Developing a Naturalistic Metaphysics for Biological Agency

Abstract

The aim of this thesis is to contribute to the project of naturalising agency, where agency can be broadly understood as the ability of an entity to control its actions on the basis of its own purposes, goals, or reasons. Such agency is central to the basic phenomenology of our everyday existence as human beings, to how we come to understand ourselves and the world around us, and to the social, moral and legal practices that structure our societies. It is also fast becoming a critical concept for addressing emerging issues in the ethics of artificial intelligence and animal rights.

Yet, despite its profound existential, societal and ethical importance, we still lack a robust understanding of the concept of agency from the perspective of the natural sciences. Part of the reason for this is a widespread and long-standing belief that agency - and related concepts such as freedom, purpose, and meaning - carry with them certain metaphysical commitments that are fundamentally incompatible with the picture of the world as it is given to us by science (what is sometimes known as a naturalistic worldview or metaphysics). This belief has led many scientists and philosophers to conclude that our concept of agency is, in fact, an illusion - a non-natural, unscientific notion that exists in our phenomenology, but not in the `real world'.

In this thesis, I push back on these conclusions by developing an empirically grounded account of biological agency. I show how several of the core features (or types) of agency that are commonly claimed to be mysterious and unscientific are in fact perfectly consistent with a modern scientific worldview. To do this, I consider a series of challenges for agency that have emerged out of the philosophy of action, behavioural neuroscience, and the philosophy of free will. Each challenge targets the scientific credibility of a particular feature of agency: specifically, (i) macroscopic, agent-level control, (ii) purposive or mental causation, and (iii) the ability to 'do otherwise' and make genuine choices. I draw on recent empirical work from across the natural sciences to argue that these features are in fact supported by modern science. I do this by developing an empirically grounded framework for thinking about living systems and their behaviour (within each of the specified contexts) which, I argue, provides an entirely naturalistic and non-mysterious account of the relevant feature of agency. That is, I argue for a novel view of naturalistic metaphysics that appears to resolve many of the existing scientific concerns about the concept of agency. I then end by considering some recent attempts to extend a similar concept of agency into the domains of developmental and evolutionary biology. I argue that, in their current form, these attempts are not successful.