1. INTRODUCTION
One of the cornerstones of contemporary social sciences is methodological individualism – the requirement that all social phenomena must be explained by showing how they emerge from the motivations and actions of individuals. Methodological individualism has strong appeal, since individuals seem to be the basic constituents of social reality. Therefore, the fact that it is the dominant approach among economists and other social scientists is not surprising.
In his book, Brian Epstein sets himself two ambitious goals. First, he aims to show that methodological individualism is false. Second, he proposes his own metaphysical framework, which, he claims, gives us conceptual tools to better understand the nature of social facts.
His approach is innovative. However, I shall argue that Epstein's attack on methodological individualism is directed against a straw man. Moreover, the suggested framework is conceptually problematic due to its reliance on vague notions of ontological dependence, which make little sense in the context of social ontology.
2. TWO CORNERSTONES OF THE SOCIAL SCIENCES
In the first part of the book (Chapters 1–9), Epstein criticizes ontological individualism and discusses its relationship to a theory of social facts, known as the Standard Model of Social Ontology.
2.1. Ontological individualism
Epstein's argument rests on Lukes’ (Reference Lukes1968) claim that methodological individualism can be construed both as an ontological thesis and as an explanatory thesis. Epstein's main target of criticism is ontological individualism, which he claims to be a (nearly) universally accepted thesis about the makeup of the social world. Epstein argues that modern ontological individualism is a thesis that ‘the social properties globally supervene on the properties of individuals’ (34). It is a weak supervenience thesis: the properties of individuals exhaustively determine the social properties, but it may not always be possible to produce a correspondence between a given social property and one or more individualistic properties. According to Epstein, weak supervenience ‘gives theorists a way to be ontological individualists, even if they are skeptical about explanatory individualism’ (34).
Epstein claims that ‘ontological individualism is false’, since ‘social facts do not supervene on the individualistic ones’ (36). In Chapters 3, 4 and 5, he offers examples that supposedly demonstrate the ‘intuitive failure’ of ontological individualism. For example, he suggests a hypothetical scenario where Starbucks’ equipment – blenders and refrigerators – gets irreparably damaged due to a late-night power spike. As Epstein puts it, ‘as the owners, employees, and accountants are asleep in their beds, Starbucks goes from being financially solvent to insolvent’. He claims that Starbucks’ ‘transition to insolvency involves property and equipment, not individuals’, and this example shows that some facts about Starbucks ‘fail to supervene on facts about the people and their interrelations’ (46).
However, Epstein's claim that Starbucks' transition to insolvency does not involve individuals is false. An asset is something tangible or intangible that individuals believe (a) belongs to some economic actor (say, a company), and (b) has positive economic value. The value of an asset depends on individual agents’ beliefs about the company's potential of using it to make a profit. This definition implies that the value of material assets can change without any changes in their physical properties, merely with changes in individuals’ beliefs. The ultimate cause of the Starbucks insolvency may lie with the (now broken) blenders and refrigerators. Yet those physical facts can only yield the insolvency the moment a sufficiently large number of individuals form the appropriate belief about the reduced value of the assets held by Starbucks. If no one formed the relevant beliefs, there would be no insolvency.
A more charitable interpretation is that Epstein's example shows some facts about Starbucks to be determined not solely by the facts about individuals, but also by facts about Starbucks’ property and equipment. However, even this interpretation does not make Epstein's argument against ontological individualism more compelling.
In Epstein's example, the set of relevant facts about individuals is restricted to facts concerning the spatiotemporal location and actions of Starbucks employees. Such an extremely restrictive version of ontological individualism cannot account for the majority of social facts: even relatively simple social facts, such as a marriage, a riot, or a trade agreement, can only be defined in individualistic terms if facts about individuals’ beliefs, psychological dispositions and interactions with their environment are added to the picture. The early proponents of methodological individualism were aware of this, and explicitly warned against the overly simplistic readings of its ontological commitments (see, for example, Watkins Reference Watkins1958: 390–395).
If the set of relevant facts about individuals were appropriately expanded, Epstein's argument against ontological individualism would become unconvincing. Epstein is aware of this option, yet he dismisses it on the grounds that some social facts, such as air force battles, ‘do not naturally break down into individual people’ (59), while offering no clear criterion of a ‘natural breakdown’.
The problem is that Epstein's intuitions seem to be fuelled by his personal opinion of what ontological individualists are committed to, rather than by what they are committed to. Epstein's argument is a critique of a straw man – an extremely impoverished version of ontological individualism, which not only does not represent the sophisticated contemporary understanding of this idea, but is also in principle untenable.
2.2. The Standard Model of Social Ontology
Ontological individualism is a view that individuals are the constituents of social reality. Football teams are indeed composed of individuals, but banknotes and traffic signs are not. How can inanimate physical objects be a part of social reality?
The Standard Model of Social Ontology, a prominent proponent of which is Searle (Reference Searle1995, Reference Searle2010), is a theory that a subset of social facts, often referred to as ‘institutional facts’, are constituted by individuals’ collective attitudes towards physical objects or persons. An institutional fact can be any person or physical object on which individuals collectively impose certain status functions. A person or an object can fulfil its functions not solely in virtue of its physical properties, but also in virtue of the collective recognition of its status. A banknote is a trivial example: Any piece of paper can fulfil the function of a banknote as long as individuals collectively believe that a piece of paper counts as a banknote, and that it can be used in economic transactions.
Epstein argues that there is a tension between ontological individualism and the Standard Model, since ‘ontological individualism holds that social facts supervene on facts about individuals’, while the Standard Model holds that ‘the social world is not an attitude in our heads, but the actual stuff in the world to which a certain status or convention has been assigned’ (59).
To understand Epstein's point, consider the category ‘non-kosher animal’. It is an animal that, due to its intrinsic properties (say, being a lobster), individuals who follow the religious norms of Judaism collectively recognize as being unsuitable for eating. It is obvious that the physical properties of the animal play an important constitutive role: that a lobster is a non-kosher animal is a social fact that would never obtain if no animal in the world had the physical properties that individuals would recognize as the properties of a non-kosher animal. The physical properties of animals are not the properties of individuals, and so this is not a fact that supervenes solely on the properties of individuals.
But this is no argument against ontological individualism. As far as beliefs of individuals are shaped by their interactions with the environment, the physical properties of objects clearly play an important constitutive role. This is, again, something that no sensible ontological individualist would ever deny.
3. THE GROUNDING AND ANCHORING OF SOCIAL FACTS
In Chapters 5 to 9, Epstein introduces his own metaphysical framework, which, he claims, offers a coherent metaphysical theory of social facts.
Epstein's framework rests on two different metaphysical relations between facts – ‘grounding’ and ‘anchoring’. According to Epstein, a fact or a set of facts is said to ground a social fact if it is the ‘metaphysical reason’ why that social fact obtains. Grounding is not a causal, but a constitutive relation between facts. A grounding fact is the metaphysical reason for the grounded fact in the same sense as flames are the metaphysical reason for fire. As Epstein puts it: ‘The flames do not cause fire; in a sense, they are fire’ (69).
Let us assume that a dollar bill is any piece of paper issued by the Bureau of Engraving and Printing. In Epstein's terms, a social fact that z is a dollar bill is then grounded by the fact that z is a piece of paper issued by the Bureau of Engraving and Printing.
The grounding facts of each social fact are established by certain rules, or ‘frame principles’. A frame principle ‘articulates what the grounding conditions are for a social fact’ (76). For example, Epstein suggests the following frame principle of a dollar: ‘for all z, the fact that z is a piece of paper issued by the Bureau of Engraving and Printing grounds the fact that z is a dollar’ (77).
According to Epstein, a frame principle ‘expresses the grounding conditions across an entire set of situations, contexts, or worlds’ (77). For instance, the aforementioned frame principle applies not only to actual dollar bills that circulate in the market at present time, but also to dollar bills that might be issued in the future.
Epstein defines ‘anchoring’ as ‘a relation between a set of facts and a frame principle’. A set of facts is said ‘to anchor’ the frame principle if that set is ‘the metaphysical reason that the frame principle is the case’ (82). To put it simply, an anchor is something that makes a particular frame principle to hold in a certain population. For instance, the rule ‘for all z, the fact that z is a piece of paper issued by the Bureau of Engraving and Printing grounds the fact that z is a dollar’ is a frame principle only as long as individuals collectively accept it.
When we put grounding and anchoring relations together, we can see that facts about individuals can play two different roles in social ontology. First, facts about individuals may anchor the frame principles. Second, facts about individuals may ground social facts.
In Chapters 7 to 9, Epstein suggests that his metaphysical framework allows us to distinguish ‘anchor individualism’ from ‘ontological individualism’. Anchor individualism is a thesis that all frame principles are ‘exhaustively anchored by facts about individual people’ (103). Epstein's own examples of anchoring suggest that frame principles are anchored by facts about individuals – their attitudes, actions and shared cultural practices that express their collective acceptance of frame principles. Yet he never endorses anchor individualism as the correct thesis about anchoring, and even claims it ‘unlikely to be right’ (105), while offering no clear suggestions of what else besides facts about individuals could be the anchors of frame principles. He instead focuses on criticizing ontological individualism on the grounds that ‘for the practice of social science, the failure of ontological individualism has a more immediate impact than the failure of anchor individualism does’ (106).
Epstein argues that, in his framework, ontological individualism can be interpreted as a view that ‘if we take the social facts that obtain, then the grounds of those facts – the full metaphysical reason for their obtaining – is some set of individualistic facts’ (109). He argues that this thesis is false: Some social facts are grounded by facts about individuals, but others are not. A dollar bill seems to be an obvious counterexample: The fact that something is a piece of paper issued by the Bureau of Engraving and Printing is not a fact about individuals.
Epstein's framework raises a fundamental question: are some social facts grounded by facts about non-individuals, or must they be grounded by facts about individuals’ beliefs about non-individuals? The difference is fundamental: a fact is something that is the case, or something that has occurred, but it is not necessarily something that individuals are aware of, or believe to be the case. The existence of neutrons is a fact, yet individuals were unaware of their existence for a very long time. Epstein claims that ‘[t]he grounding conditions involve facts about the world, which the people involved may or may not know about’ (193). Can facts about the world that individuals are not even aware of ground social facts?
In Epstein's account, the fact that z is dollar bill is grounded by the fact that z is a piece of paper issued by the Bureau of Engraving and Printing. Presumably no facts about individuals are involved. But is this really plausible? As it has been pointed out by Searle (Reference Searle1995), if no one believed, for whatever reason, that z was issued by the Bureau of Engraving and Printing, then, despite the fact that it actually was issued by the aforementioned institution, it would not function as a dollar because it would not be recognized by individuals as a dollar. The fact that z is a dollar bill is not grounded by facts about its properties per se, but rather by the fact that individuals believe that z has those properties.
It seems that, contrary to Epstein's claim, social facts must be grounded by facts about individuals’ beliefs about the world, and this means that ontological individualism lives to fight another day.
Another problem is that grounding and anchoring are vague concepts. Epstein admits that the basic characteristics of grounding and anchoring relations are ‘still unknown’, but defends his framework by pointing out that our inadequate understanding of the nature of causation ‘does not prevent the term causation from picking out a particular metaphysical relation, even though we do not quite know what it is’ (81).
This line of defence is not compelling because Epstein deals with social kinds, not with natural kinds. There seems to be a consensus that when we are talking about natural kinds, we are not talking about sets of things or properties that are arbitrarily lumped together by individuals. Realist accounts of natural kinds suggest that causal relations hold only between natural kinds. The non-arbitrariness of the natural kinds gives us a warrant for believing that causal relations are embedded in the structure of the world (Bird and Tobin Reference Bird and Tobin2015).
The same cannot be said about social kinds. If individuals can choose arbitrary frame principles, then we can be almost certain that grounding facts do not ‘cut nature at its joints’ – they are arbitrary combinations of different types of facts that do not respect the natural taxonomy of the world. So what reasons do we have to believe that social facts rest on metaphysical relations, rather than on collective beliefs and intentions of individuals? None, it seems.
4. THE ONTOLOGY OF SOCIAL GROUPS
In the second part of his book (Chapters 10 to 18), Epstein explores the grounding of social groups. This part of the book offers a number of stimulating insights into the nature of social groups. It is impossible to do all of them justice in such a short review, so I will focus on several general observations.
To understand Epstein's arguments, it is useful to view social groups as social objects. Just as we can learn various facts about physical objects, we can learn various facts about social groups – facts about their existence at different points in time, hierarchy, internal rules, size, identity and so on. Each social fact about a group has its own set of grounding facts. Some of the facts about the group, such as its hierarchy, are grounded by facts about individuals who constitute it, while other facts, such as the group's age, are not.
Epstein purports to explain certain salient properties of social groups, such as their persistence through change. For example, when the member of the US Supreme Court dies, a new member is appointed to take their place. Despite these replacements of members, the Supreme Court has existed since 1789. According to Epstein, ‘the existence of the Supreme Court at all times after its establishment is grounded by John Routledge's commission as the first member on September 25, 1789’, and that Supreme Court's ‘existence in 1789 is grounded by that fact, as is its existence in 1950, 2015, and the year 3000’ (159).
What Epstein suggests is that US Constitution should be construed as a set of frame principles, one of which is a rule that the Supreme Court comes to existence with the commission of its first member. A historical fact – John Routledge's commission as the first ever member of the Supreme Court on 25 September 1789 – is the metaphysical reason of why the Supreme Court exists at any point in time after 25 September 1789.
Epstein's explanation of group persistence has some radical implications. If a social group's existence is grounded by sets of facts that do not involve facts about their present members, it means that social groups can exist without any members.
This is counterintuitive. Social groups exist to perform certain tasks or to achieve certain fixed objectives. Without members who could do their part in achieving group objectives, a group simply loses its function. It seems reasonable to claim that non-functioning groups cease to exist. The relationship between group's existence and its functions is not thoroughly explored in Epstein's book, thus making the analysis incomplete.
5. CONCLUSION
Despite its flaws, Brian Epstein's book merits study as one of the few truly systematic attempts at answering the fundamental questions of social ontology published since Searle's great book The Construction of Social Reality (Reference Searle1995) and its later refinement Making The Social World: The Structure of Human Civilization (Searle Reference Searle2010). For its ambitious scope and innovative analytic framework, it can be recommended to scholars interested in the foundational problems of the social sciences.
ACKNOWLEDGEMENTS
I would like to thank Jason McKenzie Alexander, Jurgis Karpus, Alex Voorhoeve and Nicolas Wuethrich for helpful comments on an earlier draft.
