Tuesday, September 29, 2015

How physics is driving radical embodied cognition to extinction

Radical embodied cognition is committed to the idea that perception involves some form of "pick up" of information from the environment.  The mind constructs nothing.  Therefore the percepts we have must be direct reflections of information that is in the real world. For example, our percept of a cup is the direct result of light interacting with a unitary object that has a definite location in space.

If it turned out that in the actual world there was not unitary object corresponding to the cup and it had no definite location in space, there would be a significant mismatch between what exists in the world and what is perceived about the world.  Therefore our percepts of a unitary cup in a definite location must be a construction of the mind.

If the above assumptions about REC's claims are true and we indeed found that the world is such a weird place as in the example, then REC must be wrong.  REC is committed to a fairly standard view of a veridical space-time world.

Quantum physics has revealed that a standard space-time world does not hold at the particle level.  So if we lived in that world and took our perceptual measurements in it and if we perceived particles in definite locations (as seems to happen when we measure them with our instruments) then we would have to assume that perception is constructed and not just a direct pick up of information.  REC would not be a viable theory at the quantum level.

Recent work in physics has suggested that something like quantum weirdness applies also to the very large, that space-time doesn't exist at all.  If space-time doesn't exist then that cup is not a 3D object fixed in a particular location in space, despite the fact that my perception tells me it is.  And if this is true REC is a dead theory.

Put differently, if ecological psychology wants to be serious about the interaction of the physical world and the body giving rise to all things human, then it HAS to take discoveries in physics seriously because that is the means by which the true nature of the world can be revealed.  In this sense, REC can be falsified by developments in physics.

25 comments:

modvs1 said...

Daniel Hutto will be quite miffed!

Micah Allen said...

Very interesting. I've never been overly committed to direct pick-up or direct perception, but then i'm a fair bit more moderate than REC. I think somewhere down the spectrum from REC there is a sense in which the body strongly shapes the nature and qualities of our perception, but that this doesn't really tie into the 'realism' debate at play here. The way the body shapes the mind isn't about veridical perception, and in fact often likely goes against it. Our conscious body image is a good example of the type of construction interface seeks to explain. I could very vaguely see there being a union along the lines of: our brain essentially acts to maximally exploit the ecological niche implied by our body and it's environment. The appearance and functionality of the 'interface' is thus embodied in that it exploits these features; literally, our perception of the world is embodied because this is the optimal way to control a body. I haven't actually read IF though so just speculating based on your argument with Andrew.

On a random note, your post makes me think of a strand of analytic epistemology/ontology known as mereological realism. This view argues essentially that there are no cups, or chairs in the world but instead only clumps of atoms arranged chair-wise. Some variants of the view argue that the only 'real' objects are persons as these are indivisible explanatory/causal units, but that is obviously quite radical. Anyway I wonder what it means for interface theory if there are in fact no objects in the world. Kind of turns the realism aspect on it's head a bit doesn't it?

Andrew Wilson said...

Right. So, no.

Let me expand :)

Hoffman and I agree on one thing; perception is not about figuring out The Truth about the world. It's about coming to interact with the world in functional ways.

The ecological laws (Turvey, Shaw, Reed & Mace, 1981) governing how perceptual information about the world is created have scope. This means that they apply within a range, and that range is medium sized objects moving fairly slowly (aka the world outside your window). The claim is that if you measure the world on this ecological scale, what you end up with is a description of the world that contains things like objects, events, affordances, etc.The claim is not that these are How Things Really Are. The claim is that this is how the world is structured at an ecological scale when measured by a biological organism.

The goal of physics is not the same as the goal of a perceptual system. The goal of physics is to come up with a description of the world that covers as many possible scales as possible and has the widest possible scope. When you do this, you end up with a description that includes things like particles, waves, strings, black holes and energy. Even this is, strictly speaking, not How Things Really Are, it's just a description (with extraordinary success and scope) of how the world is structured at a different scale when measured by the devices of physics. All cool, but that's all it is.

Physics spends most of it's time trying to replace that description with something better; something with greater scope. As it does so, some of the things it talks about now will go away. But unless the thing that goes away is 'a world that can in principle be measured at an ecological scale by an organism' then ecological psychology has no need to panic about the changes.

Ecological psychology is indeed committed to the idea that there is a world that can be measured at an ecological scale. If that is not true, then we we are indeed screwed. But if that's not true, we really are screwed because what else is perception if not measurement of the world at an ecological scale? This world can also be measured at a more universal 'physics' scale; but Gibson's key insight is that that scale is irrelevant to a perceiving acting organism. Why? Because a) perceptual measurement and physics measurement are trying to achieve different things and b) there is no information in the ecological scale about things like black holes and quantum weirdness so there's nothing to drive a perceptual system to find out about those things (we needed massive equipment and theories to find this stuff). Just because quantum effects have been observed at ecological scales (which is awesome, don't get me wrong) does not (yet) entail that everything at the ecological scale is best described for an organism as a quantum effect.

On Twitter, someone mentioned chemistry. Your argument would mean that chemistry is also screwed because it's all 'really' quantum. Except chemistry ticks along just fine studying interactions that happen at a scale where quantum is not the best description.

Reduction of everything to physics was cool in the 90s but thankfully seems to have died a natural death (unless the Churchland's are still kicking this :). Ecological psychology does depend on things from physics (eg dynamics) but just because you can tunnel down from dynamics to quantum doesn't mean you have to or that it is a good idea. Like chemistry, our science needs to happen at the right scale and at that scale, there are objects and not wave/particles in our ontological tool kit.

David Poeppel said...

You really don't need physics ... REC will drive itself to extinction because its theory of concepts is so dramatically underspecified as to not go anywhere, and its inability to handle the most elementary aspects of conceptual combination will remain, until further notice, a show stopper. The physics arguments won't help either.

Ed Baggs said...

One of the stumbling blocks here is the word 'environment'. An environment, for Gibson is not synonymous with 'truth' or 'reality' or 'the real world'. An environment, for Gibson, is that which surrounds an individual animal--what constitutes a given animal's perceptual or personal world. An example: colours in the ultraviolet spectrum do not make up part of the environment for a human, but they might make up part of the environment for a hummingbird. Even though we share the same universe as the hummingbird, we have different environments because our bodies are built differently.

William Matchin said...

Andrew Wilson Wrote:

"The ecological laws (Turvey, Shaw, Reed & Mace, 1981) governing how perceptual information about the world is created have scope. This means that they apply within a range, and that range is medium sized objects moving fairly slowly (aka the world outside your window). The claim is that if you measure the world on this ecological scale, what you end up with is a description of the world that contains things like objects, events, affordances, etc.The claim is not that these are How Things Really Are. The claim is that this is how the world is structured at an ecological scale when measured by a biological organism."

So in other words, what you're saying is that organisms don't perceive things that are actually in the world. They actually perceive them according to some highly specific and abstract properties that are innate and have nothing to do with physical reality (if we believe what physicists tell us). Sounds like you're a radical nativist to me.

I think Chomsky put it best when describing the mind/body problem - there is no problem, because body (aka the physical world) is undefined, if we take physicists seriously. So you can't get anywhere without appealing to such occult notions as "medium size object". That is not a physical notion except in terms of folk psychology.

The point about chemistry seems quite irrelevant to me. It is one thing for a field to have utility by studying things at a certain level, and another thing to believe that there will be a reduction or unification of this level with another level. Even radical nativists believe that in theory a unification of cognitive science with physics is possible, but they haven't the foggiest how this will actually happen.

Greg Hickok said...

Micah Allen raises questions about the role of the body in all this. This raises a couple of interesting points. One concerns the status of not-so-radical embodiment, which still allows perception/cognition to be constructive, etc. I think the argument from developments in physics is equally damning for the idea that cognition is grounded in the environment. If the mind constructs what that environment looks like, then the grounding is occurring in a cognitive construction. In my book I put out a similar critique of the embodiment claim that cognition is grounded in perception and action by showing that the claim amounts to grounding cognition in cognition. The second interesting point concerns the perception of the body. According to ITP, the body is also in the interface and isn’t in truth what it appears to be. It’s just another useful icon. This pretty much precludes straightforward appeal to the body as constraining anything.

Micah Allen also asks about the implications of a situation where there are no objects in the world in any form. ITP is committed to there being something. It is just says that the something is not what we perceive.

Greg Hickok said...

Andrew Wilson states, “The ecological laws ... governing how perceptual information about the world is created have scope. This means that they apply within a range, and that range is medium sized objects moving fairly slowly (aka the world outside your window).” ITP says that the range of medium sized objects moving fairly slowly is an illusion constructed by our perceptual systems. In fact, the reason why it looks like ecological laws have this scope in the first place is because that is what our perceptual systems construct and reveal to us. The possibility that I am asking you to entertain (and take seriously as a topic of investigation) is that, for example, the whole of the world is, say, quantum. Objects have no fixed position, time doesn’t exist, space is not 3D. We know (or at least our best models suggest) this is true at the particle level. What if it scales up? Why are we so confident that our senses aren’t deceiving us? After all, if all we had were our measurements of the quantum world, it would look like particles have a definite position/spin, etc. But that only happens when you take a measurement. If perception is measurement, how do we now it’s not just collapsing the wavefunction of a quantum world?

Wilson: “The goal of physics is not the same as the goal of a perceptual system.” Agreed! Physics is trying to understand the nature of the physical world. Perception is trying to make babies. But you have to make babies in the physical world so the nature of physical world has to impact how perceptual systems work. In short, it has to matter to your theory of perception whether spacetime is real or not.

Wilson: “The goal of physics is to come up with a description of the world that covers as many possible scales as possible and has the widest possible scope. When you do this, you end up with a description that includes things like particles, waves, strings, black holes and energy. Even this is, strictly speaking, not How Things Really Are, it's just a description” You can’t dismiss the implications of physics for understanding how things really are in the world as mere description with no real world consequences. Case in point: Einstein’s relatively is a mathematical description that suggests the actual universe warped, time is relative to speed, etc. We don't perceive this, of course, so maybe we can write it off as just a description and not reality. But it is has real world consequences. If relativity corrections were not implemented in GPS tracking systems, they would be utterly useless. So far discoveries about the nature of the world have not impacted our theories of perception because they are discoveries about the very small or very large. My point is that physics is starting to encroach on our perceptual sweet spot of medium sized stuff. We are going to have to take these discoveries seriously.

Greg Hickok said...

Wilson: “Ecological psychology is indeed committed to the idea that there is a world that can be measured at an ecological scale. If that is not true, then we are indeed screwed.” So we have two clear and opposing predictions for ITP and Eco Psych: ITP says spacetime is doomed, EcoPsych says it can’t be, at the “ecological scale.” This prediction can be decided by research in physics. Again, you can’t dismiss research in physics as non-applicable.

Wilson: “But if that's not true, we really are screwed because what else is perception if not measurement of the world at an ecological scale?” ITP provides a framework. Perception tunes to fitness functions at whatever scale is useful. The “ecological scale” is nothing more than our user interface.

Wilson: “Reduction of everything to physics was cool in the 90s but thankfully seems to have died a natural death” This is not an exercise in reductionism. Physics is just the means to understand the nature of the information (to use Gibson’s term) that perception is dealing with. EcoPsych says that that information is veridical for medium sized objects and relatively slow speeds. What you see is (at least a portion of) what’s actually there. ITP says what you see is not even close to what’s there even for that subset of stuff that is perceived. Again, physics can answer that question by showing or disconfirming that medium sized objects are subject to the same weirdness found at different scales. If this is in fact confirmed, this doesn’t mean perception is reduced to physics. It actually makes the job of the perceptual scientists harder and more interesting: why do we perceive spacetime when it is not there? Why is it useful for making babies? What is the nature of the transformation from say quantum stuff to spacetime representations. Physics can’t solve that problem for us. They have the easy job.

Greg Hickok said...

David Poeppel: “REC will drive itself to extinction because its theory of concepts is so dramatically underspecified as to not go anywhere, and its inability to handle the most elementary aspects of conceptual combination will remain, until further notice, a show stopper.” Not to mentions illusions, consciousness, prediction in all its forms, the necessity of information processing (i.e., computation), color, emotion, decision making, and on and on. I agree. One approach to REC is to let it wither on the vine. But these ideas come back time and again. REC is a slightly more moderate version of behaviorism, for example. The idea behind engaging is that after it dies the strength of its zombie resurrection will be proportionate the strength of the arguments against it while it was alive. I’m just trying to save future generations from a zombie apocalypse.

Ed Baggs: “An environment, for Gibson is not synonymous with 'truth' or 'reality' or 'the real world'. An environment, for Gibson, is that which surrounds an individual animal--what constitutes a given animal's perceptual or personal world.” But it is veridical within that portion of the organism’s perceptual world. ITP says it’s not veridical even within the perceptual space of a given organism.

Harry Farmer said...

William Matchin Wrote:

"So in other words, what you're saying is that organisms don't perceive things that are actually in the world. They actually perceive them according to some highly specific and abstract properties that are innate and have nothing to do with physical reality (if we believe what physicists tell us). Sounds like you're a radical nativist to me."

I'm not a fully paid up supporter of REC but I think here your making the mistake of saying that anything that is not counted as part of fundamental physics isn't "real", but by this argument pretty much everything we talk about ends up not being real, since fundamental physics has no place for objects (at least as conventionally thought of) at all. You seem to make a false dichotomy in which either something is real at the level of fundamental physics or it is an abstract property without any basis in physical reality.

However there is a path in between those two options which has been outlined by a number of philosophers of science and mind, most notably James Ladyman and Don Ross who, drawing on Dan Dennett's work on real patterns, argue that objects and indeed more abstract concepts like economic systems can be considered to have ontological grounding in virtue of the fact that they allow for the tracking of patterns of relationships in a more compact amount of information than would be possible at the level of fundamental physics. On this basis the patterns that are perceived and used by an organism would count as real provided it allowed for that organism to reliably track and respond to the relevant changes in it's environment in a more efficient way than tracking patterns at the level of fundamental physics. This argument also gives an indicator as to why the status of chemistry and other scientific disciplines is relevant to this debate since on the Ladyman and Ross view all branches of science that are not fundamental physics are based on tracking patterns that exist in various subsets of the universe (note that scale isn't always the relative determinate of the subscale, for example biology is not necessarily limited in scale but rather by the fact that it operates on any region of the universe in which evolution by natural selection is able to occur. So there is a broad parallel between what makes the special sciences real and what makes information collected and tracked by agents real.

Andrew Wilson said...

The possibility that I am asking you to entertain (and take seriously as a topic of investigation) is that, for example, the whole of the world is, say, quantum. Objects have no fixed position, time doesn’t exist, space is not 3D. We know (or at least our best models suggest) this is true at the particle level. What if it scales up?
Let's take it seriously then. The world is just a seething mass of quantum whatnot, and perception is an interface we evolved to interact with said whatnot.

Interfaces are representations, and so they run right into two questions. Where do the representations get their content (ie why do we have the interfaces that we have?) and which one do we deploy at any given moment? For standard psychology, you stand a (small) chance of grounding representations so that they might work. If it's all just quantum whatnot, neither of these questions have answers (and I did notice that Hoffman et al never try to answer either of these questions in their paper).

Andrew Wilson said...

My point is that physics is starting to encroach on our perceptual sweet spot of medium sized stuff. We are going to have to take these discoveries seriously.
It did this a long time ago. It's called Newtonian mechanics and there's nothing quantum about it. Newton was good enough to land the Apollo missions on the moon, by the way.

EcoPsych says that that information is veridical for medium sized objects and relatively slow speeds.
Not quite. Veridical implies 'nearness to truth' and interface theory conflates true with 'world as measured by physics'. EcoPsych does not think information is veridical in this sense. It does claim that the information specifies certain real properties of the world at the ecological scale, but this is different.

Andrew Wilson said...

Again, physics can answer that question by showing or disconfirming that medium sized objects are subject to the same weirdness found at different scales. If this is in fact confirmed, this doesn’t mean perception is reduced to physics. It actually makes the job of the perceptual scientists harder and more interesting: why do we perceive spacetime when it is not there?
OK. This is getting closer to a good articulation of the point.

We both know that quantum effects have been found operating in biological scale activities (although iirc I think it's still at molecular levels like in photosynthesis; cool, but not yet perceivable). This is indeed cool, and indeed, if the ecological scale was revealed to be all quantum whatnot, then yes, this would be a problem. As I said; the ecological approach is committed to there being a world that can be measured at an ecological scale in such a way that things like objects, events and affordances emerge as properties of that world at that scale.

But I think we're good. To the best of my knowledge the weirdness of quantum is almost entirely filtered out by the time you reach just the molecular level and no one uses quantum mechanics to describe anything at the ecological scale because it makes no sense to. In fact, physics talks about that scale using Newtonian mechanics, a system in which things like particle/waves never even think about showing up.

So it might be particle/waves or strings at the bottom. But if you relax your magnification and look just part of the way down the scales of speed and size, all you see are surfaces and objects and the like. And both of these descriptions make sense, within the bounds of their respective scopes. Physicists know this, I think.

Greg Hickok said...

"Where do the representations get their content (ie why do we have the interfaces that we have?)" Because they are good for fitness. Nothing really changes in terms of representational/computational models. There is stuff in the world that is lawful (even it is probabilistic, say, in the quantum sense); there is no reason we have to assume it is a primordial soup. Acting on that stuff can either increase or decrease fitness. Perception evolves to represent that stuff in a way that increases the likelihood of survival and reproduction. In non-ITP representational/computational models the stuff that gets represented is truth about the world, which everyone assumes will naturally lead to better fitness. The evolutionary simulations coupled with developments in physics suggests that truth isn't what gets represented. Instead, what gets represented is fitness functions. The world looks the way it does because it is useful for reproduction. Young adult humans look attractive to other humans because that's useful for fitness, not because there is something inherently beautiful about the young human form. Ditto baboons for baboons. Beer bottles look attractive to Jewel beetles because (in the hops and barley free world they evolved in) it's useful for fitness. And on and on.

If this is correct--and signs are currently pointing in that direction--it defines a problem: why is a spacetime perceptual world useful? Do all species operate in a spacetime interface? How is a spacetime percept computed from a spacetime free world, whatever it is?

Greg Hickok said...

Re: Newton... See Hoffman's Invention of Symmetry Theorem from the ITP paper. It proves that you can land a rover on the moon using Newtonian mechanics even if there is no such thing as spacetime. That argument is off the table, by theorem.

Greg Hickok said...

GH: EcoPsych says that that information is veridical for medium sized objects and relatively slow speeds.
AW: Not quite. Veridical implies 'nearness to truth' and interface theory conflates true with 'world as measured by physics'. EcoPsych does not think information is veridical in this sense. It does claim that the information specifies certain real properties of the world at the ecological scale, but this is different.

GH: That's what I said: "veridical for medium sized objects at relatively slow speeds" -- that was your definition of the ecological scale, which you now say has "real properties." So these real properties are veridical according to EcoPsych.

Greg Hickok said...


AW: if the ecological scale was revealed to be all quantum whatnot, then yes, this would be a problem.

Quantum effect spotted in a visible object

So quantum effects hold at a scale within the ecological scale. The reason why we don't see them everyday, according to ITP, is that they are hidden by the interface which builds us a tidy spacetime representation.

Andrew Wilson said...

All this physics stuff is coming from Hoffman's reply to Objection 6 in the interface theory paper (http://link.springer.com/article/10.3758%2Fs13423-015-0890-8)

Objection 6 The interface theory says that our perceptions of objects in space-time are not veridical, but are just species specific icons. Doesn't it follow that (1) no object has a position, or any other physical property, when it is not perceived, and (2) no object has any causal powers?...

Reply The interface theory indeed makes both predictions. If either proves false, then the interface theory is false.

Hoffman then says this is all consistent with modern quantum theory, so ITP is good.The interface is then a layer on top of this world to help with the baby-making.

Greg, your move is then to say 'what if this scales up? It's quantum at the bottom, so maybe this means that the things at the ecological scale show this weirdness too.' You point to two examples.

The first is a new way to formally describe particle scattering that doesn't make reference to space-time. This just seems to imply that future theories might not invoke space-time as a player in the calculations and formalisms. This is just about physics reorganising its formalisms to suit itself as it tries to explain things neatly and efficiently.

The second involves observing quantum super position in something 40µm in size when it's cooled to 0.1K. Interesting, but hardly ecological scale, and that article also notes the cooling was critical because

a significant challenge here is to eliminate all thermal vibrations in the object, which mask or destroy quantum effects

None of this supports the claim that the weirdness ITP is committed to scales up to the ecological scale, and to the best of my (limited) physics knowledge, there are many reasons to doubt it would ever scale up, because of things like thermal vibrations, etc.

So Hoffman (and you) are in a bind. Hoffman commits ITP to the prediction that there is no irreducible way to describe the world where a property of our perceptual experience (such as the position of an object) is a real thing, ie something like quantum mechanics is the way things are at all scales, including the ecological scale. Hoffman bets the house on this - if this is false, ITP is false.

So now it is a matter of seeing if physics ever decides that the right formalism for describing the world at the ecological scale is anything like quantum mechanics rather than Newtonian mechanics. If that ever happens, REC will indeed be in trouble because as I've said, we are committed to there being a fairly well behaved ecological scale to the world. So I'll see you when physics ruins it for all of us, and in the meantime I'll just be over here getting some science done on the assumption it won't.

Oh, and if physics does ruin it for us, it'll ruin it for you too. If the world really doesn't constrain perceptual experience, even a little bit, then you will struggle mightily to come up with a good account of where interfaces get their content and why we deploy the interfaces we do when we do. This problem faces all representations, but at least by the standard view the goal is veridical mappings so they stand some chance of solving the problem. If ITP is right and veridical mappings are off the table then inferential perception is doomed for being massively, unsolvably under-constrained. Of course, if it's right about the physics, then we're all doomed, but I think the good news there is that we're here and perceiving, so ITP is probably wrong about the physics.

This was fun, as always :) I think this bit makes a nice second half to the comment on ITP I'm drafting; PB&R is the new home for everything interface, right? :)

Harry Farmer said...

@GH What I find hard to understand about your argument and the whole interface theory is that it seems to assume a dualism in which the perceiving mind has some existence independent of the quantum universe. If you (like most cognitive scientists) think that the mind arises from the activity of the physical universe then it's not clear how the perceiver would be able to form a representation of spacetime when the physical processes that would allow the formation of a perception is itself a time dependent process. So it seems like if you accept something like the idea that agents actually project spacetime onto a spacetimeless reality then you have taken a much more radical position than REC and really have to re conceptualise the whole of cognitive science.

From my, admittedly limited, understanding of the physics that suggests that spacetime is in some sense illusory the point is that there are mathematical ways to recapture the relationships that are captured by spacetime without needing to assume that it is constructed by perceiving agents.

If this is the case then we get back to Andrews argument about the signals agents can pick up from their environment being accurate enough to provide usable information to navigate through the environment regardless of whether that information represents the whole of reality or not.

Greg Hickok said...

Hi Harry. Interesting comment. You say, "If you (like most cognitive scientists) think that the mind arises from the activity of the physical universe then it's not clear how the perceiver would be able to form a representation of spacetime when the physical processes that would allow the formation of a perception is itself a time dependent process." There is no circularity here. You are stuck in your spacetime box. :) In fact, there is no reason to assume that the physical processes that drive percepts are time-dependent. Even more to the point, if there is no spacetime, they can't be. This is one of the reasons I called the theory's implications "mind-blowing" in my intro piece. It completely changes the way we think about perception/cognition.

Greg Hickok said...

Hi, Andrew. You state regarding particle scattering, "This just seems to imply that future theories might not invoke space-time as a player in the calculations and formalisms. This is just about physics reorganising its formalisms to suit itself as it tries to explain things neatly and efficiently." The idea is that a neat and efficient explanation isn't an accident. It's neat and efficient because it captures an underlying truth about the world. Here's the implication: if particle physics determines that spacetime isn't the right physical framework for the world of the very small we have to consider that spacetime doesn't exist on any scale. The current framework assumes spacetime and it is noticed that for medium size objects the world is well-behaved in spacetime whereas at the quantum scale it is weird with respect to spacetime. You can get rid of the quantum scale weirdness (or at least the apparent complexity) by abandoning spacetime. But then you don't want to say that spacetime exists on one scale and not at another. Either it does or it doesn't. Thus, the implication is that spacetime doesn't exists for large objects either, it only appears that way. This might seem like a big problem but the history of science is full of examples where appearances have misled us (flat earth, geocentrism, ...). ITP is a just a framework for trying to understand why we perceive spacetime when it doesn't exist. The answer ITP provides for why we see a different reality is that perception is constructed to represent fitness functions. Why it takes the spacetime form it does remains a deep mystery.

You don't accept the relevance of visible objects showing quantum effects because it still too small and thermal vibrations destroy the effects. Physicists correct me if I'm wrong, but I don't think the vibrations convert the quantum world into a Newtonian one. They just make the quantum effects hard/impossible to measure with our limited instrumentation.

The upshot is that despite your skepticism, Andrew, several lines of evidence from multiple fields (physics, psychology, evolutionary biology, neuroscience) are pointing to the idea that what you see isn't what's there.

(1) Particle physics is toying with the idea that spacetime doesn't exist.
(2) The weird quantum world doesn't hold for just particles, it scales up to at least very small visible objects.
(3) The entire literature of perceptual illusions, fill-in effects, hallucinations, and so on point to the fact that the mind constructs our perception
(4) The entire literature on neuromodulatory effect on perception (i.e., data from injury/disease, stimulation, drug-induced effects) point to the same conclusion as (3).
(5) And now evolutionary game theory shows that perception is not tuned to reality but to fitness.

This does not mean that ITP is right. But it the evidence is a lot stronger than that supporting the mirror neuron theory of action understanding, which many people wholeheartedly believe. ;-)

If physics indeed shows that there is no spacetime does it "ruin it" for all of us, or just embodied theorists? Answer: only the embodied folks. It is an incorrect assumption that no spacetime=no constraint on perception. That is the equivalent of saying that because my files are literally on my computer desktop there is no constraint on the relation between the icons and the literal code for the file. Computational theories are fine in a spacetime free world. What has to change is not the fundamental notion of computation (information processing) but the objects that serve as inputs to those computations.


Ed Baggs said...

I tired to write a response to the above, but ended up writing a semi-autobiographical 700-word novella, which I have posted here. It may contain croutons.

Harry Farmer said...

Hi Greg,

I think you may be underestimating how hard it is to get out of the spacetime bubble. You say that "there is no reason to assume that the physical processes that drive percepts are time-dependent." I think this ignores the question of how you even have processes if there is no time for them to occur in. You argue that computational approaches that argue we construct our perceptual reality can deal fine with the lack of spacetime but without temporal extension how would any computation (which necessarily consists of transformations from one informational state to a new state) occur?

I would argue that since we perceive the world as having temporal and spatial extensions any of the formalisms that argue there is no space-time at the level of fundamental physics will still have to have some mechanism that allows us to recapture the spatial and temporal relations that we perceive in the special science (including non fundamental physics, which I would consider to be the relevant level for REC's claims about the givenness of some physical properties in perception). One way to do that is to go with ITP and argue that somehow perceivers themselves are somehow capable of deriving or constructing the relations of spacetime however this view seems to require that the perceivers themselves are somehow divorced from the rest of the physical universe in a manner that goes against all of our best neuroscientific evidence suggesting that the mind is essentially grounded in the physical properties of the brain.

The point is there are other alternatives that allow temporal and spatial relations to be cashed out as relations with requiring a conscious perceiver to ground them. Indeed maintaining a way to reconstitute space time relations from the fundamental equations that lack them would seem essential for the equations to be testable in the world given that the all phenomena we observe have temporal and spatial extension (just as relativistic mechanics becomes equivalent to Newtonian mechanics in a particular subset of cases). Most physicists seem to think its possible to do this within physiscs without the need to bring in questions of conscious perceivers.

I guess my feeling is that accepting the ITP requires taking on some pretty heavy metaphysical assumptions about the relationship between minds and the physical universe that go against everything we've learnt in the last century or so of neuroscience. While I don't think results from the special sciences can constrain theories at the level of fundamental physics I also feel that it's not clear that theories at the fundamental level have much role in constraining theories at the level of cognitive science. Certainly it's unclear to me that the removal of spacetime from the formalism of fundamental physics would be specifically bad for REC but let computational theories carry on as before. If you think that the lack of spacetime is going to effect what theories are correct at the cognitive science level then I think basically the whole of cognitive science (and almost all other special sciences) are in big trouble.

Andrew Wilson said...

Lawrence Krauss: "There is no known physics theory that is true at every scale—there may never be."

http://nautil.us/issue/29/scaling/the-trouble-with-theories-of-everything