service that would soon come.
Pet Milk Company.
If can manufacturers introduced open display without publicizing it, equipment providers were more reflexive and explicit. In this respect, the Dayton Company deserves special mention. Equipment providers like Dayton had, of course, a distinct interest in developing solutions that would encourage grocers not only to buy fixtures, but possibly replace existing ones with new versions with supposedly greater sales efficiency. The Dayton Company promoted a new kind of metallic shelf (see Figure 7 below). The shelf’s main feature was a lack of the vertical opaque separations which were characteristic of earlier wooden fixtures. As such, the Dayton shelf provided maximum visibility to goods, an advantage that the company touted through the explicit promotion of what it called ‘Open Display’… one year before The Progressive Grocer launched its campaign under the same idea! The device seems to have been successful, since it can be seen in numerous pictures of stores published by the magazine to illustrate a variety of issues (The Progressive Grocer, 1922: 5, 44; The Progressive Grocer, 1922: 11, 11; The Progressive Grocer, 1923: 4, 2; The Progressive Grocer, 1926: 5, 17; The Progressive Grocer, 1927: 3, 13, etc.).
If The Progressive Grocer did not fully control the origins of open display, it also did not totally master its further developments. We just saw that Dayton coined the notion of open display and promoted its merits before The Progressive Grocer. Similarly, other equipment providers came after The Progressive Grocer. Though these companies acknowledged the magazine as the driving force of open display, they nevertheless promoted their own view of what the concept meant. In the following advertisements, the Sherer and Cellophane companies opportunistically seized the campaign for ‘Better Grocery Stores’ as a favourable environment for their own products and as an argument to promote not only them, but also their particular interpretations of the movement.
The Sherer Company positioned its products as appropriate means for re-agencing grocery stores according to The Progressive Grocer’s recommendations (see the top of Figure 8). Indeed, it presented its fixtures as ‘Specially Designed for the Progressive Grocery Stores’; illustrated its advertisement using the plan of the model store; explicitly acknowledged the magazine’s campaign; and provided three reasons grocers should choose its product as the best way to fulfil the programme: Grocers everywhere are responding enthusiastically to the splendid programme sponsored by the ‘Progressive Grocer’ for Better Grocery Stores. Many are using Sherer Equipment: First, because the design of Sherer Equipment fits so well with the general plans and specifications as suggested by the ‘Progressive Grocer’. Second, because Sherer is able to supply ALL the necessary equipment; And third, because the name Sherer is an undisputed guarantee of utmost reliability. (The Progressive Grocer, 1929: 3, 75).
Dayton’s open display method.
Sherer and Cellophane join The Progressive Grocer.
It is interesting, however, to note that what was presented as a perfect fit between the magazine’s campaign for better grocery stores and the supply of the Sherer Company, in fact masked opposite views about what open display should be. As we have seen, whilst The Progressive Grocer recommended abandoning glass cases for their partial fulfilment of the open display logic, Sherer did not discriminate among fixtures. Whatever The Progressive Grocer might say, Sherer promoted glass showcases because they were part of its catalogue and represented a more recent evolution than wooden alternatives… and, hence, an opportunity for replacement sales.
If Sherer partially reinterpreted open display as a better ‘visual display’, Cellophane reinvented the notion by providing additional means to meet the programme. Like Sherer, Cellophane (see the bottom of Figure 7) explicitly referred to the campaign, quoted one of the magazine’s plans, and presented its transparent package as a perfect fit for the open display logic it promoted:
Experts in grocery store layout (…) recommended open displays to bring women closer to all merchandise. (…) Into this whole plan to increase extra business, Cellophane fits perfectly. It shows every detail of a product—clean, tempting, and ready for instant sales. (The Progressive Grocer, 1931: 8, 78).
However, Cellophane did much more. Unlike Sherer, which sold existing showcases to suit the open display layout, Cellophane sold a brand-new feature, which it positioned as an extension of the concept. Cellophane’s soft transparent package merged the advantages of canned foods and window showcases, while simultaneously correcting their flaws. Like canned foods, products packaged with the Cellophane film could be handled without exposure to sanitation risks, and like goods displayed behind window showcases, they could be seen. However, unlike canned foods, Cellophane did not hide the product, and unlike window showcases, Cellophane did not prevent consumers from touching the wares. Thus, Cellophane emerged as a strong competitor to previous fixtures and as a vehicle for further promoting the ideas of open display and self-service. It invented a pico-geography within the store by re-agencing the properties of open display devices at the level of individual goods.
Last but not least, open display led The Progressive Grocer to recommend systematic price display due to the increased direct interaction between the consumer and the merchandise (1928: 8, 21). The generalization of price display launched, in turn, a further re-agencing of the elements of the grocery business. Prices migrated from price cards displayed in the store window to price tags disseminated on the shelves inside the store, and this supplementing of product labels with price information helped customers wait on themselves. All in all, the magazine continuously worked to manage the overflow of novel agencements devised by its advertisers by re-agencing them into a coherent whole, around an appropriate rhetoric aimed at modernizing and preserving independent service stores. The magazine’s efforts enticed advertisers to push further and continuously reshape the configuration of grocery devices. Paradoxically, it was this subtle, flexible and distributed combination of conservative elements (service) and progressive ones (open display) that likely eased the transition from service to self-service.
Conclusion
The case study conveys empirical, methodological and theoretical lessons. On the empirical side, we saw that the re-agencing of grocery stores around the concept of open display was heavily promoted through a scale model conducive to testing and illustrating the promises and workings of the new system. However, we also know that The Progressive Grocer did not fully control the re-agencing process. Rather, its contribution built on preceding and external initiatives that the magazine attempted to channel and orient. This said, by re-agencing the re-agencers – by aligning, combining and theorizing the ongoing transformations of the grocery business – the magazine and its open display idea proved capable of prompting new combinations of service and self-service. These ad hoc ‘agencements’ eventually eased the development of self-service, probably much more efficiently than the concept of self-service itself! Indeed, at this moment in history, chains’ competitive advantage relied more on economies of scale and buying power than on distinct merchandising features (Howard, 2015; Spellman, 2009), and self-service stores were simply isolated experiments (Bowlby, 2001).
We saw that studying such re-agencing procedures requires adopting a ‘pico-geographic’ view that pushes the micro-geographic perspective at the indoor and blueprint level. This said, macro, micro and pico scales are closely articulated. As we saw, the re-agencing of grocery stores involved changes ranging from the reshaping of the pico-geography of the grocery store to the transformation of the external macro-geography of the American mass market. A few years after the period covered in this paper, indeed, the adoption of open display re-agenced the entire geography of the grocery business and beyond: open display agencements promoted the development of self-service; self-service supported the introduction of shopping carts (Cochoy, 2009); self-service and shopping carts encouraged the use of cars and parking lots; and these latter evolutions induced the growth of sales surfaces and their migration toward the suburbs (Cochoy, 2015); and, eventually, these collective shifts facilitated a full re-agencing of our Western consumer space.
Theoretically, our study suggests moving from performativity to agencing. The performativity framework has been very useful and powerful to promote a pragmatist approach of markets and account for the spatial enactments of marketization processes. However, the linguistic roots of performativity implicitly support a somewhat simplistic view where words and things are two separate entities, with words preceding their transformation into real matters. Thinking in terms of agencing helps to overcome such problems, by showing how economic performation really works: within market agencements, words and things are not the starting points, but rather the outcomes of a ceaseless reconfiguration of various human and non-human entities. This reconfiguration – or re-agencing – is fully geographic, to the extent that it takes place in material, grounded and situated spaces, and is distributed and rearticulated spatially.
Footnotes
Acknowledgements
We warmly thank The Progressive Grocer for granting us the permission to reproduce the images upon which this paper draws. I am very grateful to Jamie Peck and two anonymous reviewers for their helpful comments and suggestions on a previous version of this article. I also sincerely thank the librarians of the New York Public Library for their assistance.
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.