What We're Reading Now

An Argument for Free Will

15 September 2026

This week Geof watched (and read the transcript for) George Ellis’ video: Free will is real: Why science cannot ignore human agency arguing that free will is not an illusion but a real feature of living systems. Drawing on Ellis’ talk for the Institute of Art and Ideas, Geof focuses on three themes he found especially compelling: agency, predictive perception, and the role of high-level conscious choice.

Tags: geof read

If you are not familiar with the Institute of Art and Ideas (IAI), I recommend checking out their site here. Briefly, IAI was founded “with the aim of rescuing philosophy from technical debates about the meaning of words, returning it to big ideas and putting them at the centre of culture. Not in aid of a more refined cultural life, but as an urgent call to rethink where we are.” They are hosting their annual “How The Light Gets In” conference this week in the UK. Their podcast features content from speakers at the conference.

IAI just released a talk from George Ellis, a South African philosopher and cosmologist, about the notion of Free Will, Free will is real: Why science cannot ignore human agency (from last year’s conference I believe). Ellis argues that free will is real, laying out a multi-themed argument. While I was interested in 100% of it, I understood about 40% of it. The parts that I did understand include:

• Agency distinguishes us (and other living organisms) from passive matter,
• Our perception is predictive rather than reactive, and
• Free will lives at the level of high‑level choice, supported by subconscious execution.

Here is a quick overview of the three.

Agency: Why a Pigeon Doesn’t Fall Like a Cannonball

In his presentation, Ellis shares a tongue-in-cheek illustration from the top of the Leaning Tower of Pisa where Galileo and Aristotle are talking physics. Galileo drops a cannonball and it falls exactly as physics predicts. “Galileo says there's no way around it. It's in the equations. It applies to everything because gravity is universal. So it doesn't matter what you drop from the top of the Leaning Tower of Pisa is going to fall like that” recounts Ellis. “Aristotle said, hang on a minute, and he dropped a pigeon and the pigeon flew up.”  

Ellis argues, with this story, that evolution has favored free will to help ensure survival. The pigeon did not fall like the canon ball because the pigeon chose not to fall. A cannonball is a passive object. It has no internal organization that can alter its trajectory. A pigeon, by contrast, is a living system with sensory inputs, internal models of the world, goals, adaptive behavior, and the ability to select among possible actions.

While physics still applies to the pigeon, gravity draws it to the earth just as it pulls on the cannonball, but the pigeon acts. It processes information, evaluates options, and chooses a response.

Predictive Processing: The Brain as an Active Interpreter

The brain does not passively receive sensory data. It actively predicts what it expects to encounter, then updates those predictions based on incoming sensory data.

Ellis shared a slide with the following text to illustrate this predictive architecture:

What I if told you
You the read first line wrong
Same the with second
And also the third

Prediction is, according to Ellis, a core operating principle of our brain, not a quirk. Our brains continuously generate expectations about what will happen. It then compares those predictions with sensory input, (what we experience), and registers “surprise,” and updates its internal model.

We therefore act based on what we think is happening, not merely what hits our senses. Predictive processing is the bridge between agency and environment, helping us anticipate, adapt, and choose.

High‑Level Choice and Subconscious Automation: Where Free Will Actually Lives

Much of human behavior is automated. Think about driving a car, typing a text, or walking. You don’t have to consciously process these rote activities. These routines run in the background, shaped by learning and repetition. Free lives in the high-level choices, not in the mindless activities. While driving a car might seem mindless, deciding to drive to a friend’s house requires free will.

Once the high‑level intention is set, lower‑level systems execute it automatically. This automation frees up our brains’ cognitive resources so we can focus on meaning, strategy, and values.

Ellis and The Institute of Arts and Ideas
Ellis’ lecture is entertaining and informative, as are most of the IAI videos and podcast. Give IAI a visit, at the least, you’re exposed to new topics and thinkers. At the most, the topics might just spark enough curiosity for further exploration.

« Return to What We're Reading Now