Structure vs Agency: A False Dichotomy? (Part 1)

Leiderdorp, a.d. III Id. Dec. MMDCCLXVIII A.U.C.,

One of the major discussions in the social sciences is the discussion whether structure or agency explains social phenomena best. On the one hand, structuralists argue that society is structured in certain ways, and that people are compelled to act in a certain way by those structures. Individuals have no choice, the structures in society determine behaviour. On the other, methodological individualists believe that people have a free will and can make their own decisions.

In the syllabus of a course I took, I found this: “Agency versus determination/constraint in the shaping of global politics: Karl Marx famously wrote that ‘Men make their own history, only not as they please…’ There is an ongoing and fundamental debate in all of the social sciences on what the line is between the  ‘makeability’ of our world and the constraints we face in the making of our world.” This implies that we have the choice between complete makeability, which would imply agency, or determination/constraints, implying structure. However, this puts the bar for agency way too high. Why should we after all expect one individual to be able to change the world completely? It only stands to reason that agents are constrained by other agents, and that only through cooperation they can achieve major changes in the world.

So, I think this discussion is based on a misunderstanding of what structures are. The problem is that some see structure as a system of norms with “equal ontological status” as agency. In other words, structure is a set of institutions which are just as much ‘out there’ as agents are. This leads to confusing discussions, in which it is not clear what this structure is (so far nobody managed to explain convincingly what a structure is), how it affects agents, how it changes, and whether this can be done by agents. In this post I will argue that there is no reason to assume that structures really exist and hold power over agents. It is merely another, simplifying, way of looking at agents, it is a way of reducing complexity by leaving out a lot of details which are thought of as irrelevant. The structure itself has no power, the constitutive agents who form the structure have power. In other words, structure is just the aggregation of individuals, and structure and agency are simply different ways of looking at individuals.

Like I discussed earlier, Hayek, argues that social science must take individuals as its basic building block, because we simply do not have an all encompassing theory of human behaviour based on quantum mechanics. I agree with him on this account, because I am compelled to believe in my own existence as a conscious being, cogito ergo sum, after all. While the assumption that other people also really exist and have a similar experience of their own consciousness is indeed merely an assumption, I think it is safer to bet that individuals exist than to bet that structures, institutions, rules or ideas exist in the same way. In other words, I am forced to believe in my own existence, your existence is very likely, but I have never seen a structure, I have only seen a lot of people interacting. So to me agents and structures do not have the same ontological status based on my personal experiences.

Moreover, Hayek argues we can only explain human behaviour by understanding how actors understand the world, rather than based on how the world ‘really’ is according to the natural sciences. In the first place this means that we assume that individuals really exist, there is such a thing as a person. Meanwhile, relations, norms, rules, ideas, beliefs and other things which connect individuals in a certain way, only exist in the minds of the individuals we study. People see other people behave in a certain way, and attach meaning to that. A police officer fining you only makes sense if you know what a police officer is, what a fine is, what laws are (or in some countries what corruption is), what money is, etc. We can only understand the behaviour of individuals by understanding how individuals attach meaning to the behaviour of other people, how individuals believe the world works, and how individuals understand the norms which they themselves both enforce, but are also constrained by.

In the second place this means that we have to think about structure as collectives of individuals and that any theory about structures must ultimately be based on how individuals behave and how they view the world: its microfoundations must be sound. This does not have to mean that a person needs to be aware of the ‘complete structure,’ something which is even fundamentally impossible (and we are back to Hayek). For instance, a police officer does not need to know how the laws he enforces are legislated by a parliament, which particular agenda setting methods it uses, and and which specific electoral laws are used to elect it, but he does need to know the law in order to fine someone successfully and legitimately. Meanwhile, a corrupt police officer does not need to know why his country is corrupt, he only needs to know which people he can blackmail and which people he can not. Nevertheless, in order to understand the police officer’s behaviour we need to understand how he thinks he is related to other individuals, even if those other individuals are related to yet other individuals, thus making the police officer part of some kind of structure.

Structures are an invention of the human mind, they do not exist ‘out there,’ they exist ‘in there.’ This by no means implies they cannot have explanatory value, but they are not real as a rock or necessarily existent like individuals (assuming that you cogitas ergo estis). Asking people to ‘translate’ their theory from how structure influences agents to how agents affect agents is not nitpicking, it is a way of gaining clarity about what people are saying. The person who argues a structure is influencing people has to show why assuming there is a structure is more useful/elegant than expressing the theory in terms of agents affecting each other. If a structuralist is unable to ‘translate’ his structure into relations between individuals, there is a good reason to be very critical about what that person is saying. This seems to indicate that he believes there is a structure ‘out there,’ distinct of human agency, an assumption which is not supported by any empirical evidence.

Thus, I think we should approach structure as the aggregation of individuals. This does not mean that individual behaviour cannot be looked at as if it is structured. In fact, often this is a very powerful tool we have to explain behaviour. The big difference between the structure-agency debate and my argument, however, is that I think structure and agency are not fundamentally opposed to each other, they are different ways of viewing the same phenomena. To give another analogy, think of the last time you were in a very crowded street. You wanted to go to the other side of that street, but you were constrained by all the other people present. You did not for a moment think that the people around you were a structure, nor that you yourself were part of a structure. Yet, when we zoom out, the crowd is a structure, and even behaves somewhat according to physical laws which also explain the behaviour of gasses and liquids. This does not mean that you did not have a clear goal, namely reaching the other side of that place, nor that you were not an actor with a free will. You were an agent that was part of a structure made up of other agents. (There is a very interesting Economist article on exactly this issue). So, just like the behaviour of light can be explained by assuming light is a particle and wave, human behaviour can be explained as people who believe they are related to others in a certain way, or as if they are a structure, when we look at the behaviour of a lot of people at the same time. This belief that all social phenomena are the result of individual actions is called methodological individualism.

Bottom Line: Structures are not ‘out there,’ they are simply another way of looking at agents. Structures are the aggregated behaviour of individuals.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

Blog at

Up ↑

%d bloggers like this: