This is part of a series on how to explain human behavior. It is based on a course developed by the late Donald Black at the University of Virginia. Parts 1 through 6 discussed the general logic of scientific theory. From Part 7 onward we cover different strategies or paradigms for developing explanations. The last installment was Part 10: Rational Choice Theory.
“Daddy,” says the boy. “Does everyone have a heart?”
“Yep,” says the father.
“Why?” asks the boy.
“To pump blood,” the father replies. “The blood carries oxygen and nutrients around your body. The heart keeps the blood circulating through arteries and veins. Altogether, they’re called the circulatory system. See, all the body parts need to get fueled, and pumping blood around is nature’s solution to the problem. So that’s why everyone has a heart.”
You’ve almost certainly heard a similar explanation before. We can divide organisms into systems and subsystems, each performing some function necessary for the survival of the whole. And we’re often content to take the function of some part as an explanation for its existence — the heart pumps blood, the eyelids protect your eyes, the liver cleans the Wild Turkey out of my system after a rough night.
Early French sociologist Emile Durkheim took a similar approach to explaining crime. Much as our hypothetical boy asked why all people had a heart, Durkheim asked why all societies had crime. That all societies have some sort of crime he took as evidence that it wasn’t just a disease of the social body, but a functioning part of it.
One of those functions, he thought, has less to do with the criminal act than with the process of labelling and condemning it. Whenever people punish crime, they are doing the important work of displaying their commitment to group morality. Punishing crime builds social solidarity. Thus he thought society need its members to periodically come together in outrage against some sort of deviance. This need is so deep that in a society of saints, where criminal behavior as we know it is absent, the least saintly person would be branded a criminal.
Durkheim’s take on crime is an example of functionalism. It was a paradigm that was prominent in anthropology and sociology from the late 19th century to the early twentieth century. It fell out of fashion in sociology around mid-century, partly from the influence of Marxian scholars who branded its proponents as political conservatives, partly because the general fragmentation of the field accelerated in the late 1960s with numerous other approaches competing for attention.
The Functionalist Paradigm
What I’m here calling functionalism is what Donald Black called systems theory, defined as a kind of theory that explains behavior with its contribution to the needs of the group.
(This is one place where my terminology differs a little from Black’s, who used functionalism as a broader category that included systems theory and NeoDarwininan theory. But most sociologists would use functionalism as the more specific term.)
In its purest form, functionalist explanation is teleological: Something is explained by the ends it serves. The heart exists in order to pump blood, the impulse to define and punish crime exists in order to create solidarity. Durkheim also thought religious rituals created social solidarity, and so even a secular society would develop something akin to them. Kingsley Davis proposed that prostitution exists in order to preserve the nuclear family by giving men a sexual outlet that doesn’t compete with family bonds. Ned Polsky had a similar theory of pornography. Jared Diamond said tribal people are chatty because it ensures they’ll share information that could be useful for survival.
The purest and simplest examples of functionalist explanation don’t have much in the way of explanatory power. But notably, even prominent functional theorists like Durkheim and Talcott Parsons saw pure functionalism as merely a starting point. Durkheim, for instance, was clear that identifying the function of a social behavior was not sufficient to explain how that behavior came to exist in the first place and was only a potential explanation for why it continued to exist. And Parsons considered the classification of society into a set of functional systems to be a good place to get started before moving onto further analysis and explanation.
Functionalist Conceptual Theory
I started with an anatomical example, and indeed functionalists sometimes conceptualize society as an organism, composed of systems and subsystems that allow it to interact with its environment and achieve goals. Writing in the 19th century, English sociologist and polymath Herbert Spencer argued that this “organic analogy” could allow for an interchange of useful ideas between biology and sociology.
A key concept in functionalism is function — the need something serves. But how to determine what the group needs? Functionalists vary in this, though often they see needs as going beyond bare physical survival of the group’s members. The group, society, or institution as such must continue to survive as a coherent pattern. According to anthropologist A.R. Radcliffe-Brown, “the function of any recurrent activity…is the part it plays in the social life as a whole and therefore the contribution it makes to the maintenance of the structural continuity.”
Something might be functional for one part of society but not for another, such as something that helps maintain a family but disrupts the larger community of which the family is part. Thus sociologist Robert Merton said any analysis of function needs to specify just what unit of society the function is for. Unfortunately, this doesn’t stop a lot of functionalist discussion from revolving around the needs of a vague and amorphous “society.”
At first glance it might be obvious what the function of something is. But Merton cautioned sociologists that the apparent or intended function of some social arrangement might not be its entire function, or even its primary function. He distinguished between manifest, or recognized and intended functions and the latent, or unrecognized and unintended functions. The manifest function of schooling might be passing on knowledge, but perhaps its latent function is day care for younger children and keeping older children and youth from competing with their elders on the job market.
Thus different observers can attribute different functions to the same behavior, and it might be difficult to settle the question of which supposed need is really being served.
Following through on the organic analogy, functionalist theorists will sometimes classify certain patterns of behavior as dysfunctional or pathological — something analogous to an injury or disease in a living organism.
Durkheim said that the criterion for distinguishing normal features of society from pathological ones was commonality. First divide up societies into different types (like modern versus tribal) and then compare societies of the same type. If all societies of a type have a social feature, it must be normal — that is, functional. After all, not all societies could be ill at the same time! This was his reasoning for classifying crime as normal rather than pathological.
I don’t think it occurred to him that it is entirely possible that a population of living organisms has a non-zero parasite load, mutational load, and so forth.
Anthropologist Robert Edgerton’s book Sick Societies argues against the nation that any firmly entrenched feature of society is necessarily beneficial — we might well be observing a society on its way to extinction. Though he argues against a common assumption of functionalists, I’d say his book is itself largely a functionalist analysis, albeit one focused on dysfunctions.
Personally, I’ve always been wary of the language of social pathology or dysfunction because, Durkheim’s attempt at objective definition notwithstanding, it’s a way to make value judgments in medicalized terms. Some of Edgerton’s examples, for instance, seem less like things that doom a society to extinction, and more like things — such as the mistreatment of women — that are moral wrongs.
A system is a group of parts that interacts to fulfill some larger function. According to Talcott Parsons, any human group or institution, from the smallest to the largest, can be divided into four systems: adaptation, goal attainment, integration, and latency.
The adaptation system interacts with the environment, gathering and distributing resources. At the national level, think of basic economic production and distribution. Goal attainment is how the group plans and makes decisions. At the national level, think political system. Integration systems keep society’s members and parts on the same page, oriented toward common behavioral rules and understandings. At the national level, think religious institutions and the legal system. And the latency system maintains patterns over time. Think institutions to socialize the youth, like a school system.
The acronym for this conceptual scheme was AGIL, leading to bad jokes among sociologists about the scheme’s agility.
One virtue of AGIL, as Donald Black saw it, was that it showed sociologists what extreme generality might look like: The AGIL concepts were abstract enough to be applied to everything from individual personality up to international global society.
But the biggest problem with AGIL is that it was never clear just what the heck one was supposed to do with it. Parsons’s work was almost entirely conceptual, and so his concepts were never tied to an explanatory theory that one could apply to solve puzzles or make hypotheses. And even as conceptual theory, it suffered from little effort at using it to classify and describe concrete empirical findings. Parsons rarely cited empirical sociology as he churned out new advances in his conceptual thinking. The so-called “grounded theory” approach in sociology, which involves forming and revising concepts based on first-hand research, was largely a reaction to Parsons’s armchair theorizing.
Functionalist Explanations
The functionalism of anthropologist Bronislaw Malinowski focused on how social arrangements meet the needs of the individuals within a society. Consider his theory of magic.
The Trobriand Islanders that Malinowski observed used magical rituals to guarantee the safety and success of their fishing expeditions out onto the open ocean. But they used no such rituals for their fishing trips into their island’s sheltered lagoon.
He observed that these people understood physical cause and effect quite well and did all within their practical power to make sure their crafts were seaworthy. But the open ocean was inherently uncertain and unpredictable — far more so than the lagoon.
He explained that the magical rituals existed to help them cope with the uncertainty, giving them the illusion of additional control beyond what their technology offered. The practice existed to meet the psychological needs of the members of society.
Black called this sort of explanation functional individualism. Others have called it psychological functionalism.
In their article “Principles of Stratification,” Kingsley Davis and Wilbert Moore began by observing two regularities in need of explanation. One was that all societies had some sort of occupational stratification — that is, in all societies, some jobs provide more wealth, prestige, or other rewards. The second was that, across all modern countries, the rough rank ordering of these occupations was similar — for instance, being a medical doctor came with rewards than being a janitor.
What, they asked, is the functional need being met by occupational inequality? Their answer was that society needs to make sure certain occupations get staffed by competent members. Society needs a certain number of janitors, doctors, priests, politicians, and so forth. Society attaches rewards to occupations so as to motivate individuals to fulfill them.
Some of these jobs are harder than others — the talent to do them is more limited, and on top of natural talent it might require a lot of time and effort to gain the skills and knowledge to perform them. Society needs to motivate the people capable of being doctors to go through the long training needed to become a doctor. Whether the ability to do a job depends on talent, training, or most likely a combination, society must reward a job more when ability is rarer, so make sure that rare ability gets distributed properly.
Some of these jobs are also more important than others — society needs police more than it needs hairdressers. To the extent an occupation is more important, it will be a priority for society to make sure it gets fulfilled first. Thus society will offer more rewards for more important occupations, to make sure they are always fulfilled.
Their explanation of occupational inequality comes down to two propositions: The rewards for an occupation increase with scarcity of qualified personnel, and the rewards for an occupation increase with its functional importance.
The first of these is an idea you can also arrive at from the rational choice logic of economics: Hold demand relatively constant, and scarcity of supply leads consumers to bid up the price.
The second proposition is distinctively functionalist. It’s also far harder to test, as one must find some way of classifying occupations according to their importance for society as a whole. And this should be a measure that’s not merely a matter of evaluation about which jobs one most respects, or which are most prestigious — after all, prestige is one of those occupational rewards we’re trying to explain.
When I was in grad school “Principles of Stratification” was one of the most frequently assigned and cited articles in sociology — but not because sociologists liked it. People assigned and cited it to dunk on it. The theory offends the moral sensibilities of most sociologists because the authors suggest inequality might have some positive effect on society, and their explanation for it has no villains. From a scientific point of view, the problem of testability is far more serious.
Another example of functionalism is Kai Eriksons’ study of deviance and social control in Puritan New England: Wayward Puritans. Erikson starts from Durkheim’s observation that defining and punishing deviance has the function of clarifying group norms and commitment to them — in Erikson’s terms, it serves as a form of boundary maintenance.
Based on this assumption, he formulated ideas about the nature and quantity of deviance. One idea is that since the social need for deviance shouldn’t change drastically from year to year, the overall volume of deviance should be relatively constant over the short to medium term. If we see a sudden concern with some particular kind of deviance — such as heresy — this ought to correspond with other kinds of deviance getting less attention.
Another of his ideas was that when we do see waves of crime and deviance, they ought to correspond to some threat to the boundaries or solidarity of the group. He went on to analyze the crime waves (including the famous witch trials) of Puritan society, arguing that each did correspond to some sort of external threat, even if the particular individuals punished weren’t the source of it.
Normal Accidents and Systems Theory
You’ll notice in the examples above not only teleology, but a lot of reification of “society” as an entity that does this or that to meet its own needs. Sociologists who are fine with teleological explanations of individuals might balk at applying them to society as a whole. At least with individuals you can ask them about their goals and believe their answers; it’s harder to question “society.”
And it’s hard to falsify ideas about whether something is serving this or that need — other than by showing something doesn’t even have the effect that is proposed as its function.
Things are a little clearer if we have a narrower conception of the system and more objective criteria for whether it is functioning or failing. Consider technological systems: An aircraft works when it is able to fly but has failed when it crashes to the ground.
And outside of what sociologists usually think of as functionalism, there are systems theories that go beyond explaining things with their functions to analyze specific properties of systems and how these predict whether or not they’ll fail.
One example is Charles Perrow’s book Normal Accidents, which analyzes catastrophic failures like airline crashes and nuclear meltdowns. The behavior of these technologies is not entirely separate from human social behavior, as Perrow emphasizes that people are part of all these system, and human behavior— particularly operator error — is always a crucial part of the disaster.
According to Perrow, catastrophic system failures stem from multiple errors within the system interacting with one another, causing further errors to ramify throughout the system’s parts until it fails. Perrow proposed that two properties of a system predict their potential for catastrophic failure.
First is their interactive complexity — the number of different parts that have some influence on one another. Second is the extent to which the parts are tightly coupled — meaning a failure in any one part very quickly has effects in others. Catastrophic potential increases with both.
Complexity produces more chances for error— more moving parts, more to go wrong. Any sufficiently complex system always has at least some fault — inflammation here, a worn gasket there, a burned-out indicator light, a maintenance crew that didn’t follow procedure and left a hatch unsealed. The greater the complexity, the greater the odds of multiple faults interacting to cause a bigger problem.
Perrow observed that, ironically, attempts to foolproof systems themselves added complexity that could become part of the cascade of failure. For instance, if the warning light burns out, operators assume that whatever it was meant to warn of couldn’t possibly be the problem, and so skip part of their troubleshooting routine that might have solved that problem before it got bigger.
Tight coupling refers to how closely connected and interdependent the parts of the system are. Tight coupling speeds up the cascade of failure, ensuring problems in one part quickly cause problems in others.
For example, in many ways the university is a complex system, with all its many departments, courses, committees, and so forth. But it is loosely coupled. If a faculty member dies unexpectedly, it hardly touches the functioning of other classes in his department, let alone in other departments. Whole departments can be liquidated, and most faculty and students will go on as before — the university as a system can limp along with large chunks of apparatus missing.
It’s a different story if you break a large chunk of a jet fighter. A jet fighter is tightly coupled, and so tends to exist in two states: Functioning near perfectly, or in tiny pieces on the ground.
There are other works out there — in biology, cybernetics, engineering, and so forth — that are often collectively referred to as systems theory. Much of it appears far afield of traditional sociological functionalism, though it shares an underlying concern with how parts contribute to the functioning of the whole.
Traditional functionalism has long been out of fashion in sociology and anthropology proper, and many of the examples I describe here suffer from a lack of testability and explanatory power. But the broader approach to studying the properties of functional systems still has value for explaining human behavior.
And in practical terms, functionalism provides a sociological reason for considering Chesterton’s fence when doing any kind of social engineering. Before you do away with some seemingly obnoxious practice, procedure, or arrangement, its best to figure out what impact it’s absence will have on the rest of the system. It may behoove you to find a more palatable functional alternative before trying to ditch the unpalatable one.
Thanks for reading! If you appreciate Bullfish Hole, please become a subscriber at the link below. You can also leave a tip of any size in the Tip Jar.