The PhD Diaries: Henry Potter & the Dermot McAleese Teaching Assistant Awards

The Role of Agency in Everyday Life

Consider the following set of seemingly unrelated questions:

• What is the difference between a wink and a blink? Is it solely to do with the number of eyes involved or would Mike Wazowski, with his singular eye, be capable of both winks and blinks? If so, what would differentiate the two?
• Why do we blame and/or punish people less severely for their actions when these actions are deemed to take place under ‘duress’ or ‘coercion,’ or in situations of ‘diminished mental capacity’? Why is it that we consider these individuals to be less blameworthy or accountable for their actions?
• How do we determine when appropriate consent has been given within healthcare contexts or in sexual relationships? Why might we even sometimes consider an individual’s given verbal ‘consent’ to be insufficient or invalid?
• Why do (repeat) viewers of Shrek 2 perceive Prince Charming’s love potion to be incapable of eliciting “true love” from Fiona? Why do the feelings of love that emerge from a love potion seem to us to be different from feelings of love that may emerge otherwise?

In one way or another, each of these questions touches on the phenomenon of agency – the subject of my PhD research.
For me, the concept of agency refers to being ‘in charge of’ one’s decisions and actions. It means having the power to act on the basis of one’s own reasons and motivations, free from extrinsic control or internal compulsion.

Such agency is a fundamental feature of our lived experience. For most of us, it feels as though we go about the world making decisions, choosing actions, and behaving in ways that (for the most part) seem genuinely and inescapably ‘up to us.’ We feel like the authors of our own story, the captains of our own ship.

As is reflected in the questions above, much of our social, legal, and moral practices are built on the assumption that we are agents of this sort, too. We generally hold people morally and legally responsible for their actions on the grounds that they need not have performed that action – that is, we assume that it was within their power to have made a different choice or to have performed a different action and, thus, we hold them accountable for the way they did choose to act. That is why, when one’s capacity to exercise such agency has clearly been impeded (as in the case of coercion), blame and punishment are typically less severe.

But is Our Agency Merely an Illusion?

However, despite the essential role that agency plays in our everyday lives, it has become increasingly common to hear it claimed that ‘science has shown that our agency is an illusion.’

These skeptics argue that, when one looks to the scientific literature, the facts become clear: we are not actually ‘in charge of’ our behaviours, nothing in the world is really ‘up to’ us, we have no real say in how the events around us unfold. It just feels like we do. But, in reality, any feeling of choice, of being ‘in charge of’ our behaviour, or of our reasons and motivations making a difference to how our lives play out, is ultimately illusory. Everything that happens, just happens. We are passive observers without any real agency.

Support for these rather bleak claims tends to take several different forms, depending on the type of (natural) scientific evidence one appeals to. The more we learn about biology, for example, and in particular about how the electrical and chemical activity in our brains gives rise to behaviour and conscious experience, the more it becomes hard to resist the impression that that is all there is to behaviour: chains of physical-chemical activity in the brain (and body), from which our feelings of agency and free will emerge as an outcome, rather than a cause. As Stephen Hawking once put it, “it seems that we are no more than biological machines and that free will is just an illusion”.

Even worse than this, perhaps, is the view that physics apparently tells us that our futures are already fixed; that the current state of the universe plus the laws of nature completely determines everything that is going to happen to the atoms that ultimately make up our brains and our bodies. And, thus, the idea that you might actually have the ability to make choices about how you behave – that is, the idea that you ever truly had the option available to you to give medical consent (or not), to eat the chocolate in the cupboard (or not), or to obey your government’s COVID-19 restrictions (or not) – is nothing more than a fantasy. The universe is going to unfold the way it is physically necessitated to unfold, and whether that involves you obeying the COVID-19 restrictions or meeting up with your friends for a chat is seemingly not something you have any actual control over (no matter how much it feels like it does).

AI, Agency, and Responsibility

This tension between the image of the world as it is given to us by science, on the one hand, and the image of the world as it appears to us in experience (and as the basis of our societies) has always fascinated me. During my MSc, where I was studying cognition, I would often become distracted with the question of whether we really do have agency (and, by extension, free will). However, the real-world consequences of these questions were not truly driven home to me until after my master’s degree, when I was working in the ethics of AI field.

In these circles, I found that questions of responsibility were regularly being discussed and debated, such as:

• Who is responsible for the outputs and behaviours of an AI-driven system?
• Is there a point at which it is no longer fair or appropriate to hold the programmer of an AI system accountable for its behaviour/output, given AI’s inherent unpredictability and black box nature? (Air Canada, for instance, recently adopted this perspective by claiming that they shouldn’t be held responsible for any misinformation their chatbot tells customers).
• Is it even possible that, in the future, we might want to hold the AI system itself morally and legally responsible for its behaviour (think of a self-driving car that makes a tragic ‘decision’ to swerve off the road)? If so, what would this form of accountability look like in practice?

These are certainly interesting questions. But I could never resist the nagging feeling that we don’t really know what it is that makes us responsible for our actions – we don’t (yet) have a (natural) scientific understanding of our own agency. And so I found it difficult to see how one could even begin to make any sort of headway on these interesting (and potentially very important) questions about AI responsibility, without first resolving the very real tension surrounding biological agency.

That, in a (rather large) nutshell, is the focus of my PhD research. My aim is to investigate whether it is possible to reconcile our view of ourselves as agents – a view which typically serves as the starting assumption for researchers in the social sciences and humanities – with the empirical evidence from the so-called ‘hard’ sciences – which, for many, seems to leave no room for anything like agency or free will to exist at all, at least not at the macroscale of humans and behaviour. (Spoiler: I think it is!).

Interdisciplinary Research & Teaching

My PhD is thus an inherently interdisciplinary project. Interdisciplinary research is something I have always found myself drawn toward (often without realising it) and it is an approach to research that I am increasingly passionate about and keen to promote.

Given these interests, I was fortunate to be given the opportunity, earlier this year, to work as a Teaching Assistant on the innovative History, Philosophy and Ethics of Science module, designed and run by the TCD Philosophy department, but delivered to science undergraduate students from across the university.
On this module, the virtues of interdisciplinarity are on full display, with multiple academic perspectives being brought to bear on a singular topic or question. Students not only learn about philosophical perspectives which they may not have encountered during their scientific training, but they also get an opportunity in the tutorial sessions to hear from and engage with peers from other scientific backgrounds too (which, in my classes, ranged from geography to geoscience to chemistry to physics).

Pictured are three of the award winners with Dermot McAleese.