There May Be No "Free Will", But That Doesn't Mean There Ain't "Agency

There May Be No “Free Will,” But That Doesn’t Mean There Ain’t Agency

Wayne Martin Mellinger, Ph.D.

In 2012, two prominent scientists announced the death of free will. In a short book titled Free Will, Sam Harris argued that our thoughts and intentions come from background causes we neither see nor control, so the feeling that we author our own actions is an illusion. A few months earlier, the evolutionary biologist Jerry Coyne had made the same case in USA Today. His argument was simple. We are collections of molecules, molecules obey the laws of physics, and so our “choices” are settled by those laws before we ever deliberate. On this view, we are not characters rewriting our parts as the play goes on. We are puppets reading a script that physics has already written.

I do not want to defend free will in the sense Harris and Coyne attack. Coyne defined it clearly: you have free will if, at the moment of decision, you could have chosen otherwise. That is a metaphysical claim about an exemption from causality. It imagines a self that stands outside the causal order and steps in to tip the scales. I think the critics are right to be skeptical of it. It is a freedom that is too free, a frontier town with no sheriff, where nothing outside us, no body, history, or tradition, has any real hold on what we do.

What I want to defend is something else, and sociologists call it agency. The debate over free will tends to erase it, and that is a mistake with real consequences.

Two ways of “acting otherwise”

Here is the irony. Anthony Giddens, whose The Constitution of Society (1984) is one of the most important modern attempts to reconcile action and structure, defined agency with almost the same phrase Coyne used against free will. For Giddens, to be an agent is to be able to act otherwise, to intervene in the world and make a difference.

The words are the same, but the claims are not. Coyne’s “could have chosen otherwise” asks whether a different outcome was possible given the exact state of the universe down to the last molecule. Giddens’s “could have acted otherwise” asks something else: whether this person is the kind of being who monitors their own conduct, grasps the situation they are in, and could have taken it up differently. The first is a question for physics. The second is a question about social life. Answering no to the first does not answer the second.

That is the core error in the free-will skepticism of Harris and Coyne. They move directly from molecules to moral responsibility and skip the level where responsibility is actually produced. That level is interaction.

Where accountability actually lives

In the Doing Modernity project I study how everyday life is accomplished moment by moment through situated practices. I draw on ethnomethodology, conversation analysis, and symbolic interactionism. From that angle, accountability is not a metaphysical property waiting in the brain for neuroscience to confirm or deny. It is something people do. We ask one another for reasons. We give accounts, excuses, and justifications. We notice when someone has breached an expectation, and we hold them to it. Harold Garfinkel showed that social order rests on this constant, mostly invisible work of making actions intelligible and answerable to others.

That practice does not fall apart because neurons obey chemistry. When I run a stop sign and get a ticket, the officer is not making a claim about my brain chemistry. The officer is holding me accountable within a shared order of rules, expectations, and consequences. That order is real. It constrains me, and I help reproduce it every time I stop at the corner.

Coyne himself concluded that the idea of moral responsibility should go the same way as free will. Harris argued that morality survives the loss. The disagreement between them is revealing. Once the self is reduced to a puppet, neither of them has an account of the practices through which people actually hold each other responsible. Sociology does.

Sociology’s own determinism problem

Sociologists should be humble here, because our discipline has its own version of the puppet. Garfinkel famously criticized theories that turn people into “judgmental dopes,” actors who simply carry out the norms and roles society installs in them. Structural determinism is the social-scientific twin of physical determinism. Both explain action by removing the actor.

Much of the best social theory of the last half century has tried to escape both extremes. Pierre Bourdieu’s habitus describes dispositions that are shaped by social position yet produce improvised practice. Giddens’s structuration theory treats structure as both the medium and the outcome of action. Margaret Archer’s morphogenetic approach and her work on the “internal conversation” show how people reflexively work out their course through the constraints they inherit. These thinkers disagree on a great deal. They agree that neither voluntarism nor determinism will do, and that the real work lies in the relation between the two.

Habit, imagination, judgment

Mustafa Emirbayer and Ann Mische, in their 1998 American Journal of Sociology article “What Is Agency?”, offer one of the clearest maps of what that relation involves. They describe agency as having three temporal dimensions. It is iterational, drawing on the past through habit and routine. It is projective, imagining possible futures. And it is practical-evaluative, judging among alternatives within the contingencies of the present.

That framework fits what I have elsewhere called reflexive praxis: the moments when we switch off the autopilot that runs so much of daily life, reflect on our habits and their unintended consequences, and change direction. Most of the time we live in the iterational mode, and determinists are right that much of what we do is not freshly chosen. But people do interrupt their routines. They imagine otherwise. They weigh what they have done against what they might do. Nothing about this requires a ghost in the machine. It requires a social being with memory, imagination, and other people to answer to.

Agency is real, and unequally distributed

Agency is not unlimited, and it is not evenly shared. The capacity to act otherwise depends on resources, time, safety, and standing. A person with a home, an income, and a network has many more ways to act otherwise than a person sleeping in a doorway, even though both are fully agents. Taking agency seriously means taking seriously the conditions that expand or shrink it. Here the puppet metaphor is not only philosophically weak but politically disabling. If everyone is equally a puppet, there is nothing left to say about why some people’s strings are pulled so much harder than others’.

The strongest case for agency is not in the laboratory. It is in the ordinary evidence of social life: resistance, improvisation, innovation, and play. People repurpose rules, invent new practices, refuse scripts, and organize to change the structures that bind them. Things could always be otherwise. That is not a claim that we escape causality. It is a claim that we are among its causes.

Harris and Coyne may be right that there is no free will. It does not follow that there is no one home.