As a neuroscientist, a certain kind of existential worry is an occupational hazard. The more we learn about the mechanisms of perception and cognition and, in particular, of decision making and action selection, the more mechanistic it all seems and the less there seems to be for the mind to do.
How can we even think that we are making choices at all, when we can see that the process is the result of just a bunch of gears turning in the machine? What reason is there to think that an entity is in charge?
Thanks to modern technology, we can actually see the figurative gears turning. Using a variety of neuroimaging tools in humans and animals to track the activity of different neural circuits or brain areas, it is possible to tease out the types of information they carry and the cognitive operations they perform as they make decisions or select actions.
Listen here:
- Listen to a full interview with Kevin Mitchell on Instant Genius, the bite-size masterclass in podcast form, brought to you by the BBC Science Focus team.
We can, for example, distinguish patterns of neural activity that correlate with (and seem to internally “represent”) the accumulation of evidence about something in the world, the degree of certainty attached to some signal, the confidence level in a belief, the adoption of a new goal, the rewards associated with a positive outcome, the learning that happens in response to such rewards, the emotional signals that accompany decision making, the gradual formation of habits, the real-time switch from habitual to goal-directed or exploratory behaviour as circumstances change, and on and on.
We can see the thinking happening.
We can even, in some circumstances, predict an incipient action before the individual performs it. There are many experimental setups using rodents or monkeys where researchers can track patterns of brain activity, observe a threshold being approached that will result in an action, and even predict (not with complete accuracy but significantly better than chance) what action it will be – whether a rat will turn left or right in a maze, for example.
In humans there is a famous example where an action was not only predicted ahead of time but also before the subject even became consciously aware of having chosen to do it.
In these experiments, performed by Benjamin Libet and colleagues in the 1980s, subjects had to randomly decide to move their fingers while watching a clock and while their brainwaves were being recorded by an electroencephalography.
The striking result: the onset of brain activity leading to a movement preceded the reported timing of the conscious awareness of the intention to act by several hundred milliseconds.
Although not relevant to truly deliberative decisions, these findings can still shake your faith in your conscious mind really being in control of your actions. Is the rest of the brain just flattering us, making us feel that we’re in charge, like a wily civil servant expertly managing his elected boss?
If pulling back the curtain to expose the neural machinery of decision making at work were not enough of a threat to our egos (in both senses of that word), it is also possible to intervene in the machine – to drive patterns of neural activity from the outside – and cause the individual to behave in certain ways.
Famous experiments carried out in the 1940s by neurologist Wilder Penfield and his colleagues in human subjects undergoing brain surgery (who were awake and aware throughout the procedure) showed that stimulating different parts of the cerebral cortex with electrodes could produce all kinds of sensations, emotions, urges, memories, or movements of various parts of the body.
This work contributed greatly to the mapping of functions across the brain and reinforced the view of a complex electrical machine producing the contents of the mind, rather than being controlled by that mental content.
Read more:
- An AI friend will always be on your side... until it threatens your free will
- Neuroscience says there’s no such thing as free will. A psychologist explains why that might not be true
- Free Will: The Greatest Illusion?
In animals, a recently developed technique called optogenetics allows researchers to activate specific subsets of neurons and study the effects on real-time behaviour. Using this technique, specific sets of neurons were identified that, when activated, drive all kinds of behaviours – from general locomotion to more subtle motor actions like reaching or grasping, from aggression to mating, from freezing in fright to lunging attacks on prey that are not present, from eating to sleeping to looking after pups, and on and on.
But this research reaches far beyond directly activating particular actions from the animal’s repertoire of behaviour. It has made it possible to dissect the cognitive machinery involved in choosing among actions, weighing options, signalling rewards and punishments, judging the reliability of sensory information, assigning a level of certainty or confidence to a decision, using past memories to guide current actions, and selecting one option while inhibiting every other possibility.
It is even possible to implant false memories in an animal’s brain that will influence its future behaviour. This is not just remote control of what the animal is doing: it is control of what the animal is thinking.
It’s hard not to look at this growing body of work and see only the machine at work. Driving this circuit or that one either directly causes an action or influences the cognitive operations that the animal – mouse or human or anything else – uses to decide between actions.
If we were dissecting a robot in this way we would apply engineering approaches to understand the kinds of information being processed, the control mechanisms configured into the different circuits, and the computations that lead to one output or another. There does not seem to be any need for something like a mind in that discussion. There is no real need for life, for that matter.
If the circuits just work on physical principles, then who cares what the patterns of activity mean? Why does it matter what the mental content associated with a particular pattern of neural activity is, if it is solely the physical configuration of the circuitry that is going to determine what happens next?
We may have set out, as neuroscientists, to explain how the workings of the brain generate or realize psychological phenomena, but we are in danger of explaining those phenomena away.
Fortunately, there is a way out of this existential abyss. The nervous system is not just an electrochemical machine. It’s a meaning-making machine.
By charting the evolution of agency and cognitive control systems, we can come to understand how it is that living organisms can act for their own reasons, and genuinely decide what to do. And we can see how we, human beings, are not just deterministic robots, carrying out our pre-programming at any moment, but entities with real autonomy and causal power in the world – in other words, beings with free will.
Excerpted from FREE AGENTS: HOW EVOLUTION GAVE US FREE WILL. Copyright © 2023 by Kevin Mitchell. Reprinted by permission of Princeton University Press.
Listen to a full interview with Kevin Mitchell on Instant Genius, the bite-size masterclass in podcast form, brought to you by the BBC Science Focus team.
Read more: