It has become fashionable, once again, to argue that sociology is an impossible science (perhaps it is a constant feature of sociology and not a fad at all). The logic, despite being cloaked in a wide range of new ideologically-drenched rhetoric, remains the same. Society and its agents are like electrons: we may know their position or their speed, but because of their complexity, we can never really know everything. Consequently, sociology should be critical, descriptive, political, or all of the above (for a strong critique of these positions, see Turner 1990; Hammersly 2005). I won’t rehash or wade into these debates, as I do strongly believe sociology is a science, has a real body of cumulated knowledge, and has produced some solid general principles that are not, in fact, provincial, and can still contribute to public discourse, civil society, and policy making. I think one of the issues, to be frank, is that the we continue to look to the past (and classical sociology) while wringing our hands about how we might professionalize and socialize new students if we stop teaching classical theory. I think this strategy continues to be misguided, and there is a better way forward.

To do this, I predictably return again to the sociology of suicide, its main protagonist (Durkheim), and the illustrative lessons we can glean from this. Though I could write a treatise on doing theory, these essays are aimed at how we teach theory as the expectations we set for undergraduates shape upstream expectations of graduate students and the obligations we, as teachers, have to pushing the discipline forward as a science. In this first essay, I would like to lay out the background for the ensuing essays that tackle the substantive issues head-on. This essay focuses on sociological theory, and the pitfalls of worshiping the classics.

Durkheim and Sociological Theory

Durkheim is (rightly) considered one of the most important sociologists from the classical (19th century) era. He is also an exemplar of the good and bad of caring this much about one’s scientific disciplinary past. On the one hand, Durkheim’s work was revolutionary in methodological innovation, theoretical ingenuity, and sociological imagination. Some of it, like the fundamentals of his interaction ritual theory found in The Elementary Forms of Religious Life, continues to be generative across social sciences as mounting evidence supports his theoretical claims (see, for instance, Rimé & Páez 2023).

On the other hand, he represents a lot of what is wrong with the discipline’s approach to theorizing. First, he is classical for a reason. Not only does his writing and examples sometimes reflect the repugnant (from a contemporary perspective) views held by (Western) white men, but they also can be quite wrong on other levels. Like all classical writers, his style, thought process, and understanding of the world are delimited by time, place, economy, polity, culture, and so on. None of these positional qualifiers makes a difference in the validity of his interaction ritual, and yet rather than hone in and extract the basic principle we continue to insist on making undergraduates read the text over and over again; even though many aspects, like the data itself, are facutally incorrect. And herein lies the problem with a backward facing sociology: instead of building on scientific principles we remain stuck between a humanities-style interrogation of its ideas and subject matter and a scientific approach to studying its phenomena. Our feet stuck to the shoulders of our giants. Thus, when we teach Durkheim, then, we run into the inevitable problems one might expect from any old text. Students struggle with the writing and the philosophic ideas that are far removed from mainstream academia and common sense; the casual misogyny, ethnocentrism, and so forth obfuscate the contribution; and in the end, we are all distracted such that instead of teaching and using theory we are embroiled in debates about the fundamentals of the discipline itself.

When we set aside this set of challenges, even with good intentions, a third problem arises: humans tend to crystallize old texts, shifting from the creative effort of producing the text and what that should inspire us to do to either orthodox interpretations or exegetical activities (Goody 1986). Both are rife with pitfalls. The orthodox approach has consequences for the principle of self-correction in science. Too many sociologists continue to treat the “canon” or their favorite works as though they are finished projects, accepted truth, and perfected works. They are not. Durkheim’s Suicide is not immune to this problem, as it remains accepted wisdom despite shakey (Breault 2001) and non-existent (Leenars 2004) empirical evidence for three of his four types of suicide. This doesn’t matter because we teach and treat it as accepted wisdom. Consequently, to study suicide “correctly” means to study it as Durkheim did, testing and re-testing his vague theoretical ideas and nineteenth century empirical examples using new data or analytic strategies (Wray et al. 2011). Nevermind creative efforts to move beyond the text or specific methodological guidelines he set (Phillips 1974), as these upset the dogma and doctrine of a beatified sociologist. Hence the second problem: if the ideas are hermetically sealed, then being creative means digging deeper and deeper into the text in search of new interpretations; reading marginal comments and correspondence with fellow travellers; and, ultimately, remaining mired not on the shoulders of giants but the ghosts of the past.

The question, then, is what do we do? Make no mistake, there is no reality in which the classics are totally ignored. In fact, a History of Sociology course has real value for a variety of reasons, including the sociology of knowledge and ideas. But, what exactly should we be teaching undergraduates? I realize that many sociologists conflate social theory with sociological theory, and thereby, see theory as oriented toward philosophical questions as much as scientific principles. Yet, there are departments that specialize in philosophy and its method, which means we need to keep Durkheim’s belief in disciplinary boundaries seriously. Thus, when we argue students must “wrestle” with the German Ideology, what we are actually saying is we must expose our students to the holy texts, have them search these texts for critical insights, and, consequently, complete their Jedi training as sociology students. The problem, though, is they leave these classes not knowing what the hell theory is, what it does, why we care so much about it, and what the relationship between theory and methods are. Marx teaches our students nothing about these things. And while Durkheim’s Suicide is instructive for a number of these questions, the irony is that his theory is so poorly constructed (by any modern standards of theory building), we end up focused on his empirical generalizations like Protestants die by suicide at a higher rate than Catholics – despite evidence to the contrary (Halbwachs 1978). What do we do?

Sociological Theory

In its simplest form, a theory is a set of interrelated statements, with each statement characterized by two concepts and their relationship. Good scientific theory is good in so far as it leaves little ambiguity regarding the definitions of concepts, the relationship between them, and the interrelationship between the several statements, as consensus around conceptual definitions can only come about where vagaries are dispatched and measurement (and testability) is possible. Theories need not be formalized into propositions or analytic models, though these often provide the clearest, tersest mode of expression (Turner 1990). In part, the terseness of a statement is due to the goal of abstraction. Theories are decidedly not hypotheses because concepts are not variables—though, the difference between the two may be more of degree than of kind (Alexander 1990).

Sociology is rife with concepts, but rarely are they placed into statements about the relationship between this concept and that concept. Durkheim’s theory of differentiation serves as a good example: social differentiation is a positive function of moral density. The two concepts “moral density” and “social differentiation” are abstract in that how we measure them, concretely, is left to the imagination of the sociologist. Moral density, for example, may be the actual spatial density of a collective and/or the multiplex nature of social ties, frequency of interaction, and tightness of culture. Meanwhile, why this relationship is so, escapes this simple statement so that it can be applied to a wide range of specific cases, such as economic niches or socioecological systems, and, therefore, be flexible re explanations focused on emergent properties at different levels of social reality.

In addition to their abstractness, theories tend to be shaped by one or more of the following five goals: description/classification, explanation, prediction, overarching framework, and control. Each of these goals represents different aspects and aspirations of theory building, with description being the most basic scientific activity. Sociology, of course, is filled with a wide range of classificatory theories, ranging from myriad Weberian typologies to Parsons’ infamous AGIL schema. The primary criteria for evaluating a classification system’s value are determined by (1) how exhaustive the system is and (2) how mutually exclusive each type is, as well as (3) how readily the system lends itself to explanatory theory-building. Explanation and prediction are typically conflated but refer to different temporal orientations. It is much easier to explain why X caused Y than to predict that X will cause Y. Sociology, typically, works best for the former hence the tendency towards causal process modeling vis-a-vis laws or axiomatic theorizing (Turner 1990). There is a second reason to separate explanation and prediction: prediction is too often the strawman critics of sociology as a science evoke when attacking positivism. We are often told that the social world and social behavior are too complex to predict, and thus sociology and sociological theory is not scientific, cannot be, and/or should not be. But, it is worth pointing out that most natural sciences are much better at explaining things than predicting them. A biologist can tell you why a leaf falls in autumn, but because of the myriad variables outside of the laboratory, cannot predict when it will fall; a seismologist can reasonably ascertain the conditions increasing the likelihood of an earthquake, but cannot predict the timing. Neither of these disciplines are criticized for being less scientific than others.

These first three goals fold neatly into the fourth goal: an overarching framework. Evolutionary theory, for example, is the theoretical shell within which the biological sciences are unified (Mayr 2001). It offers classifications (e.g., genus), explanations (e.g., natural selection), and predictions (e.g., environmental change will put pressure on phenotypes and reproductive success). Consequently, it is paradigmatic in that it becomes the orienting frame through which biologists communicate with each other, motivate their own work, and conceptualize the biotic world. In many ways, this fourth goal is the “holy grail” of science, indicative of a theoretical framework that unifies a community of scholars—even if it is not without some criticism and critics—and highlights the fact that the theory seems to genuinely make sense of the data better than the alternative, competing theories. To be sure, it may be the case that sociology, because of its varied phenomena of interest, levels of reality, agnosticism towards methods, and the panoply of theoretical perspectives, resists an overarching framework, but we can imagine a sociology that moves past the popular middle-range theories to sets of broader, more abstract theoretical frameworks that come to characterizes interrelated studies, while also overlapping in key ways. Indeed, I think we already implicitly work within this world, but whether we are ready to explicitly specify the principles at work and, more importantly, abandon the way we currently professionalize neophytes in favor of a coherent, consistent approach remains an open question.

Finally, the most controversial goal for social science: control. A tension in sociology has long existed. On the one hand, besides Spencer and Weber, the foundational theorists in Europe and, later, in Atlanta and Chicago, took as a basic fact that social science could and should alter the social environment they inhabited. From Comte’s ridiculed religion of humanity to Marx’s commitment to praxis to the treatment of urban spaces as laboratories and the adoption of a pragmatist epistemology, sociology was oriented towards application. On the other hand, eugenics, social Darwinism, Nazism, and the like present cautionary tales of applied social engineering gone awry. How sociology creates an applied wing, then, has political and moral implications as much as scientific ones, and therefore it appears to be not as “simple” as in the natural sciences. Rather than wade any deeper into this thorny issue, I will say two things about control. First, I do believe our subject matter is more difficult to control than that which many natural scientists work with. Second, like prediction, I do think control is a matter of degree, and therefore, my first point has to be qualified by the recognition that total control—e.g., harnessing fossil fuels to power engines—is unlikely in many cases, but many branches of sociology provide exquisite empirical evidence and theoretical models informing public policy. Policy is not totalistic, usually, in its reach, but it can and does shape people’s lives across a broad metric of outcomes.

The Path Forward

What, then, does this mean? In short, I propose we examine classical theory according to these five aims, teasing out what the theory does and where it falls short, and then move on to the next theory. In so doing, we provide a clearer picture of the theory itself, disassociating it from the noise of the theorist, their writing, their milieu, and their philosophic commitments. This does not preclude intellectual activities that sit on the border of the humanities, such as pouring over Durkheim’s letters to Mauss or inspecting Marx’s marginal comments in an original version of Capital, as these both may alter the fundamentals of the theory at best, and at worst, they provide depth to a history of sociology. At the undergraduate level, the process is less intensive: we present the theory and examine how it shapes current research while at the graduate level students can actually do the extraction of principles. The former are taught theory while the latter further advance their knowledge of theory while also beginning to understand how to theorize.

To demonstrate the efficacy of this method, the following paper explores Durkheim’s Suicide, scrutinizing it with three of the five aims (more on that shortly). It is an exemplar because almost every sociologist knows the theory and, arguably, most believe it is the iron-clad, definitive statement on suicide. How can I be so sure? Having been subjected to reviewers who are not suicide specialists and having reviewed a fair share of what is being written on suicide in sociology and sociology-adjacent fields, Durkheim’s theory looms large over every study. Reviewers often rhetorically ask, if they don’t like the paper, “What would Durkheim say about this?’, or worse, if you try to study suicide with only passing reference to Durkheim, there is always one reviewer confidently arguing that Durkheim’s theory needs more attention. From a reviewer’s perspective, almost every paper I read is a re-test of one of Durkheim’s classic hypotheses using some new data or analytic technique (see the recent review of sociology of suicide for more systematic support of this assertion [Wray et al. 2011]). It is low-hanging fruit to show Durkheim was only partially right or, better yet, wrong. Thus, it seems to explore what the theory is, what it actually does and does not do, and pointing to the problems can provide a clear pathway for those looking to actually extend Durkheim instead of play in his 19th-century sandbox. It makes it a living thing, while also challenging his possession over sociological common property. And by removing it from the amber in which it is fossilized and reinjecting its DNA into contemporary sociology, it encourages the very same sociological imagination that Durkheim exudes in his own writing.

In that spirit, the next entry will examine Durkheim in light of the classification system. Doing so will allow for a detailed demonstration of the theory itself, while also pointing to some of the flaws inherent in the theory.