Why do sociologists think that they are critical theorists?

One of the more frustrating features of American political culture is the homeostatic mechanism that appears to exist within the Democratic party, which ensures that no matter how unpopular the Republican party becomes, the Democrats will adjust their positioning in order to remain just slightly less popular. The latest product of this appetite for electoral failure has been the nomination of Darializa Avila Chevalier – who is basically a maniac – for a congressional seat in New York. Her political convictions would put her far to the left of almost any European left-wing party. The best part of her profile, however, is her current occupation: she is a graduate student in sociology.

The response to this, from myself and pretty much every other academic I know, has been an eyeball-rolling “of course she is.” Although this may be just the representativeness heuristic at work, anyone familiar with the current state of sociological inquiry will suspect that working in this field has been intensifying, rather than moderating, her political extremism. A student in political science, law, or philosophy might find it more difficult to maintain such uncompromising positions, because it is impossible to work in these fields without acquiring at least some exposure to contrary views. Sociology, by contrast, has developed an ideological monoculture that is becoming increasingly an object of concern (and, I should note, self-criticism).

Indeed, if I wanted to be cute, I might suggest that Chevalier’s politics lend support to the subcultural theory of social deviance. My objective, however, is not to talk about Chevalier, but rather to say a few things about how sociology came to be this way. Specifically, I want to explain how sociologists came to see their particular mode of inquiry, not as a contribution to a larger critical practice, but rather as a self-standing form of critical inquiry. At first glance this might seem strange. Obviously, if one wants to improve society, or to correct specific social problems, there are many things that one needs to know about that society. Policy studies have been dominated in recent decades by the expectation that recommendations be “evidence based” – with the implication that a major function of social-scientific research is to produce the relevant evidence. At the same time, there remains an implicit distinction between the normative dimension of policy (e.g. which outcome is better, which is worse? what should we be aiming at?) and the empirical questions that arise (e.g. what are the likely effects of making these changes?).

Traditional “positivist” sociology insisted that social scientists should stick to studying the empirical dimension – just figure out what the facts and mechanisms are – while leaving the normative questions to others. This was rightly controversial, and many people rejected the separation of “questions of fact” from “questions of value” on the grounds that, although conceptually coherent, the distinction cannot be respected in practice. No one studies crime, status inequality, or domestic violence out of idle curiosity. Others, however, went much further, rejecting the conceptual distinction as well. They claimed that certain values are constitutive of the facts, and by implication, that certain facts carry normative implications.

It is the latter view that became widely held among sociologists. Many became persuaded that merely studying society – coming to understand how things work – had transformative implications. Rather than seeing themselves as importing politics into sociology, they saw the study of society as inherently political. Specifically, they came to believe that existing structures of inequality and domination were grounded in a mistaken view of how society functions (or the conditions that must be satisfied in order for society to function), and so substituting for that a correct view would tend to undermine those structures (and would be opposed by those who benefit from them). This favoured the development of what Peter Baehr has referred to as “the unmasking style,” which treats sociological inquiry as equivalent to a form of ideology-critique. There were three quite independent lines of thinking that created this impression:

The discovery of norms. The most important discovery – that which led to the establishment of sociology as a distinct discipline – involved something that was hidden in plain sight. The greatest trick of human socialization is that it causes us to experience certain forms of learned behaviour as though they were spontaneous and natural. This leads us to lose sight of how rule-governed social life is. The most powerful corrective to this illusion is contact with foreign cultures, where one’s own lack of familiarity with the rules makes it easy to see how ubiquitous they are. The development of sociology was, in a sense, a byproduct of increased global trade and colonialism, as ongoing contact with foreign societies gave rise to greater reflexivity among Europeans in thinking about their own.

One might wonder what justifies calling social regularities rules, as opposed to just habits or patterns of behaviour. The important observation made by Emile Durkheim in the 19^th century is that certain regularities are not only expected, but also enforced. The difference between habits and rules is that rules tell you how you are supposed to behave, on pain of sanction. He introduced the term “social norm” to refer to these “sanctioned regularities of conduct.” This led to the big insight, which was in a sense foundational for modern sociology, that the orderliness of social interaction is possible because these positive and negative sanctions are internalized over the course of childhood. We start out life requiring extrinsic motives for complying with social norms, but undergo a process of socialization, which over time gives us intrinsic motives for complying with the rules of society.

The socialization process is what makes it difficult for us to see the rules of our own society as rules – having grown up with them, we have acquired a desire to act in the way that the rules require us to act (for example, wearing clothing in public), and so typically experience the behaviour as voluntary, even though it is constrained. Furthermore, because we have developed a variety of intrinsic prosocial motives, the sanctions that are imposed when we do happen to violate the rules are often just symbolic or indirect. Acting rudely, for example, is typically punished through subtle cues, such as a frown, or a cutting remark, or a joke at one’s expense. These are all instruments of social control, but the “punishment” has no intrinsic significance, it merely activates an internalized system of control (such as provoking embarrassment). This can be seen in the fact that such punishments do not work when applied against children.

Because of this, it is easy to lose track of the fact that all of our social behaviour is constrained by rules that are enforced, because we follow them willingly, and the punishments we experience are often so subtle that we are not even aware of them. Durkheim’s analysis, however, suggested that these norms were ultimately still backed by primitive rewards and punishments, and so if one were to push things hard enough – for example, breaking rules and then just ignoring the more subtle punishments that are applied – people will respond by escalating their response, shifting to punitive strategies that have greater intrinsic significance. This gave rise to a trend in the 1960s and ‘70s, among sociologists such as Harold Garfinkel and Stanley Milgram, to conduct so-called “breaching experiments,” by instructing their students to go out and violate minor social norms and then ignore the negative responses, in order to see what happens.

What happens, of course, is that things escalate very quickly. Indeed, modern research ethics protocols would clearly prohibit these sorts of studies, since they often embroiled the students in serious conflicts. (Milgram used to send his students into the New York City subway with instructions to ignore social norms regarding seating allocation. Most of the students wound up not doing it, because they were too inhibited or afraid.) What these experiments revealed, quite effectively, is that underneath the veneer of civility there remains the threat of physical violence.

This is something that we are all tacitly aware of. Some may remember, for example, this fantastic video by The Verve, which generates an entire drama from the simple premise that the singer (Richard Ashcroft) is walking in a straight line down the middle of the sidewalk:

The tension that builds over the course of the video arises from the question whether he is going to make it to the end without getting the shit kicked out of him. The fact that he is simply walking down the sidewalk, minding his own business, is presumably what led to the conclusion that this was a useful metaphor for life (as well as artistic and moral integrity).

Many people appreciated the breaching experiments for having raised to the level of conscious awareness an interesting feature of human sociality. Others, however, felt that they revealed something more sinister. By showing that everyday social norms, even seemingly minor ones, were ultimately backed by the threat of force, they revealed the structural violence underlying the system. Many felt that this had important political implications, because it showed that freedom could not be obtained within society, as conventionally understood, but required rather freedom from society. A truly emancipatory social movement should not be focused on changing the rules that we live by (as, for example, Marxism was), but rather should seek to free us from rules in general.

This idea fed into a powerful current of secular antinomianism that informed left-wing thinking in the 1960s. Antinomianism (i.e. “against law”) is a tendency that has deep roots in the Christian tradition. One of the major bones of contention between Jews and Christians, one may recall, is that Jesus gave his disciplines a new commandment, that they “love one another.” Christians have periodically interpreted this as an implicit revocation of the previous ten commandments, replacing the rules-based morality of the Old Testament with a new situational, emotionally based morality of love. This interpretation inspired periodic rebellions during the middle ages against more rule-oriented forms of religious law or morality. That tendency re-emerged in secular garb during the ‘60s (most conspicuously in the sexual revolution and the demand for “free love”), fed in part by developments in psychology and sociology that showed how much constraint people are subject to living in society.

One can find the exact same observation – social norms are backed by sanctions – articulated in Michel Foucault’s work as the claim that “power” suffuses all social interactions. He insisted, however, on couching this increasingly commonplace sociological observation in morally charged vocabulary, describing networks of social power as constituting systems of “domination” against which individuals require “strategies of resistance.” When pressed, he would always concede that no normative conclusions could be derived from his analysis, and that it was foolish to believe that we could ever free ourselves from power so understood. And yet subsequent generations of theorists impressed by his work have seldom respected these limitations. Furthermore, because Foucault wrote in such an entertaining style, many students receive their first exposure to the basic sociological insight, about how social order is maintained, not from the canonical authors (Durkheim, Parsons, etc.) but rather from Foucault, who makes it sound as though it has revolutionary political implications.

In the end, the fact that social life is governed by a vast system of social norms, backed by an equally complex and subtle system of sanctions, is an important discovery. But it has no normative implications, beyond the platitudinous observation that enforcement of the rules is good when the rules are good and bad when the rules are bad.

Social construction. The second major sociological discovery was closely related to the first, and so I will deal with it more quickly. Because people are often not aware of the norms that they are following, they also have a tendency to underestimate the amount of human behaviour that is learned, as opposed to being instinctual or natural. For example, it really is the case that during the 18^th century in Europe there was a “cult of sensibility” among upper class women that made them extremely sensitive to emotional stimuli, with reactions that included most prominently fainting (or swooning). Given that this was a physiological reaction (they really did lose consciousness), not under voluntary control, one might conclude that it could not have been learned behaviour. Many people at the time therefore assumed that it reflected a physical difference between men and women. In later years, pursuing the same theme, others suggested that some other physical cause must have been at work (such as the popular-but-largely-debunked theory that they wore corsets laced too tightly). And yet what cross-cultural analysis shows is that human societies are perfectly capable of instituting social roles that elicit an astonishing range of different behaviours, temperaments, and psychological responses from individuals, including reactive loss of consciousness.

This gave rise to the observation that, in addition to whatever natural sex differences exist between males and females, societies also construct a set of gender roles, through social norms that dictate appropriate and inappropriate behaviour for the two sexes. This role-specific behaviour comes to feel natural for individuals raised in those societies, even though it isn’t. Because of this, when people come along and challenge the rules that require certain forms of role-specific behaviour, it often leads to a debate over whether it can in fact be changed. For example, after the French revolution the exclusion of women from participation in the political system became increasingly contested. Yet many people pointed out, plausibly, that one would not want women occupying positions of political responsibility if they pass out every time they receive a piece of distressing news. In response, early feminists such as Mary Wollstonecraft spent a great deal of time (e.g. in A Vindication of the Rights of Woman, 1792) arguing that these forms of behaviour were a product of upbringing, and would change if women were raised under conditions of greater equality. (For a modern classic in this vein, see Iris Marion Young’s paper “Throwing Like a Girl.”) That turned out to be true in the case of swooning.

These debates were based on the obviously correct principle that, in order to make the case for a particular change in social behaviour, one must show that the behaviour in question can in fact be changed. At the same time, merely demonstrating that some behaviour can be changed falls far short of establishing that it should be changed. Unfortunately, because so many conservatives argue against social change by pointing to its supposed impossibility, many progressives were misled into thinking that merely rebutting this argument was sufficient to make the case for change. This was often combined with the naive assumption that the only thing holding us back from transforming society into an earthly utopia is a mistaken belief in the impossibility of such changes, arising from a factual error that people make, when they conclude that features of the world we live in are natural rather than socially constructed.

These arguments can easily generate the impression that sociology is inherently progressive, because it debunks the primary source of resistance to change. Revealing the socially constructed character of previously taken-for-granted features of the world is taken to constitute a critique of conservative ideology. (This is why, for example, there is such strong insistence in progressive scholarship that “race” must be a social construct, with no connection to any aspect of underlying biology, even though it is clearly linked to ancestry, which is a straightforwardly biological concept. The insistence arises from the belief that racists must be committed to some form of race essentialism, and so debunking the latter entails the defeat of the former.)

Finally, it should be noted that these ideas about social constructivism are often combined with the suggestion that social order is inherently coercive, to produce a somewhat more sophisticated view, which suggests that uncovering the constructed character of some phenomenon also reveals it to be a system of domination. For example, to the extent that there is any normative argument in Judith Butler’s work, this is what it amounts to. The chain of assertions is roughly as follows: The thing that you think is natural (e.g. gender, sex) is actually a social role. Roles are constituted by norms, and all norms are enforced. Therefore this thing is actually a product of coercion, not nature. So we should resist it, through transgression of the norms. Scratch a lot of “critical social science” and this is what you find as well – an amalgamation of universal social constructivism and secular antinomianism.

Sociological excuses. Adding to the confusion is a completely different line of sociological thinking, involving our attributions of moral responsibility to individuals. Back in the middle ages, it was widely thought that each human decision initiated a chain of causation, which then produced various effects (this was for complicated reasons involving Aristotle’s views on “final causation”). If one were to ask who was “responsible” for one of those effects, the answer involved finding out who initiated the causal chain that produced it. This reflected the commonsense view that, for example, the person who “causes” an accident is the one who should be held morally responsible for it. The scientific revolution obviously made a mess of this worldview, by insisting that there is no such thing as spontaneous causation, but that any causal chain can be traced back arbitrarily far. Human decision-making is no exception. The brain is a physical system, and so whatever decisions it makes are clearly caused by some antecedent conditions, which are in turn caused by other antecedent conditions, and so on.

This observation led to something of a skeptical crisis, in the 17^th century, with respect to attributions of moral responsibility. We like to hold people responsible for their decisions, which we take to be under their control. And yet for any decision that a person makes, it is necessarily the case that we can trace it back to certain causes that are not under that person’s control. This suggests that the person should not be held responsible, because they did not choose those conditions. The problem is that this is true of all decisions, which clearly implies that no one is responsible for anything. This was rightly viewed as a skeptical conclusion by philosophers, most of whom eventually responded by rejecting the underlying assumption that moral responsibility should track chains of causation.

Many sociologists, unfortunately, did not receive the memo, and so persist in believing that determining the cause of some social phenomenon has implications for how we assign blame. Specifically, it has given rise to a disciplinary commitment to showing that the “root causes” of various social phenomena lie not in individual choice, but rather in background social conditions. According to this view, conservatives believe in “personal responsibility” as a response to crime, drug use, teen pregnancy, and so on, because they mistakenly believe that the cause of this behaviour lies in individual choice. The sociological critique therefore involves finding some background condition, such as a disadvantaged upbringing, that is correlated with these choices, and proclaiming these to be the “root cause” of the behaviour. This is then taken to have the normative implication that these individuals should not be held responsible for their decisions, because they had no control over the background conditions.

This argument has been around for a long time. Its popularity among criminals was famously satirized by West Side Story in 1957. Some sociologists have earnestly suggested that the judicial system, instead of calling on psychiatrists to offer expert testimony about the defendant’s mental competence, should instead call on sociologists to describe the person’s “history, childhood, and relationship with the world” – on the grounds that this exercise would in many cases be exonerating.

Part of the appeal of this argument is that it successfully enrages conservatives, who find it difficult to refute because they share the underlying (causal) theory of moral responsibility. Furthermore, there are certain cases in which the basic point is obviously correct. For example, there is a certain type of right-winger who is unsympathetic to the plight of the unemployed, arguing that they should just try harder to find a job and that it’s their own fault if they fail. And yet it’s easy to see that if certain macroeconomic events, entirely beyond the control of the individual, produce an increase in the overall unemployment rate, just telling people to try harder and holding them responsible for their failure is both ignorant and cruel. Where things become problematic, however, is when people interpret examples like this as showing that our judgments of responsibility follow from an empirical discovery of the “true cause” of the person’s unemployment. According to this view, the conservative blames the individual because of an erroneous belief that the worker’s own laziness is causing the unemployment, whereas the progressive blames society, because of the correct belief that macroeconomic conditions are causing the unemployment. In this way, a moral question gets misrepresented as a factual disagreement.

The problem with the sociological argument, in this case, is not that it fails to exonerate the unemployed worker, but that it succeeds in exonerating everyone for everything. The argument proves that nothing is ever anyone’s fault, which is to say, it is a form of moral skepticism (or nihilism). The reason that conservatives have such a difficult time refuting the argument is that skeptical arguments are notoriously difficult to refute. Philosophers (and, for the most part, courts) avoid the problem by adopting a normative theory of responsibility, which renders the entire question of “root causes” irrelevant. (For reasons that are difficult to explain, Peter Strawson’s paper “Freedom and Resentment” is the most influential text here.) Unfortunately, no one seems to have brought sociologists up to speed on these developments (which has led to some extraordinary invective, especially in France, over the status of “sociological excuses”).

More generally, the sociological view encourages a naturalistic fallacy, by suggesting that the empirical work of sociologists, aimed at discovering causal mechanisms at work in society, has direct normative implications, such as telling us who should bear the cost of certain behaviour. This again reinforces the mistaken perception that sociology is an inherently political enterprise.

The area of study that has been most thoroughly ravaged by these normative confusions is probably criminology. There you can see all three of these arguments converging to form a mighty river. Law is a system of social control, which rests ultimately on the use of force. True! Crime is a social construct, which has varied enormously from place to place and time to time. True! The propensity to commit crime is strongly correlated with certain background conditions that are entirely outside the control of the individual. True! In all three cases, the empirical conditions that have been the object of sociological insight obtain. The problem is that, at the normative level, absolutely nothing follows from these observations taken alone. They must be supplemented with an explicitly articulated conception of justice before we can begin to explore their normative implications (which are legion). Instead, however, too many criminologists have smuggled in the highly contestable assumption that there is something necessarily wrong with enforcement, or emancipatory about deconstruction, or exonerating about the discovery of background causes. This has unfortunately turned much of the discipline into a case study of how smart people can talk themselves into a state of abject confusion.