Reading :: Markets in the Making

 Markets in the Making: Rethinking Competition, Goods, and Innovation

By Michel Callon

Michel Callon, one of the leading lights of actor-network theory, passed away in 2025. This book (French version 2017, English translation 2021) was his last, capping off decades of research involving markets. This research included the collections The Laws of the Markets and Market Devices, both of which I have recently read or reread as, to my surprise, I am once again thinking about what makes a market.

Markets, specifically as Callon understood them, are on my mind lately because I am continuing to think about how entrepreneurs understand them. Often entrepreneurs characterize markets as existing “out there.” For instance, their pitch decks might characterize the total addressable market of “all hotels in the United States” or “all potential EV buyers in North America.” They’re often characterized by industry or demographics. Yet entrepreneurs also recognize that markets emerge, shaped by entrepreneurs’ own efforts, and that’s especially true when new technologies unite people in ways they haven’t been united before. Moore famously defines a market as a collection of customers with the same basic need who reference each other when making buying decisions — and that italicized part suggests that markets are a social phenomenon. So, are markets discovered or created? 

Callon, in his preface, argues that there is no difference between “markets in the making” and “made markets”: “There are only markets in the making that are being transformed more or less intensively” (pp.11-12). (He traces his interest in markets back to his early study on the development of an electric vehicle (p.10).) Callon argues that his approach to markets, which involves “plunging ourselves into the effervescence of commercial practices,” offers two advantages (p.12):

• It provides a “more realistic understanding of how markets work,” including “everything that permits a functioning market to be formed, established, maintained, and transformed.”
• It dissociates “market” and “capitalism” by substituting the notion of “market agencement” — a term we’ll explore a little later on. (p.12)

To develop his argument, Callon draws mainly on others’ studies. But he pulls these together elegantly, allowing us to see the implications of his larger concept emerging through these different studies.

In the Introduction: The Enigma of the Market, Callon draws on Trevor Pinch’s study of Cockney street vendors. These vendors structure action in four stages:

1. Constitute the audience (building the edge): Create a space in which a flowing crowd is compelled to pause, and a diversion that allows them to coalesce into an audience (pp.16-17) who can see the goods being hawked as singular (p.18).
2. Make the good integral to the audience’s life (flashing). Invite the audience to touch and sense the merchandise. This step is material, and it includes stacking the merchandise for presentation. (pp.18-19)
3. Setting the price (ramping and batting). The price is set by starting at a high point attributed to competitors, then descending as if on a ramp (p.20). If a potential customer agrees that one price is fair, then the price goes down, the customer feels compelled to buy. “Did he decide, or was he made to decide by the disposition — the staging and the devices?” (p.20). 
4. Recruiting further buyers (twirling the edge). Finally, if the transaction has been successfully singularized, how can the seller reach a mass market? The vendor directs an assistant to people at the edge of the crowd who linger; the assistant tries to convince them to join. (pp.21-22).

Callon argues that “the twirling of the sale” demonstrates that these are neither personalized markets nor mass markets, which “should always be perceived as juxtapositions of a myriad of bilateral transactions“ (p.22). “The challenge facing all who live by commercial transactions and give life to markets is to produce the mass from singularization without losing the latter in the process” (p.22). Thus “supply and demand emerge and express themselves over the course of a continuous process”; the “eddies of the sale … give a form to the deployment of the sale that can overcome the apparent contradiction between quality and quantity, between, on one hand, bilateral transactions and the singularization they require and, on the other hand, their aggregation and massification” (p.22). Thus, although the goods are mass-produced and identical, “a multitude of separate transactions, each distinct and different, are juxtaposed in time and space” (p.23). Instead of abstract intersections between supply and demand curves, we have “the devices required to produce this intersection — I should say these many intersections — by deploying, unfolding both space and time” (p.23).

These four steps, he says, are visible “in every sector of market activity” (p.23). After giving the example of Las Vegas slot machines, Callon sets out the book’s hypothesis: 

when confronted by the enigma of the market, actors constantly invent new devices that are repeatedly modified and transformed. But these changes are not random or arbitrary. In order to be successful (sometimes), commercial activities must be organized (agencees) in a certain way. … None of these results could be obtained without the design and implementation of rigorously structured agencements. (p.30)

With that, we get to Chapter 1: From Interface Markets to Market Agencements. Here, Callon asks: What is a market? He notes that definitions of market are diverse, although covering recurring elements. “These common elements form an implicit model that gives the word ‘market’ its meaning,” he says, dubbing this model the “interface market.” In contrast, he suggests an alternate model, the “agencement model” (p.32).

He begins his discussion of the interface model by noting that the academic and general literature offer many contradictory definitions of “market,” explained by the fact that “individual experiences of the market are not reducible to one another” (p.33). He takes as a point of departure the work of Depeyre and Dumez. (Side note: I don’t know much of Dumez’s work, but he wrote a great article on case studies in 2015 that I’ve cited.) Here, Callon says he will pay equal attention to people engaging in markets and people who study markets (p.33). They follow a strategy of abstraction with three directions:

1. Sellers, buyers, and platform goods. “The strategy emphasizes the idea of an abstract space in which relationships are themselves abstractions … established by encounters between agents defined as sellers and agents defined as buyers.”
2. Competition. “These relations are [portrayed as] governed by mechanisms of competition”
3. Institutions. “For these relationships to endure and for the market to be stabilized, it is judged necessary that rules, conventions, or, more largely, institutions come into existence” (p.35).

Callon calls these assumptions the “greatest common denominator” of scholarly representations of the market in economics and sociology (p.35). And he notes the invisible work that must be put into each:

1. Sellers, buyers, and platform goods. “The material sites that allow and provide the framework for markets have to be erased so that it becomes possible to focus only on the goods and the agents prepared to engage in transactions,” and “the existence of a population of buyers and a population of sellers implies that goods have already been developed to provoke the establishment of commercial transactions” (p.36). To trouble this portrayal, he turns to “platform goods,” such as free newspapers, which exist to “articulate several separate populations of agents” (AKA “multisided markets”) (p.38). Callon argues that all goods work this way: What matters is “the capacity of these goods to put agents in relation to each other and to provoke encounters likely to end in a commercial transaction” (p.38). 
2. Competition. He quotes Braudel as saying “the market is competition” (p.39, his emphasis) — but “‘competition’ does not have a single meaning” (p.39). Regardless, “the universe of goods is left outside the universe of agents” and the role of innovation is taken to be that of reducing competition (p.41). But in an agencement model, “competition cannot exist outside of innovation strategies” (p.42).
3. Institutions. Since markets must also organize and ensure repeated transactions, they require institutions — “rules, private or public conventions, and laws or regulations that are shared and apply to all participants,” and thus the market is an institution (p.42). These rules vary, but some are constitutive, while others are regulation (pp.43-44). And he notes the relationship between the abstractions of supply and demand (p.44). The abstractions under this heading, he says, allow us to speak of the market “as a general entity” while allowing a “wide variety of descriptions” (p.44). 

“The triple movement of abstraction outlined above leads to a minimal description of market activities and organizations that I’ve proposed we call the interface model of the market” (p.44). But he sees it as too incomplete and narrow (pp.45-56), so he turns to the model of market agencements. He keeps the following from the interface model:

• The distinction between agents and goods
• The transfer of property rights
• The relations of competition that develop between agents
• Commercial monetary transactions (p.46)

But these 

• Don’t account for the material composition of market activities
• bracket the coconstructive process of creating supply and demand
• create “unrealism,” ignoring the actual mechanics
• create an impasse on processes
• precludes us from recognizing goods as processes
• “imply a description of agents that underestimates their diversity, heterogeneity, and plasticity” (p.47)

In broad strokes, the agencement model “reworks some of the hypotheses of the interface market” (p.47):

• Material component. Callon argues, citing Weber, that “the material component plays an essential part in commercial activities” (p.47) as “material assemblages” that “contribute to formatting commercial transactions” (p.48). This insight also points to emergence.
• Strategic purpose of market agencements. An agencement is a “heterogeneous assemblage” that includes material devices. Thus “markets consist of arrangements that play a double role. They contribute to imposing guiding frameworks on individual actions and provide them with motives while channeling collective action toward the conclusion of commercial transactions.” Their “strategic purpose” is “to organize and promote the establishment of bilateral transactions” (p.49).
• Competition as the struggle to establish bilateral transactions. Callon argues that in this model, “what is at stake in competition is establishing transactions” — this “is competition’s one and only goal” (p.49). In the view of the market agencement model, monopoly is “the ultimate expression of the competitive struggle” because in sidelining competitors, the company has singularized the goods it offers to individual clients (p.50). (Compare Geoffrey Moore’s advice to pick a beachhead market you can dominate before spreading into adjacent markets.) The singularization is not of the good but of the bilateral transaction: “the trio made up of the one who offers, the one who demands, and the good” (p.50). (Note: we can gloss this as Product, Market, and Organization. But we also see slippage here between “market” and “customer”; I think Callon would agree that these two terms are not clearly separable.)
• The proliferation of transactions requires innovation. “Competition does not stimulate innovation any more than it paralyzed it. Innovation is the essence of competition” because “the greater the singularization [i.e.] scope of the innovation, the more intense is the competition” (p.51). And in some cases, such as the tech sector, “the rhythm and depth of innovations, by which I mean the proliferation of bilateral transactions, grow spectacularly”; this is “a regime of intense innovation” (p.51). 
• Process goods. “If the singularity of a good must be established for each and every transaction, then supply, demand, and the qualities of products do not drive market activity, but are the result achieved by endless activities that constitute and reconstitute an active market agencement” (p.53). And he introduces the concept of qualification: “the coproduction of goods’ characteristics, as well as those of buyers and sellers, into the activity I refer to as coprofiling.” Qualification is “a central feature of all commercial activities” and at the heart of the establishment of markets (p.53). A good is requalified throughout its life: a car, once sold, can become an object of social distinction, can be rented or resold, can be wrecked and sold for spare parts, or could be requalified as a classic car (p.54). Thus a product is a process, “made up of a series of actions, a succession of operations that transform and displace them, passing them from hand to hand through a series of metamorphoses that work out forms judged attractive by economic agents who pay to enjoy their benefits” (p.54). (Compare this view to Lusch and Vargo’s Service-Dominant Logic.)
• Proliferating agents, plastic identities, and networking. “The process good articulates supplies and demands” — “an action that involves a set of modifications and adaptions that connect design, production, and circulation activities to one another. The action must mobilize multiple agents to collectively engage in the creation manufacturing, sale, and use of these products” (p.55). These agents go well beyond supply and demand, including researchers, civil servants, investors, users, consumers, and certification bodies (p.55). (This wider set of stakeholders is more visible, I think, in early-stage technology entrepreneurship!) Thus, rather than referring to buyers and sellers, Callon says “the essential notion here is networks” in which “goods, agents, and material infrastructures are caught up in the same dynamic, where they are subject to being transformed while coevolving,” with their qualities “reworked until they get adjusted to one another” (p.55). Thus “market relations engender social relations, but not just any social relations. They also manufacture impersonal relations, but not exclusively. One of this book’s objectives is to further characterize this double process” (p.56).

Callon concludes the chapter by characterizing markets as framings: “devices that orient and structure action toward its strategic goal” (p.57). He proposes five framings, which he discusses in detail throughout the rest of the book:

1. “Rendering goods pass(act)ive” (p.60), formatting “the entities designed to be marketed” (p.57). 
2. “Activating agencies capable of evaluating and transforming these goods” (p.60)
3. “Organizing their encounter” 
4. “Ensuring the attachment of the goods to the agencies” 
5. “Obtaining consent to pay”
6. “Setting a price and compelling payment actions that combine and interweave with each other, with possible feedback loops and iterations” (p.60)

He also promises to examine reagencing, in which forces provoke market agencements to reformulate themselves; the dramatic expansion of market agencements; and the question of political engineering (p.60).

In Chapter 2: The Process of Making Goods Mobile and Alienable, Callon argues that commercial activity relies on the distinction between commercial agents and goods, which is a reflection of the distinction between people and things (a distinction that ANT famously questions). “Things are [understood as] incapable of intentional actions and can be defined by their passivity; they are ready to become goods that are exchanged” (p.63). This distinction is fundamental to interface markets, yet is questioned in practice, more recently by AI (p.64). Callon, predictably, argues that the distinction is the result of processes: agents and goods are separated and endowed with rights and responsibilities via processes (i.e., trials). This is central in market agencement, which involves activities that create fragile problematic disentangling of people and markets (p.64).

In this chapter, Callon seeks “to understand how the activity of a thing can be directed and controlled while being rendered mobile and alienable” (p.64). 

First, he identifies an assumption in interface markets: that “goods are given, and their properties are stable.” But “this perspective is untenable” because commodities are contested and sometimes become goods only at the end of a long process (p.65). 

He notes that things participate in market transactions only at certain moments; sometimes the “social density” of things (not his term) — resulting from “a set of associations that, when they are strong, numerous, and durable, weigh a thing down” and “condemn it to a form of immobility, thereby imposing a stable status” (p.72). Think in terms of inalienable possessions such as heirlooms. Callon adds that calling this “social density” misleads us because the material properties of an object are also important here (p.74). Associations are social, symbolic, material, and physical, all at the same time (p.75). 

For a good to be mobile, then, it has to be transferable. Sometimes that requires reformatting or disentanglement (p.75). The good has to be part of an associated milieu (here he cites Simondon). For instance, a car has to be associated with “highways, garages gas stations,” oil companies, insurance companies, etc. without which the car could not function; it would “simply disappear” (p.75). This is true of services as well (p.76). “Without its associated milieu, a good or service is nothing” — and yet to “travel from one agent to another” it must be “separated from its milieu” (p.76). It is a feat to offer “a good that can function autonomously, even as this functioning requires a milieu on which they generally cannot have much influence” (p.77). So successful disentanglement “must make it possible for an equilibrium to be reached, an always fragile and delicate equilibrium between autonomy and heteronomy. Only under this condition can the good be transferred and change hands” (p.77). 

The good must also undergo reentanglement, reinserting itself “into the life of its addressee” (p.77). Drawing on Akrich’s work on scripts, he argues that these technical and moral scripts format interactions with an object while leaving room for improvisation (p.79). Reentanglement involves transforming a thing into a usable good that can participate in the buyer’s world; it must balance the amount of autonomy given the object (p.80). Callon nominates the — in my view, unfortunate — term “passiva(c)tion” to name “the strange process that leads an entity to be capable of living its own life, to be active, to act and make act, but not too much, and always in a specific manner within preestablished frameworks” (p.81). This continuous process “leads the good to enter the life of its addressee”; the good and addressee co-evolve and entangle, something Callon calls singularization (p.81). To become a commodity, the good must further be “detached from the entity that puts it into circulation” (p.81) — unlike a gift. So market-oriented passiva(c)tion “ensures detachment” (p.86). 

After some discussion that I’ll skip over, Callon adds that “In order to become a commodity, a thing must go through transformations, enter into the process of passiva(c)tion that accompanies the detaching and reattaching,” a process that Callon sees as “a form of production in its own right“ (p.104). 

Skipping again, let’s get to this point: Callon argues that a good must go through detachment. This involves disassociating from its milieu while maintaining relations between the two. This, when successful, is market-oriented passiva(c)tion (p.143). 

This brings us to Chapter 3: Agencies and their Qualculative Equipment. I’ll briefly air a minor complaint with French theorists, which is that they seem to love their portmanteaus such as “passiva(c)tion” and “qualculative.” I don’t. But for you, my reader, I will struggle through these. Callon argues that goods, or “pass(act)I acted objects or things, qualified and requalified, must, in the end, have a value in the eyes of agents who must then decide whether or not to pay a price in order to gain their possession” — but for the agents to recognize how the goods can help them, to evaluate these goods, “they must have the skills that allow them to introduce an order into the universe of goods that surrounds them” (p.149). To buy, they must evaluate; to evaluate, they must have a way to conduct the evaluation. (Note: In pitch decks, this work is facilitated with a competition grid.)

Callon introduces the notion of a “qualculative agency as the active force of ordering the world of goods. A qualculative agency is a collective entity that has equipment and instruments at its disposal that will assist in the task of ordering by establishing lists, identifying similarities and differences, and producing comparisons and hierarchies between goods.” This equipment “actively participates in the constitution of classifications and the shaping of rankings” (p.150). Since “the dynamic of market agencements is entirely structured around the interactions between agents and goods,” “the qualification and evaluation work conducted by agents transforms the goods,” meaning that “goods do participate, in their own way, in their own classification and the classification of agents!” (p.159). Callon adds that “Sociotechnical prostheses do not simply provide tools for the action; insofar as they participate in it, they shape and structure it from the inside” (p.159).

Here, Callon introduces the term “calculation,” which he broadens beyond math to address how one evaluates and orders the world before taking action (p.160; note: some applications of genre may fit into this definition). In qualculation (which he attributes to Franck Cochoy), qualitative and quantitative dimensions are combined; “at each step, words and numbers are combined” (p.161). Quoting Guyer, he adds that “an unnamed good cannot exist” (p.161). Callon draws on Latour’s centers of calculation to identify calculation as a three-step process:

1. Detach entities
2. Connect them through associations
3. Extract a result (p.163)

He says that qualculation is a material activity, “even and especially when it is described as mental” (p.165). “The process of transformation through evaluation acts on goods by reconfiguring the relations into which they enter and continues uninterrupted throughout the good’s career” (p.165). Qualculative agencies are heterogeneous beyond individuals’ capabilities, therefore a collective that acts: reflects, analyzes, and behaves strategically (p.167). The agency “acts in the name of agents,” e.g., when “the agency designates … a company, an administration, a union, or a user group,” but even when it works for an individual agent (p.167). Examples of qualculative agencies include comparison tables that rank competitors by criteria (p.173 — we can add those competition grids that show up in pitch decks). 

In startups, Callon draws on Doganova & Eyquem-Renault to point out the central role of the business model, which both calculates ROI and shows that value can be created (p.178).

Callon concludes that “goods and agencies codefine each other over the course of qualculation tests” (p.199); “It is the goods that put us into relation, and in return, through the choices we make, we put goods into relation with one another” (p.200).

In Chapter 4, Market Encounters, Callon turns to the chain of actors who bring something to market (p.201). “A market agencement can be seen as a device that incessantly organizes the collective labor of producing encounters and progressive transformations in order to lead to this specific event: a good is offered by an agent to another agent, who accepts to pay in order to obtain it” (p.202). The interface market model does not consider how to structure such encounters (p.203) — encounters that are staged (p.204). “Associating, which requires mutual adjustments and adaptations, carries value for the agents. If the association breaks or fails, this value disappears” (p.207).

Moving on, in Chapter 5. Affectio Mercatus: Attachments and Detachments, Callon turns to attachment: “the (successful) process of reconfiguring an agency such that having been transformed and wanting to persevere in a new way of being, it has only one option left — namely, to appropriate the good and pay to obtain it” (p.245). Callon uses the term affectio mercatus to refer to “the set of mechanisms that push agencies to engage in a collective action that concludes, at some point, with a bilateral transfer in return for monetary compensation” (p.246). After some discussion, Callon distinguishes three types of attachment devices: “devices for listening and dialoguing, devices for coproducing, and devices for addicting” (p.269). 

In Chapter 6, Price Formulation, Callon argues that price level sets the conditions for the experience; the price “is a quality of the good. It does not measure the good’s value” (p.288). He urges us to think in terms of price formulation, in which price is “the result of a process that often begins very early, when a good is still only a vague idea in a design studio or a research lab,” and is “a strategic variable around which power relations are formed and through which the establishment of bilateral monopolies is at stake” (p.288). (Think in terms of how an early-stage tech startup might identify a market for its innovation and shape its R&D efforts around addressing the needs of that market.) Callon writes a lot on price formulation — the chapter stretches from pp.287-353 — but it’s not that interesting to me, so I’m skipping the rest.

In Chapter 7. How do Market Agencements Evolve? Callon centers his previous work in Foucault’s concept of dispositif, or heterogeneous ensemble or apparatus (p.355). This term has been translated into English as “assemblage” or “arrangement,” (p.358), and in so doing, has tended to drift in meaning (p.359). “These two notions risk erasing what is innovative in dispositif and especially in agencement, as Deleuze defines it, revisiting Foucault” (p.359). Callon advocates for reviving the term agencement, positing that “an agencement acts” (p.361). He adds that “there are no market agencements without nonmarket agencements” (p.362): through alignment and articulation, “we can describe market agencements as multisite networks while accounting for their dynamism oriented toward the institution of bilateral transactions and the competition that supposes in view of creating attachments and imposing prices on the transactions” (p.363). For alignment, he turns to the translation model (pp.363-364); for articulation, he invokes Foucault’s networks and Deleuze’s lines of flight (p.366). 

In Chapter 8: The Role of Theorization in Transforming Markets, Callon considers how his theorization could help us understand the dynamics of agencements. He provides a simplified diagram describing a loop, “the elementary cycle of the political engineering of market agencements” (p.388):

1. Expressing concerns
2. organizing concerns
3. theorizing by experts or laypeople
4. controversies, alliances, confrontations
5. site transformations
6. how new framings spread

He summarizes the book, adding in a section on climate change, and ends with another diagram, this one on the task of theorizing.

So what did I think? 

Overall, I found this book to be very rewarding. It pulls together Callon’s earlier work on the market as well as the sociology of translation, providing an understanding of market that is well grounded in sociotechnical theory. It also resonates with what I know about startups in particular, since startups have to work with so many uncertainties and erect loops, produce intermediate descriptors and artifacts, design and test prototypes, and link criteria and competition. In the startup world, this work starts immediately and in its early stages, it is highly provisional even though it often is presented in a way that appears highly confident and conclusive (to those who are not in on the game). Reading this book has highly influenced how I think about markets and helps me to frame a lot of work that I’ve read this year regarding “market emergence.” (This is one topic in a literature review I’ll be publishing in the IEEE ProComm conference proceedings next month.)

On the other hand, I think the book could have easily been far more concise. Yes, this book wrapped up so much of what Callon had studied over his long and storied career. But a few rounds of ruthless editing could have made it more direct, less abstruse, and clearer. Certainly Callon is much more readable than I’ve found Foucault or Deleuze to be, but I think he could have accomplished his goals in 200 pages rather than 426. As is, the book is a bit of an undertaking. 

Still, if you are interested in markets or sociotechnical theory, I highly recommend this book.