How Do We Define the Social?

Ontology is the study of what exists. Philosophers have devoted most of their attention to epistemology of scientific research; but it is perhaps even more important to think carefully about the ontology of the domain to be studied.

The social sciences study “social phenomena” - but what are social phenomena? Philosophers have paid less attention than is needed to the question of social ontology; and yet good social science theorizing requires good thinking about the fundamental nature of the social. So an important contribution that philosophers of social science can make is to return to the most basic questions about the nature of social entities. (Ian Hacking has done some important thinking about some of these questions of social ontology; see Hacking 1999, 2002. Other original contributions include Gould (1978) and Gilbert (1989).)

How have social scientists defined the nature of the subject matter of their investigations? Social scientists have singled out a wide variety of types of social phenomena for study. Consider this range: comparison of social types; crime and deviancy; urbanization; social movements and collective action; the workings of legislatures; voter behavior; and consumer behavior. And, clearly enough, it is not necessary to have a comprehensive definition of all social phenomena in order to define a domain for description and analysis.

Some social theorists have sometimes treated social constructs as unified macro-entities with their own causal powers. Structuralist theories main­tain for example that “capitalism causes people to value consuming more than family time” or “democracy causes social cohesion.” Likewise, some theorists have held that moral systems and cultures cause distinctive patterns of behavior - “Confucian societies produce cohesive families.” Each of these claims places a large social entity in the role of a causal factor.

Is this a coherent way of talking? Can large structures and value systems exercise causal influence? The problem here is that statements like these look a lot like “action at a distance.” We are led to ask: How do capitalism, democracy, or Confucianism influence social outcomes? In other words, we want to know something about the lower-level mechanisms through which large social facts impact upon behavior, thereby producing a change in social outcomes. We want to know something about the “microfounda­tions” of social causation (Little 1998).

One point seems obvious - and yet it is often overlooked or denied. Social behaviors are carried out by individuals, and individuals are influenced only by factors that directly impinge upon them (currently or in the past). Consider a particular voter's process of deciding to support a particular candidate. This person experienced a particular history of personality formation - a particular family, a specific city, a work history, an education. So the person's current political identity and values are the product of a sequence of direct influences. And at the moment, this socially constructed person is now exposed to another set of direct influences about the election race - newspapers, internet, co-workers’ comments, attendance at political events, etc. In other words, his or her current political judgments and preferences are caused or influenced by a past and current set of experiences and contexts.

This story brings in social factors at every stage - the family was Catholic, the city was Chicago, the work was a UAW-organized factory. So the individual is socially influenced and formed at every stage. But here is the important point: every bit of that social influence is mediated by locally experienced actions and behaviors of other socially formed indi­viduals. “Catholicism,” “Chicago culture,” and “union movement” have no independent reality over and above the behaviors and actions of people who embody those social labels.

Consider this formulation of a better ontology for the social sciences. Social phenomena are the manifestation of the actions, choices, thoughts, and behaviors of individuals. Individuals interact with each other in a variety of ways. Social outcomes are the aggregate result of these interactions and choices. Social outcomes can be observed and described, so we can provide a set of empirical social facts (e.g., demography, rates of suicide, or urban conditions). With a body of empirical observations in hand, the social scientist can ask a variety of questions: Are there patterns among these data? Are there causes that can be identified that lead to variations in the data? What sorts of causal mechanisms can be identified at work among social phenomena? And the social scientist can consider a variety of ways of further investigating these data: surveys of participants, ethnographic research, collection of data such as prices and quantities of commodities, formula­tion of ideas about social structures, and so on. A social science might then be loosely defined as a research tradition or paradigm that takes a particu­lar definition of the empirical scope of the research; the methods that appear most illuminating; some core theoretical and explanatory ideas; and some specific criteria of empirical evaluation of claims in this area.

4.