Are Theories Politically Flexible?

Abstract

Social theories are politically flexible if people use them to support opposite political claims. But is this even possible? And what kind of theories have such a property? Moving beyond epistemological debates about neutrality and value-leadenness, this article defends an empirical approach to the study of flexible and rigid political uses of social theories. I identify two main sources of flexibility: endogenous properties of theories, notably their generality, ambiguity, and neutrality, and exogenous features of the contexts in which they are used and of the individuals who use them, notably differences in political and epistemic culture.

The Possibility of Political Flexibility

If we accept, as most sociologists do, that cultural representations and ideas are not merely an epiphenomenon but play an active role in shaping social reality, then social theories are no exception. Under certain conditions, social theories can affect how people behave, what laws are in place, and how societies look. As Keynes ([1936] 2013:383) famously quipped, whether they do it consciously or invertedly, “practical men who believe themselves to be quite exempt from any intellectual influence, are usually the slaves of some defunct economist” or of “some academic scribbler of a few years back.” And to the extent the actions theories inspire are contentious, their use can be political. Politicians, revolutionaries, managers, judges, terrorists, and many others regularly invoke theories in support of a political goal. Some take their claims as clear indication there is something logical and inevitable about the pairing between a theory and a political project. But is that so? Does a theory compel its users to act only in a certain way, to defend only one political project, or can a theory be used to make opposite political claims? In other words: are theories politically flexible?

The answers to such questions are consequential, on practical and theoretical grounds. The approval or rejection of a theory (and of its authors) often depend on who might use it and with what purposes. ¹ Fearing guilt by association, it is not uncommon for social scientists to complain their theories are being misused for political purposes and to dissociate themselves from unwelcome adherents, such as when a leading proponent of democratic peace theory complained in 2005 that the theory had been “perverted” by the Bush administration to justify the U.S. invasion of Iraq (Russett 2005:396). If it turns out theories are politically rigid, that is, in practical disputes the same theory can only be consistently used by one side, this would make the idea that social scientists are “moral philosophers in disguise” (Abend 2008b:87) much more plausible, potentially undermining the objectivity of social knowledge. It would also support the Kuhnian view that theories are complex entities formed around a conceptual core that include normative assumptions. On the other hand, if social theories turn out to be politically flexible, this would mean social scientists pursuing research on topics deemed politically inappropriate have an additional argument to defend their theoretical choices. It would also vindicate a relational and antiessentialist view of cultural products, according to which ideas acquire their meaning through their use and through their relation to other ideas in the same context of interaction (see Abbott 2001; Emirbayer 1997; Skinner 1969).

Two antithetical approaches are commonly adopted to make sense of the political dimension of social science theories. The first is the critique of social scientific ideology (e.g., Blakely 2020; Gorski 2013; Haraway 1988; Taylor 1985). According to this approach, theories are loaded with normative assumptions that guide those who accept them toward a limited set of political options. Seeing the world through the lens of a theory has clear political consequences: certain groups are portrayed in a better light than others, and certain political projects are dismissed as useless or dangerous. Critics of ideology warn against trusting and legitimizing theories that can be used, often covertly, to serve the interests of harmful and powerful groups.

The other approach is the defense of social scientific value-freedom (e.g., Black 2013; Campbell 2014; Collins 1989; Nagel 1961). Its proponents claim that although normative elements are often combined with social theories, there is nothing necessary about this relation. As Schumpeter (1949:353–54) said about Adam Smith’s economic thought, “there is some semiphilosophical foliage of an ideological nature but it can be removed without injury to his scientific argument.” According to defenders of value-freedom, a theory can be broken down to its basic scientific components, and all normative addition is to be considered a contingent evaluation that lays on a logically distinguished plan. The building blocks of the theory are independent from political judgment and do not bind one to support certain policies or practices. Defenders of value-freedom argue it is more fruitful to put political disagreements aside and focus instead on improving the scientific dimension of theories.

I argue that both these approaches offer only limited help for understanding whether social theories are politically flexible. The critique of ideology correctly recognizes the politically consequential character of social theories and hinges on the sensible sociological notion that society is divided into groups that have specific interests and beliefs. However, it regularly ends up undermining the complexity of cultural practices: the way people use ideas in practical situations is much messier and opportunistic than this model assumes. Instead of studying how theories are used in concrete settings, such as parliaments, courts of law, board rooms, advisory committees, and ordinary conversations, this approach often postulates that such uses will stream straightforwardly from the content of a given theory. The defense of value-freedom is in principle more receptive to the variety of practices associated with the use of theories. However, because its emphasis is on defending the norm of neutrality, this approach usually disregards how theories are actually put to use and how they are associated with various political projects. The fact that a theory is used more by some political groups than others is seen as a mere accident that does not undermine its neutrality.

In contrast to both these approaches, I argue that the political flexibility of theories should be treated like any other social phenomenon: whether and when it occurs presumably depends on several factors that can be empirically observed. Following recent developments in the sociology of ideas (see Camic, Gross, and Lamont 2011), I argue that we need a much wider array of evidence that theories are ideological or neutral, grounded not only in textual analysis, but also in the study of actual practices in a variety of settings. Only an approach of this type would allow for the possibility and explanation of the fact that a given theory is used to make antithetical normative claims and support opposite political projects. In contrast with the critique of ideology, which assumes the relation between ideas and politics can only be rigid, and with the defense of value-freedom, which assumes all relation is purely accidental, an empirical approach allows one to identify and explain varying degrees of rigidity and flexibility. By adopting such a stance, we might discover not only that social theories are politically flexible, but also that certain theories are more flexible than others, and that certain factors increase or limit such flexibility.

As we shall see, various scholars, and even whole scholarly traditions, have hinted at the possibility that theories are politically flexible. But in most cases, these have been unsystematic allusions that need to be analyzed and expanded. Drawing on this diverse literature, I have three main goals in this article. The first is to lay the foundation for a better understanding of political flexibility by showing that how we answer such questions largely depends on how we define several key concepts (see the Terms and Conditions section). The second is to offer some preliminary remarks on what factors increase the political flexibility of a theory. In particular, I distinguish between “endogenous” and “exogenous” sources of flexibility, where the first are related to intrinsic properties of theories, notably their generality, ambiguity, and neutrality, and the second are related to different features of the context in which they are used and the individuals who use them, notably differences in political and epistemic culture (see the Two Sources of Political Flexibility section). The third goal is to show how considering political flexibility as a sociological puzzle can open new avenues for theoretical and empirical research (see the Conclusion).

Terms and Conditions

As a provisional definition, we might say political flexibility occurs when the same social theory is used in support of opposite sides of a political conflict, either in the same context or in different contexts characterized by relatively similar political cleavages. To unpack this formulation, we need to define at least the following components: (1) What counts as a social theory? (2) What does it mean to use a theory? (3) What does it mean to use a theory politically? (4) What does it mean that two sides in a political conflict are opposed? and (5) What counts as a context? Each of these questions would require a paper by itself to be properly addressed. I can offer only some remarks here to clarify how I see these issues.

Social Theories

The crucial concept, of course, is “social theory” itself. Although there is little consensus on what “theory” means in sociology and cognate disciplines (see Abend 2008a), a common view defines theories by contrasting them to factual statements: factual statements describe how the world is (or was, or will be), whereas theories offer an explanation or interpretation of why the world is as it is. Theories are complex explanations or interpretations of puzzling social phenomena that are not merely descriptive (and presumably not merely metaphysical either). ² Concepts are often seen as occupying an intermediate position between theories and factual statements. But if one follows the distinction between description and explanation, it is common to lump them together with descriptive propositions: concepts are not, properly speaking, the explanations of puzzling phenomena, but rather the variables that need to be explained, although they might be used as shorthand to refer to an explanation and might have other methodological and heuristic functions (see Swedberg 2016).

Even if we adopt such a wide definition, we still face several challenges in identifying a theory. For example, is Émile Durkheim the author of a single theory or of many theories? Are this theory or these theories located in specific texts? In his oeuvre taken as whole? In a Popperian “third world”? If there are several theories, are these compatible with each other? If there is only one theory, does this mean all of Durkheim’s writings are perfectly consistent with each other? Philosophers and social theorists regularly distinguish incompatible theoretical views in the same author: an “early” and a “late” Wittgenstein, a “first” and a “second” Foucault, a “young” and an “old” Hume. These demarcations are contested, and even their proponents are often uncertain about the boundaries of each phase. Dispensing with authorial issues is no easy solution: even if we purport to reconstruct an author-free “theory of social solidarity” rather than “Durkheim’s theory,” in all probability we would still be faced with a variety of different claims that cannot be easily reconciled.

Adding to the confusion, the status of an account of social reality as a “theory” is perpetually contested, as theorists and interested parties compete for recognition and legitimacy in a highly saturated intellectual attention space (see Collins 2002). Intellectuals struggle to make their accounts of the social world recognized as theories, and their claims are often dismissed as ideology, conspiracy theory, or mere drivel. This is a congenital issue of the sociology of intellectual life (and arguably of all social sciences that deal with cultural products), which no operational definition will ever solve in a fully uncontroversial way. The approach I suggest is a methodological nominalism: theories are whatever most scholars in the social sciences and humanities (itself a contested category) would call “theories,” supported by a reflexive awareness of what makes us see certain cultural products as more legitimately theoretical.

Theory Uses

What does it mean to “use” a theory? Most of the time, when scholars say they are using a theory, they mean they are drawing on it to explain, interpret, or give an account of a certain phenomenon. In this case, we could say the theory is used to make an epistemic claim, or it is being put to a conceptual use. But theories can also be applied in practical settings, where decision makers use a theory to inform or justify their choices, and where academics, experts, and other epistemic authorities draw on a theory to give advice to decision makers. In this case, we might say the theory is used to make a practical (or at any rate nonepistemic) claim. The history of the social sciences is rich in admonitions on the danger of conflating epistemic and practical uses of knowledge. Weber ([1919] 2012:346) famously warned that the words of

a person who speaks about democracy at a public meeting . . . are not instruments of academic analysis, but tools for political campaigning to win over others to his position. They are not ploughshares for lightening the soil of contemplative thought, but swords to be used against the opponent: weapons in the struggle. But in a lecture or an auditorium, it would be an outrage to use words in that way.

Many would find these statements reminiscent of nineteenth-century positivism. Not only Latourians, but historians, sociologists, and philosophers of science of many persuasions today believe that “winning over others to [one’s] position” is the bread and butter of what scientists and academics do qua scientists and academics. Yet, one can recognize ideas are put to different uses in different contexts while avoiding a strong metaphysical dichotomy between science and politics. Although practical uses often occur in “extra-academic sites” (Camic, Gross, and Lamont 2011:20), decisions potentially inspired by social theories regularly take place in academic institutions (e.g., when arguments from queer theory are used to implement unisex public toilets). ³

Scholars working on research utilization generally distinguish between “instrumental” and “symbolic” uses (along with “conceptual” ones, which correspond to my “epistemic” category): if you use a theory instrumentally it means you discover what is to be done thanks to the theory; if you use it symbolically it means you merely justify a preexisting practical decision by referring to the theory (see Beyer and Trice 1982). Actors can be more or less explicit in using a theory: they can make direct references and citations to a theoretical work or they can hint at it more discreetly (see Belorgey et al. 2011:9–11). Not all actors are candid enough to fetch a copy of a book and slam it down to the table saying “this is what we believe,” like Margaret Thatcher apocryphally did with Friedrich Hayek’s Constitution of Liberty during a policy debate (Campbell 2007:365).

Moreover, theory users can refer to (or discuss) a theory in a positive, neutral, or negative way. Berman and Milanes-Reyes (2013) show that in the U.S. Congress during the 1980s, the “Laffer curve” theory was discussed approvingly by Republicans, whereas Democrats tended to refer to it mockingly or disparagingly. This means there is a trivial sense in which social theories are used by opposite sides in a political debate, namely because both their proponents and critics talk about them. However, while one might want to say an actor is “using” a theory even when the theory is being mockingly dismissed, it is clear that for the purposes of establishing whether a theory has been used in a politically flexible way it makes sense to limit the concept to situations in which actors refer to a theory approvingly (whether instrumentally or symbolically).

Another important distinction is between explicit practical uses of a theory and practical unintended effects of a theory. Scholars have pointed to the ways theories can inadvertently become self-fulfilling by shaping assumptions and frames through which actors interpret social reality, so that actors change institutions, norms, and actions without even being aware they are “using” a theory (see Ferraro, Pfeffer, and Sutton 2005). This is particularly insidious when theories are false or inaccurate. The kind of flexible use I have in mind is the intentional one, in which actors refer more or less explicitly to a theory. In principle, flexibility might also occur in unintentional uses, although most empirical accounts of self-fulfilling effects show these reinforce one clear behavioral or policy pattern rather than many, potentially opposite, ones (e.g., Tilcsik 2021).

Political Uses

But what exactly does it mean to say the use of an idea is “political”? If there is a concept that is “essentially contested,” in Gallie’s (1955) felicitous expression, it is “politics” and the definition of something as political. Few would deny that events such as the 1789 Tennis Court Oath are political, but boundary work increases as one moves away from such centers of conceptual gravity. Judgments about the political nature of things are highly variable and correlated to various social factors (see Fitzgerald 2013). This makes definitions especially tricky. A useful conceptual distinction to narrow down the scope of politics can be found in Frickel and Gross’s (2005:207) theory of scientific/intellectual movements: they state these movements are inherently political because they involve “a desire to alter the configuration of social positions within or across intellectual fields in which power, attention, and other scarce resources are unequally distributed.” They distinguish this sense of “political,” which they call “Weberian,” from a more specific sense according to which some scientific/intellectual movements explicitly aim to alter the “distributions of power or recognition in society as a whole.” The kind of political use of social theories I have in mind is related to this latter notion of “political”: namely, a relatively explicit effort to shape the future outlook of society as a whole.

This does not mean only professional politicians use ideas politically, or ideas are used politically only in places like parliaments or party conferences. Political debates take place in a variety of institutional and organizational settings, including judicial forums, state agencies, media outlets, and even ordinary conversations. An editorial in a mainstream newspaper can reasonably be described as political if it takes side in a struggle over how (a sufficiently significant portion of) society should look. In this respect, the organizational literature on knowledge use often defines too narrowly decisions informed or justified by theories. When Mussolini stated in a 1924 speech at the headquarters of the Italian National Fascist Party that what a “German philosopher” (Nietzsche) had said, that is, to “live dangerously,” should become the watchword of the party, he was using an idea politically: he was drawing on it to advocate a certain view of how people should behave and how society should look (as he clarified immediately afterward: “this must mean to be ready for everything, any sacrifice, any danger, any action, when it comes to defending the fatherland and fascism” [quoted in Sznajder 2002:254]).

It is important to make two further clarifications. The first is the distinction between political uses of a theory versus political reactions to and judgments of a theory. Social theories solicit a wide range of emotional reactions in people who are exposed to them, such as anger, sadness, frustration, hope, and hilarity. Some reactions are more elaborate and produce complex judgments about whether and why a theory is or is not politically valid (see Black 2013:769). Such reactions and judgments can either remain at the level of subjective mental states or be expressed as public speech acts, but they should be distinguished from political uses of a theory. That the same theory elicits highly divergent emotions or political judgments in different people seems perfectly normal, whereas the fact that the same theory is used by different people to defend highly divergent political views is much more puzzling. The second clarification is that political uses are not the only nonepistemic (or practical) uses of a theory one can make. For example, one might speak of a “prudential” use of a theory when individuals draw on it to decide how to conduct themselves in their daily lives without any clear intention to transform society, or of a “managerial” use when the administration of an organization does the same.

Political Opposites

What does it mean to say the same theory is used to support opposite political views? To be sure, politics requires conflict. Insofar as decisions are made unanimously and no grievances are raised, politics fades into the background. Conflict is often conceived by political actors and laypeople as a field where two sides face each other, and the sides are often either ideologies, such as the left against the right, or social groups, such as progressives against conservatives. Seeing things like this, it might seem easy to find out whether an idea is politically flexible: one simply looks at whether the two sides are using it, and if that is the case, then it means the idea is politically flexible.

Unfortunately, things are not so simple. Scholars in political science and other fields have long conceived of political attitudes as positions that can be put on a scale between two extreme points, an approach that allows researchers to measure their object with greater precision. Accordingly, instead of two sharp sides facing each other, a whole series of positions can exist along the same axis, and it is not obvious whether two distinct positions on the axis should be considered “political opposites.” It would thus be inaccurate to claim that for any position on a political axis there is one that is its opposite, because all other positions are in a sense opposite to it. The opposite of Bernie Sanders’s political views, for example, is not simply Donald Trump’s political views, but everything that is not Sanders’s political views that is relevant in that political context, including views, like Joe Biden’s, that diverge from Sanders’s for other reasons or to different degrees. This poses a problem for a theory of political flexibility. Should we consider a theory politically flexible only when it has been used to support very distant views (e.g., Sanders–Trump) or also when it has been used to support not so distant views (e.g., Sanders–Biden)?

An additional difficulty arises from the fact that “maps” and “spatial analyses” of political conflict often reveal that conflict can take the form of highly complex structures in which several axes intersect (see Ostiguy 2017). These accounts depict a highly fragmented field composed of a multitude of political regions, which might have different kinds of conflictual or harmonic relations between each other, and which would be reductive to define simply as “political opposites.” Take a popular version of a two-axis “political compass,” where authoritarianism is opposed to libertarianism and the (economic) left is opposed to the (economic) right. Here again, even assuming only four positions exist (so positions cannot be put on a scale), all positions in the diagram are still opposing each other, albeit perhaps to different degrees (a “left libertarian” is presumably more opposed to a “right authoritarian” than to a “left authoritarian”).

There is no magic recipe to solve this problem, other than resorting to the metaphysical principle that natura non facit saltus (in the case in point, it does not jump between flexibility and rigidity): flexibility comes in degrees, and the more staunchly disputants in a political conflict are opposed, the more the ideas they simultaneously use in the conflict should be seen as flexible, and are thus interesting for our purposes. One should thus pay careful attention to the contexts in which political disputes take place and identify the main political cleavages that appear. ⁴

Contexts

I have suggested that political flexibility occurs when the same social theory is used in support of opposite sides of a political conflict, either in the same context or in different contexts characterized by relatively similar political cleavages. Now I must clarify what is meant by context and what it means to say contexts can have similar political cleavages.

As we shall see, scholars who point to some kind of political ambivalence of a theory often refer to different contexts in which the theory is used. The sociology of ideas has arguably moved away from a focus on sweeping macrosocial factors to an emphasis on “local institutional settings” (Camic and Gross 2001:247), but these accounts take a wide variety of contexts as units of analysis, including countries, epochs, regimes, organizations, and controversies. If political flexibility is something that occurs in different contexts rather than in the same context, it is clear one cannot simply say theory t is flexible because of the context c in which it is used. Instead, one should say theory t is flexible because of the difference d between context c₁ and context c₂ in which it is used. In other words, one has to compare the features of the contexts in which the theory is used in support of opposite political claims.

Contexts might have different features, but there should be a certain degree of similarity between them for us to be able to say the theory is used to support opposite political claims. Suppose a theory is used in country c₁ to promote right-wing policies and in country c₂ to promote left-wing policies. If the meaning of right-wing (and left-wing) in country c₁ has nothing to do with the meaning of right-wing (and left-wing) in country c₂, then we would not be able to say the theory is used to make opposite political claims, but simply to make different political claims. Instead, if right-wing in country c₁ is present in some form (and perhaps under a different name) in country c₂, this would make it much easier to establish whether the theory is truly flexible. This leads us back to the difficulties of establishing when two political positions are opposite, and to the inevitability of treating political oppositions as a gradual phenomenon, whereby the distance between different politicians, ideologies, and policies can be compared and judgments of greater or smaller flexibility can be made.

What if the same theory appears to be used to make opposite political claims in the same context, for example in the same parliamentary debate? This is a common scenario: one side in a dispute uses a theory to make a political claim, which leads the opponents to challenge how the theory was used and propose a different interpretation that supports their side. Faced with such rival claims to the right interpretation of the same theory, it seems the most fruitful approach from a sociological perspective is to avoid making final judgments about the soundness of each interpretation, and instead grasp what each party is trying to obtain by invoking the authority of the same theory or theorist, and what might explain their different uses.

A similar situation occurs when the same individual uses the same theory with different political aims at different life stages. Claims that a theorist is engaging in this type of flexibility are often made by rivals or former disciples: one way of applying a theory is consistent with the author’s original intentions; other ways are inconsistent and are attributed to ageing, treason, or some kind of mistake. Biographical accounts also sometimes make these assessments of whether intellectuals were politically consistent with their theories. For example, in his exploration of “moral consistency,” Abbott (2019:24) claims that some of Durkheim’s political interventions during the Dreyfus Affair “directly contravene[d]” his theoretical positions developed earlier in lectures and books, and reveal “a complete empirical contradiction between Durkheim’s own moral position and his theory of morals.” Abbott’s account points to inconsistency rather than flexibility: certain political positions are compatible with Durkheim’s theory (e.g., opposition to divorce, or the reintroduction of occupational corporations), whereas others, such as supporting Dreyfus in the name of individual rights, simply are not, and are advanced (intentionally or not) out of opportunism or wishful thinking. Instead, if one assumes theories can have some degree of flexibility, factors such as intrinsic features of the theory or changing historical circumstances become more relevant. I now turn to the analysis of these factors.

Two Sources of Political Flexibility

Adopting an empirical approach in the study of the political use of theories leads to a search for factors that increase their political flexibility and rigidity. I argue that we can distinguish two main sources of flexibility: endogenous sources related to intrinsic properties of theories, and exogenous sources related to different features of the context in which they are used and of the individuals who use them.

Endogenous Sources: Generality, Ambiguity, and Neutrality

The view that intrinsic properties of theories can increase their political flexibility rests on the assumption that there are not just plenty of theories, but also many different types of theories. We can classify social theories according to various principles, and the different types that are thus distinguished might be relevant for our purposes: perhaps only certain types of theories are politically flexible, or certain types are more flexible than others.

One reason a theory could be used by opposite political forces is simply that it works for both sides. This interpretation resonates with classic works in sociology of knowledge that posit an evolutionary dynamic to ideational change. Mannheim ([1928] 1952:222, 226), for example, claimed that “whenever groups compete for having their interpretation of reality accepted as the correct one, it may happen that one of the groups takes over from the adversary some fruitful hypothesis or category—anything that promises cognitive gain. . . . [T]he stream of history, then, tends to sift out in the long run those contents, patterns, and modes of experience that are of the greatest pragmatic value.” There are two problems with this interpretation. First, it would be difficult to isolate what makes a theory useful from the fact that it is actually used, which seriously undermines the explanatory power of this approach. Second, in this case we would not say the theory is used politically, because it is not used to make a political claim that diverges from claims of the opposite side, but rather instrumentally to bring about whatever result both sides are striving for.

In contrast to this perspective, I identify three principles of differentiation that seem to be connected to variation in political flexibility: generality, ambiguity, and neutrality. The more a theory is general, ambiguous, and neutral, the more likely it will be used to support opposite political claims. I shall consider these three properties in turn.

Generality

As we have seen, theories are often conceived as sitting somewhere in between two extremes: the purely metaphysical and the purely descriptive. In addition, most authors share the view that theories can have different degrees of generality: the phenomena they are supposed to make sense of can be more or less particular, more or less concentrated in space and time, and can deal with bigger or smaller portions of the social world. ⁵ A theory of why Italy experienced a decline in mortality when it transitioned to modernity is more particular than a theory of why countries in general experience a decline in mortality when they transition to modernity. ⁶

More general theories inevitably deal with phenomena that are increasingly dissimilar. Theories of social movements, for example, can be used to understand events of varying size that took place in different countries at different times, and also to interpret phenomena that resemble social movements without being generally conceived as such, like scientific/intellectual movements (Frickel and Gross 2005). This is seen as a problem when taken too far, as “conceptual stretching” (Sartori 1970) and “procrustean concepts” (Jasper and Young 2007) create confusion and make it difficult for researchers to test theories, but this is clearly a regular feature of social scientific theorizing: a more general theory is able to explain different things and to connect the dots between disparate pieces of information. As Swedberg (2020) points out, abstraction is part of the broader epistemic value of simplicity.

It seems reasonable to think more general theories are more flexible: they can be applied to a greater range of social groups, events, and phenomena that might have a different political resonance for different people. For example, consider “social contagion,” a theory according to which individuals in crowds are more likely to imitate unreflectively others’ behavior and act impulsively. Peyrot and Burns (2001) report that this theory was used as a defense strategy in a trial in which two black men were charged with beating a white truck driver during the 1992 Los Angeles riots. It is easy to picture a conservative commentator being outraged by this use of the theory. Yet, it is also easy to picture a conservative commentator advancing a similar line of reasoning to make sense of other situations, for instance, the 2021 storming of the U.S. Capitol or an episode in which police lose control and react disproportionately against an unarmed suspect. Since the work of Gustave Le Bon, social contagion theory has taken the riotous crowd as its favorite exemplar, but its core principles are general enough that the theory can be applied to explain both violent behavior by black protestors, a group that, at least in the contemporary United States, has the sympathy of the progressive left, and brutal behavior by police officers, a group that has the sympathy of the conservative right. ⁷

However, not all theories have this degree of generality. Some refer not to recurring situations in which all sorts of people might find themselves, but to concrete social groups with relatively stable characteristics: the power elite, same-sex parents, authoritarian personality types, Native Americans, the West. These theories can support a narrower range of political judgments. An example of a less general theory that has some complementarity to contagion theory is “black rage” theory, according to which white supremacy and systematic discrimination under certain conditions cause black people to commit violent acts (Grier and Cobbs 1992). This theory was also used as a defense in criminal trials, to the dismay of conservative commentators (e.g., Dershowitz 1994). What seems to differentiate this theory from “social contagion” is that “black rage” is less general: it can only be applied to a certain group of people characterized by relatively stable traits, that is, black people in a highly racist society.

Of course, one could argue that the principles of black rage theory could be abstracted to make it more general, so the final result would resemble a kind of “strain theory,” according to which any individual experiencing heavy and sustained racial discrimination is more likely to commit violence. ⁸ Or the theory could be applied to a different population: right-wing ideologists might explain white supremacist violence as “white rage,” resulting from what they perceive as “reverse racism” and “white replacement.” ⁹ But regardless of the validity of any of these theories, the point is that such moves, that is, both the generalization of the theory and its application to substantially different objects, would presumably alter the initial theory to a point where we would speak of a different theory. Much depends, of course, on the criteria one uses to identify theories and draw boundaries around them, a problem we also encounter in the next section.

Ambiguity

A second principle of differentiation is the degree of ambiguity (as opposed to univocity) of a theory. A theory is ambiguous when it is composed of multiple elements that are not perfectly consistent with each other—they are not merely logical steps in a linear argument—and that can be emphasized selectively when the theory is interpreted and used. By contrast, a univocal theory is highly coherent and meaning, so to speak, holds together. Ambiguity does not primarily mean theories are interpreted in very different ways by their readers and users. Rather, that is a consequence of ambiguity: highly ambiguous theories possess an intrinsic multiplicity of voices, which is what allows divergent interpretations and uses. This view of ambiguity arguably has roots in literary theory, especially in Bakhtin’s concept of polyphony. Bakhtin ([1929] 1984) famously read Dostoevsky’s novels as incorporating a plurality of voices, each conveying a different ideological and existential perspective. These voices, which are in constant dialogue and tension with each other, cannot transcend to a higher synthesis of what Dostoevsky really meant, unlike the works of other writers that are dominated by a single voice (and are thus “monologic” in Bakhtin’s terminology).

This insight has proved fruitful within sociology. For example, while recognizing that “a cultural object has no meaning independent of its being experienced,” Griswold (1987:1082, 1079) noted that “cultural works vary in their coherence and their multivocality . . . [that is,] the ability of the cultural object to carry multiple messages.” Sociologists of science have advanced a similar view that scientific norms constitute a “repertoire” from which scientists can draw selectively (Gieryn 1983; Mulkay 1976).

Can such an approach be applied to social theories? A long-standing scholarly tradition urges us to read scientific theories, and social science theories in particular, as novels or literary works, and to consider theorists as artful storytellers who use various rhetorical devices to persuade their audience (e.g., Blakely 2020; Gusfield 1976). As Hallett, Stapleton, and Sauder (2019:548) make clear, “In many ways, social science ideas are cultural objects.” From this perspective, a considerable literature investigates the ambiguity, ambivalence, and multivocality of social theories. Levine (1985) identified no fewer than ten different meanings of anomie in Merton’s writings and explained Durkheim’s own vagueness about anomie by referring to divergent concepts of society in his thought, which produce antagonistic emotional and cognitive needs. Recognizing the value of the ideal of “univocality” in science (because “the point of any secular effort to surpass commonsense understandings is to achieve some determinate cognitive mastery of our various environments”), Levine (1985:217) argued that in its radical version it is impossible to realize. “Human realities,” Levine (1985:14) wrote, “are intrinsically ambiguous” and “so is the language that social scientists use in attempting to represent them.”

Ambiguity fulfills several functions, including the “protection of one’s meanings and intentions through ambiguously opaque utterance,” the “evocative representation of complex meanings,” and the “bonding of a community through diffuse symbols” (Levine 1985:218). Along similar lines, Davis (2008:67) suggested the concept of intersectionality is better conceived as a “buzzword” characterized by “ambiguity and open-endedness,” whose different components appeal to groups with divergent political and academic projects (e.g., the critique of social inequality and the postmodern deconstruction of binary categories). Other scholars argue that multifaceted theories maximize the chances for an author to appeal to different audiences and reach prominence in different social contexts (e.g., Davis 1999; Lamont 1987).

However, the question of how boundaries should be drawn around theories leads to difficulty in establishing that a theory is ambiguous. One might be uncertain as to whether a given theoretical inconsistency should be imputed to the fact that the theory is ambiguous or to the fact that we are actually dealing with two discrete (and potentially univocal) theories. A difference between theories and novels should be noted here. A novel is more clearly bounded by textual support. A text might be translated, revised, prefaced, and commented on by the author and other people in very different ways. It might contain many different “voices,” but there is at least a reasonable agreement, say, that the novel Don Quixote is contained in a text, the reference edition of Don Quixote. This is not always (or even frequently) the case for social science theories. In many cases, theories are developed by several authors in many different publications over multiple generations. To be sure, some theories are strongly associated with a single text written by a single author, such as Durkheim’s theory of suicide and his 1897 book, Le Suicide: Étude de sociologie. But even these theories are seen as comprising not only that single text but also what the author wrote and said elsewhere, and what followers and commentators wrote and said about it. It is clear that the broader a theory is defined, the more likely it will be found ambiguous.

But putting these difficulties aside, it should still be possible to affirm provisionally that theories can be more or less ambiguous. And it seems obvious that the more ambiguous a theory is, the higher its capacity to accommodate different political uses. An example might be helpful here. In his analysis of the diffusion of Keynesian theory across countries, Hall (1989) claims that the highly heterogeneous political use of the theory in the postwar period had something to do with its inner ambiguity. Keynesianism was composed of at least three dimensions: the idea of a balance between aggregate demand and supply; an activist view of the state in relation to the economy; and a set of policy prescriptions known as countercyclical demand management. As Hall (1989:367) puts it, “The very ambiguity of Keynesian ideas enhanced their power in the political sphere. By reading slightly different emphases into those ideas, an otherwise disparate set of groups could unite under the same banner.” Of course, this is a case of relatively mild ambiguity; Hall does not go so far as to claim that Keynesianism was internally inconsistent. Still, policymakers and experts who drew on Keynesian theory faced an ambiguous object that easily lent itself to different interpretations.

Neutrality

A third principle of differentiation is the degree of neutrality (as opposed to normativity) of a theory. It seems obvious that the more normative a theory, the less it will be possible to use it in support of opposite political projects. But what does “normative” mean here? A common view defines theories as normative when they include (and neutral when they do not include) explicit political judgments, invectives against hostile politicians and parties, and justifications of favored ones. Many classical and contemporary authors couch their theories in highly polemical tones, not only against intellectual opponents but against political figures and institutions. However, when theories are interpreted, taught, disseminated, and used, this explicit normative coating can be easily removed. Sometimes this cleansing is part of an explicit theoretical project, as was arguably the case for analytical Marxism. Other times, it is simply a practice guided by the belief that what makes a theory relevant is primarily its explanatory dimension. To go back to the case of Keynesian theory, one could easily ignore the policy prescriptions and consider it exclusively as an explanation.

But there is another, deeper definition of normativity according to which even if theories do not contain explicit value judgments, they still possess a latent normativity: they establish a hierarchy between different possible states of affairs, dismiss certain negative features as side effects we should accept to attain a greater good, and reject certain political projects as unfeasible. Scholars with some familiarity with major debates in the social sciences intuitively tend to assign a political color or ideological leaning to many theories. The archetype of such claims in sociology is the pairing of structural functionalism and conservatism (Gouldner 1970), but similar associations are widespread across disciplinary fields (see Myrdal [1954] on economics and Fabian [1983] on anthropology). For Dupré (2007:33), theories of human behavior provide at least a vague indication about “the appropriate policy responses” to a given social issue, which are likely to be politically controversial as people have different interests and will be differently affected by them. As much as one might try to cleanse a theory from all explicit value judgments and isolate its analytic bits, the result will still contain, in Taylor’s (1985:89) words, “some, even implicit, conception of human needs, wants, and purposes,” which will determine the range of available polities and policies. It is only by radically altering the theory itself that one can alter such political implications. According to Taylor, theories of democracy and authoritarianism developed by postwar U.S. political scientists simply cannot be reasonably used to advocate authoritarianism: these theories cast democracy in such a light that it is impossible not to favor it.

But even staunch critics of neutrality would presumably concede that some theories have at least less clear practical implications than others, or that normativity comes in degrees. For example, Tilcsik (2021:99, 96) argues that the statistical theory of discrimination, which states that using ascriptive characteristics in hiring decisions increases firm profits, offers a “relatively clear and specific behavioral lesson,” in contrast to other economic theories that “are simply too abstract to have any clear behavioral or policy implications.” If employers want to increase profits, the theory suggests, then they should make sure to take ascriptive characteristics into account when making hiring decisions. One might then conclude this theory is not politically flexible: it makes little sense to use the statistical theory of discrimination to argue against using ascriptive characteristics in hiring decisions. In contrast, theories that have less clear behavioral or policy implications might be more amenable to antithetical political uses.

A caveat is in order, however. Imputing political implications to a theory involves plenty of assumptions about how people use theories. Going back to the example of statistical discrimination, in the same way as one might point to the relation between discrimination and profit to justify discrimination, one might well point to the same relation (thus accepting the theory as valid) to condemn profit-maximization as a firm goal. The formulation “a happens because of b” can easily lead one to think that the only possible political implication is “if you want a, make sure b happens” (e.g., “if you want to increase profits, make sure to take ascriptive characteristics into account”). But while this might be the most common way a theory is used to infer political or behavioral prescriptions, this does not exclude that the same theory could also be used to prevent a from happening by opposing b. As Barnes (1995:46) argued in criticizing accounts of functionalism that portrayed it as inherently conservative, “a theory of system stability is, after all, ipso facto a theory of instability as well: a source of insight into the integration of a system is also a clue to the conditions that could lead to its disintegration, differentiation or change” (see also Stinchcombe 1968:91–93).

This illustrates the limits of a purely endogenous analysis of what makes a theory politically flexible. Generality, ambiguity, and neutrality are intrinsic properties of a theory, and we have good reasons to think these properties matter: only a certain degree of generality, ambiguity, and neutrality allow one to use a theory in a way that is highly politically flexible. But theories are politically flexible only if they are actually used in certain contexts, by actors who are engaged in concrete political struggles. It is to this crucial dimension that I now turn.

Exogenous Sources: Political and Epistemic Cultures

Exogenous sources of political flexibility are all conditions external to a theory itself that make it more likely the theory will be used in support of opposite political projects. As we saw, such uses either occur in different contexts or, if they occur in the same context, they are performed by different individuals, so the exogenous conditions should be found in the differences between contexts and between individual theory users. In this section, I first consider differences between contexts by discussing promising theoretical approaches and I look at various examples. I then turn to political flexibility in the same context, and I argue that differences between individuals can also be interpreted in terms of differences between contexts, because the individuals involved most likely do not share significant social characteristics and belong to different contexts. I conclude by arguing that two main kinds of differences between contexts can be identified: the political and the epistemic.

Scholars from widely different traditions have articulated the idea that theories “travel” (Said 1983: chap. 10): they can be brought into different contexts in which they are interpreted and used in novel ways. By showing that audiences actively decode the meaning of the messages they receive using different frames of reference, cultural theorists (e.g., Hall 1980) have demonstrated the untenability of linear models of communication. No matter how careful a “sender” is in avoiding misunderstandings and trying to “get the message across,” there is always the possibility audiences will use alternative frames of reference. Drawing on the work of Bourdieu and others, the sociology of translation (Sapiro, Santoro, and Baert 2020) similarly emphasizes the break between the production and the reception of literary and scientific writings. As Bourdieu (2002:4) remarked, “texts circulate without their context”: cultural products, including social theories, are transferred by strategizing gatekeepers from the field in which they are produced (which for historical reasons has often been a “national” field) to other fields that may be structured differently.

A common way of describing this approach is that fields in which ideas are produced are in a position of “structural homology” with similar fields in other countries, or with different fields in the same country. One can find similar oppositions between diverging interests: the powerless and the powerful, autonomy and heteronomy, left and right. For example, the dominant businesspeople in a country share many features (including ideas and attitudes) with dominant politicians and intellectuals in the same country and with dominant members of all three fields in other countries. If one stops here, there seems to be little space for political flexibility: idea importers have strong affinities with idea producers, and theories associated with certain political claims in one country will be associated with similar political claims in other countries. However, Bourdieu also claimed that homology between countries is far from perfect, and frequently there is a “structural discrepancy” (décalage structural) between contexts. Countries have different political and cultural institutions that reflect their distinctive histories. The political claims viable in a given context are influenced by such things as the balance of power between different parties and interest groups. The opposition between left and right, for example, might be skewed to the right in one country because of its electoral system or because its working class is more strongly divided along religious or ethnic lines (Lipset and Marks 2000). Major events, including wars, regime changes, technological developments, and natural disaster, can produce a realignment of political cleavages in the places where they occur. Because theory producers and importers navigate contexts that are structured differently, they might have different motivations and interests that explain their different uses of the same theory. Audiences at the receiving end might be unaware of this complex chain of mediations and are often oblivious to the specific conditions and debates in which the original theory emerged. ¹⁰

There are abundant examples of how theories change their political connotation when they travel across countries. During the 1970s and 1980s, John Rawls was identified as a leftist in the United States, but in France he was seen as a neoliberal theorist, and often lumped with authors such as Friedrich Hayek, because he used methods and concepts that free-market economists were introducing in France at the time (Belorgey et al. 2011). During the 1950s, in Britain, Keynesianism was primarily seen as “a relatively noninterventionist doctrine that promised hands-off economic management in contrast to the nationalization program, wage controls, and industrial planning of the 1945–1951 Labour government.” In the United States, in contrast, Keynesianism was mostly took to be “a relatively interventionist doctrine, as a result of its association with wartime controls and proposals for national economic planning” (Hall 1989:269). In Italy and Germany, in the 1920s and 1930s, Nietzsche was frequently invoked as a forerunner and spiritual guide of Nazi and Fascist revolutions by leaders and ideologists, and most people, including in other countries, thought there was an intimate affinity between the philosopher and Nazi ideology. But starting in the mid-1950s, translators and theorists in France and the United States reinterpreted Nietzsche in support of “radical, leftist, or even Marxist formations” (Aschheim 1994:191, see also Lang 2002). In Apartheid South Africa, white intellectuals, including several anthropologists, justified racial segregation by appealing to the theory of cultural relativism, according to which “cultural differences were to be valued” and “if the integrity of traditional cultures were undermined, social disintegration would follow.” Yet in the United States, this “classical Boasian discourse about culture” was associated with efforts to dismantle racial segregation in the South on grounds that nurture, not nature, was the important factor in making an individual (Kuper 1999:xii–xiii).

But what if the same theory appears to be used to make opposite political claims in the same context? This is what happened in a case I plan to explore in a forthcoming publication, which deals with flexible uses of Merton’s theory of the self-fulfilling prophecy in early U.S. desegregation legal cases. Once the Supreme Court ruled segregation was unconstitutional in 1954, southern states claimed desegregation should be implemented at a slow pace because doing otherwise would cause widespread civil unrest, and they cited sociological evidence suggesting police officers would not react properly to disorder because they would assume civil war was inevitable. Civil rights organizations retorted that such studies had the potential to become self-fulfilling prophecies, whereby the sociological view that desegregation was incompatible with the attitudes of southern Americans, far from being an objective assessment of the situation, had the potential to create the reality it was supposed to explain. Thus, Merton’s theory was used to support both immediate integration of schools and a much more gradualist reform program. To be sure, such flexibility was made possible by certain features of the theory, notably its generality and ambiguity. But exogenous factors were also important: the context of use was the same legal case, but the actors participating in it were not interchangeable, and they carried with them dispositions acquired from the distinctive milieu they inhabited. In particular, Lewis M. Killian, a sociologist who authored an expert report by request of the Florida attorney general, was a professor in racially segregated Florida, where white people, including intellectuals, tended to be much more conservative—especially on racial issues—than in New York, where civil rights organizations involved in the case hired most of their experts. Progressivism in postwar New York involved arguing for the radical dismantling of any form of racial discrimination. Progressivism in postwar Tallahassee meant advocating a much more modest program of reform.

This example shows that differences between individuals can be interpreted in terms of differences between contexts, broadly conceived: even if different individuals are in the same situation when political arguments are exchanged (e.g., a parliamentary debate), presumably they still have different characteristics in terms of age, status, education, and networks. They have navigated more or less different environments to reach that situation. In other words, they have different trajectories, which means they come with different sets of embodied dispositions.

Speaking in terms of contexts, social characteristics, trajectories, and the like is of course vague. We need a more precise idea of what factors play a role in sustaining political flexibility. There are two main kinds of differences between contexts we should watch closely: differences in political and in epistemic cultures. Political and epistemic cultures are the symbolic structuring of political and epistemic claims-making. They provide people rules and tools to think about what is good and bad, politically or epistemically (see Knorr Cetina 1999; Lichterman and Cefaï 2006). I already alluded to differences in political cultures: compared to a country with a historically strong socialist party, for example, people in a country with a historically weak socialist party have a different range of political options they will consider viable, so they will form different opinions and make different claims. But what about epistemic cultures? The idea is that theory users regularly rely on different assumptions of how a theory should be applied to arrive at a policy prescription. Even if the political culture is the same, different epistemic cultures might lead applications to go in different directions. For example, concerning economics, Reay (2012:64) writes that flexibility occurs because theories and data are combined with “less formal assumptions—‘subframes’—concerning likely causes, plausible parameter sizes, and standards of proof.” Drawing on work by constructivist epistemologists such as Fleck, Polanyi, and Knorr Cetina, Reay (2012:72) argues that flexibility in deciding how a theory should be applied to reach a practical decision (or in interpreting evidence to choose which theory to apply to a case) is common to all scientific disciplines because “all scientific knowledge involves subjective norms about reasonable assumptions and standards of evidence.” Similarly, Hall (1989:370) claims that when individuals deal with “complex sets of ideas [that] are ambiguous and far from immediately comprehensible,” interpretation becomes “a necessary prerequisite to understanding; and to make such interpretations, individuals tend to refer to an existing stock of knowledge that is generally conditioned by prior historical experience.”

In summary, the incorporation of a theory into different political cultures creates the conditions for the same theory to be advantageous for political forces that, notwithstanding differences between the two contexts, would be placed at opposite sides by most observers. Similarly, the use of a theory in different epistemic cultures creates the conditions for users to apply different standards when interpreting the theory and thus reach different political conclusions.

Conclusion: Studying Political Flexibility

When Rupert Cadell, the professor in Alfred Hitchcock’s 1948 thriller Rope played by James Stewart, discovers that two of his former students have committed a murder inspired by his teachings of Nietzsche’s theory of the superman, he protests and denies responsibility:

But you’ve given my words a meaning that I never dreamed of! And you’ve tried to twist them into a cold, logical excuse for your ugly murder! Well, they never were that, Brandon, and you can’t make them that. There must have been something deep inside you from the very start that let you do this thing, but there’s always been something deep inside me that would never let me do it—and would never let me be a party to it now.

When the same theory is used to make opposite political claims, a common concern, exemplified here by Cadell’s words, is that certain uses of the theory should not be considered “uses,” but normatively criticized as gross misinterpretations or opportunistic abuses. In this article, I argued that if we want to understand the political power of social theories, it is more sociologically fruitful to assume, following Barnes (1974:129), that “any particular belief may be made to serve any particular interest.” Moreover, whether a given theory is made to serve a particular interest is an empirical question, there might be profound differences in the degree and flexibility of different theories, and flexibility is not tantamount to randomness or arbitrariness. In other words, rather than deciding once and for all whether social theories really are flexible or rigid, I suggest we treat this as an empirical, not a philosophical, question. Allowing for variation in this manner, we pave the way for empirical analyses of the political flexibility of social theories, guided by a healthy dose of nominalism (because the main way to identify instances of theory use is the use of certain words, although references might be allusive) and agnosticism (because the empirical investigation of theory use should not be replaced with normative critique or dismissal, even when the appeal to a theory may seem preposterous).

That being said, the disputes and negotiations to determine the right interpretation of a theory, including its political implications, are interesting phenomena that need further study. How do social theorists deal with the fact that their intellectual creations are appropriated to support political positions different from their own? What kind of conceptual strategies and institutional frameworks do actors develop to limit the appropriation of social theories? Walter Benjamin ([1935] 2008:20) famously claimed that what distinguished his materialist theory of art from that of other cultural critics was that it was “completely useless for the purposes of fascism.” Are such concerns common in the formation of social theories? To what extent do they influence theoretical developments? One could speculate that intellectuals and professional organizations might try to institute regimes of political rigidity by punishing and censoring politically dissident interpretations of dominant social theories.

There is something deeply disturbing about a regime of perfect rigidity between how we comprehend and how we judge the social world. Barthes ([1953] 1970:23–24) alluded to such a dystopic overpoliticization of thought in his vivid account of Stalinism: “in the era of triumphant Stalinism,” he claimed, Marxist writing had lost its explanatory force and had merely become “a language expressing value-judgments,” where “there is no more lapse of time between naming and judging,” and in which “the closed character of language is perfected, since in the last analysis it is a value which is given as explanation of another value.” ¹¹ But the very existence of institutional enforcements of political rigidity reveals that theories can, in principle, be used to defend highly contradictory political projects. As I have shown, this happens because of intrinsic features of theories and because theory users inhabit contexts characterized by different political and epistemic cultures.

Further research should determine whether theories are more or less flexible according to a variety of other factors. Potential candidates include their origin (Were they developed following a precise request by a patron or client? Or did they develop independently of such a demand?), their social incorporation (Are they associated with the name of an author? Or has their creator been “obliterated by incorporation”?), and their intelligibility to the public (Are they easy to grasp and have a catchy appellation? Or are they highly complex, require background information, and cannot be summarized in a memorable aphorism or name?).

The political flexibility of social theories entails a series of broader practical and theoretical implications that need to be explored further. I will limit myself to a few scattered remarks here. First, the fact that theories can be used in a politically flexible way could suggest that attacks against politically problematic scholarship are unwarranted because ideas that appear to have a certain political connotation might eventually end up helping the other side of the barricade. As we saw, it is not uncommon for scholars and scientists to be accused of providing arguments to politically illegitimate groups, or to be reminded that if their ideas are the same as those of politically illegitimate groups, then it means there is something wrong with them. ¹² People subject to this criticism often retort that their opponents are committing a “guilt by association” fallacy and the validity of theories is independent of their use. They sometimes invoke something akin to the political flexibility of knowledge as a defensive strategy, such as when Winegard and Carl (2019), two controversial researchers who were fired from their academic institutions, suggested that “under some value systems, such as ‘luck egalitarianism,’ confirmation of a genetic contribution to racial IQ gaps could actually strengthen the case for redistributive taxation.” In this view, we should “let a hundred flowers bloom” and condemn only unacceptable political uses of theories, rather than the development of potentially problematic theories, because any theory can be put to problematic use. ¹³ Acknowledging political flexibility might shift the focus of debates back to questions related to the logical coherence and empirical validity of a theory, rather than its alleged political implications, and might strengthen social scientists’ epistemic authority by dismissing worries they are merely “moral philosophers in disguise” (Abend 2008b:87) or deceitful ideologues. ¹⁴

A second point concerns the relationship between theoretical and political commitments: is it theory users’ political views that determine why they use a theory, or is it their theoretical views that determine why they use it to make a certain political claim? In settings characterized by strongly adversarial procedures, such as a political debate or a politicized trial, it is easy to interpret the acts of theory users as the search for arguments (including social scientific ones) to support preexisting normative views. The question is whether this reflects a universal asymmetry between practical (or political) commitments and scientific (or theoretical) views. The assumption behind this would be that political beliefs are closer to social interests, preferences, and practical concerns than abstract theories, and therefore exert more causal power. This view is well supported by research on motivated reasoning (e.g., Kahan 2012), as long as one is ready to accept the analogy between political commitments and “motivations” on the one hand, and between theoretical commitments and “reasoning” on the other. But what if the causal relation goes the other way around, or if the process simply cannot be comprehended as a simple cause-effect relation? After all, if it intuitively makes sense to think, for example, that Marxist engagement is more readily explained by a preexisting aversion to inequality rather than by a belief in the validity of historical materialism, it is only the latter theoretical framework that gives meaning to one’s scattered observations of unequal social conditions.

A third and final point relates to the implications of political flexibility for a sociological conception of culture. If the same social theory is alternatively a right-wing and a left-wing theory, and if the political right supports and opposes a given theory, depending on social circumstances, then this means the nature of political ideologies and intellectual products is not clearly established once and for all, but is rather constantly negotiated and acquires meaning through social uses and through contrast and comparison with other cultural products in the same context of practice. Although this seems to be a well-established principle within cultural sociology, it is often neglected in the study of theorizing in the social sciences.