The market ideology conception of fetishism: An interpretation and defense

When Charles de Brosses first coined the term ‘fetishism’ in On the Worship of Fetish Gods (1760), it was in a rather misled attempt to demonstrate the immaturity of ‘primitive’ religious cults (de Brosses, 1760; see Iacono, 1992, 51). Yet a little more than a hundred years later, Marx had turned the concept into one of the most deep-probing tools which social philosophy can bring to the study of capitalism.

The German philosopher-cum-economist had noticed the way in which his European contemporaries would still sneer at the West African religious habit of treating social objects as ‘independent figures endowed with a life of their own’, and he realized he could turn the joke on them: they themselves did the same with their own ‘immense collection of commodities’ (Capital I, 165 and 125; see also Iacono, 1992, 79–80, Heinrich, 2012, 179–81 and Graeber, 2005).

The ‘joke’, importantly, was a rather pointed one, and has remained so to this day. Just as Marx hoped to spur his contemporaries out of capitalism, the contemporary literature uses the concept of commodity fetishism to mount a radical critique of the capitalist market. Two main conceptions can be distinguished. According to the first conception, the concept of commodity fetishism alerts us to a form of market ideology that plays a crucial role in the reproduction of market domination (Cohen, 2000; Elster, 1986). On the second conception, by contrast, commodity fetishism refers to market domination itself, understood as a form of structural domination with a specific profit-maximizing logic (Roberts, 2017, Vrousalis, 2017, Ripstein, 1987).

In recent years, mainly thanks to the efforts of Roberts (2017), drawing on Arthur Ripstein (1987), the market domination conception seems to have taken precedence. This is unfortunate, I believe. Granted, the market domination conception has the undeniable benefit of emphasizing the profit-maximizing logic that distinguishes market domination from other forms of structural domination. But at a time when the detrimental effects of this logic have become well-known, the concept can provide a better ‘basis for resistance’, as Sally Haslanger would put it (cf. 2020, 36), by focusing less on market domination itself than on the exact workings of its ideological reproduction.

In this paper, therefore, I attempt to go against the grain. As we will see, this requires developing an innovative theoretical framework for understanding ideology—one which not only adapts to the market the influential account which Haslanger and others have offered in relation to racism and sexism (2012, 2017c; see also Celikates, 2016, Einspahr, 2010), but which also refines this account by showing how acknowledging the influence which structural domination may have on ideology helps it solve an important problem.

The proponents of the market domination conception ground their claim that fetishism is not market ideology in two main objections. First, that while fetishism may involve certain representations, these are not ideological since they are not epistemically deficient, at least not by inclusion of falsehoods (Roberts, 2017). Second, that anyway we should understand fetishism as concerning activity in the market, not representations of the market, lest we lose sight of its practical significance (Ripstein, 1987).

My first step is to argue that the market ideology view can escape both, provided it departs from the ‘theoretical’ interpretation of the analytic Marxists (Cohen, 2000; Elster, 1986) and favors instead the more ‘practical’ understanding of ideology that we can find most prominently in Halslanger's work (2017c). On this understanding, ideology refers to cultural schemas which are liable to epistemic deficiency by omission of truths, rather than by inclusion of falsehood, because they are involved in, indeed, crucial to, social activity and its practical failings. The first feature rescues the market ideology conception from the first objection, and the second from the second.

In my view, however, the real problem with the market ideology conception lies elsewhere: that is, in its apparent inability to explain how market ideology can play its central role, namely contributing to the reproduction of market domination. This is because it is unclear why, unlike other forms of ideology, it should be expected to be widespread enough across social milieux to impair agents' ability to break free from it. Ideology, after all, is never total (Haslanger, 2017b; see also Geuss, 1981).

To solve this problem, I draw on the market domination conception of fetishism, and specifically on its analysis of the profit-maximizing logic of market domination (Ripstein, 1987, Roberts, 2017, Vrousalis, 2017). I argue that this logic makes market ideology pervasive (see Sewell, 1992), which deprives agents of the symbolic-material discrepancies between social milieux that would help them break free from market ideology.

Thus, on the interpretation I offer, fetishism should be understood as the pervasive ideology of the market, where its pervasiveness is due to what can be called the standardizing influence of market domination across social milieux.

This new interpretation clarifies one of the central mechanisms—indeed, perhaps the most fundamental mechanism—of the ideological reproduction of structural domination in the market. As such, it provides a better basis for resistance to structural domination in the market than the market domination conception of fetishism, the analytic Marxists' interpretation of the market ideology conception of fetishism, and the influential conception of ideology we owe to Haslanger and others.

In the first section below, I focus on the market ideology conception, its analytic Marxist interpretation, and the two objections leveled against it by the proponents of the alternative, market domination conception. In Section 2, I build on Haslanger's influential account of ideology to offer a more ‘practical’ interpretation of the market ideology conception that escapes both objections. In Section 3, I turn to what I believe is the central problem with the market ideology conception of fetishism. Finally, in Section 4, I draw on the market domination conception to suggest a way to rescue the market ideology conception from this problem.

To set the scene, I begin with the relation between class domination, the market, and commodification. Then, I focus on the market ideology conception, its analytic Marxist interpretation, and the two objections.

Class domination refers to the situation of the worker who, forced to sell her labour power to some member of the capitalist class to make a living, finds herself unable to exit a relationship in which the relevant capitalist can get her to do more or less what she wants her to do for as long as she has bought her ability to work (see, for example, Gourevitch, 2018). Importantly, class domination depends on the market, which organizes the systematic transfer of resources from workers to capitalists that impairs the ability of the former not to work for some capitalist or other. As such, the market explains why there can be class domination without any legally or normatively sanctioned class distinctions (see, for example, Young, 1990, 47).

The market is a social practice of commodity production and exchange, that is, a collective solution to a coordination problem: the problem is for private producers to make a living in the absence of some form of central planning, and the solution the commodification process which generates market exchange and competition (Cohen, 2000; Roberts, 2017, for example, 78n97; Sewell, 1992; Vrousalis, 2017). Those who enact the market participate in commodifying things—from means of production to products to their own ability to produce—by repeatedly reducing them, by means of money and other measuring devices, to units which can be counted, added, compared, and converted into one another. This in turn enables and encourages them to produce and exchange things not as favors or as gifts, but as commodities, viz. in such a way that no one gives more than exactly what she gets (Graeber, 2001, 55–56 and 2011, 103–05, Sewell, 1992, 25–26, and Cohen, 2000, 416–23; see Ripstein, 1987, 736, Heinrich, 2012, chapter 3, and Marx, Capital I [1976], chapter 1).

Crucially, from the fact that things are commodities, not naturally, but because those who enact the market participate in commodifying them, we should not conclude that it is entirely up to those who enact the market whether to commodify or not: as Marx suggested, the commodification process tends to occur ‘behind their backs’, as an unintended result of their actions (Marx, 1976, chapter 1; see Sewell, 1992, 22 and 25). How exactly does this work? This, and its consequences for the entrenchment of structural domination in the market, is the focus of the market ideology conception of fetishism.

Ideology, on a broadly Marxist conception, has three related features: first, it entrenches structural domination, second, it is epistemically deficient, and third, it has a ‘tainted origin’ in structural domination (Geuss, 1981, 21; see Shelby, 2003, and Celikates, 2016). Indeed, ideology is often said to entrench structural domination because it is epistemically deficient, and to be epistemically deficient because agents learn it from dominating practices (see, for example, Haslanger, 2017a, Einspahr, 2010, Celikates, 2016). In line with this account, the market ideology conception of fetishism defines fetishism as an epistemically deficient view of the market, which agents infer from the way things appear to them, and which leads them to reproduce the market and entrench class domination (Cohen, 2000, 115, Geras, 1971, 78–79, Celikates, 2016, 14; see also Haslanger, for example, 2017c).¹

The view in question is epistemically deficient insofar as it does not alert agents to the fact that the things they exchange with their employers and clients are commodities because they themselves participate in commodifying them (see Cohen, 2000, 116, Elster, 1986, 57; cf. Torrance, 1995, 165 and 112–20). This epistemically deficient view, however, is not like a ‘hallucination’ attributable entirely to a failure of perspicacity on the part of agents, but is rather ‘like a mirage’, insofar as the external world itself is misleading (Cohen, 2000, 115; see also Elster, 1986, 56 and 177, Geras, 1971, 78–9). In particular, the ‘foundation’ of exchange-value ‘in labouring activity is not [visible]’ (Cohen, 2000, 116). This epistemically deficient view, finally, explains why commodification occurs ‘behind the backs’ of agents, as an unintended results of their actions, and so why they fail to change the market for the better and entrench class domination as a result: agents will not even think of organizing production and exchange differently if they fail to realize that things are commodities because they themselves participate in commodifying them. Thus, as Cohen puts it, ‘[f]etishism protects capitalism’ (2000, 129).

This is the core of the market ideology conception of fetishism, and as such it is common to the two main interpretations of this conception: the dominant, analytic Marxists's interpretation put forward most famously by Gerald Cohen (2000) and Jon Elster (1986), and the socialization-based account which I will adapt from feminist and antiracist political theory in the next section (Haslanger, 2012, 2017c; see also Celikates, 2016, Einspahr, 2010). The difference between the two interpretations lies in the exact way in which the ideological view in question is epistemically deficient.

According to the analytic Marxists, this view is not epistemically deficient merely by omission of part of what commodities are, viz. their being commodities because agents participate in commodifying them, as it will be on my alternative interpretation. Rather, it is epistemically deficient by inclusion of something which commodities are not, viz. their being commodities ‘as an inherent property’, as a matter of ‘substance’, or, more generally, ‘naturally and inevitably’ (Elster, 1986, 57 and Cohen, 2000, 116 and 127, respectively). Thus on the analytic Marxist view, and in Cohen's words, there are ‘two phases in commodity fetishism: (1) separation of exchange-value from its material basis; (2) attachment of exchange-value to the substance of the commodity’ (2000, 117; cf. 116). On my alternative interpretation, by contrast, only the first phase obtains.

Insofar as including such ontological precisions as an account of the substance of commodities is the mark of a theoretical attitude, one on which agents flesh out their view of things beyond what is ‘practically necessary’ (Torrance, 1995, 48), we may say that for the analytic Marxists, the ideological view in question is a theoretical view. 2 On such a view, commodities are endowed with a natural dimension, where this means not only that their social dimension is stripped out from the way they appear, as it is on my alternative interpretation, but also that their appearance is more elaborate than it is on my alternative interpretation: to those who hold this view, commodities appear not just as things whose exact value must be paid if they are to change hands, but also as things which possess value by virtue of their physical properties, ‘just as they have weight’ (Elster, 1986, 57).³

This interpretation of the market ideology conception of fetishism has recently come under sustained pressure from the proponents of the market domination view of fetishism. William Roberts (2017), in particular, has built on Arthur Ripstein (1987) to argue that ‘[f]etishism ought to be understood as a form of domination rather than a form of false consciousness’ (2017, 85). I will describe the form of domination they have in mind in Section 4. Here, I focus on their central objections.

The first is Roberts's. As we saw above, to define fetishism as market ideology involves claiming that the view which agents infer from the way the things they exchange appear to them is epistemically deficient. But this view is not epistemically deficient, Roberts argues, since things appear as what they are to those who hold this view. As he puts it, the market ideology conception of fetishism ‘trips over Marx's explicit claim that, in fetishism, “the social relations between [the producers’] private labours appear as what they are”. Where social relations are mediated by commodities, exchanges are the real relations between the producers of commodities' (Roberts, 2017, 86–87, quoting Marx, Capital 1, 1976, 166, to which he adds the emphasis).

For those of us less interested in exegesis than Roberts, the fact that this was ‘Marx's explicit claim’ may not carry additional weight. Yet Roberts is also interested in theorizing fetishism further, and he makes it clear that he considers Marx to be right on this point: fetishism concerns social reality as it is constructed by ‘the social practice of exchange’ (2017, 88n133). I agree, as suggested above. In fact, this point is widely accepted in the literature on fetishism. The analytic Marxists themselves concur, insisting, as we saw above, that fetishism is ‘like a mirage’, insofar as the external world itself is misleading (Cohen, 2000, 115; see also Elster, 1986, 56 and 177, Geras, 1971, 78–9).

This, however, makes the analytic Marxists's interpretation vulnerable to Roberts's objection. On their ‘theoretical’ reading, recall, the ideological view at the heart of fetishism fails to alert agents to part of what commodities are (viz. their social aspect) not just by omission of it, but by inclusion of something which commodities are not (viz. their purported natural aspect). To those who hold this view, therefore, commodities do not appear as what they are: they appear to be naturally commodities while, really, they are commodities by social construction. Thus the analytic Marxists' theoretical reading of the market ideology conception is vulnerable to the first objection.

The same goes for the second objection, viz. that the ideological view at issue is irrelevant since fetishism is best understood as concerning social activity rather than theoretical understanding. This objection is raised by Ripstein, who insists, against Cohen and the classical proponents of the market ideology conception, that fetishism ‘is the failing associated with practical involvement in the world’, not with ‘the virtue of knowledge’, because activity and understanding are distinct modes of engagement with the world, and fetishism is about activity, not understanding (Ripstein, 1987, 743).

Like Roberts above, not only does Ripstein take this to be Marx's claim, but he also insists that Marx was right. For to focus primarily on theoretical understanding rather than social activity, he suggests, is to risk concentrating on ‘how knowledgeable or error prone’ people are in the market, to the detriment of the fact that ‘[t]o have one's activities entirely shaped by [the market] is to be enslaved […]’ (1987, 743 and 747).

I agree that Ripstein's—and Marx's—primarily practical orientation is crucial. The concept of fetishism should help us focus on structural domination in the market, be it to draw our attention to the phenomenon itself, as Ripstein believes, or to alert us to the exact workings of its ideological reproduction, as I think would be more useful at this stage. I also agree that the risk of losing sight of this practical predicament is real on a ‘theoretical’ interpretation of the market ideology conception of fetishism. The analytic Marxists testify to this. Elster, for instance, concludes that ‘[fetishism] is a cognitive illusion arising from market transactions, not a morally deplorable feature of markets’ (1986, 58, my emphasis). Likewise, Cohen, mired in theoretical debates about the labour theory of value, offers only an ‘explanation without defence of Marx's views [on fetishism]’ (2000, xii, his emphasis). For fear of losing sight of social activity and its current predicament, therefore, we should not take fetishism to concern theoretical understanding. This is to say that the analytic Marxists' conception of fetishism falls prey to the second objection, as it did to the first.

I conclude, with Roberts and Ripstein, that fetishism should not be conceived as the ideology of the market if ideology is understood as the analytic Marxists understand it, that is, theoretically. Things are different if ideology is understood practically, as I argue in the next section.

The ‘practical’ understanding of ideology has recently become influential in feminist and antiracist studies, largely thanks to the efforts of Sally Haslanger (e.g. 2017c). In this section, I adapt it to the market, in order to offer an interpretation of the market ideology conception of fetishism which escapes the two objections encountered in the previous section.

We saw above that, according to the analytic Marxists, the ideological view at the heart of fetishism is epistemically deficient not merely by omission of part of what commodities are (viz., their social aspect), but also by inclusion of something which commodities are not (viz., their purported natural aspect), which I suggested was the mark of a theoretical mode of engagement with the world.

By contrast, on the conception of ideology put forward by Haslanger and others, ideology is much more ‘practical’. Developed as part of a critico-theoretical, socio-constructivist account of gender and race (see, for example, Haslanger, 2012), this conception offers an account of ideology as the schemas of dominating practices insofar as agents are socialized into them, where socialization is understood as the process through which agents learn from the milieux they frequent the schemas of the practices enacted in these milieux (Haslanger, 2017b; see also Celikates, 2016).

While Haslanger does not apply this account to specifically capitalist and classist issues, it can be fruitfully put to the service of the market ideology conception of fetishism. Seen in this light, commodity fetishism can be understood as a schema of the market insofar as agents are socialized into it. According to the broadly Bourdieusian account of socialization Haslanger relies on, this suggests that to fetishise the market is to keep applying the commodity schema to things, because in this milieu things ‘actualise’ or ‘incorporate’ this schema, and so ‘teach’ or ‘inculcate’ it in turn (Sewell, 1992, 12–13, quoted in Haslanger, 2017c, 22; see also Einspahr, 2010, and Bourdieu, 1977).⁴

Importantly, socialization is, on this account, oriented to helping agents enact social practices, understood as ‘collective solutions to coordination or access problems’ (Haslanger, 2016, 126). This means that the schemas of these practices, being geared primarily to helping agents solve collective problems, often focus on those aspects of things which are useful in this respect, to the detriment of other, less immediately relevant aspects (cf. Torrance, 1995, 45–47). Given this, the commodity schema can be understood as the schema that things are commodities full stop, not that they are commodities naturally (or not), this latter aspect falling outside the immediate concern of agents aiming to make a living.⁵

In other words, on the interpretation we can build from Haslanger's account of ideology, the commodity schema is epistemically mistaken not by inclusion of something which commodities are not, as it is for the analytic Marxists, but instead by omission of part of what commodities are. As Haslanger would put it, it only ‘leave[s] out’ the fact that commodities are such because we participate in commodifying them, and this is how it ‘obscure[s] [their] social dimension’ (Haslanger, 2012, 18 and 467; see also Celikates, 2016). In other words, it does not give them a more elaborate appearance by representing their (mistaken) ontological status, as it does on the analytic Marxists' interpretation: here, commodities only appear as things whose exact value must be paid if they are to change hands, not as things which, in addition, have their value in this way or that.

This point is crucial, because it enables this alternative interpretation of the market ideology conception of fetishism to escape the two objections from the previous section.

The first objection, recall, was that the commodity schema is not epistemically deficient, since things appear as what they are to those who hold it (Roberts, 2017). As noted above, I agree with Roberts—and Marx—that fetishism is about the social construction of reality by the market: as we may now say with Haslanger, things on the market ‘actualise’ or ‘incorporate’ the commodity schema, and to this extent appear as what they are to those who hold this schema.

But from this, crucially, we should not conclude that the commodity schema cannot be epistemically deficient as well, and, therefore, ideological as well. From the fact that commodities do appear as what they are to those who hold the commodity schema, it does not follow that they appear as everything that they are. It may well be the case that everything that commodities appear to be (viz., commodities) is true of them, but that the schema still leaves out some crucial fact about them (viz., their social dimension).

In other words, Roberts's objection applies to the analytic-Marxist view, as we saw above, but not to the alternative interpretation I just offered. To agents whose ‘theoretical’ view of commodities fails to alert them to part of what commodities are (viz. their social dimension) by inclusion of something which commodities are not (viz. their alleged natural dimension), commodities do not appear as what they are. But to agents whose ‘practical’ schema of commodities is epistemically mistaken only by omission of part of what commodities are (viz. their social dimension), commodities do appear as what they are (viz. as commodities), if not as everything that they are. Therefore, the alternative interpretation, unlike the analytic Marxist view, escapes the first objection.

The same goes for Roberts's objection, viz. that the commodity schema is irrelevant since fetishism should concern market activity rather than theoretical understanding, on pain of neglecting the structurally dominated character of this activity. As I wrote above, I agree with Ripstein that a focus on structural domination in the market is crucial. But I also hold that, unlike the analytic Marxists' interpretation, my alternative interpretation can focus on understanding without running the risk of losing sight of this practical failing.

Indeed, from the fact that fetishism should not concern theoretical understanding, it does not follow that fetishism cannot be about understanding at all. For even if fetishism is not about the kind of theoretical understanding the analytic Marxists focus on, it can be about the kind of practical understanding on which the alternative interpretation insists. Ripstein's distinction between activity and understanding is too stark: it misses the fact that activity involves understanding in the form of the schemas (e.g., the commodity schema) which enable agents to enact their various social practices (e.g., the market).

Put differently, from the fact that ‘all social life is essentially practical’ (Marx, Eighth Thesis on Feuerbach, quoted in Torrance, 1995, 45, original emphasis), we should not conclude with Ripstein that fetishism ‘is the failing associated with practical involvement in the world’ rather than with any kind of understanding (1987, 743). Instead, we should conclude with Torrance that ‘what people observe and how they experience and describe their surroundings depends on their purposes and the problems they face’ (Torrance, 1995, 45)—that is, more precisely, on the schemas which help them enact the social practices in which they participate and which, geared as they are to helping them solve collective problems, may fail to alert them to more than they strictly need to understand in order to do so (cf. Torrance, 1995, 47).⁶ The commodity schema is a case in point: as we saw above, it leaves out the social dimension of commodities, this otherwise crucial fact being of little immediate interest to agents focused on making a living. Thus, while it would indeed be a mistake to take fetishism to concern the kind of theoretical understanding which the analytic Marxists have in mind, the concept can be about the kind of practical understanding involved in social activity on which the alternative interpretation operates. Therefore, this interpretation escapes the second objection, as it escaped the first.

I conclude, contra Roberts and Ripstein, that fetishism can be understood as the ideology of the market, provided that ideology is understood practically rather than theoretically. The main problem with the market ideology conception of fetishism lies elsewhere, as I argue now.

The problem is the following: the market ideology conception of fetishism seems unable to explain how the commodity schema can play the central role which it is meant to play, viz. contributing to the reproduction of market domination. This is because it is unclear why it should be expected to be widespread enough across social milieux to impair agents' ability to break free from it, when, in general, ideologies should not be expected to be so widespread.

On the practical interpretation I offered above, it is agents' socialization into ideology, understood as the process through which they learn ideology from the social milieux in which it is embodied, that explains ideology's domination-reproducing effect. But as is widely recognized, including by Haslanger herself (2017b; see also Celikates, 2016 and Geuss, 1981), ideology is never total. In the case at hand, things should be expected to embody and teach the commodity schema with employers and clients, but the gift schema with friends or family, and the favor schema with neighbors. Given this fragmentation, it is unclear how agents can fail to correct their schema of commodities by learning that things are commodities at work because they are treated as such in this milieu but not in others—and, therefore, how they can find themselves in the grip of anything like an ideology of the market that has them participate in commodification ‘behind their back’ and so contribute to the reproduction of market domination.⁷

To make this argument in more detail, the first thing to emphasize is that agents are socialized into a multiplicity of social practices, some of which are not capitalist. As Haslanger insists, ‘[t]here are multiple reasons to avoid the idea that ideology [e.g., capitalist schemas] functions as a total system governing society as a whole’ (2017b, 161). Indeed, Robin Celikates emphasizes, we should bear in mind Raymond Geuss's famous claim that ‘a society of happy slaves, content with their chains […] is a nightmare, not a realistic view of a state of society which is at present possible’ (1981, 83–84, quoted in Celikates, 2016). In other words, we should expect ideological schemas to be actualised in things in some of the milieux which agents navigate, but not in others.

Yet if so, crucially, we should also expect agents to learn that things actualise a schema because they themselves incorporate it in things in some milieux but not in others. Agents are ‘knowledgeable’, after all (Giddens, 1979, 5; see also Celikates, 2006, Boltanski & Thévenot, 2006), and should be expected to realize that if things actualise different schemas in different milieux, it is largely because they themselves apply different schemas to them. Indeed, it is ‘a leading theorem’ of Anthony Giddens that ‘every social actor knows a great deal about the conditions of reproduction of the society of which he or she is a member’ (1979, 5, emphasis removed; see also Scott, 1985, 319).

This is particularly clear with the commodity schema. If socialization is fragmented in this way, we should expect the commodity schema to be incorporated in things when people exchange them with their employers or clients, but not when they exchange them with their neighbors, friends, or family. As Graeber (2011) and Cohen (2009) insist, commodity exchange is very different from the kind of exchange that in principle occurs between neighbors, friends, or family. The schema of commodity exchange is not just ‘give as good as you get’, as in neighborly exchange, but rather ‘give exactly as good as you get’, the price being the expression of this mathematical equivalence made possible by the reduction of things to countable units (Graeber, 2011, chapter 5). The discrepancy is even more striking with close friends and family. There, the schema is the ‘baseline communist’ one, ‘to each according to their needs and from each according to their means’ (Graeber, ibid.; cf. Cohen, 2009, 39–45). Thus, in principle, the things which agents exchange with neighbors, friends, or family do not actualise the commodity schema, but the favor schema or the gift schema (as the case may be). When plumbers fix the bathrooms of their relatives for free, for example, their ability to work does not incorporate the commodity schema. It is not standardized but personalized, and the plumbers' relatives are not presented with a bill.⁸

Yet, crucially, if, as I just argued, the things which agents exchange with their neighbors, friends, and family do not incorporate the commodity schema in the way that the things which they exchange with their employers and clients do, then it is unclear how the commodity schema can leave out the fact that these things are commodities largely because people participate in commodifying them. People, again, are not ‘judgmental dopes’, in Garfinkel's famous phrase (1984, 75; see also Celikates, 2006, 30 and Roberts, 2017, 95), and if things incorporate different schemas in different milieux, the discrepancies between things or milieux should help them realize that they themselves participate in the incorporating. When a plumber fixes her neighbor's kitchen sink because he previously babysat her children and she owes him one, or repairs her parents' bathroom for free, she can hardly fail to infer that if her ability to work is a commodity at work, it is partly because she participates in commodifying it by selling it to her employer or clients for an hourly wage.⁹ In other words, we should expect her commodity schema to include the fact that her ability to work is a commodity at work in part because of herself and her employer, just like we should expect her gift schema and her favor schema to include, respectively, the facts that her ability to work is a gift in family milieux partly because it has been made so by herself and her parents, or that its being exchanged as a favor in neighborly spheres is due, to a significant extent, to her and her neighbors' making it so. To expect any less of them would fail to do justice to what we can call, with Giddens, their ‘penetration’ (1979, 71–72).

If socialization works in this fragmented way, then, we should expect things to actualise the commodity schema with employers and clients, but the gift schema with family and friends and the favor schema with neighbors. But this suggests in turn that we should not expect agents' commodity schema to limit itself to the fact that at work things are commodities, omitting the further fact that, if they are so, it is largely because agents themselves participate in commodifying them in this milieu. If so, however, then it is unclear how the commodity schema can explain how commodification may ‘work “behind the backs” of the social actors who produce and reproduce [it]’, as Giddens would put it (1979, 71), and thus prevent them from changing the market for the better and from entrenching class domination as a result.¹⁰

Put differently, the market ideology conception of fetishism fails to adequately explain the ideological reproduction of structural domination in the market.

To rescue the market ideology conception, I draw on the market domination conception: specifically, on its analysis of the profit-maximizing logic of market domination (Ripstein 1986, Roberts, 2017, Vrousalis, 2017). I argue that this logic makes market ideology pervasive (see Sewell, 1992), which deprives agents of the epistemic discrepancies between milieux that would help them break free from it and, therefore, explains why commodification may occur ‘behind their backs’. On this modified interpretation, fetishism should be understood as the pervasive ideology of the market, where its pervasiveness is due to what can be called the standardizing effect of market domination across social milieux.

As we saw above, the ideology conception of fetishism has recently come under pressure from Roberts (2017), building on previous work from Ripstein (1987). I have already pushed back against their objections above. Now I want to focus on what I take to be the central insight of the market domination conception of fetishism they offer. This is not meant to constitute an endorsement of their view, however, but merely an acknowledgement that the phenomenon they focus on is of the greatest importance in the analysis of fetishism, if not what fetishism is.

On the market domination conception, fetishism is a form of domination that is distinct from, even as it influences, class domination: it is the impersonal domination suffered by both workers and capitalists because of market competition (see Roberts, 2017, 88 and Vrousalis, 2017, 379). As Roberts insists, ‘the dominant class in modernity, the class of capitalists, is as subject to impersonal domination as are the laboring classes’ (2017, 102). Vrousalis agrees: Roberts, he writes, is right that ‘capitalist [or impersonal] domination is not equivalent to class domination’ (2017, 379). Ripstein concurs: according to him, the central feature of this domination is that ‘the options of all are limited by the market’, not merely those of workers relative to capitalists (1987, 747).

The complete definition of this impersonal domination is the object of some debate between Roberts and Vrousalis, who disagree as to who is to be held responsible for it, and whether it is arbitrary in some sense (see Roberts, 2017; Vrousalis, 2017). But let me bypass these disagreements to focus on what both Roberts and Vrousalis agree on, and Ripstein as well. This is that both workers and capitalists are dominated into maximizing profit. As Vrousalis insists, market competition affects radically the power of the capitalist class over the working class. Without competition, each capitalist can act as an ‘absolute monarch’ over their workers, unconstrained in the wage they offer and in what they ask them to do. But as soon as they face (perfect-enough) competition, they can no longer determine wages arbitrarily, or have workers do whatever takes their fancy: at this point ‘[each] is constrained, on pain of competitive disadvantage, to maximise profit, which in turn requires paying [workers] a market-clearing wage’ and exploiting them (2017, 380–81). Roberts concurs, writing for instance that ‘[t]he capitalist, dominated by market imperatives, is compelled thereby to exploit labor’, which by definition entails making a profit (2017, 102). Ripstein agrees as well, emphasizing both that the worker ‘must make himself marketable and once sold, direct his activity to whatever his employer demands’, and that ‘[t]he employer's options are broader but still limited: On pain of bankruptcy, this demand can only take a single form: produce what is profitable’ (1987, 748).

In other words, all the proponents of the domination conception of fetishism agree that in an important respect the domination at issue is ‘non-arbitrary’ in the sense of ‘regulated’—indeed, regulated by one key rule: ‘maximize profit’ (Vrousalis, 2017, 381).¹¹ If Roberts downplays this aspect, it is only to emphasize that this domination is in another sense arbitrary, viz. whimsical. For the two aspects seem to clash, and it is on this apparent clash that Vrousalis insists when he characterizes this domination as ‘non-arbitrary’. But the clash is only apparent, as it is quite compatible for the domination that on each view constitutes fetishism to be both arbitrary and non-arbitrary in these senses. The market may have movements that are quite difficult to anticipate and, at the same time, still channel capitalists (and workers in their wake) in one definite direction: that of making profit. As Sewell emphasizes, what characterizes capitalism is precisely both a ‘chronic instability or unpredictability’ and ‘a continuous dynamic of capital accumulation […]’ (1992, 25–26).¹²

Now, Vrousalis also insists, against Roberts, that the agents of this domination are not ‘markets or market imperatives’, but ‘[c]apitalists who dominate each other by jointly constituting the ‘external coercive necessities confronting the individual capitalist’ (2017, 3, quoting Marx, 1976, 381). Here, I do not engage in this debate, however, for it would take us too far afield. Instead, I focus on the profit-maximizing logic that they all agree on. For it offers a solution to the problem identified in the previous section.

The reasoning is this: if capitalists and workers are dominated into maximizing profit, then they are dominated into incorporating the commodity schema beyond the workplace, in as many milieux as competition requires. This in turns deprives them of the discrepancies between milieux that could help them realize that they themselves are incorporating the commodity schema in things, which explains how this schema can fail to include this fact.

As Sewell emphasizes, ‘the commodification of things’ is at the core of the ‘continuous dynamic of capital accumulation’ and profit maximization we saw him mention earlier (1992, 25). Capitalists, on pain of competitive disadvantage, have a strong incentive to ensure that as many things are commodified as possible, and so, therefore, do their workers. Only commodities are ‘opportunities for profit’ after all, and competition requires capitalists to maximize profit. This is why, in Sewell's words again, ‘the commodification of things’ is now ‘pervasive’, that is, ‘present in a relatively wide range of institutional spheres, practices, and discourses’ (ibid., 25 and 22): the ‘chain of commodity exchange’ is ‘vast’, as ‘the commodity form […] organizes a virtually universal intersection of resources’ (ibid., 26). In this respect, it is no surprise that Marx's first description of capitalism in Capital is as ‘an immense collection of commodities’ (1976, chapter 1). The domination identified above turns commodity exchange into a practice that is pervasive in a way few practices are.¹³

Now, Sewell seems to take such pervasiveness to result from the fact that the commodity schema is ‘exceptionally transposable’ (1992, 25), rather than from the profit-maximizing logic of the domination which Ripstein, Roberts, and Vrousalis insist on. Sewell nowhere mentions domination, but instead insists that the commodity schema itself ‘knows no natural limits’ as ‘it can be applied not only to cloth, tobacco, or cooking pans, but to land, housework, bread, sex, advertising, emotions, or knowledge […]’ (ibid., 25–26). But this seems to me mistaken. Exceptional transposability is not exceptional transposition, and a motivating force—‘a force that requires it’, as MacKinnon puts it (1982, 540)—is crucial for the actual rather than merely virtual pervasiveness of the chain of commodity exchange. This motivating force is the domination with a profit-maximizing logic that affects both capitalists and their workers in the market.

Thus, while capitalists and workers might try to keep the things they exchange with their family largely uncommodified, they will often expand the commodity schema to the things they exchange with friends and neighbors. The magazine Plumbing Connection provides a telling example of this phenomenon in an article entitled ‘The Deal with Mates Rates’, by Brad Fallon (2014). Fallon begins by noting that ‘people with trade skills are always faced with the old ‘mates rates' dilemma’. He frames the dilemma as follows: ‘with a client […] I just roll off the invoice – job well done’; ‘however, add a stressed friend, relative or neighbour into the scenario – someone who I see all the time, whether in my street, at school, or socially – and suddenly, I have this overwhelming need to become the not-for-profit, happy to spend my weekend plumbing “for free” emergency plumber’. And so he goes on to offer ‘some of [his] most helpful tips for ensuring that [people with trade skills] are adequately paid for the work [they] do’. The tips can be grouped in two categories: making excuses, and commodifying, the only exception being the family (‘Obviously […], if it is your Mother-In-Law knocking on the door, throw all the rules out and do the job straight away for free’). One good tip, in particular, consists in ‘booking the job in during standard work hours with one of [your] staff members'. As Fallon emphasizes, this ‘change[s] the dynamic of the relationship back from a personal favour to a professional plumbing service’. Other tips go in the same direction.¹⁴ As this example illustrates, capitalists and their staff members are dominated into extending the commodity schema beyond the workplace to as many milieux as competition requires – even with their friends, if not with their mother-in-law.¹⁵

But this standardizing influence of structural domination on socialization, crucially, offers a solution to the problem affecting the market ideology conception of fetishism. The problem, recall, was that if things incorporate different schemas in different milieux, which they should do if socialization works in the fragmented way it is widely recognized to work in, then the discrepancies between these milieux should prompt agents to realize that they themselves participate in the incorporating, and to include this fact in their commodity schema. If our plumber fixes her neighbor's kitchen sink because the neighbor previously babysat her children and she owes him one, or her parents' bathroom for free, we should expect her schema of her ability to work in the workplace to tell her that it is partly commodified by herself and her employer. Brad Fallon is a case in point: he complains that ‘because I like to “help” my friends’, […] it feels wrong to charge them’ (2014). But the standardizing influence of structural domination offers a solution to this problem. For if the plumber, or any other market agent, is dominated in such a way that, with her neighbors or friends, she does not exchange her ability to work as a favor or as a gift but as a commodity, in just the same way as she does with her employer and customers, then the move from one milieu to the next will be so natural that it will blunt her critical consciousness: specifically, she will not realize that she herself is playing an important part in actualising the commodity schema in her ability to work, and her version of the schema will not include this fact. Fallon's version of the schema, because he has reflected on it, may be sufficiently critical, but he is an exception. Indeed, he insists, ‘most friends, neighbours and relatives […] don't actually want a discount or preferential treatment’. In fact, usually they do not even ask him ‘to discount [his] prices' (ibid.). Unlike him, they may well fetishise commodities.

More generally, I want to suggest the following explanation of the mirage at the heart of fetishism. If the commodity schema is incorporated not just in the things agents exchange with their employers and customers, but also in the things they exchange with their ‘friends, relatives or neighbours’, as Fallon puts it, then the schema they infer from the commodities they exchange with their employers and customers will leave out the fact that they themselves are partly responsible for the commodification of commodities. For when a schema is pervasive in this way, there are no hitches between the spheres of activity agents navigate, and they stop noticing that they are applying it themselves: as Sewell puts it, schemas that are incorporated in a relatively wide range of milieux tend to become ‘relatively unconscious, in the sense that they are taken-for-granted mental assumptions or modes of procedures that actors normally apply without being aware that they are applying them’ (1992, 22; see also 24 ff.). Agents in such a standardized environment are deprived of the prompts that could have helped them realize that they are following the standard, and end up doing so without thinking (cf. Graeber, 2005, 431).¹⁶ One might say that the commodity schema, in particular, is reflected back to them by so many things that they become ‘naturals’ relative to it, applying it ‘naturally’ to the things that reflect it, aware only that they are exchanging them as commodities, not that they are contributing to commodifying them as they do. Their critical consciousness is blunted, and this is why the commodification process occurs ‘behind their backs’, as an unintended result of their actions.

The commodity schema does not alert capitalists and workers to the fact that it is because they participate in commodifying them that the things they exchange are commodities, and it does not alert them to this fact because they participate in the incorporation of this schema in things without thinking. They do this, in turn, because they are dominated into incorporating this schema in things, not just in the workplace, but pervasively, for instance in friendly reunions and neighborly encounters, if not in family settings. In other words, the impersonal domination at issue deprives them of the standpoints from which to realize that they are participating in the construction of social reality in such a way that they entrench, in a vicious circle, their own domination (see Lahire, 2001; MacKinnon, 1982).¹⁷

Thus fetishism, or the ideology of the market understood in terms of socialization into the market, entrenches structural domination because structural domination, in the form of the impersonal domination Vrousalis, Roberts and Ripstein emphasize and of the class domination it affects, has a standardizing influence on agents' socialization.

Put differently, if the ideology of the market is understood to be both practical (against the theoretical interpretation of the market ideology conception) and rendered pervasive across social milieux by what I have called the standardizing influence of market domination (against the current interpretation of the practical conception of ideology), then the market ideology conception of fetishism can shed some new light on one of the key mechanisms of the ideological reproduction of structural domination in the market. This, I conclude, would provide us with a better ‘basis of resistance’, in Haslanger's phrase (2020, 36), than the market domination conception of fetishism, the analytic Marxist interpretation of the market ideology conception, and the influential account of ideology developed by Haslanger herself.