I feel that I’ve never understood what the word ‘agency’ is supposed to mean.
In social science it commonly appears as the opposite term of ‘structure’ as in the phrase ‘structure vs. agency‘. This duality connotes the supposed opposition between social forces and individual autonomy in causing the actions of human beings.
I have no trouble understanding how human action is socially caused, i.e. the ‘structure’ side of the duality. But I cannot even imagine what the other side of the equation is supposed to refer to.
Let me explain:
Assume an imaginary world in which all human beings have the same motivations and the same perceptions. For the sake of simplicity let’s assume also that the world is objectively real and that people perceive it objectively.
In this world, human action would be caused entirely by the availability of opportunities for people to realize their motivations.
In this world, if there were no probability or chaos then understanding human nature and knowing objective reality would enable us to predict human action with perfect accuracy. We would know what a given person or any number of people would do next on the basis of what opportunities were available to them. Economists and rational choice theorists build abstract models aimed at precisely this kind of prediction. Of course, they tend to base their models only on instrumentally rational motivations, i.e. motivations for personal gain defined in some objective, quantifiable way. But even without those limiting assumptions, the principle would still hold that, in our imagined world of universal fixed motivations and perceptions, knowledge of human and nonhuman nature would allow us to explain human action as a function of the opportunities available for people to realize their motives.
In other words, we could explain human action entirely by studying the distribution and movement of opportunities.
Let me use the word ‘structure’ to refer to this distribution and movement. In particular, let me shed the notion that ‘structure’ refers to something fixed and static, like the structure of a building; let us use the word ‘structure’ to embrace a wide range of concepts of emergent order, including relation, process, figuration, system, and so on.
To the extent that opportunities are produced by human interaction, and not simply by nonhuman nature alone, then opportunity structures are social structures.
To sum up: if human motivations were constant and universal, and if the opportunity structure were entirely a product of human interaction, then all variation in human action would be determined by social structures.
Let me clear up a few misconceptions that might arise:
- The causal importance of social structure in this model has nothing to do with political oppression or liberty; i.e. ‘structure vs. agency’ does not map onto ‘authoritarianism vs freedom’. The actions of a person who has all the freedoms of the modern white male heteronormative bourgeois subject are still just as socially structured as those of a slave or a prisoner in a concentration camp.
- Similarly, the causal importance of social structure in this scenario has nothing to do with the extent to which people exercise individual creativity or intelligence in deciding their actions. This is because we are assuming (for the sake of illustration) that all people have basically the same cognitive facilities, so that variations in how they use those cognitive facilities are caused by variations in their social situation.
- Social causality is not the same as societal causality. Social structure exist on any level of scale from the global to fleeting interactions between two people. The actions of a person who violates societal norms because the local structure of opportunities makes doing so the more effective path to realizing their motives are, in this model, just as socially determined as those of a normative conformist. In the former case, they are caused by local social structures (maybe even features particular to one interaction between two people) which happen to conflict with more dominant or widely distributed social structures.
Furthermore, as far as I can tell, complicating this model by introducing confusion or ignorance into people’s knowledge of the opportunities available to them, or by making the consequences of action dependent on chance (probability) or chaotic (stochastic) causality does not modify the conclusion: as long as human nature is constant, then variations in human action must be explainable by variations in the structure and system of opportunities for the realization of people’s motives.
* * *
Obviously this account of human action as completely determined by social structure depends on highly artificial and unrealistic assumptions. I’ve made these assumptions to clarify what I mean by social structure, not because I think they obtain. So let’s open things back up.
As soon as we allow that people’s motives vary, as they obviously do in real life, things get considerably more complicated. Then we must ask: where do motives come from? What causes people to have the motives they do?
From the perspective of a naturalistic social science – that is, an investigation that refuses to invoke the soul or any other theological or metaphysical notions – any explanation for human motivation must lie in one or more domains of the natural universe. Broadly, these can include physiological factors (genetics, hormones, etc.), psychological factors (cognitive dispositions, concepts, the unconscious, etc.), social factors (relationships, institutions, norms, etc.), and ecological factors (the living and inanimate nonhumans to which people relate) — recognizing that these three domains overlap considerably. Explanation in any of these domains necessarily invokes some concepts of structure.
Whether my motives arise from my physical environment, biology, my cognition, or my social relations, or some complex interaction of all four, a scientific explanation of my actions necessarily explains my motives in terms of some combination of structures.
This is still the case even I observe that I act reflexively on myself (e.g. through meditation or therapy or whatever) to change my own motivations. I am still acting out a motivation to change my motivations, and that motivation came from somewhere.
If my actions are caused by my motivations relative to my opportunities, and my motivations themselves are caused by some combination of ecological, biophysical, psychological, and social structures (or, more precisely, my motives are themselves physical, psychological, and social structures simultaneously), then there seems to be no meaningful sense in which my actions are ‘self-determined’ for the purposes of scientific explanation. ‘Agency’ defined as ‘self-determination’ seems
So … what is agency supposed to be? What is left over?
(Addendum for system theorists: is ‘agency’ just that portion of my action which is determined by the personality system as distinct from the social or cultural systems? Is that really all that the word is supposed to mean?)