Under the Umbrella: Goal-Derived Category Construction and Product Category Nesting

Abstract

Categories are organized vertically, with product categories nested under larger umbrella categories. Meaning flows from umbrella categories to the categories beneath them, such that the construction of a new umbrella category can significantly reshape the categorical landscape. This paper explores the construction of a new umbrella category and the nesting beneath it of a product category. Specifically, we study the construction of the Quebec terroir products umbrella category and the nesting of the Quebec artisanal cheese product category under this umbrella. Our analysis shows that the construction of umbrella categories can unfold entirely separately from that of product categories and can follow a distinct categorization process. Whereas the construction of product categories may be led by entrepreneurs who make salient distinctive product attributes, the construction of umbrella categories may be led by “macro actors” removed from the market. We found that these macro actors followed a goal-derived categorization process: they first defined abstract goals and ideals for the umbrella category and only subsequently sought to populate it with product categories. Among the macro actors involved, the state played a central role in defining the meaning of the Quebec terroir category and mobilizing other macro actors into the collective project, a finding that suggests an expanded role of the state in category construction. We also found that market intermediaries are important in the nesting of product categories beneath new umbrella categories, notably by projecting identities onto producers consistent with the goals of the umbrella category. We draw on these findings to develop a process model of umbrella category construction and product category nesting.

Product categories are embedded in larger category systems. Within these systems, categories are organized vertically in a “‘stem and branch’ type hierarchy” (Wry and Lounsbury, 2013: 120) in which higher-level, umbrella categories (Kuilman and Li, 2009) nest lower-level product categories. This nesting of categories is important, as meaning can flow from umbrella categories to the product categories below (Durand and Boulongne, 2017). For instance, shifting the position of a product category nested under one umbrella category to another can profoundly affect the meaning of that lower-level category, as in the case of certain traditional Indian art whose meaning was transformed when art critics nested it under the “modern art” umbrella category (Khaire and Wadhwani, 2010). The emergence of a new umbrella category can potentially have an even more profound effect on the categorical landscape, as it “deliberately involves [re]positioning certain entities or goods that were previously disparate” in a way that imparts new “value potential” to those products (Durand and Khaire, 2017: 96).

While extant organizational research has shed much light on the construction of new categories, for the most part it has focused on the product categories that reside at the lowest levels of the category structure rather than on higher-level umbrella categories. New product categories are said to arise when a group of “products or services are perceived to be of the same type or close substitutes for each other in satisfying market demand” (Navis and Glynn, 2010: 440). As entrepreneurs develop novel products, their audiences may apprehend distinctive attributes or features that come to define a category prototype against which category membership is then assessed (Pólos, Hannan, and Carroll, 2002; Vergne and Wry, 2014). Entrepreneurs may further the perception that a group of objects constitutes a new category by defining and rendering salient the attributes that distinguish the nascent category from incumbents (Kennedy, 2008; Santos and Eisenhardt, 2009). Entrepreneurs do so by “emphasizing similarities across the firms that claim membership to the category,” often in opposition to incumbent peer categories (Navis and Glynn, 2010: 442). The imagery is one of creating sharp distinctions between emergent and existing categories based on product attributes (Wry and Lounsbury, 2013; Grodal, Gotsopoulos, and Suarez, 2015), a process that is often assumed to involve a “division of labor” (Grodal, 2018: 785) whereby producers claim membership to categories and their audiences evaluate these claims.

There is some indication, however, that the construction of new umbrella categories may differ significantly from the process described above. Whereas the prototypical view that dominates organizational theory emphasizes similarity via shared attributes, umbrella categories may subsume under the same umbrella products that ostensibly share few attributes. For example, the umbrella category of “transportation” encompasses distinct modes of transport such as planes, trains, and automobiles that share few common attributes (Loken and Ward, 1990). Moreover, Durand and Khaire (2017) suggested that the reordering of the categorical landscape may be of less interest to entrepreneurs and more so to actors further removed from production, and it may follow very different category construction processes. ¹ The latter set of actors may, in fact, prefer more ambiguous categories, in contrast to the sharply defined categories sought by entrepreneurs (Pontikes, 2012). Instead of assessing membership using category prototypes, these actors may make use of ideals whereby “category members are items that most suit an ideal, even if they greatly differ from each other—category structure is driven more by goal pursuit than by family resemblance” (Durand and Paolella, 2013: 1109). Termed “goal-derived categories” (Barsalou, 1985; Barsalou and Hutchinson, 1987; Barsalou, 1991), such categories reflect a form of categorization distinct from the prototypical approach that dominates organizational theory.

Despite these potential differences, there is little empirical research on the construction of new umbrella categories and in particular how these come to subsume product categories. A goal-derived lens can be useful in this regard, yet much of the discussion of goal-derived categorization in organizational studies has remained conceptual. In particular, we lack evidence of goal-derived categorization processes unfolding at a collective level and how these processes can produce enduring structures. This is because goal-derived categories were originally conceived as ad hoc categories (Barsalou, 1983)—ephemeral structures created when individuals “face unique situations for which they have no comparable previous experiences or expertise” (Durand and Boulongne, 2017: 651), leading them to creatively recombine existing categories to meet their immediate needs.

To address these lacunae, we adopt a goal-derived lens to study how umbrella categories are constructed and come to nest product categories. While prior conceptual work has suggested that umbrella category construction may differ in important ways from product category construction, these differences have not been elaborated, and moreover, we know little about the linking mechanisms that nest product categories under umbrella categories. We study how such processes unfolded in the context of the Quebec terroir products umbrella category and the Quebec artisanal cheese product category between 1977 and 2013. This context represents a revelatory case (Yin, 2009) because both artisanal cheese production and the notion of terroir were new to Quebec, and the meaning of the former seems to have been transformed as it was nested under the latter. Artisanal cheese had emerged across North America in the 1980s guided by a rejection of industrial production (Paxson, 2010, 2012), but in Quebec, artisanal cheese became representative of Quebec culture and traditions in the way evoked by notions of terroir.

Over our study period, the Quebec terroir products category came to link foods produced in the Canadian province of Quebec to its land, culture, and traditions, valorizing them as symbols of the Québécois identity, as indicated in this quote from a food writer:

Let us be proud of our roots, our terroir and what can grow here despite the northern climate. The Québécois culture is made up of what we produce and eat. We need to discover or rediscover what comes from our soil, from our own production. (Samson, 2010: 11)

“Quebec terroir products” is an umbrella category that includes an array of foods as subcategories, among which the best known is Quebec artisanal cheese. But despite its association with Quebec’s culture and traditions, the Quebec terroir products category entered public discourse and began to be used only in the late 1990s. It reordered the categorical landscape of food products in the province, as diverse categories of foods were nested beneath it. As a prominent food critic wrote in 1999, of all the categories that would become known as Quebec terroir products, “[artisanal] cheese is a star of the Quebec terroir” (Kayler, 1999). Quebec artisanal cheese was considered virtually emblematic of Quebec culture and history, and yet artisanal cheese production was also relatively new to the province, dating only to the 1980s. Cheese producers had initially been motivated by anti-industrial beliefs, without subscribing to notions of Quebec having a cheese-making heritage, and their product category had received little attention prior to it becoming nested under Quebec terroir.

This paper aims to offer contributions in two main areas. First, we engage with recent work suggesting that category construction can differ dramatically at different vertical levels (Khaire and Wadhwani, 2010; Durand and Khaire, 2017). Specifically, we distinguish between diverse sets of actors—producers, market intermediaries, and macro actors more removed from the market—and trace their involvement in the category construction process at different vertical levels. Notably, our findings reveal that the state can be a direct participant in the category construction process, a largely overlooked phenomenon given the prevailing view in institutional theory that the state’s role is one of raising or lowering entry barriers through regulatory change or financial incentives (Georgallis, Dowell, and Durand, 2019; Grandy and Hiatt, 2020). We also show that while product category and umbrella category construction may begin independently, the two processes may be linked through an identity attribution process whereby market intermediaries project onto producers an identity consistent with that of the umbrella category.

Second, we contribute to a burgeoning literature in organizational studies on goal-derived categorization (Durand and Paolella, 2013; Durand and Boulongne, 2017; Glaser, Krikorian Atkinson, and Fiss, 2020). Whereas goal-derived categories have mostly been conceptualized as ad hoc categories—created, used, and then discarded by individuals (Barsalou, 1983)—we show that goal-derived categorization can unfold collectively and with enduring effects. Specifically, we show that participants in the category construction process may collectively define a goal and a set of ideals for an umbrella category and only afterward populate the category with a wide diversity of products that appear to approach those ideals. One corollary of this is that goal-derived categorization may begin absent material exemplars and as such can support higher levels of ambiguity than prototypical categories. In this way, we show how goal-derived categorization may give macro actors a means to reorder the categorical landscape despite being removed from the production of marketable goods or services.

Umbrella Categories and Goal-Derived Categorization

Categories are social constructions that differentiate among entities, such as products, people, and organizations (Lamont and Molnar, 2002; Giorgi, Lockwood, and Glynn, 2015). Categories locate an entity within a broader system of meaning or classification (Glynn and Navis, 2013) and in this way organize an otherwise amorphous and bewildering landscape into identifiable clusters. The dominant approach in organization studies conceives of categories as defined by category members’ essential attributes and encapsulated by their category prototype (Durand and Paolella, 2013; Vergne and Wry, 2014). In this attribute-based approach, a category prototype is used by audiences as a “baseline to assess family resemblance” for membership into a category (Glynn and Navis, 2013; Durand and Boulongne, 2017: 651).

Considerable research has discussed how entrepreneurs construct new market categories. This research shows that entrepreneurs deploy value-laden (Glynn and Navis, 2010; David, Sine, and Haveman, 2013) and coherent accounts that define distinctions with incumbent categories to give meaning to their own (Wry, Lounsbury, and Glynn, 2011; Lo et al., 2020). As Verhaal, Khessina, and Dobrev (2015: 1466) explained, new market categories may coalesce around “producers who define themselves as categorically opposed to the dominant market logic employed by incumbent firms in existing product space.” The emergence of craft beer in the United States provides a well-known example, as it emerged in direct opposition to industrially produced beers—microbreweries and brewpubs defined “the specialty beer segment in ways that exclude major brewers and contract brewers” (Carroll and Swaminathan, 2000: 731). Similarly, producers of grass-fed beef “solidified the set of binary oppositions” with respect to the production, exchange, and consumption of traditional beef (Weber, Heinze, and DeSoucey, 2008: 558). In this way, producers stake “a reactionary claim about who they are in direct reference to a set of established incumbents” (Verhaal, Khessina, and Dobrev, 2015: 1467). The dominant imagery of this research is thus that new categories emerge as producers define the central and distinctive attributes of their product categories in contradistinction to horizontal peer categories, such as craft beer versus mass-produced beer, grass-fed beef versus traditionally raised beef, and soft drinks versus alcoholic drinks.

However, this emphasis on distinctions with horizontal peers tends to overlook that category systems have a vertical dimension as well (Durand and Boulongne, 2017). In addition to considering horizontal arrays of peer categories, such as different kinds of feature films (Hsu, 2006), wines (Negro, Hannan, and Rao, 2011), software firms (Pontikes, 2012), or restaurants (Kovács, Carroll, and Lehman, 2014), we must also consider categories as being vertically nested under larger, umbrella categories. For example, grappa is part of the larger spirits category (Delmestri and Greenwood, 2016), and sports cars, sedans, trucks, SUVs, and minivans are all part of the larger automobile category (Wry and Lounsbury, 2013). Umbrella categories are significant as they impact the meanings attributed to the subcategories nested under them. As Wry and Lounsbury (2013: 120) explained, “relevant distinctions may sit at higher or lower levels of analysis . . . one possibility is that categories may cohere in configurations that are associated with a higher level aggregate.” For example, Jones and colleagues (2012) described how two new architectural categories, “modern organic” and “modern functional,” were later grouped into a larger, more amorphous category of “modern architecture,” which also included a third category that blended the two styles. Likewise, in their study of the emerging category “modern Indian art,”Khaire and Wadhwani (2010) described how art historians and critics recategorized twentieth-century Indian art under the “modern art” umbrella, which already existed in the West. In both examples, the nesting of focal categories under higher-level categories affected both their meaning and valuation.

Yet, there has been little research on the construction of these higher-order umbrella categories. One reason for this lack of attention may be that the attribute-based approach to categorization tends to see new categories largely as increasingly specific subtypes of the larger categories below which they are nested. For example, in their characterization of a vertical category system, Kuilman and Li (2009: 231) stated that “in such a system, one can define a subpopulation as a subset of a general population by combining all the membership-defining features of the general population with some more specific features idiosyncratic to the subpopulation.” Such a view may preclude the study of the construction of new umbrella categories, as it seems to presume that subcategories are constructed as ever more specific subtypes of umbrella categories in a temporally sequenced process. While such a sequence may reflect the emergence of many vertically nested categories, it certainly does not reflect all. Take for example the umbrella category of cleantech (or greentech), which encompasses a wide range of product categories such as solar energy, biofilms, or biodiesel (The Economist, 2007). While the cleantech category clearly informs the understandings of the subcategories that it encompasses, these subcategories share few visible attributes, and moreover, many of the subcategories existed well before being labeled as cleantech. Thus one cannot readily presume a temporal sequencing whereby the category of cleantech first emerged and then fragmented into related but more specific subcategories.

Compared with the attribute-based (i.e., prototypical) approach to categorization, the goal-based approach may prove more fruitful in the study of umbrella categories. ² At higher vertical levels of categorization, an attribute-based approach may become inapplicable as subcategories may no longer exhibit similar attributes (Loken and Ward, 1990). Goal-derived categorization, in contrast, may better describe umbrella categories that comprise several quite different (prototypical) subcategories that serve the same (often ambiguous and idealized) goal (Ratneshwar, Pechmann, and Shocker, 1996; Ratneshwar et al., 2001). Unlike in the attribute-based approach, where category membership is assessed on the basis of the possession of certain attributes or features (Vergne and Wry, 2014), in the goal-derived approach category membership is assessed based on approaching ideals. Whereas prototypical attributes are visible among the exemplars of a category, ideals refer to qualities that category members may approach but never fully attain in the pursuit of the category goal (Lynch, Coley, and Medin, 2000; Voorspoels, Storms, and Vanpaemel, 2013). In the classic example provided by Barsalou (1985: 630) for the goal-derived category of “foods to eat on a diet,” an ideal is zero calories: “the fewer calories an exemplar has, the better it serves the goal associated with its category, namely, lose weight” (emphasis in original). Returning to the example of cleantech, each (prototypical) subcategory may include highly specific and agreed-upon attributes, but they share little in common across subcategories apart from serving the goal of environmental sustainability. Among the ideals relevant to this category may be that the included technologies be carbon neutral or even carbon negative. In both examples, goals and ideals, not product attributes, provide the lens that unites the objects under the umbrella category.

Moreover, the goal-derived approach can help disentangle the construction of umbrella categories and the subcategories that they encompass. As mentioned above, prior work on the construction of vertically nested prototypical categories suggests that the umbrella category is constructed first and increasingly specific subtypes then follow. But with goal-derived categorization, umbrella categories may be constructed independently of their member categories. In this perspective, category construction “does not rely on exemplars that capture the central tendencies of the category, but on ideals” (Durand and Boulongne, 2017: 652). This point has two implications. First, umbrella categories can begin as empty sets—i.e., with nothing more than a goal—which challenges the dominant view that “categories emerge as audience members recognize similarities among groups of producers and come to regard and label them as members of a common set” (Negro, Koçak, and Hsu, 2010: 7). In effect, goal-derived categorization can begin absent material exemplars. Second, the construction of umbrella categories may be independent of and potentially even follow that of the prototypical product categories they encompass. In fact, the very importance of goal-derived categories derives from their use in creatively recombining existing categories. For example, various prototypical categories of small-scale electricity producers (e.g., wind, solar, biomass) had long existed in relative obscurity until legislators grouped them under the new “alternative energy” umbrella category as part of their effort to reduce U.S. reliance on imported oil (Sine and David, 2003). In so doing, legislators not only regrouped existing categories but also reinterpreted them along the lines of the goal of increasing national security. In this way, umbrella categories can reorder the categorical landscape and give new meaning to the subcategories that come to be nested under them.

However, much remains to be learned about goal-derived categorization. A goal-derived lens has historically been used to study ad hoc categories: temporary categories created by individuals to meet their immediate goals (Barsalou, 1983). For example, corporate legal services are classified according to sharply defined prototypical categories, yet sophisticated clients for legal services may define their particular goals and populate their goal-derived categories with firms classified under diverse prototypical categories (Paolella and Durand, 2016). Each user can thus potentially construct their own unique (personal-use) goal-derived category without it impacting the category structure prevailing in the field. We know little about how goal-derived categorization may operate at a collective level and whether it can produce enduring, collectively recognized categorization schemes. By extension, we lack understanding of how goal-based categorization can serve as the basis of market exchanges (Glaser, Krikorian Atkinson, and Fiss, 2020). Moreover, an important aspect of considering goal-derived categorization at the collective level is the potential distinction between the goals of category users and those of the actors who construct the category. In the case of ad hoc categories, this distinction is not relevant because users construct their own categories. But in the case of enduring, collectively shared categories, the situation is more complex as users’ goals may differ from those of producers. Returning to Barsalou’s (1985) example of foods to eat on a diet, which he conceived of as an ad hoc category constructed by individuals desiring to lose weight, this category differs from the diet foods market category. The latter reflects a degree of consensus among market actors, yet the goals of users (to lose weight) differ from the goals (or interests) of food and dietary supplement producers or authors of books on dieting, who stand to benefit economically from consumers’ interest in dieting. Accordingly, we refer to the goals of the category from the point of view of users of the category, and we distinguish these goals from the interests of other actors who might benefit from the category’s use.

Finally, building our understanding of goal-derived categorization can help reconcile debates in the extant literature about the role of producers and non-producers in category construction. Grodal (2018: 785) lamented a “division of labor” portrayed in prior work, whereby producers are said to strategically and symbolically make claims and non-producers (often described as the producers’ audiences) evaluate these claims. ³ In contrast, she argued that non-producers “might do more than merely evaluate claims” (Grodal, 2018: 785). Similarly, Pontikes and Kim (2017) argued for more attention to non-producers’ ability to engage in strategic categorization. By helping us elucidate non-producers’“high capacity to influence” categories (Durand and Khaire, 2017: 95), a goal-derived categorization lens can help us better understand their role in the construction of new categories.

Methods

Empirical Context

We study the construction of the Quebec terroir products umbrella category and its most prominent member category, Quebec artisanal cheese. At the start of our study period, the concept of terroir as applied to local food products was entirely new to Quebec; similarly, the Quebec artisanal cheese product category was itself new, dating only to the 1980s. That the latter became the most prominent member of the former despite being initially unconnected to terroir—and that the Quebec terroir category became a shared, enduring “category in use”—makes this a revelatory and exemplary case (Yin, 2009) in the study of umbrella categories.

The construction of these two categories took place during a period of great social upheaval, which must be viewed in the context of Quebec history. Social, political, and economic inequalities between English and French Canadians had spurred the rise of a Quebec nationalist movement in the 1950s (Gundlach and Neville, 2012). The movement produced a change of identity among the French-speaking population of Quebec from “French Canadian” to “Quebecer”—Québécois in French—and intellectuals and activists promoted the Québécois identity to widen the perceived gulf between French- and English-speaking Canadians. Their discourse emphasized Quebec’s pre-conquest past and, by extension, Quebec’s historical links to France (Maclure, 2003; Oakes and Warren, 2007). Ties between the colony of New France and its mother country had been broken after the British Conquest of 1759, and the two societies had evolved along different paths for the next 200 years, resulting in a break with French traditions, including culinary traditions. Intellectuals celebrated the early life of French settlers, claiming that any traditions appearing after the 1759 defeat were of foreign origin and suspect (Maclure, 2003). The nationalist movement culminated in two referenda on secession, the first in 1980 and the second in 1995, the latter resulting in a secessionist defeat by only 0.6 percent. In 2006, the Canadian Parliament passed a motion recognizing the Québécois as a distinct nation within Canada, and secession remains an ongoing debate in the province.

Although it is plausible that Quebec society, by virtue of its French origins, might have already had a tradition of fine cheese consumption, this was not the case. Prior to the founding of the first artisanal cheese producers in Quebec in the 1980s, there had been no culture of fine cheese production or consumption in the province. While there had been some cheese production in the French tradition during the early years of New France, dairy consumption in the colony had been largely limited to milk and butter (Lambert, 2006; Desloges and De Courval, 2009). After the surrender of New France to the British and the ensuing influx of British Loyalists from the south, cheddar production grew sharply, particularly fueled by demand in England (Fournier, 1994). Even among the cheddars produced in Quebec, the stronger tasting variants were exported, and Quebecers preferred the milder variants for local consumption (Fournier, 1994; Dubuc, 1996). Save for a few cheeses produced by monastic orders, mild (i.e., bland) industrially produced cheddar remained the norm at the start of our study period.

Rather than being a longstanding tradition, and similar to what happened in other parts of North America (Paxson, 2010, 2012), artisanal cheese production in Quebec emerged as part of the back-to-the-land movement, and artisanal cheese producers were highly motivated by a rejection of industrial capitalism. Whereas in “the 1960s and 1970s, such persons sought self-realization through the counterculture; in the 1980s they came to seek self-fulfillment through self-employment” (Paxson, 2012: 67). As we show below, the first artisanal cheese producers appeared in Quebec during the same period, starting in the 1980s. The number of artisanal cheese producers over time is plotted in Figure 1.

Figure 1.

Annual Count of Quebec Artisanal Cheese Producers

Similarly, the concept of “terroir” was itself new to Quebec, entering popular discourse only in the late 1990s. Terroir is described generally as “the taste of place” (Trubek, 2008) and is well established in Europe, having contributed to food becoming a vehicle of national identity on that continent (DeSoucey, 2010). Perhaps best known in relation to wine (e.g., Bourgogne or Bordeaux), it is also applied to a wide variety of gastronomic product categories such as Feta cheese, Kalamata olives, and Serrano ham. The concept elevates the importance of place of origin, drawing attention not only to regional natural environments but also to regional cultures and traditions (Barham, 2003). In Europe, the concept has developed over the past 100 years and is codified and regulated, for example as set out in European Commission Rule 1898/2006 on the protection of designation of origin. In contrast, at the start of our study period, Quebec society did not consider its food as particularly distinctive in any way (Bizier, 2012), and the term terroir was a novelty (Lemasson, 2012). No rules or regulations comparable to those in Europe were in place, and the category was referred to only in highly general and aspirational terms. Notably, Quebec terroir was frequently referred to in the singular, ignoring the fact that the large territory of this province contains a number of distinctive regions and micro-climates.

Data Collection

Table 1 provides a summary of the data collected. We initially collected data on Quebec cheese producers, a product category we understood to be a terroir food. However, we soon realized that this category had not initially been connected to the emergent terroir category and that cheese producers had not been the primary exponents of the terroir category. We thus expanded our data collection both outward in scope and backward in time by collecting data first on the market intermediaries surrounding artisanal cheese (retailers, food critics, journalists) and then on actors more removed from market interactions—macro actors—who our initial data collection suggested were active in constructing the terroir category. Each group of actors required a different set of data, as we describe briefly below.

Table 1.

Summary of Data Sources

Actor Group	Description
Macro-level actors
Chefs and culinary institute	Books and newsletters (7), internal course materials and working documents (10)
Government of Quebec	Interviews: Project leader for the Quebec regional foods program (1), artisanal cheese sector assignee (1). Archival materials: Meeting minutes, correspondence, and working documents (20: 202 pages), press releases and publications (13), legal and regulatory documents (17), program evaluation reports (4: 242 pages)
Solidarité rurale du Québec	Annual reports (5), reports and books (13: 764 pages), conference proceedings (3)
Market intermediaries (i.e., journalists, writers, retailers)	Interviews: Cheese retailers (6), cheese distributor (1), food authors and tourism experts (3), dairy producers’ association (1) Books: Quebec cheeses (18), Quebec terroir products (17) Newspaper articles: Articles containing Quebec cheese mentions (585), articles containing Quebec terroir mentions (1,051) Other: Book about role of media (1)
Cheese producers	Interviews with cheese producers (17), reports (5)

Our data on cheese producers derive primarily from semi-structured interviews. Informant selection began by gathering the names of all producers from a government-issued list of small-scale milk-processing permit holders. We verified the websites of all the firms and dropped from the list those that were not artisanal cheese producers. We then sent invitations to all the names on the list and ultimately interviewed 17 artisanal cheese producers located across the province, in each case visiting them at their facility or home. (Producers generally lived next to or near their production facilities.) Interviews generally lasted between 60 and 90 minutes and were transcribed and manually coded.

Our data on market intermediaries comprise books, newspaper articles, and semi-structured interviews. We collected all books pertaining to Quebec terroir products and Quebec cheeses published between 1980 and 2013. Of the books on terroir products, all discussed Quebec cheeses, which frequently represented the most important category of foods discussed in the books in terms of the number of pages. We also collected newspaper articles mentioning Quebec cheeses and Quebec terroir. Using the Eureka database, we searched Quebec’s largest French language daily (La Presse), which was the earliest indexed French language newspaper (starting in 1988). We conducted semi-structured interviews with cheese retailers, food writers, and one of two dedicated artisanal cheese distributors in the province.

Finally, the macro actors active in the construction of the terroir category produced considerable archival data. We began with Solidarité rurale du Québec (SRQ), a not-for-profit activist organization that had been a vocal proponent of terroir during the late 1990s and early 2000s. We realized that the SRQ’s efforts had been an extension of an earlier government program led by the Quebec Ministry of Agriculture, which itself was an extension of a project first implemented by the province’s main culinary institute—Institut de tourisme et d’hôtellerie du Québec (ITHQ)—and chefs’ association—Société des chefs, cuisiniers et pâtissiers de la province de Québec (SCCPQ)—in the 1970s. As the culinary institute is a public entity, its older archives, as well as those of the Quebec Ministry of Agriculture, were stored at the Bibliothèque et Archives nationales du Québec (BAnQ), the provincial archives. We also submitted access-to-information requests to the Quebec government for documents stored in its internal archives and visited the Ministry of Agriculture’s library. Our searches yielded meeting minutes, strategy reports, course notes, newsletters, and other documents pertaining to the culinary institute’s and the government’s activities as early as 1977. We also collected publicly available government documents, including those related to the development and implementation of laws and regulations. Meanwhile, the SRQ had been prolific, and its publications, including its annual reports, were collected from university and public libraries across the province. We also interviewed the individual who had led the government program that we discuss below.

Data Analysis

Our analytic approach was one of abduction, as surprising and unexpected initial findings forced us to suspend our initial notions and motivated our subsequent inquiry into the phenomena at hand (Shepherd and Sutcliffe, 2011). As we explained above, we had initially focused on cheese producers but soon expanded our data collection and analysis to include market intermediaries and ultimately macro actors. For each actor group, we analyzed the data in Table 1 following the steps recommended by Langley (1999) and used by Delmestri and Greenwood (2016). We began by constructing an event history to identify the actors involved and the major events that occurred at various points in time, starting from the first references to the notion of a Quebec terroir (as a term or concept) we could find. Next, staying as close as possible to the words and phrases of our 30 interviewees and the archival materials, we proceeded with open, in-vivo coding (Corbin and Strauss, 2014) of our data for each actor group, looking particularly for references to cultural heritage and geography. ⁴ Following the advice of Grodal and Kahl (2017), we sought to remain sensitive to the diversity of actor groups involved in the category construction process and to take into account the distinct contexts and interests of these actors.

Concurrently, we engaged in temporal bracketing to identify any “discontinuities” in the process of category construction (Langley, 1999: 703). This revealed four critical junctures in the flow of events: the successive entries of three macro actor groups—hospitality/culinary organizations (ITHQ and SCCPQ), the Quebec Ministry of Agriculture, and the SRQ—as the primary proponents of the Quebec terroir category followed by the entry of market intermediaries (and to a lesser extent cheese producers) in the nesting of artisanal cheese as a terroir product.

We then sought to group our in-vivo codes into second-order themes that pertained to our general research question, a process that proceeded through several iterations to discern groupings of first-order codes that were both meaningful and discriminant. In the third step of data coding, we considered whether the emerging themes suggested concepts that might help us describe and explain the phenomena of interest (Gioia, Corley, and Hamilton, 2013). We did this by juxtaposing our second-order themes with existing theory so as to generate more-abstract aggregate dimensions (Langley, 1999; Delmestri and Greenwood, 2016). The result was a three-tiered data structure for each actor group (Gioia, Corley, and Hamilton, 2013).

Two aspects of this data coding process are notable with respect to macro actors. First, as we followed the coding process outlined above, we were struck by strong similarities in our codes across the three macro-actor groups. For example, as each group successively took the lead in attempting to define and implant the terroir category during its respective period, each called for and subsequently engaged in “inventory taking.” This led us to realize that rather than successive stages of a longer process, we were witnessing recurring cycles of a process repeated three times. As such, our development of second-order themes for macro actors involved the constant comparison (Corbin and Strauss, 2014) of codes within each cycle and across cycles. For example, we asked ourselves questions such as “taking inventory of what?” and “why again?” for each set of macro actors.

Second, juxtaposing emerging second-order themes for macro actors with extant theory in order to develop third-order dimensions led us to revise and expand our theoretical scope. For example, the term “inventory taking” stood at odds with how extant theory tells us that categories are constructed, namely that categories emerge as a function of audiences’ observations of similarity among ostensibly like objects (Negro, Koçak, and Hsu, 2010). Instead, in our context, we were seeing macro-level actors first defining (or redefining) the category at the start of each cycle and only afterward scanning the environment (i.e., taking inventory) to fill the predefined category. We came to the realization that while the extant theory in organization studies focuses almost exclusively on prototypical categories, the sequence we were witnessing appeared consistent with theory on goal-derived categories (Barsalou, 1991).

With our coding structure for each type of actor in place, we sought to induce “a process model that is both grounded in and sensitive to our case’s particular historical evolution while still being generalizable to other settings” (Croidieu and Kim, 2018: 11). We developed this model by iterating between theory on goal-derived categories and our coding structure, all while taking into account the broader cultural environment within which the process unfolded (Grodal and Kahl, 2017).

Findings

In line with our research question of how umbrella categories are constructed and come to nest product categories, we present our results from the perspective of the Quebec terroir products category. As we explained above, we identified three sequential attempts at constructing this category led by three different sets of macro actors. We refer to these attempts as cycles because they repeated the same series of actions: envisioning, populating, and encouraging adoption of the nascent umbrella category. The first cycle was led by Quebec chefs, who were represented by a hospitality institute and a culinary association that sought to establish a cuisine that would “taste of Quebec.” The second cycle was led by the provincial government, which picked up where the chefs left off. Both these attempts garnered few adherents and largely failed to get traction outside of policy circles. The third cycle was initiated by the SRQ, whose efforts resonated with market intermediaries. We describe how these intermediaries nested artisanal cheese under the terroir umbrella and how their efforts were ultimately accepted by cheese producers, resulting in the Quebec terroir products category becoming a material “category in use.” We present our accounts of these unsuccessful and successful attempts at umbrella category construction, which we draw on later to develop a theoretical model of umbrella category construction and product category nesting.

Unsuccessful Attempts at Constructing the Quebec Terroir Umbrella Category

Hospitality institute and chefs’ association begin the process

The first attempt to construct a Quebec terroir products category was initiated by Quebec chefs. Table 2 provides the coding structure and code descriptions for these and the two subsequent macro actors, and Table A1 in the Online Appendix provides illustrative evidence pertaining to these macro actors. (The appendix is structured to provide evidence for the same codes in each of the three cycles.) Historically, French Canadians did not conceptualize their foods as fine or distinctive but rather as rich, hardy, working-class fare (Bizier, 2012). Fine restaurants were virtually nonexistent in Quebec, and those that did exist were rarely staffed by French Canadians, who generally believed cooking was done in the home by the mother (Bizier, 2012). Practicing chefs looked down on traditional Quebec dishes and rarely served them, preferring instead to mimic fine dining in France. Quebec chefs, with the active support of journalists and the provincial government, embarked on a professionalization project in the 1960s. The Institut de tourisme et d’hôtellerie du Québec (ITHQ) was founded in 1968 by the provincial government as a training institute for the hospitality industry. Chefs sought to embed their professionalization project in the rising tide of Quebec nationalism, arguing that traditional Quebec cuisine was both distinctive and on par with that of every other great nation—which is how Quebec society was beginning to regard itself (Bizier, 2012).

Table 2.

Coding Structure for Macro-Level Actors Constructing a Goal-Derived Category

First-Order Codes	Second-Order Themes	Aggregate Dimensions
Motivating the goal “tasting Quebec”: Elaborating the meaning and means of “tasting Quebec.” The category of foods deemed to be reflective of Quebec culture developed and evolved over time as new macro-level actors became involved.	Motivating the goal-derived category	Envisioning category: A forward-looking stage of category development in which the goal of the category is being elaborated. While some general orienting principles may be proposed, membership in the category is still ambiguous, and the category remains largely empty.
Elaborating ideals for category membership (i.e., qualities of foods that taste of Quebec): Providing lists of ideals that potential members of the category should approach. The ideals served as a general orienting tool to help guide the identification of products that should be included in the category. The ideals often lacked specificity and were discussed in terms of general values.	Elaborating ideals
Collecting: Any act of building an inventory of existing organizations or products that appear to be candidates for the category. Vetting: Efforts to classify or select between the artifacts identified during the collection phase. Inventing: Conducting research and development activities to create entirely new artifacts that can serve as exemplars in the new category.	Compiling	Populating category: Efforts directed at filling the empty category with exemplars identified from organizations or products already in existence, or through the development of entirely new exemplars. The exemplars serve to educate practitioners and the public in the next phase.
Involving stakeholders in collecting: Gaining the involvement of a broader range of stakeholders in the scanning for potential category members. Involving stakeholders in vetting: Gaining the involvement of a broader range of stakeholders in the approval and recognition of the results of the compilation process. The actors involved in the vetting process were also later involved in encouraging the adoption of the category.	Engaging stakeholders
Educating practitioners: Informing the industries involved about the new category and how it is to be used. Educating ranged from formal training programs to the publication of documents listing exemplars to encourage reproduction. Educating the public: Informing the public about the new category by promoting exemplars in advertising.	Educating	Encouraging adoption: Efforts to encourage the use of the goal-derived category by existing market actors or through the founding of new organizations. Often, the descriptions of the category remained obscure and relied on the presentation of exemplars to provide meaning.
Certifying: Development or application of certification programs. Certification was seen as a mechanism to incentivize the participation of the industry, as it would provide certified organizations a means to differentiate. Celebrating: Development and granting of awards that celebrated the best performers.	Incentivizing
Subsidizing: Any effort to reduce entry costs for producers. Building infrastructure: Any effort to develop or encourage the development of the supply chains and relationships required for the sector to develop.	Enabling

However, the ITHQ and the Société des chefs, cuisiniers et pâtissiers de la province de Québec (SCCPQ), the major association representing chefs, recognized that this initiative was hampered by the lack of recognition of traditional Quebec dishes. For example, an ITHQ professor explained in a book that “it would be sad if, due to snobbism, we refuse to recognize the undisputable merits of our cuisine. Our hoteliers and restaurateurs should make it a point of pride to include each day on their menus a Quebec dish that reflects their region” (ITHQ, 1977). Likewise, another instructor at the ITHQ explained that “in all the countries of the World, regional cuisine plays a primordial role in the tourist industry. It is the same here. Except, we do not give it the importance it deserves. What is worse, we cheat on its authenticity in a desire to present a more continental [i.e. European] cuisine” (ITHQ, 1980). As part of their professionalization project, chefs sought to take traditional recipes and rejuvenate them with an increased use of fresh, local ingredients, while maintaining their apparent authenticity.

The chefs thus envisioned a category of foods that would have a distinctive taste of Quebec and sought to develop recipes that would incorporate local agricultural products. Chefs were inspired by French nouvelle cuisine, which used lighter modes of preparation with fresh, local ingredients, permitting the flavors of the local environment to shine through (Rao, Monin, and Durand, 2003). The reinvented Quebec cuisine would have a distinctively Quebec flair:

What the tourist is looking for here is the Quebecer, his manner of living and eating. Give him what he wants, while being purely and simply what we are, Quebecers. I believe that Nouvelle-Cuisine fills this lacuna, because it is regional by the provenance of its products, but, most of all, by who prepares it. (ITHQ, 1977)

The idea was to develop a cuisine that would represent “the Quebecer” through the use of ingredients from Quebec’s regions. This cuisine would be an expression of Quebec culture (“what we are”), and chefs would be (“most of all”) the purveyors of that culture. The dishes served in each region would be based on local food traditions and use distinctive local crops. ⁵

It is important to emphasize that the distinctive local dishes and ingredients associated with the concept of terroir were still yet to be defined. For example, an ITHQ instructor lamented the general ignorance of the foods that should be considered representative of Quebec: “for a few years we are noticing that in the elaboration of new recipes, we find apples and cider in 80% of them. Before our Nouvelle-Cuisine becomes overloaded with apples, I think it is time to look at other possibilities” (ITHQ, 1977). Quebec is a large producer of apples, and the instructor bemoans the over-reliance on that single ingredient by chefs wishing to project cultural distinctiveness. Similarly, the attributes of the dishes were yet to be defined, and the dishes were described using ideals and general statements, such as “representative of our culture” or using “fresh” ingredients.

The chefs therefore began work to populate the new terroir-based Quebec cuisine with recipes. Without actually delimiting Quebec’s regions according to differences in its terroirs (i.e., based on differences in local climates, soils, and traditions), chefs simply divided Quebec according to its administrative regions as proxies for its different terroirs. The ITHQ then conducted multiple, expanding efforts to catalogue the traditional recipes and ingredients used in each region. The first effort to catalogue traditional recipes was conducted internally by ITHQ professors and students. Later, the entire population of Quebec was called upon to submit traditional recipes, which were reclassified according to region. This was a major undertaking, and 30,000 recipes were received. ITHQ students were even sent to old age homes across the province to gather traditional recipes. Concurrently, ITHQ’s food research and development center became involved by inventing new dishes and modifying heavier traditional dishes to make them lighter and using fresher ingredients.

The ITHQ and the chefs’ association (SCCPQ) then engaged in a number of efforts to increase awareness and encourage adoption of this terroir-based categorization of Quebec cuisine. A course on Quebec cuisine was added to the ITHQ’s curriculum for the first time in 1978, and its research center developed a database of new recipes that was made available to practicing chefs. Cookbooks were published containing selections of the so-called traditional recipes collected during the previous phase. (These were not the original recipes as received but modified versions, lighter in preparation and utilizing ingredients said to be representative of the associated regions.) And notably, the SCCPQ sought to develop a certification program for restaurants across the province: restaurants would be certified if at least 70 percent of their ingredients originated from Quebec and 50 percent originated from their local regions.

However, the challenges proved too great, and the SCCPQ was unable to launch the certification program on its own. A lack of interest in Quebec cuisine among practicing chefs was identified as a major concern, but there was also the practical matter of a lack of access to local ingredients. Historically, there was little variety in Quebec agriculture, and what fresh ingredients did exist were generally unavailable during winter months (Bizier, 2012). Moreover, given the structure of supply chains, at times it was easier to access certain ingredients in urban centers than it was in the regions where they were produced (ITHQ, 1977). These obstacles were too large for the chefs to solve on their own, and they approached the Quebec government for help (Coulombe and Jutras, 2014), leading to a second cycle.

Quebec government becomes involved

The control of the certification program was transferred to the Quebec government’s Ministry of Agriculture in 1988. With the SCCPQ as an equal partner, the Ministry envisioned the category in much the same way as the chefs did, as a cuisine that tasted of Quebec through a combination of distinctive local products and culinary traditions. But the government brought a much greater focus on the ingredients that would be used in this cuisine. Whereas the SCCPQ would continue to work on encouraging chefs to prepare dishes representative of Quebec culture, the Quebec government would focus on identifying the local ingredients deemed representative of each region and resolving supply chain issues to ensure chefs could gain access to them. From the government’s point of view, the “chef/producer partnership is essential for the development of a regional identity” (Ministère de l’Agriculture, des Pêcheries et de l’Alimentation, 1997). For the government, this was important because of the potential economic impact of increasing demand for local agricultural products:

The cuisine of the terroir is made from products, habits and the culture of a given region. It is the cuisine of regional traditions. . . . The regional cuisine [program] was born out of a desire on the part of the Ministry of Agriculture to encourage among restaurateurs and chefs of the entire province, the use of Quebec products IN PREFERENCE TO THOSE COMING FROM ABROAD. In this choice of definition of regional cuisine, based primarily on the utilization of products from here, the producers are implicated more than the chefs. (Ministère de l’Agriculture, des Pêcheries et de l’Alimentation, 1997; emphasis present in the original text)

As the quote suggests, agricultural producers would have an elevated importance in the process of developing a Quebec cuisine. As in the first attempt to define the terroir category, however, the attributes of the agricultural products that would be considered representative of Quebec culture were largely undefined and were described using ideals such as having high quality or local distinctiveness.

Like ITHQ and SCCPQ had done in the previous attempt, the Ministry thus established committees across the province to populate the category. These committees worked to identify the agricultural products said to be representative of each region and to suggest how these products could be used to make dishes distinctive. Similar to what had been done previously, the government simply divided the province according to administrative regions rather than attempt to identify distinct terroirs according to physical or cultural elements (i.e., regions having distinct soils, climates, and cultures). ⁶ A structure was established comprising regional committees working under a provincial committee. The regional committees worked with local actors to develop specification sheets identifying and classifying the agricultural products of each region. The regional committees then submitted their lists to the provincial committee for approval. Figure 2 illustrates a provisional specification sheet for one such region. As shown, products were distinguished between targeted, important, and limited categories. Those listed under the “targeted products” category were not only produced in large quantities but also described as representing local traditions and with good potential in the hospitality industry. At this stage, however, there was still no mention of artisanal production. Even the products in the targeted category were generally industrially produced. As can be seen in the figure, this particular region had a well-developed dairy sector, to the point that the regional committee identified many cheeses as targeted products, all of which were industrially produced.

Figure 2.

Sample Specification Sheet*

* The document was translated from French, while keeping the appearance of the original.

The provincial government took a number of measures to raise awareness and encourage adoption of the proposed category of terroir foods and in particular the use of the ingredients it deemed representative of each region. While the emphasis was put on ingredients, dishes remained the primary vehicle for tasting Quebec, and chefs remained at the forefront of the new category. The first cohort of restaurants participating in the certification program initially created by the SCCPQ chefs was announced in 1990. Coinciding with the launch of the certification program, the Ministry of Agriculture worked with a Quebec grocery store chain to install kiosks in its stores to present targeted products. The kiosks were intended to introduce consumers to the products said to be representative of their own regions. Somewhat ironically, given that the products in question were said to derive from local traditions, the kiosks were staffed by chefs who would show customers how they could prepare recipes using these products. The Ministry of Agriculture also enlisted the Ministry of Education to raise awareness of regional cuisine among agricultural producers.

In 1993, the government also launched the “Development Projects Concerning the Value-Added of Local and Regional Products,” a subsidy program designed to encourage the development of new local agricultural products that included advisory support from governmental regional development officers. ⁷ Relatedly, in 1996, the government enacted a law allowing for reserved designations. Modeled after the reserved designation system in Europe, whereby names of product categories (e.g., Champagne wine, Feta cheese) are protected, the law was intended to permit producer associations in Quebec to obtain legal protections for the names of their product categories. (Regulations allowing the implementation of this law were not put into place until over a decade later.)

As in the case of the first attempt to define the Quebec terroir category, this cycle led by the government also ended with little tangible impact. An evaluation of the restaurant certification program conducted in 1993 concluded that after three years, the program “has so far generated little economic impact, either for the restaurant members of the network or for the Quebec and regional bio-food sectors” (Ministère de l’Agriculture, des Pêcheries et de l’Alimentation, 1993: 37). This was because chefs either did not know of, or still had no access to, the agricultural products designated by the Quebec government. What the two unsuccessful cycles also appear to have in common is that they both elevated the role of chefs as the carriers of Quebec cultural traditions, whereas the agricultural producers, upon whom chefs depended for fresh ingredients, remained largely anonymous. In both cycles, macro actors sought “to adapt [agricultural] production to the needs of the restaurant and hotel industries . . . [and] to showcase the professions of chefs and pastry chefs” (Société des chefs, cuisiniers et pâtissiers de la province de Québec et Ministère de l’Agriculture, des Pêcheries et de l’Alimentation, 1990: 3). Agricultural products were not valued in and of themselves but only for what chefs could make of them. In other words, even though chefs relied on agricultural producers for greater availability of fresh, high-quality ingredients, chefs would be the ones to capture any increase in recognition and status deriving from use of the category. Without the engagement of a larger array of actor groups, the category Quebec terroir foods remained largely at the policy level and little used by market actors.

Successful Attempt at Constructing the Quebec Terroir Products Umbrella Category

Solidarité rurale du Québec becomes involved

Following the second attempt to gain traction for the notion of a category of Quebec terroir products, SRQ became the primary proponent of the nascent category. The SRQ had been formed in 1991 when chambers of commerce, tourism boards, trade unions, and activist groups across rural Quebec joined forces to critique the Quebec government’s policies on rural development. Its inaugural general assembly was attended by over 1,200 delegates concerned about the depopulation, economic stagnation, and social decay of rural communities due to globalization (Solidarité rurale du Québec, 1991). In 1997, the Quebec government named the SRQ an official advisor on the topic of rurality, an interesting development given that group’s vocal criticism of the government.

The SRQ envisioned the concept of terroir somewhat differently from how it had been envisioned during the earlier cycles. It placed Quebec artisanal products and producers at the center of attention and specified that terroir food products should be artisanal in nature. By 1998, the numbers of artisanal food producers had increased across the province, and the SRQ saw in them a new means to protect against the dangers of globalization: “terroir products are a response to the standardization of mores and products, a reaction to the effects of globalization” (Solidarité rurale du Québec, 2000). As were culinary dishes before them, artisanal food products were described as expressions of Quebec identity and culture, but this time artisans became the purveyors and custodians of that culture: “in the face of the standardization of everyday consumer products, the comparative advantage of traditional know-how is certainly its uniqueness. The production of terroirs is emotionally charged, because it is a matter of identity” (Proulx, 2001). The SRQ promoted the idea that artisanal food products were terroir products expressing local traditions and described them as having been produced using know-how passed down through the generations—even when this was not necessarily the case.

Despite the increased focus on artisanal production, however, the attributes of Quebec terroir products remained largely undefined. At a conference organized by the SRQ in 1998, it was apparent that most of the so-called traditions were, ironically, yet to be created:

Tradition is to be unpacked, ingenuity to be cultivated and palates to be seduced. . . . We can give ourselves the objective to take all the products of the soil and try to Quebecify them, that is, give them a local flavor . . . to give an identity and distinctive attire to products, resulting in a different taste, a unique mode of production and packaging representative of here. (Solidarité rurale du Québec, 1999: 30)

Despite the language of tradition and cultural distinctiveness, the Quebec population would need to be convinced to actually consume ostensibly traditional products. Moreover, the products themselves were yet to be defined and were described using ideals and general statements: they resulted from “traditional know-how,” had a “relation to identity,” and were “authentic.”

As in earlier cycles, new rounds of environmental scanning were undertaken to populate the category, this time focusing entirely on artisanal products. The SRQ conducted a sweeping scan of all artisanal foods produced in the province in an effort to identify exemplars that met their criteria of being both culturally representative and distinctive to place. To classify artisanal products, SRQ distinguished between simply small-scale productions and terroir products, describing a terroir product as one that “valorizes the potential of local natural and cultural resources. Its form or use results from the transmission of traditional know-how and a chain of production. The terroir product combines a unique resource, unique manufacturing processes and is associated with a delimited and named territory” (Solidarité rurale du Québec, 2002a).

In its effort to populate the terroir category, the SRQ turned to Quebec artisanal cheese. Although many types of artisanal products were being increasingly produced in Quebec at the end of the 1990s (i.e., wines, meats, fruit preparations, etc.), artisanal cheeses had become the best known. Quebec artisanal cheese would become a recurring theme in SRQ publications as an exemplar of the Quebec terroir. For example, in one document, the SRQ defined the emergence of terroir products as a “cultural reflex” having “a strong symbolic charge and a marker of identity that expresses the specific relationships to the resources of a place, a region, a country” (Solidarité rurale du Québec, 2002a: 12). To illustrate this complex point, the document turned to artisanal cheeses as exemplars:

A few examples perfectly illustrate our point: cheeses with a long tradition, such as Bleu Ermite, Oka, and those of the Ile d’Orleans, are the result of artisanal processes developed in their own environment, which were strictly passed down through time and which remain respectful of the raw material and the quality of the cheese. (Solidarité rurale du Québec, 2002a: 12)

The document went on to state that “it is common knowledge that the cheese-making know-how currently being developed in Quebec is the fruit of tradition” (Solidarité rurale du Québec, 2002a: 35). In another document, the SRQ would devote 14 pages to artisanal cheeses, beginning the discussion with the statement “it is almost impossible to trace the first cheese made in Quebec. But there is no doubt that this product has been linked for a long time to the eating habits of the inhabitants” (Solidarité rurale du Québec, 2002b: 65). Artisanal cheese became the primary exemplar to illustrate the ideal of terroir products as representing Quebec culture and tradition, thus allowing users to “taste Quebec.” But the SRQ’s claims of tradition stand in contrast to the assertions from cheese producers that they had created a market from scratch, struggling against consumers’ ignorance. Whereas the SRQ argued the existence of terroir by drawing an unbroken line of cheese making know-how stretching across centuries of Quebec history, there was in fact no tradition of fine cheese making in Quebec past the Conquest of 1759 apart from a few monastic cheeses, and cheese producers saw themselves as pioneers facing general ignorance of and even resistance to their products, as we discuss further below.

Finally, the SRQ worked to encourage adoption of its conceptualization of the Quebec terroir products category. The SRQ provided training to governmental regional development officers on how to identify and promote the artisanal products available in each region. The SRQ’s efforts to popularize the terroir concept coincided with a second subsidy program launched by the Quebec government in 1998 to encourage market entries of firms producing terroir products. This time the subsidy program was named the “Measure of Support for the Development of Terroir Products.” ⁸ But even though the SRQ was mandated by the Quebec government to instruct regional business development offices about how to identify and best promote terroir products, its language remained highly conceptual and abstract. As in the previous cycle, the report evaluating the subsidy program concluded that it had produced few tangible impacts at the market level. The funds allocated to the subsidy program had not been entirely disbursed, and even in 2004, despite the SRQ’s efforts to disseminate the concept of terroir, the report concluded that the meaning of terroir remained obscure to many working within the field (Cimon-Morin, 2004).

Intermediaries project goal-consistent identities

While macro actors in each cycle produced a slightly different iteration of the umbrella category, only the third iteration was embraced en masse by market intermediaries: journalists, food critics, retailers, and distributors. As Figure 3 shows, media mentions of terroir rose dramatically starting in the year 2000. The early 2000s were also a period that saw numerous entries of retailers and distributors of artisanal food products. For example, one of the best-known stores dedicated to selling Quebec terroir products in Montreal, the largest city in the province, opened its doors in 2000. Previously, it had been quite difficult to purchase artisanal food products in urban centers. In an interview, the store’s founder stated that she was motivated by the idea that “every nation should have a gastronomy of its own” (retailer, interview 17). ⁹ Much of the excitement surrounding terroir products was driven by excitement about artisanal cheeses, as indicated by the content of the books published on the topic of terroir. Our searches revealed that all terroir books contained a section (often the largest section) on artisanal cheeses, while they varied greatly in terms of the other categories of foods they covered. As Figure 3 shows, the rise in interest in Quebec terroir went hand in hand with an interest in artisanal cheese. Likewise, virtually all stores selling terroir products sold artisanal cheeses.

Figure 3.

Annual Numbers of Books and Newspaper Articles Published on Quebec Terroir and Quebec Artisanal Cheeses*

* The dramatic increase in the number of newspaper articles mentioning Quebec cheeses in 2013 was due mainly to a free trade agreement being negotiated at the time between Canada and the European Union. As part of the agreement, the importation of French cheeses into Canada was expected to rise, and there was fear that this could threaten the Quebec artisanal cheese sector.

Our analysis of media accounts and interviews with retailers and distributors revealed how intermediaries linked the abstract language of terroir proposed by the SRQ to the artisanal food products that were to be subsumed under the umbrella category. The coding structure and supporting evidence for this analysis are provided in Table 3.

Table 3.

Evidence for Intermediaries Projecting Goal-Consistent Identities

Representative Quotations	First-Order Codes	Second-Order Themes	Aggregate Dimensions
“Do yourself the pleasure of meeting them. They will tell you about their craft, their products, their values, their terroir.” (Foreman, 2012)	Inferring producers’ attachment to Quebec soil	Ascribing motives to producers	Projecting goal-consistent identities
“This [cheese] route is a tribute to the producers who transmit through their cheeses the multiple and unique terroirs of Quebec.” (Arnaud, Boudreau, and Fortier, 2010)
“Over the past few decades, we have seen the emergence of a variety of high-quality cheeses that have made Quebec one of the leaders in the art of cheese making in the world. All the artisans who contribute, through their creativity and their talent, to composing this heritage and spreading our culture deserve to have their products recognized and integrated into our culinary tradition.” (Tendland, 2012)	Inferring producers’ attachment to Quebec culture
“While travelling around, I noticed that Quebec cheesemakers are somewhat ADVENTURERS. Adventurers, because it is in their genes to set out on adventure, to achieve, to discover! . . . Of course, they are not the only ones who want to change things. They are only some among thousands of others who wish to make our land a unique place on the continent. And I am convinced that they are succeeding.” (Roiseux, 2002)
“In all this effervescence of small-scale cheese production, we can understand that there was a desire by the Quebec economy to protect itself from globalization, where the only way to survive is to offer mass-market products like [industrial producers], or a niche product, so this was a way of defending against global markets.” (Retailer, interview 27)	Inferring producers’ attachment to fellow Quebecers
“If [a region of Quebec] defines this notion of terroir well, assuming that they are able to say what makes this region so special, I think someone in the city will give it more credibility. So it’s associated a little bit with wealth. I call it creating wealth. If we can sell this identity, say we are proud, we are from that region, because it is defined by a clay soil, and a flower typical to our region, I extrapolate, but someone who is more of an urbanite will say that ‘I will buy it because it corresponds to an ideal,’ or there is an emotional attachment at different levels for this consumer who is trapped in a zone made of concrete . . . but yes I think it can lead to certain qualitative concepts that are justified and also marketable.” (Retailer, interview 20)	Activating consumers’ attachment to Quebec soil	Activating attachments of consumers
“People come in to buy local . . . local for them means Quebec. Consumers know that when they buy local, they are protecting their terroir. They want a product that is specific and represents the taste of Quebecers and not a universal product that would please everybody.” (Retailer, interview 27)
The “international reputation [of Quebec cheeses] does not cease to grow and, with this well-deserved success, the ‘Land of Québec’ is on the way to becoming the high place of fine cheese production on North American soil.” (Bizier and Nadeau, 2004)	Activating consumers’ attachment to Quebec culture
“We made cheese all along the [Saint-Lawrence] river. And we lived through the first death of our cheese heritage with the arrival of the [British] loyalists . . . here, we always had a desire to create, to make our own cheeses, and we came back to that . . . and why did it come back? It has partly to do with evolution, the emancipation from the Anglo-Saxon culture. . . .” (Retailer, interview 21)
A retailer explained the importance of telling clients about Quebec cheese makers: “To show to people, ‘yes!’ without becoming chauvinistic, to say ‘yes, we do things, we do beautiful and good things.’ There are people who have talent, and then it’s just a matter of giving them a chance, period. That was the only thing they were asking for these people: ‘Give us a chance. Period.’” (Retailer, interview 18)	Activating consumers’ attachment to fellow Quebecers	Activating attachments of consumers	Projecting goal-consistent identities
“Pride is to pay . . . to invest in integrity, origin, in the regional, and to know that ‘when I buy, I vote. I buy, I encourage the economy, and I know who I help. I know who benefits from my money.’” (Retailer, interview 21)

Intermediaries nested artisanal cheese under the terroir category using two main strategies. First, they ascribed to artisanal cheese producers an identity consistent with the discourse surrounding the umbrella category. Specifically, they ascribed to producers a patriot identity, which was consistent with the nationalist discourse surrounding terroir. Second, they activated the same national identity among consumers to reframe the act of purchasing artisanal cheese as a patriotic act. With respect to producers, intermediaries inferred producers’ attachment to Quebec soil, Quebec culture, and Quebec people (see Table 3). For each theme, intermediaries framed the producers as altruistic and ascribed to them motives that the producers themselves seldom discussed during interviews. First, producers were described as motivated by a deep-seated love for their soil and particularly by a desire to capture and express their soil’s unique qualities in their products. In the following quote, a cheese retailer represents cheese producers as entirely motivated by capturing the unique qualities of their terroir:

All these people, when we speak of artisans, what they want to showcase is the specificity of their terroir. In other words, the region where they produce, and what they want to produce, in other words, the finished product, which is the cheese. If you look at the goat cheeses, they will have their own goats, they will create their own cheeses to capture an aspect of taste that represents the region. (Retailer, interview 20)

During our interviews with cheese producers, those who raised their own animals consistently explained that doing so gave them greater ownership over the production process and was a source of differentiation with industrial producers. In contrast, the retailer quoted here reinterpreted the same practice so as to infuse the cheeses with a notion of territorial distinctiveness and ascribe to producers an overriding love for the land.

Second, intermediaries described producers as motivated by a desire to contribute to Quebec culture. In this highly evocative description of the daily life of agricultural producers, the author describes producers as not only working the land but also its culture:

To all those farmers who, every morning, with their hands, respectfully, knead the earth, the tradition and the landscape, living as they are by the demands of the terroir; the terroir that gives to the village a face, to the land a song of us; the terroir that escapes time, melding past and present, to feed all imaginations. (Dorion, 2003: 3)

The overlapping understandings of terroir as referring to soil and culture are evident in the quotation, as the author describes producers as respectfully kneading both the land and traditions. In doing so, the author suggests that producers are contributing to Quebec culture by adding to the distinctive “song of us.” It was particularly with respect to the cultural theme that artisanal cheese stood out among all other artisanal foods. Virtually every book discussing artisanal cheeses began with at least some reference to cheese production centuries ago in New France, whereby “from the beginning of the French colony” (Foreman, 2012), “inhabitants of New France produced all kinds of artisanal cheeses” (Tendland, 2012) that were “based on the recipes and traditions of their country of origin [France]” (Serre, 2005). These references to New France were consistent with the origin myth of society as an authentically French colony prior to the British conquest (Maclure, 2003). This was despite artisanal cheese producers telling us that they had started their market from scratch and had struggled to overcome the population’s biases against fine cheeses given the lack of a tradition of fine cheese consumption in the province.

Third, intermediaries represented producers as motivated by a desire to support their villages, even suggesting this as the original reason for the founding of their firms. Here, the founders of one firm are described as being motivated by a feeling of injustice for seeing their village not reap the benefits of the milk produced in the region:

The idea of opening a cheese dairy had been in the minds of the farming couple since the closure of the region’s dairy. Finding it foolish to see the milk go to the major centres of the province and then return processed to their village, the couple decided to embark on the adventure of cheese making. (Tendland, 2012: 76)

In contrast to our interviews with producers, which highlighted their desire for self-actualization, in this quote producers are characterized as selfless and working for the good of their compatriots, driven by the purpose of contributing to their community.

Having ascribed patriotic identities to cheese producers, intermediaries used the same themes of Quebec’s soil, culture, and people to activate consumers’ own patriotic attachments, as shown in Table 3. First, artisanal cheeses were described as representative of Quebec soil and thus a means of defining one’s place of belonging in the world. This place of belonging referred to Quebec in general and, more specifically, to the particular villages or regions within Quebec where consumers could trace their roots. According to a cheese retailer:

It is the signature, when we speak of cheeses of the terroir, it is the agricultural signature of a region. . . . There is pride when we speak of terroir. There is a pride to say it is where I belong, it is my place, in a sense, where I grew up, recognize myself. And the food we have in front of us is a reflection of us. It is like poetry. In other words, if I eat a cheese from [a region of Quebec] and I am from [that region] I say WOW! The animals ate the herbs, produced the milk, and today I have this cheese . . . people have this bond to the Quebec terroir. (Retailer, interview 19)

Intermediaries thus advanced the notion that, as per the notion of terroir, artisanal cheese carried the “taste of place” (Trubek, 2008) and thus strongly represented territory and provided the sense of belonging for consumers—a notion absent from the discourse of cheese producers themselves.

Second, artisanal foods were described as representative of the cultural distinctiveness of Quebec. Quebec society had long regarded itself as a distinctive Latin society surrounded (and threatened) on all sides by an Anglo-Saxon North America (Maclure, 2003). Intermediaries emphasized the uniquely Latin quality of the notion of terroir, defining Quebec society’s love for food as intrinsic to its culture while subtly denigrating the public’s relationship to food in the rest of North America:

There is no word in English to translate “terroir.” They use “local food,” but it is not the same. The Anglo-Saxons are interested in it largely from a perspective of health. That has nothing to do with the soil . . . terroir products have a soul and a cultural value. It is a very Latin concept. (Bérubé, 2009: PLUS4)

In the case of eliciting consumers’ attachment by virtue of cultural distinctiveness, artisanal cheese again stood out among all artisanal products. The growth of the artisanal cheese sector was itself presented as proof of the cultural distinctiveness of Quebec, as artisanal cheeses were described as the “family jewels [that are] the image of Quebec” (Deglise, 2005: D6). The following evocative quotation attributes a patriotic motive to cheese producers and activates the patriotic attachments of consumers. The author ascribes to cheese producers a motive to contribute to the heritage of Quebec and then, by deploying patriotic themes, calls on the public to support cheese producers in their quest.

Thanks to the perseverance and courage of its cheese artisans, Quebec is today recognized as the master of cheese in North America. . . . If importation remains high, Quebec products are taking a special place on the shelves of boutiques, and as long as the artisan is supported in his quest for a gastronomic heritage to leave to the land, we see coming a time when cheese from here will reign master in its home on all our tables. (Belmont, 2008: second cover)

The phrase “master in its home” (maître chez lui) is a variant of a phrase (maîtres chez nous, translated as “masters in our own home”) used during the rise of Quebec nationalism in the 1960s (Kaplan, 1989), indicating support for a wide range of provincial government policies designed to wrest greater independence from English Canada. Here, the phrase is used to rally support for Quebec artisanal cheese.

Third, the consumption of artisanal cheeses was described as supporting fellow Quebecers, be they the producers themselves or their villages. This quotation evokes the different notions of “local” contained in the concept of terroir:

We’re going to buy terroir products because they travel less, because we encourage people from here. Those who think about it will say, “Well, I preserve jobs, I preserve know-how.” It can go further, but I think when people buy locally, for them—or in small productions—it is a product of the terroir. I think it is in the imagery. (Retailer, interview 5)

The retailer begins by suggesting the consumption of terroir products is better for the environment due to geographic proximity, but he then invokes notions of identity as he describes the need to support “people from here.” These two understandings of local—geographic and identity-based—need not coincide, as many agricultural producers in the neighboring province of Ontario are geographically closer to consumers in Montreal than those dispersed across the vast province of Quebec. As such, the retailer valorizes the consumption of terroir products because doing so helps support fellow Quebecers and not necessarily because it is better for the environment.

Cheese producers accept the goal-consistent identity

Table 4 provides our coding structure and illustrative evidence for cheese producers’ discourse. Contrary to our initial expectation that cheese producers had been the primary proponents of the idea that their products reflected Quebec’s terroir, we were struck by cheese producers’ lack of mention or apparent interest in the terroir concept. Instead, consistent with Weber, Heinze, and DeSoucey’s (2008) account of grass-fed beef producers, we found that cheese producers largely defined their category in opposition to industrial cheese producers, not unlike artisanal cheese producers across North America (Paxson, 2012). Producers openly rejected the existence of any long-standing tradition of cheese consumption in the province and instead had a highly coherent discourse that emphasized two “oppositions”: opposition of craft to industrial practices and opposition of artisanal self-expression to the impersonality of industrial production. Cheese producers emphasized the manual labor involved in craft production, the use of quality ingredients, and direct contact by the artisan with every batch of cheese. In addition to these central and distinctive features of their craft practices, producers spoke of the experiential qualities of artisanal production, describing the values motivating their production practices, the emotional experiences of craft production, and the notion that artisans imparted aspects of their unique selves onto their products.

Table 4.

Evidence for Cheese Producers Claiming Oppositional Identity

First-Order Codes and Representative Quotations	Second-Order Themes	Aggregate Dimensions	General Theme
Raising and milking animals “There is high demand right now for [artisanal] products so [farmstead producers may] say to themselves they always ‘want to get bigger,’ they always ‘want to get bigger.’ So they start buying milk, but then, after that, they start processing. Let’s say I have goat’s milk, I’ll make cow’s milk, so that there I start buying cow’s milk from, I don’t know, the neighbor, then goat’s milk, sheep’s milk, so that at some point you become [an industrial producer like] Saputo. What makes you different from big industry? Is it because you process fewer liters? That’s not true. You lose the value of the work.” (Producer, interview 13)	Personal ownership of processes	Opposing craft and industrial practices	Claiming oppositional identity
Making cheese by hand “Definitely handmade makes a big difference. . . . Sometimes it doesn’t make a difference in the quality of the product, but it does make a difference in how much time has been invested in a single product. I think that’s the beauty of [artisanal production] too.” (Producer, interview 1)
Choosing ingredients “I think it’s mostly in your ingredient list. If you’re so big that you have to put milk powder to make your cheese, then I think it’s no longer all that authentic.” (Producer, interview 3)
Informing buyers “People need to see what kind of farm we are, and after that they don’t care about organic or not. . . . When they need this information, my customers ask me. At the beginning, at start-up, yes, I had a sign [stating] where I am located, what kind of production I do, the number of goats there were, and all that. Now we’re at eight years later. Most of them know me.” (Producer, interview 12) “People can see how we work, how our animals are, where they are. They can come and see us anytime, how we breed, they can ask us the question also.” (Producer, interview 14)	Transparent supply chains
Knowing suppliers “A little bit like is done in France with the ‘fruitières’ model where one [artisanal] cheese producer purchases the milk of a number of milk producers. So multiple milk producers permit the cheese producer to survive, and the cheese producer permits the milk producers to survive . . . it is not a co-op, but nevertheless I have a system of royalties that I installed. Even if the milk producers are not financial partners of the cheese company, for me it is important to give back when the company makes a profit. It is basically a bonus.” (Producer, interview 29)
Caring (for people and product) “I doubt [an industrial] cheese producer racks his brain or does everything from one end to the other. You are a number there. . . . The people that work with me are not employees you replace from one day to the next.” (Producer, interview 5)	Assertion of values
Simplicity “We propose to people to consume less, to be less consumerist, be less materialistic, and more into the art of living together, live together, fulfill our basic needs.” (Producer, interview 11)
Honesty “Authentic for me is that you also do what you say. I’m not going to say that I make raw if I pasteurize.” (Producer, interview 3) “We never say we’re organic; we say we go in that direction, but we’re not certified. We always tell the truth. People trust us.” (Producer, interview 1)		Opposing artisanal self-expression and industrial production
Passion “I think [the public] come here a lot, and they see me and [my wife], and they’re interested to see our, you know, why we do this. We love what we do, and you can say that it’s like they eat a little bit of our passion, it’s special. So it’s not just a product, but also the whole atmosphere. So they like it. It’s an experience [for them].” (Producer, interview 3)	Liberation of emotion	Opposing artisanal self-expression and industrial production	Claiming oppositional identity
Love “I had fallen in love with the goat, so starting up with cows didn’t interest me. . . . [My partner] found it interesting but not me. The cow, I don’t like that animal. It’s an animal I don’t think I have any interaction with. The main thing is really a love of the animal and of processing [milk into cheese].” (Producer, interview 13)
Creating unique products “[Others would say] Ah! You should make Chèvre Noir (a popular brand of cheese), but I don’t want to make someone else’s cheese, I want to make my cheese.” (Producer, interview 13) “We love being surprised by flavors. In my case, that is why I became a cheese maker, why I became interested. There is a panoply of incredible flavors, and [the idea] of becoming master or artisan of that interested me.” (Producer, interview 15)	Imbuing product with the self
Controlling own destiny “I am a meticulous person, and I like to do everything, I wanted to control everything, from production to processing. . . . For me [an artisan] is to master everything. So to take 100% responsibility, be able to have a personality that says okay: I manage the risk, and I organize things to make it work. That’s the most important thing. It’s about being independent. Artisanal for me it is independent.” (Producer, interview 12)
Extension of producer “There’s the whole relationship, I find, of identity and soul behind a plate of food. You know, just with my children, I often try that. I find it interesting to have a plate and say, ‘These are Mr. So and So’s vegetables, these are Josée’s fruits.’ I find that we eat differently; we don’t appreciate food in the same way if we know people compared to when we don’t know them, it doesn’t have the same meaning. The notion of the work in the background, the notion of respect, I think it’s very important. But [society] always had it, but we no longer have it. Because I know all the other producers, we see each other a lot, and we see all the work, all the commitment, and all the difficulties that they have.” (Producer, interview 1)

But while most producers did not directly broach the topic of terroir on their own, they did not reject it when directly asked about it in interviews. In contrast to the coherence of their discourse surrounding production practices and their quest for self-expression, the subject of terroir produced more nuanced and idiosyncratic responses. Most producers rationalized the concept of terroir with respect to their artisanal practices, while a smaller subset actively embraced the concept. First, confronted with the suggestion that the media tended to frame artisanal cheese as a terroir product, many producers creatively rationalized such claims without necessarily refuting them. After some thought, producers would frequently fold the concept of terroir into their anti-industrial discourse, stripping it of any notion of historical or cultural distinctiveness as evoked by macro-level actors. For example, producers often associated the term “terroir” with being small or sourcing their feed from local suppliers, sidestepping the richer meanings of cultural and historical significance. When asked whether one could taste the distinctive flavors of local herbs in their cheeses (a common assertion when speaking of terroir), producers might suggest that a nutritious diet would inevitably enhance the quality of the milk, but they sidestepped the question of any distinctive flavors representative of the land.

Second, only in a few cases did producers actively embrace the notion of terroir. In each case, we found that doing so was consistent with commercial interests. During our interviews, only three producers made strong claims about tasting the soil in their products or the importance of tradition. Interestingly, in two of those cases, the producers sought to circumvent retailers and sell directly to urban consumers, while in the third case the producer was the president of the farmstead cheese producers’ association and was presenting a case for increased government support to the sector. In all three cases, it appears that the producers were particularly attuned to how urban consumers apprehended artisanal cheese. Similarly, we witnessed a cheese producer at a conference exclaim that raw-milk cheese production was “the only way to taste the terroir!” Raw-milk cheeses are produced without the use of pasteurization, which kills all living microorganisms in the milk. This method of cheese production is highly regulated, and the cheese producer was evoking terroir as an argument for more lenient regulations (despite the fact that raw-milk cheeses were scarcely known or consumed in Quebec, much less a valued tradition). Thus, while the patriot identity ascribed to cheese producers by intermediaries appeared divorced from their core identity as artisans toiling in opposition to industrial production, most producers accepted the identity in at least a pro forma fashion, and some occasionally made use of it strategically to advance their interests.

Categorical consensus

Macro-level actors worked for decades to construct a category of foods that would allow consumers to “taste Quebec.” Rather than identifying the specific attributes shared among category members, these actors first elaborated abstract ideals and then tried to fill the category with tangible exemplars that would approach these ideals, in this case being representative of Quebec’s land, culture, and traditions. In sequential iterations, different groups (chefs, government, and the SRQ) sought to populate the category with disparate products (dishes, agricultural products, artisanal products) that ostensibly served the goal and then tried to encourage adoption of the category by market actors. But macro actors’ efforts were largely divorced from what was occurring on the ground among producers, and the category “Quebec terroir products” remained little known and used until market intermediaries embraced the construct. These intermediaries ascribed to producers an identity consistent with the terroir umbrella category and activated a related identity among consumers. Producers accepted the identity projected onto them, and consensus emerged within the field (macro actors, intermediaries, producers) about the meaning of Quebec artisanal cheese as a Quebec terroir food. Artisanal cheese gained a new meaning, and the Quebec terroir category gained the visible exemplar it previously lacked.

Many items would become “hastily re-baptized ‘products of the terroir’” (Lemasson, 2012: 112), but artisanal cheese was and remains the most prominent member of the category, as illustrated by the cover of a Quebec terroir guidebook shown in Figure 4 (Association de l’Agrotourisme et du Tourisme Gourmand, 2011). ¹⁰ While the breadth of the book is large, covering hundreds of producers, restaurants, and inns, cheese is displayed prominently on the cover. Uniting all the products and places featured in the book is the Quebec terroir umbrella category, which the introduction claims allows consumers to taste Quebec: “Come and discover the more than 685 business owners who want you to experience and taste Quebec in a very distinctive and authentic manner” (Association de l’Agrotourisme et du Tourisme Gourmand, 2011: 4). Quebec terroir products, and especially artisanal cheese—both seldom discussed only two decades prior—had become emblematic of Quebec’s culture and history, reminiscent of the invention of traditions such as the kilt in Scotland (Trevor-Roper, 1983), Rumba in Cuba (Lindholm, 2008), and country music in the United States (Peterson, 1997).

Figure 4.

Sample Book Cover

Conceptual Model

Based on our analysis, we develop here a conceptual model of how umbrella categories are constructed and instantiated as a material reality, including how they come to nest subcategories. As we described above, macro actors had worked in successive cycles to construct an umbrella category, but it did not become a “category in use” until market intermediaries reinterpreted producers’ identities to fit with the goals that had been elaborated by macro actors. Figure 5 summarizes this process and shows how different groups of actors participate at different vertical levels and in different ways during the category construction process. Specifically, producers work to construct categories in opposition to incumbents, as shown on the bottom left of the figure. In contrast, macro actors can suggest new umbrella categories in an attempt to reorder the categorical landscape. Market intermediaries play a bridging role (Weber, Heinze, and DeSoucey, 2008) by nesting the former under the latter.

Figure 5.

Process Model of Umbrella Category Construction and Product Category Nesting

In phase one of the model, macro actors engage in a process of goal-derived category construction comprising three stages. Envisioning the category involves collectively elaborating a goal and a set of ideals that category members should approach. The category is largely empty and aspirational at this stage, as indicated by the large dotted circle at the top left of the figure. The ideals voiced by macro actors act as orienting principles that guide discussions rather than as clear attributes used to evaluate category membership. Populating the category then involves filling the empty and aspirational category with exemplars through environmental scanning. Armed with ideals rather than specific product attributes, macro actors suggest a wide array of prototypical categories that can appear to have little in common with one another, shown as small dotted circles inside the proposed umbrella category in the figure. Finally, encouraging adoption involves disseminating the category to market actors, principally market intermediaries.

The goal-derived category construction process pursued by macro actors can proceed through any number of cycles or iterations without necessarily ever succeeding. At the end of a failed cycle, i.e., one that does not see widespread use by market actors, the mix of macro actors involved may change as new actors enter and others leave, with the process restarting anew (or possibly being abandoned). In our case, we saw the process repeat three times, but three is not a magic number.

Much of the phase one activity of macro actors takes place at a distance from what is occurring on the ground among market actors, such that the umbrella category and the prototypical categories are disconnected in this phase. For example, in our case study, agricultural producers were largely oblivious to macro actors’ efforts to define a cuisine or products representative of Quebec culture and traditions.

In phase two of the model, the umbrella and product categories are linked. The goal-derived category construction process led by macro actors exits the cycle in phase one when the nascent category resonates among market intermediaries who begin adopting it en masse. Market intermediaries validate the nested category structure proposed by macro actors by projecting onto producers new identities that are consistent with the goal proposed by macro actors. This implies ascribing to producers motives that may be somewhat removed from their original motives and activating among consumers related identities to enhance their interest and engagement with the category. Producers, for their part, may be complicit in this redefinition of their products and identities. In our study, no producers refuted the terroir concept, which was the basis of much of the consumer interest in their products. Instead, producers may rationalize the projected identity, connecting it with their own practices by selectively interpreting its meaning in a way that feels comfortable to them. Alternatively, they may actively embrace the new identity and promote the umbrella category. Through the mechanisms of projecting goal-consistent identities (market intermediaries) and accepting goal-consistent identities (producers), the goal-derived category proposed in phase one becomes a material reality. As shown by the solid circle encompassing multiple smaller solid circles in phase two, the umbrella category becomes filled with tangible exemplars, while the product categories that get nested under the umbrella category acquire new meaning.

Discussion

Understanding the category construction process requires looking beyond a single category and set of producers. We suggest that attention to the vertical dimension of category systems can shed light on how new umbrella categories reorder the categorical landscape and give new meaning to product categories. To this end, we offer in this paper a conceptual model of how umbrella categories are constructed and come to nest product categories, in the process going from empty categories to “categories in use.” Our work builds understanding in two related areas: (1) how vertical nesting occurs and (2) the origins and nature of goal-derived categories and their relationship to prototypical categories.

Umbrella Categories and the Nesting of Product Categories

Extant research on category construction has generally examined single-product categories, focusing on the role of producers (e.g., Glynn and Navis, 2010; David, Sine, and Haveman, 2013) or professional classifiers (e.g., Khaire and Wadhwani, 2010). In that stream of work, category construction turns on defining the key attributes of the products or services that are to fill the category. Consistent with this prior work, we found that cheese producers made category claims about their products. As seen in prior research, such as on craft beer (Carroll and Swaminathan, 2000), grass-fed beef (Weber, Heinze, and DeSoucey, 2008), modern architecture (Jones et al., 2012), and management consulting (David, Sine, and Haveman, 2013), the producers in our case employed persuasive discourse to frame their activities in opposition to dominant incumbents and to thereby project an image of themselves “as a coherent category with a meaningful label and identity” (Wry, Lounsbury, and Glynn, 2011: 450). But in our context, a singular focus on cheese producers would have provided an incomplete understanding of how the Quebec artisanal cheese category developed. Producers’ theorization efforts were only one part of a larger story. The meaning of the category arose from the “interaction between multiple self-interested market participants” (Grodal and Kahl, 2017: 157) across vertical levels. Groups of non-producers that we call “macro actors,” such as government agencies and professional associations, constructed an umbrella category. Other groups of non-producers—intermediaries such as journalists and retailers—embedded the product category under the umbrella category and in the process altered its meaning.

Key among our findings was the role of macro actors in constructing a broader umbrella category with the potential to reorder the categorical landscape. Prior research has generally conceived of non-producers as either evaluators or reclassifiers, rarely as category constructors. The literature on market categories has often suggested a “division of labor” (Grodal, 2018: 785) wherein producers theorize and non-producers evaluate those theorizations. As Pontikes and Kim (2017: 83) pointed out, non-producers may be reluctant to engage in new category construction because “developing categories based on original insights, coining labels, and having them catch on broadly, is a risky process that often fails.” We suggest that this observation is less applicable to macro actors, who are further removed from market exchanges and thus less exposed to market risk. As Edman and Ahmadjian (2017: 112) explained in their study of the construction of the ji-biru (local beer) category in Japan by macro actors (they used the term “external actors”), the interests of such actors “lie not in producing, buying or consuming the outputs of an industry” but rather in furthering objectives such as town revitalization, regional economic growth, and tourism. The macro actors in that case, much like in our case, were regional associations and governments removed from market risk, and like in our case, the ji-biru category was initially empty. We conjecture based on our findings that macro actors are actually well placed to conceptualize and propose umbrella categories.

This finding has particularly important implications for research on the role of the state in category construction, as the literature has tended to confine the state’s role to one of regulation, deregulation, and/or financial support, not active meaning construction. For example, Georgallis, Dowell, and Durand (2019) described how financial and regulatory support for the solar photovoltaic industry in Europe led to its expansion. Similarly, railroads (Dobbin and Dowd, 1997), farm wineries (Swaminathan, 1995), wind power (Sine and Lee, 2009), and private liquor stores (Dowell and David, 2011) required changes in policies and regulations to develop. This regulatory role is in fact a form of evaluation, whereby regulators and policy makers respond favorably to nascent categories. We show a much richer, more active role for the state in the category construction process, whereby the state actively constructs meaning.

Nevertheless, our findings also suggest that macro actors, the state included, may not be well placed to construct lasting umbrella categories on their own. Their timelines are longer, and their work is not tied to the success of any single organization or even category of organizations. As a result, macro actors often work at the policy level, involving a wide array of stakeholders, and the categories they propose are unlikely to survive if they are not elaborated upon and filled in with concrete exemplars. Market intermediaries may play an important role in conferring to the umbrella category a semblance of material reality by nesting product categories, while transforming the meaning of those product categories in the process.

Our findings on how market intermediaries do so also help unpack and extend the reclassification role typically attributed to them. Durand and Khaire (2017: 89) explained that intermediaries involved in category construction engage in a “form of rearrangement or cognitive reinterpretation of existing cues to benefit the actors that initiate these processes in material and symbolic ways.” Similarly, Khaire and Wadhwani (2010) showed how auction houses (among other market intermediaries) introduced constructs and criteria imported and adapted from an existing umbrella category (Western modernism) to delineate the nascent category of modern Indian art and provide the basis of valuation and exchange. We extend these findings to show that intermediaries’ reclassification efforts can involve projecting new identities onto the actors involved in the focal category. This includes ascribing new motives to producers to align their actions with the goals of the larger umbrella category, while activating similar attachments among consumers. Intermediaries thus work at the level of identities—in our case encouraging “patriotic” consumers to support “patriotic” producers—to reclassify a set of products under an umbrella category. This points to identity as a lens that serves a sensemaking function that elevates the importance of certain product or organizational attributes over others (Kim and Jensen, 2011; Smith, 2011). For example, viewed through an environmental sustainability lens, proximity would take precedence over origin in food production. But when viewed through a patriotic lens, origin takes precedence over proximity, leading to the “positive collective identity” that is “central to market functioning” (Weber, Heinze, and DeSoucey, 2008: 546).

Goal-Derived Categories

In the case analyzed here, a goal-derived lens fruitfully described the umbrella category constructed by macro actors. Whereas prototypical categories encompass ostensibly like objects that are evaluated according to their resemblance to a category prototype, goal-derived categories may encompass highly dissimilar objects serving a similar goal that are evaluated according to the degree to which they approach ideals that no category members may ever fully manifest (Barsalou, 1985, 1991; Ratneshwar et al., 2001). “Quebec terroir products” illustrates a goal-derived category comprising highly dissimilar products (e.g., cheeses, meats, wines), all evaluated according to the seemingly unattainable ideals of representing Quebec’s land, culture, and traditions such that they allow consumers to “taste Quebec.”

This finding expands the conceptualization of goal-derived categories, as it reveals that they can be enduring cognitive structures. Scholarly work to date on goal-derived categories, whether in psychology (e.g., Barsalou, 1983, 1985; Ratneshwar et al., 2001; Voorspoels, Storms, and Vanpaemel, 2013) or organization studies (e.g., Durand and Paolella, 2013; Durand and Boulongne, 2017), has focused on ad hoc, individual-level categorization processes, such as how an individual selects “foods to eat on a diet” the first time one decides to diet. Consequently, goal-derived categories have been discussed as “temporary categories constructed around specific goals in a given context” in order to solve immediate problems (Durand and Paolella, 2013: 1109). Our findings suggest that goal-derived categories can, in fact, be enduring and shared, and our conceptual model thus responds to calls for theory to explain how goal-derived categories might become institutionalized (Glaser, Krikorian Atkinson, and Fiss, 2020). Moreover, processes of goal-derived category construction at the individual level seem to be similar at the collective level: just as individuals “first define a goal, and only afterwards do they observe and organize the reality into categories of objects likely to help them reach their goal” (Durand and Paolella, 2013: 1109), we found that the Quebec terroir category was collectively envisioned by macro actors prior to their collectively scanning the environment to populate the category with exemplars.

Our findings suggest that goal-derived categories can begin absent material exemplars and progressively take shape over time, and as such they can support higher levels of ambiguity than prototypical categories. Whereas prototypical categories do support some degree of ambiguity—because category attributes define only the minimal requirements for category membership (Glynn and Navis, 2013)—the ambiguity of the Quebec terroir products category appeared to be much greater. Even as late as the early 2000s, when the term “Quebec terroir” was in widespread use, many actors in the field could not define the concept (Cimon-Morin, 2004). Rather than a hindrance, this ambiguity appeared to facilitate collective action among a diverse set of actors, a finding consistent with work showing the relationship between ambiguity and diffusion (David and Strang, 2006; Giroux, 2006). Lacking tangible exemplars to guide them, macro-level actors began by expressing ideals, which may have also proven less contentious than defining specific product attributes. Moreover, ideals can remain stable even while the attributes of the member products evolve dramatically over time. This extends our understanding of how goal-derived categories can provide a basis for market exchanges through a process of “goal-based attribution” whereby market value is determined more by perceptions of attaining an ideal than by possessing particular attributes (Glaser, Krikorian Atkinson, and Fiss, 2020: 938). It also helps us to understand how category ambiguity can persist, not necessarily because actors apply multiple frames to a category (Chliova, Mair, and Vernis, 2020) but because the product attributes that are thought to lead to an ideal may not be clear or may change over time. As a point of comparison with our case, one may consider the category cleantech (or greentech), which encompasses a wide range of product categories under an ideal of carbon neutrality. The member categories of cleantech will likely change over time as new and better technologies emerge and the environmental impacts of older technologies become better known, but the guiding ideal can remain stable over time.

Thus goal-derived categorization may prove particularly relevant in contexts of changing societal values and beliefs, as a means to reimagine markets and guide entrepreneurial behavior. Extant organizational research has often focused on the role of entrepreneurs who attempt to legitimate their novel offering by drawing on changing values and beliefs (David, Sine, and Kaehr Serra, 2017). For example, soft-drink producers referenced the values of the Temperance Movement at the birth of the soft-drink category (Hiatt, Sine, and Tolbert, 2009), just as management consultants deployed the values of the Progressive Movement at the birth of management consulting (David, Sine, and Haveman, 2013). However, in other cases, the reframing may be done prior to the innovation itself (Durand and Khaire, 2017) as a means of guiding subsequent entrepreneurial activity. For example, Rao and colleagues (2003) showed that French nouvelle cuisine emerged not because practicing chefs introduced new recipes but because certain French culinary elites translated the anti-authoritarian values of the May 1968 protests into the sphere of cooking. These (macro) actors set forth ideals such as transgression and acclimatization that gained meaning as they were increasingly instantiated by practicing chefs across France (Rao, Monin, and Durand, 2003, 2005). Thus actors removed from markets may turn to goal-derived categorization to set forth new value-laden goals and ideals that guide subsequent entrepreneurial activity.

Implications and Future Research

Where might processes similar to the ones uncovered here be found? We would look first for contexts in which macro actors take an active approach to markets. We know that there is variance (in geography and over time) in how active states are not just in regulating markets but also in trying to shape them (Hall, 2015), and our findings might be more generalizable to other contexts where states take a more active role in markets. More generally, contexts with more scope for collective action in the economy by macro actors—not only governments but also professional associations and non-governmental organizations—may see more frequent attempts at umbrella category construction. We might also look for contexts with active social movements circulating new sets of values and beliefs across society. Without the Quebec nationalist movement promoting the belief that Quebec constitutes a distinct nation, the goal of “tasting Quebec” may have never formed the basis of an enduring umbrella category. Similarly, the widely circulating value (and related social movement) of environmental sustainability, for example, underpins the umbrella category of cleantech, an ambiguous category that nests a variety of prototypical categories.

Naturally, certain aspects of our context may limit the generalizability of our findings and thus call for future research. First, with respect to umbrella category construction, we found that new macro actors entered after each failed attempt. It is possible instead that existing macro actors could reconceptualize the umbrella category in subsequent attempts, although in our case, each new macro-actor group breathed new life into the umbrella category, which otherwise could very well have been abandoned. Second, not all umbrella categories are fruitfully seen through a goal-derived lens. Higher-order categories could be prototypical in nature, and in such cases macro actors may have more limited scope to reorder the categorical landscape.

With respect to producers, we found indifference among many artisanal cheese producers in fueling the beliefs that artisanal cheeses were representative of Quebec culture and traditions, such that this role fell largely to intermediaries. Cheese producers may have been particular in their desire to remain small, and it is possible that producers in other categories may be more assertive in nesting their categories under the umbrella category, particularly if they ascertain (and desire) economic gain. Moreover, once the umbrella category is well established, new categories may subsequently be “born under” that umbrella. At the same time, it is interesting to consider how the nesting of an existing product category (such as artisanal cheese) might affect its original identity. Will we see a layering of identities, in this case anti-industrial and patriotic, or a displacement of the original identity in favor of the projected one?

This research opens new doors into the study of goal-derived categories. We show that such categories are not always ephemeral structures; they can persist over time and form the basis of collective action. Yet this does not imply stasis. Future research could investigate the co-evolution of higher-order umbrella categories and the prototypical categories nested underneath. Might the accumulation of prototypical categories under an umbrella category alter the latter’s meaning? And how might the meanings associated with prototypical categories change as a result of their nesting or the nesting of other prototypical categories? In nascent markets characterized by an evolving and diverse set of organizations and products (Grodal, Gotsopoulos, and Suarez, 2015), goal-derived categorization may be common as it readily permits evolution and change in the member organizations given the inherent ambiguity of the ideals that define the category. Over time, we may find that goal-derived umbrella categories take on more prototypical attributes as the two modes of categorization co-exist, the former defining the value-laden goals served by the category and the latter employed to identify the attributes used to assess category membership.

Authors’ Biographies

Johnny Boghossian is an associate professor of strategic management at Université Laval, 2325 Rue de la Terrasse, Quebec City, Quebec, Canada, G1V 0A6 (johnny.boghossian@fsa.ulaval.ca). He received his Ph.D. from McGill University. His research interests include the emergence and evolution of industries and organizational fields. He is currently studying the state’s role in protecting industries targeted by social movement activism.

Robert J. David is a professor of strategy and organization and Academic Director (Research) at the Desautels Faculty of Management of McGill University, 1001 Sherbrooke Street West, Montreal, Quebec, Canada, H3A 1G5 (robert.david@mcgill.ca). He received his Ph.D. from Cornell University. His research focuses on the emergence and evolution of markets, industries, and organizational fields. His prior empirical contexts include management consulting and alternative energy, and he is currently studying the market for local food products.