Chapter 12: The Body in Action and Embodiment Thesis

Welcome to Last Minute Lecture.

This free chapter overview is designed to help students review and understand key concepts.

These summaries supplement, not replace, the original textbook and may not be redistributed or resold.

For complete coverage, always consult the official text.

Welcome back to The Deep Dive.

Today we are taking a really long look at something we think we know well, the central nervous system.

Right, but we're not looking at it as this, you know, mysterious computer locked away inside the stall.

Exactly.

We're seeing it as just one component in a much, much larger system.

A dynamic system that includes our hands, our environment, and yeah, even our culture.

We're diving into a really central debate in cognitive science right now.

It's about how to square two very powerful ideas.

One is this super successful brain -first view called predictive processing.

And the other is a more radical body -first perspective, which is known as the embodiment thesis.

And our source for this deep dive is a chapter by Michael D.

Kirchhoff.

And our whole mission today is to show that these two frameworks, they're not actually enemies.

Not at all.

In fact, we're going to argue that the embodiment thesis, this idea that the mind literally extends beyond the brain, is actually made stronger.

It's clarified when you look at it through the lens of a wider, more embodied version of predictive processing.

So we're looking for unity where, you know, a lot of people just see conflict.

And to really set the scene for this idea of a dynamic interwoven system, let's start with a classic analogy.

The potter at the wheel.

Yes, the ceramist at the potter's wheel.

And I want you to just picture that for a second.

That continuous,

fluid, almost seamless action.

You've got the ceramist pressing, shaping, guiding the wet clay, and the clay is spinning, its form is just constantly emerging.

It's only sustained by the momentum of the wheel and the tension in the potter's hands.

It's so delicate.

It's this moment of constant high stakes feedback.

I mean, if the potter's hand moves just a millimeter too far, the pot collapses.

Right.

Or if the wheel speed changes, the potter's whole body has to adjust instantly.

The body, the hands, the eyes, the neural circuits, they're not spread things acting on the clay.

They're all completely linked.

And this intense continuous dance is what the philosopher Susan Hurley called a dynamic singularity.

Which is just a great term.

It's a single complex system defined by these constant feedback loops.

You have the external loops, the clay, the wheel gravity, and then you have the internal ones like the neural and muscular processes.

But the whole point is they form one unified system in action.

You can't just point to the brain and say,

that's the cause of the pot shape.

No way.

And there's this fantastic quote from the archaeologist Lambros Malaforos.

He says that trying to separate the brain, the body, and the material world in an activity like this.

Is like trying to construct a pot, keeping your hands clean from the mud.

It's a perfect image.

It's just not possible.

And that profound impossibility is really the starting point for the embodiment thesis or ET.

Right.

So for anyone listening, maybe we should just clarify the key difference between ET and the sort of older cognitive science we all grew up with.

The main difference is that ET, which really came from inactive thinkers like Varela, Thompson, and Rush, it insists that cognition is not about the brain sitting back and building this detailed internal picture of the world.

So it's not a detached reconstruction of reality.

No.

Instead, cognition is a suite of dynamic, world engaging processes, and their whole purpose is to enable embodied activity.

It's all about action.

That's the heart of 4 -E cognition, right?

This shift from thinking about representation to thinking about action.

Exactly.

The key takeaway is that action isn't some secondary output that happens after the thinking is done.

It's the other way around.

Precisely.

Action enables perception and cognition to happen in the first place.

This entire view, it really finds its home in the 4 -E framework.

You know, embodied minds, embedded environments, enacted through action, and sometimes even extended into the world.

Okay.

So when proponents of the embodiment thesis make their case, they usually build it around four foundational claims.

And these are the pillars for our deep dives today.

Right.

This is our roadmap.

The first and maybe the one that causes the most arguments is the constitutive thesis.

The big one.

This is the claim that minds are expensive.

They are realized in sensorimotor activity.

That is, and this is a key phrase, nonlinearly coupled with the environment.

Which, in simple terms, just means the mind is not skull bound.

The body, and sometimes the world, is literally part of the machinery of thought.

Okay.

Second is the non -representational thesis.

So if the body is part of the system.

Then the organism's dynamic ability is its physical structure.

That's often enough for a lot of cognitive tasks.

You don't always need to build these complex, expensive, internal mental maps of the world just to, you know, walk across a room.

Right.

And third, the cognitive affective inseparability thesis.

This one is about experience.

It argues that affect, so, feeling,

valuing, and cognition, thinking, perceiving, they're all inseparable.

You don't perceive the world and then decide if you feel good or bad about it.

The feeling, the value, is part of the information from the very start.

And finally, the fourth one.

The metoplasticity thesis.

Which takes the classic idea of brain plasticity, the brain's ability to rewire itself, and just blows it up.

It says that plasticity happens across the entire brain -body world network.

It's shaped by our actions and even our culture.

So that's a really beautiful unified picture.

A dynamic system, always learning, feeling, acting.

It sounds great.

But.

But.

Here comes the complication.

The theoretical threat.

And that threat is predictive processing, or PP.

This has become, you know, probably the dominant theory in computational neuroscience today.

It offers a single story for everything from perception and action to learning.

At its heart, it says the brain is a statistical machine.

Its main job is to minimize prediction error.

Which is just the mismatch between what the brain expects to sense and what it actually senses.

So the brain is basically a hypothesis testing machine.

It's constantly guessing what's about to happen.

And any sensory input that violates that guess is an error, it's a surprise.

And the brain works relentlessly to get rid of that surprise, either by updating its internal guess, or by acting on the world to make the world match its guess.

But this model seems to immediately threaten the embodiment thesis, especially in its most common form.

Yes, what we call internalist PP.

Let's be really clear about what the internalist view argues, because this is where the conflict starts.

Okay, so the internalist view, championed by people like Jacob Howey, it admits that the body is essential.

You need a body and an environment to get sensory input.

Sure.

But, and this is the crucial part, it insists the body's role is merely causal.

It causes the input, but it's not part of the cognitive process of minimizing the error.

So in that view, the actual number crunching, the internal model that's doing the predicting,

that all happens exclusively inside the skull.

Exactly.

The body is like a really high tech sensor and a motor, but it sits outside the cognitive boundary.

Cognitive states are not extended, the body is external.

And if you accept that the brain's whole purpose is to minimize its own internal prediction errors, it really does seem to point to a very secluded, skull -bound mind.

It's a very powerful argument.

So that's the conflict.

ET says the mind is extensive, internalist PP says it's skull -bound.

And our deep dive today is going to show you exactly why we think this conflict is a false one.

We're going to follow Kirchhoff's argument and show that if you adopt a wide or embodied view of PP, these two theories actually merge beautifully on all four of those theses.

And this doesn't just save ET, it actually drives this really powerful shift away from orthodox cognitive science towards a view of minds realized across this whole dynamic system.

Okay, so let's dig into that first thesis, the constitutive thesis.

The idea that the mind is extensive, that the body and world are literally part of the cognitive system.

This is the real battleground.

It's the cornerstone of 4E cognition, and it's where the internalist critics launch their most powerful attack.

And that attack is built around something called the causal constitutive fallacy.

Right.

This is a major philosophical hurdle for anyone defending an extended or embodied mind.

The critics say, look, just because your notebook causes a change in your memory process, that doesn't mean the notebook is part of your memory.

It's a necessary tool, a causal influence, but the two things remain separate.

Exactly.

And this fallacy is the bedrock for internalist PP.

Howie is basically saying, yes, the body is essential.

It generates the sensory signals, but the computation that minimizes the error in response to those signals that happens only in the brain.

So the brain is defined as epistemically secluded, a black box.

It's shrouded from the world.

Everything outside that box, your body, the room you're in is hidden from view.

And since the brain can't know the outside world directly, its only option is to make its best guess, to generate inferences about what's out there.

Which makes total sense from a statistical point of view.

If you're trapped in a dark room, your only option is to build the best possible simulation of what's outside.

And this leads directly to the formal argument for a skull -bound mind, which Howie calls the self -evidencing brain hypothesis.

It's a really tight logical loop.

Let's just walk through that logic.

Premise one is that if PP is correct.

The brain is fundamentally self -evidencing.

Its primary job is to maximize the evidence for its own existence and its own internal models.

Premise two.

If the brain is self -evidencing, then the mind must be secluded from the body and the world.

Because everything outside the skull is just the source of uncertainty that it has to infer.

And premise three, PP is correct.

So the conclusion is inescapable.

The brain is self -evidencing, therefore the mind is skull -bound.

It's a very compelling argument.

And the engine driving this whole secluded system is something called the explanatory evidentiary circle or the EE circle.

This is pure Bayesian inference.

The brain's internal models, its hypotheses about the world generate top -down predictions.

I predict I will feel my foot on the floor.

Right.

And when that prediction meets the incoming sensory signal, any error, any mismatch is explained away by the model.

And by explaining away the error, the brain maximizes its own evidence.

It proves itself right.

And this entire loop unfolds inside the skull.

The body in this model is just a parameter.

It's something represented in the model, but it's not a constituent part of the cognitive engine itself.

The boundary is the physical skull.

That seems airtight.

So how do the embodiment proponents break out of that circle?

They do it with what's called the wide PP rebuttal.

They accept the basic mechanism prediction error minimization, but they challenge that restrictive physical boundary.

They say the generative model doesn't have to live only in the brain.

Precisely.

Thinkers like Friston, Clark, and Thornton argue that generative models can have wide realizers.

The predictive architecture isn't just neural tissue.

It can be made of parts of the entire organism, its physical shape, its body plant, its capacity for action.

This is that crucial shift in perspective, moving from the brain having a model of the world.

To the agent being a model of its world.

Friston says this directly.

An agent does not have a model of its world.

It is a model.

And if the whole agent is the model, then self -evidencing can't just be about the brain proving itself right.

It has to be understood relative to the entire organism's engagement with its niche.

Minimizing error isn't just about neat internal calculations.

It's about coordinating and maintaining the viability of the whole system, sensory, motor, and neural in the real world.

So the goal shifts from internal accuracy to external practical coordination.

And to make this really concrete, we can use that classic thought experiment.

The outfielders problem.

Ah yes, baseball.

How does an outfielder manage to run across a field and arrive at the exact spot where a fly ball is going to land?

If you try to solve this the old -fashioned way, the skull -bound way, the math is just impossible in real time.

You'd have to calculate the ball's trajectory, factoring in air resistance, wind, spin.

The brain would have to be a supercomputer running a complex physics simulation on the fly.

But the real -world embodied solution is so much simpler and more elegant.

It's called optical acceleration cancellation, or OAC.

And the player doesn't actually calculate the landing spot at all, do they?

No.

They just execute a simple dynamic strategy.

They run in such a way that the image of the ball moving through their visual field stays at a constant speed.

They cancel out any perceived optical acceleration.

So if the ball looks like it's speeding up, they run faster.

If it looks like it's slowing down, they slow down.

And here is the absolute crucial insight for the constitutive thesis.

The player has to be moving for this to work.

A stationary player cannot do this.

Their prediction error goes through the roof.

So the movement itself is part of the calculation.

The movement is the computational mechanism that minimizes prediction error from moment to moment.

The OAC strategy isn't an inner plan sent to the legs.

It's a dynamic strategy realized by the brain, the legs, and the eyes, all working as one single integrated prediction unit.

So the brain alone cannot minimize the prediction error in this case.

The process of minimization literally breaks across the neural and bodily systems.

Exactly.

The body isn't just causing the input.

It is constituting the successful predictive process.

And this shows that the constitutive thesis isn't just compatible with PP.

When you view PP widely, it actually provides the mathematical machinery to explain how the body constitutes cognition.

The mind is not skull bound.

Okay, so we've broken the skull boundary.

We've established that the body can be a constitutive part of the mind.

That was a huge step.

It's the biggest hurdle.

Now let's move on to the second thesis.

The non -representational thesis.

The idea that if the animal and the environment are in this constant coupled loop, do we even need internal representations?

This is a core idea in embodiment.

The argument from people like Silberstein and Camaro is that if there's this constant real -time coupling,

then the so -called gap between the organism and the world.

The gap that representations were supposed to bridge.

It just doesn't exist.

Minds, in this view, are primarily for action, for getting things done in the here and now, not for building detached, detailed internal descriptions of reality.

And this connects right back to the idea of efficiency in the predictive brain.

The internalist's view tends to focus on representational fidelity, getting the internal model as accurate as possible.

But that descriptive accuracy comes at a huge cost, especially in a time -pressured environment.

It's computationally expensive.

The trade -off.

It's a fundamental trade -off.

Prediction error, formally, can be expressed as a balance between accuracy and complexity.

Organisms are always trying to maximize the accuracy of their predictions while, at the same time, aggressively minimizing the complexity involved.

And this is where the body is just a genius at minimizing complexity.

Yes.

The body provides these brilliant, low -cost shortcuts.

And that brings us to the concept of synergies.

Synergies.

Okay, let's define that carefully.

It's not just, you know, a bunch of muscles firing together.

No, it's more specific.

A synergy is a temporary assembly of different processes that are kind of enslaved to act as a single, coherent unit for a specific task.

They're compensatory, low -dimensional relations.

And crucially, they're not static, pre -written structures like the old idea of motor programs.

Exactly.

That's a huge philosophical difference.

A motor program implies the brain writes up a detailed blueprint and then just tells the body to execute it.

A synergy implies the system dynamically organizes itself on the fly in response to the immediate context.

It's a much more fluid concept.

Much more.

Let's look at a couple of examples.

First, central pattern generators, or CPGs.

We usually think of these as controlling simple, rhythmic things like walking or breathing.

Right, these circuits that seem to run on their own.

But the source uses this great example of a mollusk, which shows how dynamic and non -representational this can be.

Okay.

So when a mollusk is just chilling out, it's using one set of neural circuits for slow movement.

But if a predator suddenly appears, a specific neurotransmitter gets released, and it causes this instant functional shift.

It recruits other neurons.

It recruits other interneurons to form a brand new synergy, an orchestrated circuit that enables a sudden, rapid escape.

So that high -speed escape isn't based on the brain calculating a new plan.

It's a physical re -orchestration of the circuit that acts instantly as one unit.

The computational work is basically offloaded to the physical reorganization.

Now think about something that seems really simple.

Postural control.

Just standing upright.

We think of it as just standing still.

But we know that's an illusion.

We're actually in constant low -amplitude movement.

It's called postural sway.

And that sway isn't a bug, it's a feature.

It's the embodied solution.

It requires these ultra -fast compensations across whole -body synergies.

This ceaseless, subtle movement is an active dynamic strategy for staying balanced.

By embracing this tiny bit of instability, the body minimizes the massive control complexity it would need to stay perfectly rigid and still.

It's continuous self -correction, not a static plan.

And these embodied shortcuts show how the system can get the job done effectively.

It can maximize its evidence without having to build those complex internal representations that the old view required.

But how does this more dynamic, non -representational view deal with the formal PP architecture?

What about that boundary, the Markov blanket?

Well, we can revisit that formal boundary.

Friston himself argues that the way the Markov blanket works creates a circular causality.

It perfectly mirrors the action perception cycle that the embodiment people have been talking about for decades.

So external states cause changes inside the organism via sensory states.

And the internal states couple back to the external states through active states, like movement.

It's a continuous reciprocal loop.

That's it.

And that reciprocal circular causation totally undermines the internalist claim that the internal and external are these separate domains that have to be bridged by representations.

If the causation is circular, there's no gap to bridge.

And Clark supports this.

He argues that when you look at low -level free energy minimization, the processes that get picked out as the model, like the self -organization and synergies, they don't seem to have representational content at all.

They're just physical processes that manage uncertainty.

So if the connection is maintained by dynamic action and not internal pictures, then synergies are the perfect candidate for how we anticipate the world, which is the core of PP, without building these laborious representations.

Which proves that PP can fully support the non -representational thesis of ET as long as it embraces the incredible efficiency of the embodied system.

OK, so we've broken the skull boundary and we've shown the system prioritizes action over calculation.

Now we get to the third thesis.

The inseparability of cognition and effect.

Yeah, this is a really important part of forecognition.

The idea that our engagement with the world is fundamentally effective.

It's value -laden from the start.

We encounter the world through what some philosophers call sense -making.

So the cognitive effective inseparability thesis is saying that feeling, thinking, and acting are all just woven together into one single process.

Which means if you're following the logic of predictive processing, that feeling has to be an intrinsic part of minimizing prediction error.

But the narrow internalist PP view really doesn't see it that way, does it?

Kind of deflates the role of affect.

It really does.

In the standard internalist view, effect is secondary.

The ability to judge a situation, to decide if a hypothesis is good or bad, that's just a purely neural, incurential process.

So affect is just a cognitive byproduct.

A little signal that tells the brain how well its model is performing.

Right.

And the body in that view becomes a mere mechanical and functional vessel.

It generates inter -receptive signals like hunger or proprioceptive signals about body position.

But it isn't constitutive of the feeling itself.

But there's a much richer, less intellectual way to think about this, right?

By linking PP with homeostasis and sense -making.

This is where the two theories really start to sing together.

Homeostasis, the body's drive to maintain internal stability, is the key.

And inactivists like DiPaolo link that drive directly to sense -making.

And sense -making is.

It's the creation of a perspective, right?

A point of view from which things in the world are intrinsically meaningful.

Exactly.

An event isn't neutral.

It's either good for the organism because it helps it sustain itself, or it's bad because it threatens it.

And that perspective is rooted in the basic metabolic drive to stay alive.

And that sounds an awful lot like prediction error minimization.

If your expectations are met, you're stable, error is low.

If they're violated, you're surprised, destabilized, and you have to act.

It's an almost perfect one -to -one mapping.

Let's use the simplest possible example, the E.

coli bacterium.

This is like the most minimal form of self -evidencing you can imagine.

Right.

E.

coli swims around looking for glucose to maintain its homeostasis.

It has a very simple generative model that basically predicts there will be high levels of glucose around here.

If that prediction is met, prediction error is low.

The E.

coli keeps swimming smoothly.

It's a state of low surprise.

You could even say a state of relative goodness for that simple system.

But if it swims into an area where there's no glucose, its expectation is violated.

Big surprise.

High prediction error.

And it immediately initiates active inference, which for the E.

coli is a tumbling motion that reorients it to search for something else like lactose.

So the whole concept of goodness or badness, the effect of valence, it's just relative to how well the E.

coli is modeling its little niche.

Expected states, error minimizing states, are inherently valuable because they maximize self -evidence.

They keep the organism alive.

So affect isn't some judgment that happens after perception.

It's baked into the process of self -evidencing itself.

Exactly.

If organisms are models of their environment, then affectivity is an inherent feature of that entire embodied relationship.

Affect and perception aren't two separate steps in a sequence.

They are simultaneous effects.

They're two sides of the same coin, the same underlying strategy of minimizing prediction error.

And there's a formal hypothesis for this, isn't there?

The effective prediction hypothesis from people like Lisa Feldman Barrett.

Yes.

And they argue that effective responses, the signals telling you if something is important, relevant, or valuable, don't happen after you identify an object.

They actually support vision and perception from the very beginning.

So before my brain has even fully identified what I'm looking at, my body and brain are already configuring a felt response, a sense of its value.

Your generative model is already being configured with an effective flavor.

And this brings PP into this really deep constructive alignment with the inactivist idea that organisms enact their world.

We're actively constructing what we experience based on our predictions.

And if that construction, that sense -making is inherently effective, if the very process of staying alive is imbued with value, then the process of minimizing prediction error is itself effective.

We enact our world effectively.

Which perfectly confirms the cognitive effective inseparability thesis from within the wide PPE framework.

All right.

We've arrived at the final and I think most expansive thesis.

Metaplasticity.

Yeah.

This is the capacity for the entire system to continuously reorganize itself.

Traditionally, you know, plasticity was a brainbound concept, synaptic plasticity.

Right.

But the embodiment thesis, especially drawing on material culture studies, just radically expands that idea.

The ET view, and particularly the work of Malafors, argues that metaplasticity is an emergent property of the inactive, constitutive, intertwining between neural and cultural plasticity.

Okay.

That's a mouthful.

So what does that mean in practice?

It just means that our minds, as they act in the world, are constantly being reshaped, rewired, and remodeled across the brain, the body, and our cultural practices.

So learning isn't just about the brain storing new files.

It's about the whole system changing its physical architecture and its repertoire of actions through long -term engagement with the world.

And we have some absolutely stunning empirical evidence for this kind of extensive bodily plasticity.

I'm talking about the study by Iriki and colleagues on Japanese macaques using rakes.

This is a fantastic study.

It's so memorable, and it really drives the point home.

It really does.

So the researchers trained macaques to use little rakes to pull food towards them, and they were measuring changes not just in behavior, but in the animal's body steamer.

Which is the brain's dynamic map of the body, right?

The one it uses for motor control and self -awareness.

Exactly.

And they focused on these fascinating neurons called bimodal neurons in the premotor and parietal areas.

Bimodal because they respond to two different things.

Right.

They respond to touch on a specific body part, like the hand, and they respond to seeing something near that body part.

They basically help map the space immediately around the body.

So before the training, the visual field for a hand neuron would just cover the space right around the monkey's actual hand.

Exactly.

But after two weeks of training with the rakes, the results were incredible.

The visual receptive fields of the neurons associated with the hand had expanded to include the entire length of the tool.

Wow.

And other neurons, ones associated with the shoulder, expanded their receptive fields to cover the entire space the monkey could now reach with its arm plus the tool.

So at the neural level, the rake literally became part of the macaque's felt body.

Its body schema was physically reconfigured by an external object through embodied action.

It's use -dependent assimilation.

The plasticity isn't just a change in the brain's filing system.

It's a change of the entire systemic whole.

The tool, once external, is now constitutively integrated.

OK.

That's incredible.

So how does predictive processing account for this radical, extensive kind of learning?

Well, for humans, our embodied activity almost always unfolds within cultural practices.

And these practices have strong predictable regularities, patterned ways of doing things, from how we speak to how we drive a car.

And in PP terms, these patterns are what establish our priors or hyperpriors.

Exactly.

They are our long -term, high -level expectations that shape how we perceive and act over our entire lives.

Our culture basically stacks the deck, making some things way more likely and easier to predict than others.

And this leads us to another critical PP concept, precision estimation.

Right.

If the brain's job is to minimize error, it needs a way to know how much to trust any given error signal.

Is this error signal important noise I should pay attention to, or is it just random static I should ignore?

And precision estimation is that volume knob.

It's the volume knob.

And it's absolutely constitutive of PP.

A model without precision estimation just wouldn't work.

And the argument here is that the setting on that volume knob isn't just determined by the brain.

No.

The patterned regularities in our cultural practices play a huge role in shaping how we assign precision.

Therefore, the process of precision estimation itself breaks across neural and cultural parameters.

Our culture literally teaches us which sensory signals to trust.

And this becomes painfully obvious when those culturally tuned predictions totally fail, like in cases of culture shock.

Culture shock is just the state of extreme negative log evidence.

It's massive unmanageable prediction error.

The source uses Eva Hoffman's experience of moving from Poland to Canada as an example, her feeling of distress and alienation.

You can't explain that with purely neural parameters.

Because her entire set of deeply tuned, culturally learned, generative models, her social expectations, her linguistic priors, were just constantly being violated by this new environment.

She describes her old country as having fed me language, perceptions, sounds.

And when she arrives in Canada, she feels this profound emptiness because her whole predictive machine is just generating costly unresolvable errors over and over again.

So the lesson, which really ties everything together, is that any explanation of our immediate experience has to be based on this long -term process of error minimization.

And that long -term process has to make reference to the cultural practices that structure our embodied lives.

To think otherwise is to commit what Hurley called the internal endpoint error.

The mistake of assuming that once our brains mature, they alone explain our experience.

So the capacity to learn, to adapt, to even feel the distress of culture shock,

it has to be understood as part of this extended dynamic system.

Which means the metoplasticity thesis is strongly supported by a wide embodied view of predictive processing.

This has really been a deep dive.

It feels like we've shown how the core philosophical ideas of the embodiment thesis, which are so foundational to forecognition, are not just compatible with predictive processing.

They're actually explained and strengthened by it.

As long as you adopt that wider interpretation, the whole landscape just shifts once you let go of the idea of the skull -bound calculator.

So let's just quickly synthesize the remarkable points of convergence we've uncovered today.

First, on the constitutive thesis, we saw that things outside the brain, the body, the dynamics of the world, like in the outfielders problem,

can literally be part of the process of minimizing prediction error.

The body is part of the solution.

Second, the non -representational thesis.

The wide PP view confirms that the mind is fundamentally for action, not for detailed description.

It uses these dynamic efficiencies, like synergies, to minimize complexity and just get the job done.

Third, with cognitive affective inserability, we saw that affect and cognition are really just two sides of the same coin.

They're simultaneous aspects of the organism's basic drive to maintain homeostasis and make sense of its world.

Feeling is the value inherent in prediction.

And finally, the metoplasticity thesis shows that learning and development are these continuous processes of remodeling across the entire brain -body -world system, and they're decisively shaped by our culturally learned priors and precision mechanisms.

So we've moved so far beyond the brain as this detached, secluded computer.

It's more like a dynamic manager of uncertainty, and it's deeply embedded in and constituted by the active life of the whole organism.

And the final provocation that really emerges from all this, I think, is a profound one.

By bringing PP and ET together, we're really committing to understanding the basic conditions that unify life and mind, the life -mind continuity thesis.

If the very definition of what counts as evidence, that drive to minimize surprise, is rooted in the entire organism's active, effective engagement with its niche,

then what boundary really separates a living system from its cognitive processes?

The line between self and world might be less like a solid wall and far more like a semi -permeable membrane, constantly negotiating and integrating across a single dynamic surface.

And that unification, I think that's the future of cognitive science.