Methodological Individualism: Social Facts Emerging from Individuals and Interactions, Study notes of Social Sciences

Download Methodological Individualism: Social Facts Emerging from Individuals and Interactions and more Study notes Social Sciences in PDF only on Docsity!

The Ant Trap: Rebuilding the Foundations of the Social Sciences

Brian Epstein

Table of Contents

• PART ONE — FOUNDATIONS, OLD AND NEW INTRODUCTION iii
• 1 Individualism: a recipe for warding off “spirits”
• 2 Getting to the consensus view
• 3 Seeds of doubt
• 4 Another puzzle: a competing consensus
• 5 Tools and terminology
• 6 Grounding and anchoring
• 7 Case study: laws as frame principles
• 8 Two kinds of individualism
• 9 Against conjunctivism
• PART TWO — GROUPS AND THE FAILURE OF INDIVIDUALISM
• 10 Groups and constitution
• 11 Simple facts about groups
• 12 The identity of groups
• 13 Kinds of groups
• 14 Group attitudes: patterns of grounding
• 15 Group action: more than member action
• 16 Group intention...............................................................................................
• 17 Other theories I: social integrate models
• 18 Other theories II: status models
• LOOKING AHEAD
• Bibliography

Individualism: a recipe for warding off “spirits”

In the middle of the 20th^ century, a battle raged between “individualists” and “holists” about the nature of the social world. At the center of these battles was the philosopher Karl Popper. Popper is best known, nowadays, for his work on scientific hypotheses and how they can be falsified. But in his heyday, he was much more famous for his book The Open Society and its Enemies, one of the bestsellers of 1945. In that book, he railed against what he took to be a cancer in social theory: the idea that societies have collective minds that direct their activities. The Open Society describes a menagerie of philosophical villains, all supposed opponents of human freedom. Karl Marx was probably the most wicked, but following close behind was G.W.F. Hegel. Popper ascribes to Hegel a disturbing view of political systems. He argues that Hegel privileged states, like Germany, Britain, and the Roman Empire, over individuals. Hegel believed, according to Popper, that the interests of states are more fundamental than those of individuals, and that the lives of individuals are in service to the state, rather than the other way around. Behind this, says Popper, was a twisted metaphysics of the state: the idea that states are autonomous, thinking organisms, something above and beyond the people that comprise them. Instead of just taking states to be aggregates of people, this metaphysics ascribes conscious “spirits” to states: The collectivist Hegel…visualizes the state as an organism… Hegel furnishes it with a conscious and thinking essence, its ‘reason’ or ‘Spirit’. This Spirit, whose ‘very essence is activity’… is at the same time the collective Spirit of the Nation that forms the state. 1 By Popper’s account, the trouble begins with the distorted idea that the state is a real thing or substance, a whole that exists separately from the people that comprise it. Hegel’s first error, he argues, was to “reify” the state, treating it as a real and separate thing. Once that mystical metaphysics was underway, it became natural to see the state as having interests and values of its own. It was clever for Popper to choose Hegel as an enemy. The link

(^1) Popper 1945, p. 41

INDIVIDUALISM: A RECIPE FOR WARDING OFF “SPIRITS”

between Hegel and Marx made Popper’s book a bible for anti-communists, whose ranks were swelling in the late 1940s. It did not matter much that Popper’s depiction of Hegel was a caricature. Philosophers nowadays do not look kindly on Popper’s scholarship. Even so, we have to give Popper credit for capturing a genuine anxiety of the time, and also a genuine philosophical problem. The issue Popper gives voice to — that many social scientists make use of a worrisome metaphysics — makes sense. Popper was right that social science has found it hard to avoid speaking of society’s “spirit,” and that it was never particularly clear what that spirit was supposed to be. Over the years, many theorists struggled unsuccessfully to avoid social “spirits.” I will mention two, one from the nineteenth century and one from the twentieth.

1. Leopold von Ranke was arguably the premier historian of the nineteenth century. Ranke was the first truly modern historian, pioneering rigorous historical research. Moreover, he was a fierce opponent of Hegel. He rejected the idea that you could tell a single unified history of the world. The histories of different societies and times, he argued, do not belong in one narrative. And, to be sure, history does not unfold in any particular direction. Ranke did admit the possibility of moral and cultural progress, but he did not see a logic to the history of the world. Instead, he professionalized history, transforming it into a specialized field based on empirical investigation. The job of a historian, according to Ranke, is to craft explanations from the detailed investigation of historical particulars in their contexts. Despite this, Ranke could not figure out how to elude “spirits” in historical explanation. Social epochs, for Ranke, are marked by particular tendencies, and tendencies differ from epoch to epoch. The goal of social research is to uncover those tendencies, as they manifest in particular events. Together, these tendencies reflect a state’s spirit, which is the driving force of human history, much as a person’s soul animates his or her life: If we now ask, ‘What is it which enables a state to live?’ then it is the same as with the individual, whose life incorporates both body and spirit. So too with the state. Everything depends on spirit, which is the pre-eminent of the two. 2 Even the scientifically-minded Ranke was drawn into this position. Nations, states, and societies perform actions. They go to war, they are party to trade deals, they rise and fall. These actions are not coincidences, but occur at least

(^2) von Ranke [1836] 1981, p. 112

INDIVIDUALISM: A RECIPE FOR WARDING OFF “SPIRITS”

enormously improved upon. Even by Ranke’s time, social theory had disposed of a number of worrisome commitments, and by the end of the nineteenth century, many of the assumptions Ranke had made were also abandoned. For example, Ranke did not manage to shake off the idea that the agents moving the course of history were nations or states, acting something like organisms, with their own unified interests and directions. In the late nineteenth century, theorists gave up on these commitments. Moreover, the religious overtones of social progress had withered as science matured over the nineteenth century. Even the focus of social science on states themselves had faded with the rise of theories of class interests. By 1900, the term ‘spirit’ had fallen from fashion in the social sciences. Yet even in the middle of the twentieth century, when Popper and Parsons were both writing, the lurking presence of some incomprehensible ectoplasm lived on. Parsons did not talk of “spirit,” but nonetheless he did not manage to find metaphysically secure ground, free of mystical social unities. To many theorists, Parsons’ theory is as imbued with mysticism as Ranke’s, especially in its apparent suggestion that societies somehow direct, or are autonomous from, individual action. For instance, the sociologist Harold Garfinkel, in a biting critique, accused the theory of insulting individual autonomy and individual intelligence. Every day, as individuals we talk with our friends, families, and therapists about how we should live our lives, Garfinkel pointed out. We seem to be guided by our own goals, thoughts, and imaginings. But Parsons’ theory seems to assign genuine agency to social structures, not to individual people. Garfinkel argues that Parsons’ theory portrays us as little more than puppets, who play our role as dictated by society. As he famously put it, the theory turns people into “judgmental dopes.” 4 Much like Ranke, Parsons found it indispensable, when giving social explanations, to appeal to some kind of social unity that was different or separate from individual people. The reasons for this are clear: how we act and what we do differs radically from one society to another, and from one time to another. These differences are a product of the social contexts in which we are embedded. Those social contexts have properties, they change, they affect us. When we talk about them, we are talking about something.

(^4) Garfinkel 1967

FOUNDATIONS: OLD AND NEW

Both Ranke and Parsons seem to have “reified” the social world in order to construct explanations about social phenomena. Ranke found that historical explanations were best given in terms of social tendencies, and Parsons held that explanations for individual action were best given in terms of the social functions that those actions played. Was this sort of move sound, or was it mystical, as Popper argued?

The trouble with social entities

This debate is an instance of a familiar pattern in contemporary philosophy. It does not just arise in connection with individuals and society. Rather, it is one example of a common problem in inter-level metaphysics: the discipline which studies the nature of “high-level” or “macroscopic” things, and how they are related to “low-level” or “microscopic” things. (Macroscopic and microscopic contrast things at intuitively different levels of organization: for example, to contrast macroscopic things like economies and governments [the things treated in macroeconomics] with microscopic things like individuals, households, and firms [the things treated in microeconomics]. Or to contrast macroscopic things like climates and oceans, with microscopic things like clouds and waves. Or things like proteins and strands of DNA, with things like protons and electrons. 5 ) Is there anything to society above and beyond individuals? The structure of this problem is similar to that of other problems in inter-level metaphysics. We see it in the relation between biology and chemistry: is there anything to life, or to living organisms, above and beyond chemicals? Or in the relation between minds and brains: is there anything to the mind, or to thinking, above and beyond the firing of neurons? In each of these domains, there seem to be two different stances one could take. First, there is the dualist stance: Yes, there is something to living things over and above interacting chemicals, some “vital essence.” Yes, there is something to the mind, over and above interacting neurons, some “soul” or “thinking substance.” Yes, there is something to society over and above interacting individuals, some “spirit” or “social substance.”

(^5) It is problematic even to divide the sciences into “levels.” See, for instance, Thalos 2013; Wimsatt 1976, 1994.

FOUNDATIONS: OLD AND NEW

individuals? And how can we tell what “structures” are in the first place? Despite the considerations motivating Ranke, Parsons, and their ilk, it seems like an enormous step backward for us to reify the social world.

Methodological individualism as the alternative

If “reifying” social entities amounts to mysticism, what other choice do we have in understanding them? The alternative that developed in the long running battle is “individualism.” Put very roughly (with the promise of refining it as we go), this is the view that the social world is made up of nothing more than individual people. In The Open Society, Popper approvingly cites John Stuart Mill, who insists that social phenomena are nothing more than the thoughts and actions of individual people: Thus ‘all phenomena of society are phenomena of human nature’, as Mill said; and ‘the Laws of the phenomena of society are, and can be, nothing but the laws of the actions and passions of human beings’, that is to say, ‘the laws of individual human nature. Men are not, when brought together, converted into another kind of substance…’ 6 Popper does not follow Mill slavishly. In fact, much of his book is dedicated to criticizing Mill’s view that we should fold social theory into the discipline of psychology. But Popper does endorse Mill’s thinking about what society is. There is nothing more to facts about society, above and beyond facts about individual psychology. One of Popper’s students, J.W.N. Watkins, became the dominant voice in favor of individualism in the mid-1950s. Watkins elaborates the claims of individualism in much more detail than Popper did. One of his sharpest polemics is against the sociology of Talcott Parsons. Watkins accuses Parsons of taking a view that is basically theological, and just layering a secular veneer on it. According to Watkins, Parsons does not distinguish his view much from the view that history is guided by divine providence. After all, Parsons does not object to the idea that the actions of individuals are guided by some social entity. All Parsons does, says Watkins, is replace the divine plan with something that seems more scientific. Parsons, like Hegel, is a “methodological holist”:

(^6) Popper 1945, p. 101. Though Popper endorses the metaphysics of “psychologism,” he does not endorse it as a theory of social explanation. See Gellner 1973; Udehn 2001, 2002; Wisdom 1970.

INDIVIDUALISM: A RECIPE FOR WARDING OFF “SPIRITS”

On this view, the social behavior of individuals should be explained in terms of their positions within its cultural-institutional structure, together with the laws which govern the system… This is what is called methodological holism. 7 In contrast, Watkins proposes “methodological individualism”: 8 It is people who determine history, however people themselves are determined. This factual or metaphysical claim has the methodological implication that large-scale social phenomena like inflation, political revolutions, etc., should be explained in terms of the situations, dispositions and beliefs of individuals. This is what is called methodological individualism. 9 Worries about dualism remain, Watkins points out, even if we replace talk of social spirits with social structures, and even if we moderate theories like Parsons’, so that people are not quite so dopey. So long as we talk about social structures and social functions and cultural systems, about bases and superstructures and frameworks of oppression, even about nations and institutions and corporations, we risk treating them as if they are real objects or agents, with intentions, plans and goals, and governed by their own laws or logic. From a metaphysical perspective, all these seem questionable, unlike the more sensible view of those who take the social world to be nothing more than individual people. As the name suggests, methodological individualism is a view about the proper methodology of the social sciences. Methodological individualists take a certain attitude toward theories, explanations, or models, in the social sciences. They argue that these are best given in terms of individual people: a “holist” theory, like Parsons’, is a bad theory, a poor explanation. In the 1950s, debate grew feverish between methodological individualists and methodological holists. In that period, a number of promising individualistically inclined theories flowered, with some especially exciting developments in economics. 10 And individualists like Popper and Watkins were persuasive, not only in their insistence that holist theories were predicated on a mystical metaphysics, but also in their claims that holism threatened individual freedom.

(^7) Watkins 1955, pp. 179- (^8) This term was coined by Joseph Schumpeter in 1908, but only gained currency later on. (^9) Watkins 1955, pp. 179- (^10) Particularly important were game theory (Nash 1950; von Neumann and Morgenstern 1944) and general equilibrium theory (Arrow and Debreu 1954).

INDIVIDUALISM: A RECIPE FOR WARDING OFF “SPIRITS”

the claim that social facts are best explained in terms of individuals and their interactions. That is, theories and models in the social sciences should be individualistic. They should model the properties of individual people and the interactions among individual people. This, however, is a separate thesis from ontological individualism. Ontological individualism is a thesis about the makeup of the social world. It holds that social facts are exhaustively determined by facts about individuals and their interactions. Ontological individualism says nothing about theories or models or how best to construct explanations in the social sciences. All it says is that there is nothing more to societies, their composition and their properties, above and beyond individual people. Explanatory individualism is a stronger thesis than ontological individualism. Even if societies consisted of nothing more than people, it may be impractical or impossible to construct social explanations individualistically. Even if explanatory individualism is false, in other words, ontological individualism need not be. Lukes, for his part, rejects the need for (and often, the possibility of) explaining social phenomena in terms of individuals. But he accepts the ontological thesis that social phenomena are fully made up of individualistic ones. In other words, he sides with Parsons and Mandelbaum on explanation, while agreeing with Popper and Watkins on ontology. Lukes begins with a statement of individualistic ontology: Let us begin with a set of truisms. Society consists of people. Groups consist of people. Institutions consist of people plus rules and roles. Rules are followed (or alternatively not followed) by people and roles are filled by people. Also there are traditions, customs, ideologies, kinship systems, languages: these are ways people act, think, and talk. At the risk of pomposity, these truisms may be said to constitute a theory (let us call it “Truistic Social Atomism”) made up of banal propositions about the world that are analytically true, i.e. in virtue of the meaning of words. 15 Lukes’ argument thus begins with a concession about the nature of social facts, about what they “consist of.” He points out, however, that this does not imply that explanations can be given individualistically. He notes that there are many forms of explanation, among which there are perfectly good ones that do not involve individuals at all. And Lukes uses Mandelbaum-style examples to show that in many cases, we should not expect to be able to give

(^15) Lukes 1968, p. 120

FOUNDATIONS: OLD AND NEW

strictly individualistic explanations of many social phenomena. Even if facts about bank withdrawals are exhaustively determined by facts about large numbers of individual people, that does not mean that we can construct individualistic explanations of banking. In the minds of most philosophers, the distinction between explanatory individualism and ontological individualism divides the uncontroversial issues from those that are to be debated. The ontological thesis is settled: there is no more to society, over and above individual people. What remains open is how best to construct social theories and explanations. Are the best social explanations individualistic? Once we sever the link between ontological individualism and explanatory individualism, we can endorse the former and debate the latter. Having separated these two theses — a controversial thesis about explanation and an uncontroversial one about ontology — philosophers were on their way to a consensus about dualism in social theory: There is no need to be a dualist. Instead, we can be ontological individualists, and still debate explanatory individualism. They were on their way, but still not quite there. The Lukes paper is not particularly precise. He only paints the distinction between ontological and explanatory individualism in broad brushstrokes. For a time, this was good enough for social theorists, but it soon became clear that we could do better.

FOUNDATIONS: OLD AND NEW

With this chapter, I want to underline that emergentism is little different from the prevailing, consensus view. When, in subsequent chapters, I challenge the consensus view, I also challenge emergentism. This book is just as much a challenge to the theory that society “emerges” from individuals and their interactions, as it is to other versions of the consensus view.

Reduction

Let’s begin with the idea of a “reductive” explanation. Consider the relation between “macro” and “micro” phenomena in a different sort of system. Consider the behaviors of a school of fish — a huge school of herring, for instance. Atlantic herring live out most of the year in the North Atlantic, feasting on microscopic creatures. In early winter, these creatures sink into the depths of the ocean, so billions of herring migrate to the Norwegian fjords. (They stay there until January, when they turn back across the Atlantic to spawn.) The fjords are about as protected a place as the herring can find, but their stay is not altogether peaceful. Where herring gather, pods of killer whales follow, attacking in coordinated multi-flank maneuvers. To stay alive, the herring need strategies to protect themselves. In November 1993, the biologists Leif Nøttestad and Bjørn Axelsen sent a small boat out into the fjords to study these strategies. Using a sonar imaging system, they tracked schools of herring as they responded to whale attacks. A single herring school is enormous, consisting of as many as 50 million fish, and stretching for a quarter mile. Imaging entire schools, Nøttestad and Axelsen found that a school behaves as if with one unified mind when whales approach. 1 When a whale swims at it, a school may split in two, half the fish going left and half going right. Or the school may take a sharp turn in one direction. Or, if it is a particularly large school, it may create a moving vacuole around the whale as it passes through, the herring gathering in propagating waves of density. Or the fish may cluster together in a tight defensive ball, herring packed shoulder to shoulder as if they were already in a jar of cream sauce. A school of herring exhibits a wide range of “defensive”

(^1) Nøttestad and Axelsen 1999

GETTING TO THE CONSENSUS VIEW

patterns, parrying in clever ways with each thrust of the predators. With these countermeasures, the herring can largely escape mass slaughter. How do the herring manage this melee, knowing as a group whether to cluster, or split in two directions, or rapidly reverse? What makes the fish at the edges of the ball join the cluster, where they are most likely to be eaten, rather than swim off on their own? What makes them so apparently altruistic, sacrificing themselves for the safety of the group? Back in 1971, the biologist W.D. Hamilton had offered a rather depressing hypothesis to answer questions like these. In “Geometry for the Selfish Herd,” he argued that the members of a herd need not be altruistic for the herd to exhibit apparently coordinated defenses. 2 A number of theorists in the 50s and 60s had used schooling behavior to defend theories of “group selection” — theories in which the evolution of individual traits is explained by the advantages they provide to the group of those individuals. 3 They argued that the gregariousness of individuals, clustering as they do under threat, is mutually protective for the group. Hamilton argued for a simpler mechanism. When a member of a herd or school is threatened, Hamilton suggested, it does one simple thing. It tries to hide. Unfortunately, the only thing for a herring to hide behind is another herring. So every herring tries to dart behind the others. And as they do so, the packing of the school as a whole generates patterns, such as splitting in two, creating a vacuole, or forming a ball. One simple behavior generates a variety of macroscopic patterns in the aggregate. To evaluate this hypothesis, Nottestad and Axelsen used their sonar to record detailed measurements of the whales and the schools of herring. They recorded the reactions of the schools and the countermeasures they took under what circumstances, comparing the results with computer simulations. To the dismay of many an altruist, they found that Hamilton’s theory was all they needed. No self-sacrifice, no coordination mechanisms, no instinct to cluster. Just a swarm of selfish herring, acting according to a simple rule. All the complex behaviors of the school “emerged” from a large number of individuals interacting with one another. This sort of theory, should it turn out to be correct, allows us to talk

(^2) Hamilton 1971 (^3) E.g., Wynne-Edwards 1962, Pitcher and Parrish 1993

GETTING TO THE CONSENSUS VIEW

Adam Smith’s famous statement, “It is not from the benevolence of the butcher, the brewer, or the baker, that we can expect our dinner, but from their regard to their own interest.” 7 Public welfare, it is suggested, can arise from private selfishness. Mandeville, in his 1705 essay The Fable of the Bees, makes the point even more vividly. Where there is a class of vain and lazy nobles, there is demand for herdsmen, weavers, tailors, furriers, and more. If these vices of vanity and laziness were replaced by honesty, temperance, and toil, we might live in a society with greater individual virtue, but collapsing prosperity. In Mandeville’s view, the “diseases” of lust, sloth, avarice, and pride in individuals are essential for a healthy society. 8 This echoes Hamilton’s “Geometry of the Selfish Herd,” where selfishness at an individual level produces gregariousness in the herd. 9 It is a striking fact that properties of individuals can produce their reverse in the aggregate. But the real insight of explanations like these is not that group benefits arise from individual selfishness. Rather, it is that complex and varied properties of aggregates can arise from simple properties of their members in combination with one another. Looking down at swarms of pedestrians from atop a skyscraper or at troops of ants from a picnic blanket, we perceive macroscopic regularities that seem to manifest a kind of group coordination or intelligence. The appeal of the analogy between humans and insects is that it helps us to see social phenomena as mere abbreviations for the complex patterns that emerge from potentially simple interactions among individuals. And this defuses anxiety about “reifying” the social world.

Nagel on theory reduction

If Hamilton’s and Mandeville’s accounts are successful, they give individualistic explanations of group properties. Ideally, the aim of such accounts was to mathematically derive the “geometry of the herd” from the behavior of individuals. This was exactly the kind of work that mid-century

(^7) Smith [1776] 2006 (^8) Mandeville [1714] 1934 (^9) To be precise, Mandeville does not quite present what I have called an “aggregation reversal.” In his essay, it is a subset of the population being dissolute that triggers effects in the rest of the population.

FOUNDATIONS: OLD AND NEW

philosophers of science saw as the central quest of the sciences. Explanations like this formed the heart of their program of “the unity of the sciences,” one of whose aims was to put dualism to rest once and for all. The centerpiece of the program was the notion of theory reduction. Ernest Nagel articulated the classic model of reduction in his 1961 book The Structure of Science. 10 Reduction is a relation between two scientific theories. A theory, in Nagel’s view, is a set of sentences expressing a set of causal laws, both experimental and theoretical. 11 A theory about schools of herring, for instance, might consist of causal laws about the behavior of the school under attack from predators:

1. When the school is in configuration A, and a predator attacks from direction B at speed C, the school will pack into a ball.
2. When the school is in configuration D, and a predator attacks from direction E at speed F, the school will form a vacuole around the predator as it swims through the school. Etc. A theory of individual herring, in contrast, might consist of laws about the behavior of individuals when they encounter disturbances in their immediate environment:
3. When a herring perceives such-and-such a threat, and is in the presence of some other object, it hides behind that object. Etc. To reduce the herring-school theory to the individual-herring theory, we need a set of “bridge laws,” which satisfy two conditions. First, they connect the vocabulary of the first theory to that of the second theory. For every term of the herring-school theory (e.g., ball, vacuole, etc.), a corresponding term is defined in the individual-herring theory. For example, the term ‘ball’ in the school-theory corresponds to a term in the individual-herring theory, defined as a set of configurations of individual herring. The new term is defined so that whenever the individuals are in one of those configurations, the school is in a ball. Second, if we combine the bridge laws together with the laws of the individual-herring theory, it has to be possible to derive from them all the laws of the herring-school theory.

(^10) Nagel 1961. For a more detailed exposition, see Suppe 1977. (^11) See Nagel 1961, pp. 79-105.