Yeah. But the brain isn’t literally minimising free energy is it? It is minimising information surprisal. — apokrisis

Friston appeals to more or less exactly the same thing Schrodinger is talking about in his original motivations for the free energy principle.

So Friston is talking about the modelling relation just like the biologists. An epistemic cut has to be involved — apokrisis

Well this role is taken on by Markov blankets but is much more general than what is implied by Patee. It isn't special in any way, and markov blankets are nested within markov blankets ubiquitously. "Observers" then just reflect the slowly evolving regularities of their components on smaller scales.

Something unphysical is going on even if it must also have its physical basis. And whether you fine grain or coarse grain the physics ain’t going to make no difference. — apokrisis

"Unphysical" just seems like a misleading word imo when you are just talking about the utility of high level explanations that trace over and present what we observe in a nice, useful way.

And the higher level explanations are coarse-graining over physics if they supervene on it, e.g. the self-regulatory dynamics of homeostasis can plausibly be re-described or re-modelled in terms of the behavior of particles under the laws of physics.

Imo, the fact of the matter is that there is only one reality. Nothing about observation or coarse-graining due to observation changes that, but we can only maximize information about reality by being able to make distinctions at the smallest possible scales.

On the other hand, toss a Bayesian inference engine into the mix - armed with the need to repair and reproduce itself — apokrisis

I believe that in the free energy perspective, the describing of things in terms of the former effectively follows from the latter, tautologically.

In a sense, yes, they would describe the behavior of the interactions. But whereas the 'bottom-up' perspective says that conservations law are 'contingent consequences' of the behavior of interactions, the 'top-down' picture (i.e. interactions are more fundamental) says the reverse. — boundless

Not sure I agree. I've only started to think more deeply into this after a conversation I am having in another thread and it is beginning to dawn on me there are potentially a number of ways to view this kind of thing. Maybe too off point to go into those thoughts though; so, to skip to the point:

The kind of "behavior of interactions" I had in mind would be effectively equivalent to the conservation principles. Just alternative descriptions of the same thing though this I imagine depends on the nature of specific statements, formulations, descriptions. An example special case could be:

https://www.engineering.com/whats-the-similarities-between-these-principles-1-dalemberts-principle-2-law-of-conservation-of-energy/

It seems to me that whatever is conserved is always implied in the described behavior of the interactions. Obviously you might be able to apply these principles as a blanket description of various systems of different sizes and claim holism in virtue of the fact you could be talking about large spatially separated systems. Thinking about it then; for me, I would accept a holistic explanation if say, the forces and displacements in the above link were non-local. But if they are solely local or mediated locally, then I don't see the need for a holistic description. Sure I may not be able to directly explain why these descriptions apply, but if everything interacts only locally then I don't see the need for holistic descriptions. The blanket description for the system would not be distinct from compatible descriptions applied to all the sub-components of a system.

Obviously, you this may seem to not apply for quantum non-locality and so holism seems the case there. My reply again would be that quantum non-locality is not a real example of the exertion of forces over space and time, but a correlation whose origin is local.

I don't see how this isn't some kind of 'non-realism', thought. It seems to imply this rejection of 'unicity' — boundless

Based on the Stanford article, I would say the stochastic interpretation manages to fulfil unicity in the sense of: "a single point represents the exact state of a system at any given time" ehich applies to particles but not the wave-function.

You keep saying you don't see there is a problem. But the sciences of life and mind exist because physics can't even model physics with a computational notion of laws and initial conditions, let alone jump the divide once semiosis enters the chat. — apokrisis

Because what current physics can or can't model has nothing to do with my point, otherwise I wouldn't be defending it with a simulation that can't be built. At the same time, I'm not entirely sure how far biology has actually got with actual successful models of these things.

Have you read Schrodinger's classic What is Life? — apokrisis

Are you aware I advocated the free energy principle and active inference a few posts ago?

But sure. You don't care. The rate independent dynamics is the whole of the story according to your preferred metaphysics. Anything beyond that is just another model at a different level you protest in epistemic plurality as you fall back on that familiar reductionist ontology that all systems are essentially a collection of atoms in a void. — apokrisis

Its all models from physics upwards to anything else. No part of science is any less of just a model than any other.

The point is that any kind of observation or perhaps description about the smallest scales of reality will have more information about reality than all the scales upwards simply by the fact that descriptions on higher scales necessarily coarse-grain over details, while at the same time all the observations on higher scales are effectively redundant in terms of how they would correspond to a mind-independent reality. Doesn't matter what the descriptions are, which is why in previous posts I tried to make an effort to not mix up physics and smallest scales of existence. If you were to take a correspondence view of truth, then obviously the smallest scales would carry the most information about distinctions one could make about the mind-independent reality beyond one's senses. Because if higher scale descriptions are coarse-grained over, they lose information about correspondences.

And how does that pan out given Heisenberg uncertainty? — apokrisis

Heisenberg uncertainty principle is referring to constraints on probability distributions regarding the behavior of statistical systems. Uncertainty relations like this are generically derivable for all stochastic systems including classical ones like Brownian motion. Purely incidentally (I am not attributing the discovery of this fact to him), Karl Friston actually derives it in his free energy principle papers A free energy principle for a particular physics and parcels and particles as a generic property of the non-equilibrium steady state when discounting solenoidal flow, and it is also responsible for non-quantum energy-temperature uncertainty relations in thermodynamics.

I want to go back on the idea I think I have been alluding that this grounding issue has something to do with ontology in a way which is contrasted to or independent of explanation. Because, thinking about it, I don't think it is.

I think what I am talking about is just some very general idea of decomposition when it comes to all our empirical observations about reality. We observe reality at different scales where the windows of space and time are expanded or contracted, details fine-grained or coarse-grained. We can divide up and model the reality constructed from observations with various boundaries plausibly, and perhaps not in mutually exclusive ways. Following that, notions of causality are welded to particular descriptions or models. Insofar as there is a plurality of possible descriptions there is no downward, upward or horizontal causation between frameworks; after all, they are all engaging with the same reality, just under different purviews from observers.

My position becomes weaker then when I think about what it means for one scale to depend on the other in a way which is inherently asymmetrical. In my mind, implying such a relationship suggests that different scales are independently manipulable in a way that one can test a direction of causality like you might in statistical modelling. Clearly we cannot do this because they are just different views on the same reality.

Rather, maybe the importance is in the coarse-graining of our observations and conceptualizations of reality - simply, when we zoom-out we lose information about reality. There is redundancy if you allow different levels of zooming out / in simultaneously. But the more you zoom-out, the more information is lost regarding a mapping to some mind-independent reality.

Maybe the degree to which information is lost is what I mean about fundamentality here.

Describing wouldn’t be explaining. Simulating wouldn’t be capturing the causality in question. — apokrisis

Or how a non-linear system can be reduced to a linear model. — apokrisis

Coarse graining is needed because fine graining can’t deliver. — apokrisis

But my point isn't about explaining. As I have said multiple times, many explanatory frameworks are important. I don't expect all explanations to be reduced to or replaced by fundamental physics because those aren't the only explanations we need or find useful; at the same time, their incompleteness and issues of complexity prevent such things pragmatically.

The point is though that such simulations as alluded in the first quote above should be possible in principle if we had the computational power, and able to reproduce all possible events of reality above the fidelity of its description. If all biological processes are composed of things like particles moving in space then this should be plausible. I don't see why not. We may need better explanations, but that doesn't preclude the fact that in principle the lower resolution descriptions are undergirded by the higher resolution description.

Yours becomes a really odd position when physics can’t even settle on an agreement of how a classical realm emerges from a quantum one. — apokrisis

Under my preferred interpretation of quantum mechanics, the emergence of the classical from quantum is extremely straightforward and therr are no fundamental issues here

The systems view of causality is that nature is all about global constraints shaping up the local degrees of freedom — apokrisis

Like which constraints?

Atoms emerge due to the constraints of top-down topological order being imposed on quantum possibility. — apokrisis

source?

You can only assure me you could reconstruct the world as some kind of simulation of its shaped material parts. Some set of atoms arranged in space and moving “because” of Newtonian laws. — apokrisis

But in principle, why wouldn't that be enough to demonstrate my point? The only thing I am saying is that everything else you can possibly explain or describe in principle can be instantiated in that description at the lowest level, until you can find an even lower level.

The fact that it doesn't provide comprehensible explanations of your concepts in special sciences or everyday life is irrelevant if reality is indifferent to your ability to comprehend it.

As a reductionist, you can’t in fact reduce at all. You can only enumerate parts. You can’t speak to the causality of the whole. The only compaction of information you can offer is a mechanics of atoms. The offer to simulate is given in lieu of what is meant by a causal account. — apokrisis

I am asking you to ground your account in its causal principles. — apokrisis

You mean reality resolves into its fundamental atomistic detail at the level of the Planckscale? Of the quantum foam? Of quantum gravity?

Yeah. How is that project going exactly? — apokrisis

Again, you confuse my point. My point has nothing to do with explanations that we might find attractive or necessary as observers.

Its about the idea that in principle all of the possible information about reality is only attainable if it is maximally resolved, if it isn't coarse-grained, if details are not ignored.

If a description misses out details that one knows to exist by utilizing higher resolutions of observation, how can that description be considered more fundamental to reality?

What you’re calling the lower level physical description, the irreducible ground floor for the understanding of all higher order descriptions (chemical, biological, psychological and cultural) has evolved over the history of philosophical and scientific inquiry. And it has evolved in such a way that all of the higher order resources of cultural knowledge arebrought to bear on redetermining in each era of inquiry the nature of the lowest level. Another way of putting it is that the very highest level of cultural understanding is inextricably intertwined with our models of the very lowest level. This may not seem like an objectionable claim in itself, but what if I were to suggest that it often happens in the historical course of scientific inquiry that insights gained from scientific and philosophical investigations of phenomena seemingly far removed from the subject matter of physics, that supposed ground floor level of study, can point the way toward paradigm shifts in the models describing the nature of that lowest level? — Joshs

Yes, this is a fair point. I would just question if it ever seems reasonable to say that events observed at the larger scale of existence do [not] depend on and are decomposable in terms of the smaller scales or higher resolutions.

This was true in the early days of the social and psychological sciences — Joshs

It still is true because no matter how complicated we view biology or social sciences, a description at the level of physics would be orders of magnitude more complex to describe the same phenomena and it may not even make a lot of sense.

1) You are not actually treating the higher order psychological account as consistent with the lower order one, but you are just assuming without examining the details that the higher must be reducible to the lower since of course the physics has been rigorously validated empirically. — Joshs

No, this is nothing specifically to do with physics, it is about whether it logically makes sense that observations on larger scales do not depend on lower scales. If the difference between higher and lower scales amounts to expanding the spatial and temporal scales and coarse-graining on a single reality, then I am not sure an alternative in principle makes sense for any kind of description. It would result in a radically different view of reality which would seem strange to me.

2) You interpret the higher order as subsumed by the same theoretical logic as the lower one, and so miss the radical departure of the former from the latter’s grounding assumptions. — Joshs

I have made the distinction between issues that are concerned with the actual structure and reducibility / derivability of different explanatory frameworks (what you are saying is part of this issue) vs. more general issues of ontology. I have not been making a point about the former.

Edit: first reply changed

But you didn't get the memo about categories. I'm afraid the news is that there are many different kinds of existence. — Ludwig V

Not sure what you're implying or what you are referring to in what I said. Categories are things we apply enactively like any other knowledge. They reflect things we do, rather than something inherent about fundamental ontologies.

Including physics. — Ludwig V

Yup, already said this.

Oh, to be sure they are. My brain is heavily involved. But the point is that my brain is not the whole story. Same applies to plus tasks. — Ludwig V

Well, all I can do is point you back at what I already said about why I think this kind of perspective is only rather superficial significance and doesn't really contribute much. At the same time, I have no trouble saying that a calculator does plus tasks.

You seriously mean that you live in your head? — Ludwig V

If you are not a dualist, and there is something like an isomorphism between experiences and how the brain functions, and there is only a single realm of existence, then it is clear that consciousness is in the vicinity of the brain. How could it be any other way? Yes, we all conceptualize ourselves in terms of an extended person in the physical world (or perhaps any other way you choose) but all of these concepts emerge, are constructed directly, are vicariously engaged with in experience. Experience is inside your head. Thoughts, feelings, perceptions are tied to neural activity in your head. However you may conceptualize the world, it is via experience, and experience must be situated in the vicinity of your brain if you take the isomorphism of consciousness and brain descriptions seriously, if you think there is only one realm of existence.

The idea that the self or the person is another creature like us inside our heads was the founding mistake of dualism. — Ludwig V

Well this is nothing like what I have said. What are you then if you are not a dualist?

And yet you defend your brain tirelessly. So it must be important to you even if it is not big. — Ludwig V

Not sure this makes sense. You must have misunderstood something but it doesn't seem like a significant point.

So the concept of ontological grounding is not perspective-dependent? H'm. — Ludwig V

Well no; but, by observing the world, we can make the conceptual distinction between a world that in-principle exists in a mind-independent way and the frameworks we use to engage with it. I think that the fact that we cannot talk about anything in a perspective-independent way shouldn't necessary preclude us from talking about the concept of a perspective-independent world. There may not be much at all that can be said; nonetheless, I think such a concept is important in how we see the world.

We can then note the difference between issues of derivation and reduction between different explanatory frameworks as opposed to the empirical observation of how the world seems to decompose when we zoom in at different scales. Most people's objections seem to be really preoccupied with the former. To me, objections to the latter seem to require a radically different conception of reality which frankly I know wouldn't seem plausible to me.

Looking back on my description, I think introducing the notion of "behaviors at different scales" does actually introduce perspective in a stronger way than I had desired with regard to ontology - because scale is a perspectival concept. But I still think the idea of how the world seems to decompose as we zoom-in with out observations is reasonably independent of any specific kind of field of knowledge you can invoke.

The calculator neither knows not cares whether it is correct. It cannot evaluate its own answer, in the sense of trying to correct wrong answers. — Ludwig V

True, but this just identifies a difference between people and calculators. I don't see it as necessarily meaning much for whether we should say a plus task is being performed. After all, the calculators knows almost as little about how it performs a plus task as we do.

Try stopping your heart or draining your blood. Same result. — Ludwig V

Yes, but it only does this vicariously through your brain. There is nothing in the universe that could get the same result without being mediated by your brain.

Physics, from my perspective, is not unique in any respect. — Ludwig V

But what about the smallest scales of existence. You think events observed at large scales are not grounded on what can be observed at a higher resolution?

How can I refute that in the face of your refusal to engage with the question of how physics - coarse or fine - accounts for the functional structure of a neuron? — apokrisis

I replied here:

Alot of the details are probably out there in the field of biology in terms of things like gene translation and cellular development. Is any of this not mediated through fundamental physics? Seems implausible. Does any of these descriptions require the notion of "biological information"? I doubt it. — Apustimelogist

What I was implying is that all of the events that led to the development of neuronal structure- whether on an evolutionary or developmental scale - can be in principle described purely in terms of particles and how they move in space and time. In principle, such a thing could be simulated using a complete model of fundamental physics - it would just obviously be orders of magnitude too complicated to ever be possible to do.

From this, it would follow that higher-order descriptions are both in principle: redundant, in the sense that they are describing behavior that could be described purely in terms of smaller scales; and also incomplete, in the sense that any higher-level description would have to be missing out on details that actually occur in reality on the smaller scale but are not included in the higher-order description.

Obviously that doesn't mean we don't need the higher level description - but clearly, higher level descriptions will be grounded on the details of smaller scales. How could it not be?

Why use the higher-level description then? Obviously it is required because it is less complex and doesn't require precise resolutions, maybe it is also closer to our everyday levels of descriptions. The reasons for using the higher-level description or a lower-level description are clearly about epistemic, explanatory needs, not ontological ones - this makes the following quote from your post clearly ass-backward:

I’m talking about ontology rather than epistemology. Life and mind as a further source of causality in the cosmos. The stakes are accordingly higher. — apokrisis

When you are asking about neuronal structure, you are clearly asking for an explanation that you can understand. Because obviously, in principle one could describe the entire process of cell development and the entire history of the world in which evolution occurs in terms of particles moving in space - it would just not be tractably comprehensible by yourself.

So I think you are talking about explanation, not ontology. Redundancy is acceptable, even useful and required, when it comes to explanation. I disagree that it is when it comes to fundamental ontology.
I don't want to crystallize descriptions from physics too much as some kind of in principle absolute perspective independent view of reality but clearly any treatment of reality that misses out on the smallest scales misses out on details that are fundamental to reality in the sense of having observable consequences which undergirds observations from a less resolved perspective of larger scales.

Smaller scale descriptions don't give us all our required explanations, but clearly a model of reality could only be in principle complete at the highest resolution, other resolutions being redundant. Our observations about reality are grounded on and instantiated in the most zoomed-in scale, fully resolved, fully decomposed - higher scale observations reflect coarse-grainings of that over space and time.

Sure. You've certainly said how it seems for you. But as a biologist and neuroscientist, I see this as question-begging reductionism. — apokrisis

Why is it reductionist if I explicitly talk about the importance of higher level explanatory frameworks?

When are you going to refute the idea that all coarse-grainings of behaviors over larger scales are grounded on higher resolution details at smaller scales of space and time?

This just shows that you haven't read or understood the stuff. — apokrisis

I feel like I read enough of what you sent to get a gist.

All the parts of the puzzle that come together to form a general theory of life and mind. — apokrisis

Show us the state of the art papers of this theory then rather than one from 2001.

The question you deny is even a question is a question I've been academically engaged with for a long time. — apokrisis

Well, unfortunately that doesn't guarantee anything. I have no doubt Bernardo Kastrup has been academically engaging in utter drivel for some time. I even would say Qbism and many worlds are as bad.

Seems you are trying very hard to do exactly what biologists complain about. Failing to understand the epistemic cut. — apokrisis

I have said a couple times in the thread I see the importance of different explanatory frameworks on different levels but just seems to me all complex behavior are grounded on and emerge from the smaller scales as described by more fundamental, simpler physical laws or descriptions. As just said, I don't think that precludes higher level frameworks but they just aren't as fundamental.

I just don't find biosemiotics compelling as some kind of foundation for biology. I don't have an issue with studying something like that, but I don't see it as fundamentally necessary to describe how things work in biology. This is partly because I am already very biased against attempts to reify meaning and against views that seem inherently strongly representational. The idea of symbols or signs in biology then seem to me something like an additional level of idealization and approximation that is another way of telling stories about biology, perhaps more intuitively - similar to teleology. But it doesn't seem fundamental to me compared to notions like blind selectionism which does not necessarily require things to be packaged up in terms of neat symbols and meanings.

I personally find ideas like active inference and the free energy principle have more clarity, eloquence and mathematical grounding than the Howard Patee stuff, in addition to being prima facie simpler to couple with my enactive inclinations. The epistemic cut idea also seems to draw from ideas in quantum mechanics which I just do not believe to be the case

So are you saying that mathematical objects don't really exist? What is your criterion for existence? Is it, by any chance, being physical? I don't think Quine's slogan "to be is to be the value of a variable" is perfect. But it's not bad as a slogan. — Ludwig V

I don't have a criterion for existence but my assumptions from what science and philosophy seems to say to me is that: there is a single realm of existence; everything is grounded on behavior at the smaller scales of that existence; there is no alternative platonic realm where mathematical objects exist. All I know is that my ability to use math comes from my brain, - and my brain and all my biology and behaviors are grounded in the behaviors of the smaller scales of existence.

I think we construct mathematical objects and impose them on the world enactively, which is not really any different from any other concepts or knowledge we use. I don't really have a problem saying mathematical objects exist, but I would not see any mystery to their existence beyond how our brain functioning has allowed us to use math.

But notice that in the latter case, the hormones do not map one to one with my emotions. — Ludwig V

Because emotions are much more than just hormones.

Nonsense. They know perfectly well how to count. Maybe they can't explain how they count very well, but that's a different know-how. So we say they act blindly. But the point is that they act correctly. — Ludwig V

Well then the only criteria I see for the plus task is that it is performed correctly in the way regular people deem it correct. A calculator can plus correctly imo.

I never said it was. All I'm saying is that what I do is not what my brain does - except by synecdoche. — Ludwig V

I just don't really understand what practical consequence saying this has when, even if I don't identify what I am doing as what my brain does, clearly everything I am perceiving and experiencing and all my acts are direct consequences of brain behavior interacting with the environment. I don't see any interesting consequences for what has been said so far by maintaining this distinction. You may not want to say a brain is doing what you are doing but lets see what happens when we stop the brain doing what its doing and knockout that occipital lobe - how that affects what you are doing.

From my perspective anyway, everything I am experiencing is literally what it is like to be some kind of higher level, higher scale functional structure in the vicinity of that part of existence which we might label my brain. So the distinction does not seem so big from my perspective. Even if I identify as a whole person in some sense that is something different from my brain, the whole person embedded in its external environemnt is still as much an inferred construct that I effectively would be experiencing from a perspective within the insulation of the brains sensory boundaries - given what I said in the first sentence of this paragraph. The self arguably might be seen as an inference like any other. Different brains may then have effectively different models or perceptions of persons or even self. And you can get hints into its constructed nature through how people perceive things like this:

https://youtu.be/9Tt7aqHFUCU?si=yHjzV0Mvr_YQJLQQ

Some people just struggle to understand these clips much more than others - they struggle to make the same inferences others do, suggesting how such concept are imposed and not directly apparent a-prior-i (perhaps in something like the Bayesian sense) from the moving images. But I digress!

A brain may not do what a person does in some sense but making the identification is where I am drawn and I personally find concepts flexible enough to allow that.

Quite so. But it doesn't follow that we can in principle describe my behaviour in terms of the same levels. You can describe my running in physical terms. But physics has no equivalent to an intention or to the rules of athletics, so you can't describe my running and winning a race in terms that physics would recognize. — Ludwig V

Yes, I get that and I have never excluded those things, after all that is the level at which we engage with the world in everyday life. But I think a distinction can be made between: the use of different explanatory frameworks and ways we engage with the world that are perspective-dependent for various reasons; and then the concept of ontological grounding in principle - that behaviors described at one scale will be grounded in those on smaller scales, even if I require different explanatory frameworks to make sense of the world in any pragmatic way.

So how do you derive the structure of a neuron from the laws of physics? — apokrisis

Sure, the laws don’t forbid the structure. But in what sense do they cause the structure to be as it physically is? — apokrisis

Well do so then. Tell me how the physical structure of a neuron is the product of fundamental physics. Tell me how neurons appear in the world in a way that does not involve the hand of biological information. — apokrisis

Alot of the details are probably out there in the field of biology in terms of things like gene translation and cellular development. Is any of this not mediated through fundamental physics? Seems implausible. Does any of these descriptions require the notion of "biological information"? I doubt it. At the same time, you're asking about deriving neuronal structure from physics but I don't really see where you would derive neuronal structure from "information processing which entropically entrains the physical world" either any more than you can from physical laws.

and interpreting them as a 'faithful portrait' of reality is wrong. — boundless

I wouldn't say that its not like the portait cannot possibly in principle be faithful (where it does not have wrong predictions); but that it cannot tell us anything about reality intrinsically beyond tools that are used by us to essentially anticipate what comes next or came before or what could happen in some scenario.

I think that 'non-representationalist' interpretations of QM — boundless

I feel like my point should be interpretation-independent.

But IMO the 'reductionistic' picture takes conservation laws as accidental properties of interactions — boundless

I disagree. They would still be an inherent part of the descriptions of those interactions, it just doesn't have to be anything more than local to that picture.

In the thermal interpretation, as I understand it, the wave-function is a pure fiction — boundless

Not entirely sure this is the case. Hard to tell. Imo, the 'holism' can be explained away given that the wave-function isn't real and entanglement depends on local entangling interactions ans locally incompatible observables.

But you said all the complex behaviours of neurons emerge from lower level physics which is quite wrong. They emerge from the information processing which entropically entrains the physical world in a way that brains and nervous systems can be a thing. — apokrisis

I am not sure this makes sense. A neuron is characterized as a physical object made up of particles that behave according to the laws of physics. All neuronal behaviors follow from this and we put information processing on top of it. Not the other way round.

I don’t favour computer analogies but what do you think causes the state of a logic gate to flip. Is it the information being processed or the fluctuating voltage of the circuits? — apokrisis

It is obviously the latter. The most advanced models of functioning neurons are characterized in exactly the same way. It explains how neurons seem to process information.

The physics of neurons is shaped by the top-down needs of Bayesian modelling. Bayesian modelling isn’t a bottom-up emergent product of fluctuating chemical potentials. — apokrisis

You can always in principle describe whatever a brain is doing in terms of more fundamental physics. You may choose a higher level of explanation for what a brain does but that will still more or less have a grounding in and depend on the fundamental physics of brain components. The only way to dispute that is to dispute what the brain is composed of which no one would sanely do. Given that, we can always in principle describe the brain behavior in terms of those more fundamental levels. Exactly the same goes for how it got there, whether developmentally or evolutionarily - even if such things are more desirably explained in terms of higher level explanations regarding things like selectionism, canalization, gene-environment interactions or whatever.

You rephrased the question. Surely, applying math to the smallest scales of existence implies that physics and math exist independently. — Ludwig V

But this is just the application of a tool which is nothing more than what is enacted in behavior or cognitive states. In fact, applying math to the smallest scales of existence is basically physics. (though physics isn't exclusively or fundamentally about describing the smallest scales of existence).

It is then important to make the conceptual distinction between the smallest scales if existence - which grounds everything - and our use of math or physics as a tool. The caveat is that the only way we can intelligibly describe or grasp the smallest scales of existence effectively is by applying the tools. But the tools have no inherent ontological existence beyond our enacting of them.

Sure, you could argue that the objects in math are not reducible to objects in physics... they are more general and perhaps abstract than physics... but we can make any sets of arbitrary tools we want that are not inherently related or reducible in a hard way to other tools or descriptions. They are, after all, just constructs. There is no hard reductions without bridges and assumptions anywhere in knowledge. At the same time, there can be multiplicitous applications of tools - e.g. fields of math can be applied to many different areas / different areas can have strong parallels. Nonetheless, if you want to talk about ontology and existence, then it seems that all behaviors of objects must be grounded in behaviors at smaller scales.

The fact that maths does not reduce to physics - in the way that arguably certain physical theories can be reduced to each other - has no ontological significance. What is significant is that mathematical behavior, mathematical intellectual ability, information processing is grounded ontologically in the smallest scales.

Would a Popperian ontic triadism be better? I doubt it. I suppose it is time to come out. I do have a view of this. I see your claim as the classic philosophical mistake of thinking that a grammatical device, which is purely rhetorical, has some philosophical significance. "Brains do plus tasks" is synecdoche for "People do plus tasks". You may not know what synecdoche is (I had to look it up to be sure). — Ludwig V

But is what a person does independent of what a brain does? No.

Is a brain how a person does a plus task? Yes.

Are persons and brains exemplifying constructs we use whose relations come from / are about different epistemic perspectives rather than inherently about ontology? Yes.

Then - while your issues about whether brains or people do plus tasks may have semantic validity in terms of how we characterize things relying on concepts or definitions or aquaintances from different perspectives - what are the consequences it has in terms of what actually happens in existence? Absolutely nothing, but for the caveat that we cannot view existence in a perspective-independent way. Nonetheless, empirical evidence about isomorphism of experience and brains is convincing. Philosophical arguments too like Chalmers' dancing / fading qualia, etc, are also convincing. Neither do I feel the need to qualify that humans have bodies and exist in both physical and cultural environments or niches every single time I talk about them. We can acknowledge the conceptual divides between different perspectives but I think we also must acknowledge that if different perspectives map up to each other substantially, like the brain and mind, then its simply seems impossible to me to not talk about those mappings in terms of some kind of underlying commonality. To say that a person can do a plus task but a brain cannot is an epistemic clarification that, if taken too absolutely, completely obfuscates a valid ontological clarification with tangible consequences...

Unless you are a hardcore dualist.

At the same time, I question whether your distinction between brains and humans doing plus tasks is even that interesting or valid. For the ways I would talk about a plus task, I see no issue with saying something like 'a calculator performs plus tasks'. I therefore see no issue with the notion that a brain can perform a plus task either. You could get an population of neurons, put it on a petri dish, wire it up to some computer or other apparatus and teach it how to do plus tasks. I have no inherent problem with saying it was performing a 'plus task' - we set up a 'plus task' with criteria on success, and the petri dish sarisfies them.

You could say that well this petri dish doesn't satisfy some kind of unique human criteria of doing a plus task experientially... but what is that exactly? I really have no idea because I don't think anyone knows exactly how they count or do plus tasks. They just do. Mental arithmetic is almost like a brute ability (from our own perspectives). The answers just come. I don't see a good criteria that makes human plus tasks special in some way, and from the third-person perspective, watching a human doing a plus task on a computer is not necessarily qualitatively different from the petri-dish.... they both share the fundamental resemblance of neurons hooked up to a computer. One might be immensely more complicated than the other but they are doing the same task. One we might want to characterize in terms of more complex experiences but I don't think experience is necessarily important for characterizing plus tasks. At the same time, I am inclined to say that questions about experience or if other things (e.g. neurons in a petri dish) have experiences may be both intractably meaningless and meaninglessly intractable.

Yes, I understand that. So the language that you use to describe the brain process excludes the possibility of describing a plus task. So in what sense can it explain or cause a plus task? — Ludwig V

Why would I use the language of brain processes to describe a plus task? On the other hand, I can get a brain to perform a plus task.

I don't think there is anything to 'meaning' beyond enaction. So there is nothing special. 'Meaning' is entirely enacted... in the mechanistic flows of experience.

However, the synchronization that is involved here (mirroring) is not obviously the same as the one that Apustimelogist is concerned with. But I don't know what the active inference/free energy principle is, so I could be wrong. — Ludwig V

It is the same kind of synchronization.

This neatly inverts things. An informational mechanics is precisely what biology and neurology impose on the physical world. Not the other way around. — apokrisis

Well this just ignores the context about which of two things is more fundamental. In any case, under my view of enactivism, a "semiotic modelling relation" cannot be fundamental because it is the kind of representationalism that my enactive views would prefer to actually explain away.

This ignores the fact that organisms are organised by codes and so exist in a semiotic modelling relation with the world. — apokrisis

More like what a modeller imposes on the world, including their models of modelling.

But the world itself is not a machinery of linear cause and effect — apokrisis

It is if you can unmix all the interactions that would make causality non-linear.. something that generally does not occur but arguably we infer or observe in experiments... but the point is that we generally do not study the fundamental nature of the world in terms of its full complicated mixture of effects... we separate out simpler, fundamental principles latent in theory complicated behavior. When causality seems non-linear, it is because of how different parts of reality interact. "Non-linear cause and effect" therefore emerges and isn't fundamental.

For all practical purposes, we may regard a wave function as collapsed as some probe with a switch mounted on its end has been heard to flip state. Holism can be considered localised. Another bit of thermodynamic history has now definitely been added to universe's equation of state. — apokrisis

Wave functions are not real objects nor do they physically collapse.

If neurology relied on ions crossing membranes as its deep explanation, then it would be getting us nowhere — apokrisis

And that is an epistemic issue not an ontological one and I have more or less explicitly alluded to this in recent posts. Again, just because it may not be your preferred level of explanation, does not preclude it from being more fundamental or at least perform a role of grounding the other more preferred explanation so that preferred explanation itself would in principle be explained by and depend on this more small scale perspective.

Yes, these are very good examples I think.


Alot of that seems to be detailing experiments so I don't see how it can be contradictory to the view you offered in this post.

“relations are not secondarily derived from independently existing “relata,” but rather the mutual ontological dependence of “relata”—the relation—is the ontological prim­itive. The notion of intra-actions constitutes a reworking of the traditional notion of causality.” — Joshs

Just sounds like unnecessary obfuscation to me.

I’m arguing that the full implications of the non-linearity of complex systems in living beings makes it impossible to derive them from physical models as they are currently understood. — Joshs

I feel like you can in principle, it would just be extremely complicated - and that is probably an understatement.

about non-linear systems. — wonderer1

Like Hodkin-Huxley neurons!

Actually, the most complex scale of existence grounds our use of math. Mathematics is a conceptual invention. — Joshs

Mathematics is something we enact, and in that sense it is grounded in the smallest scales of existence because our behavior and brain obviously is.

The billiard ball model of causality — Joshs

The fact that you may need higher levels of explanation to make a dynamic system intelligible doesn't negate the fact that it may be at the most fundamental level just a consequence of simple billiard ball causality. It could not be any other way since such explanations you talk about are by their very nature not fundamental. Brain message passing entirely functions by molecular interactions which is comparable to billiard ball causality. It doesn't mean that this behavior doesn't result in extremely complex or even chaotic behavior.

I just don't feel the need to qualify at every moment that the brain has extremely complex, non-linear, recurrent dynamics. This is how I inherently think about the brain all the time. That fact is simply not relevant to the point I am making in this context; nor does it even make sense to me to identify this kind of thing as some kind of different, special form of causality in any meaningful, non-trivial way. Clearly, whatever sense of 'mechanistic' you are thinking about is just much more narrow than mine.

I suggest that such non-linear reciprocal affecting between cause and effect is more fundamental than the mechanistic billiard ball or domino form of description we might try to foist onto neural processes as their ‘real’ basis. — Joshs

This doesn't make any sense since all of the complex behaviors neurons do are emergent from very simple ones at smaller scales - described by more fundamental laws of physics - such as ions crossing a membrane barrier.

It by definition cannot be more fundamental even if you would never want to describe what brains do purely at the level of ions crossing membrane barriers and molecules interacting with receptors.

The problem is that it seems that there are no properties present in the insentient matter (that we are aware of) that might be able to explain in an intelligible way the arising of consciousness — boundless

Yes, I don't think so either. My desire to just get rid of an inherent conflict between our direct aquaintance of experience and our descriptions of ontologies in physics. I think there is much less conflict by getting rid of this notion of a bottom to the universe with a fixed set of objects just arranged in different ways. Already, the conflict is weakened somewhat imo if it is emphasized the way that physics can be seen as models or tools that describe or trace functional aspects of the universe rather than intrinsic things.

Ultimately, I do not think it is actually possible to give an informative, coherent characterization of fundamental ontology or intrinsic nature of reality. I would even go as far as saying that ontology and 'being' are empty concepts in regards to characterizing fundamental ontologies.

Why these interactions behave in the precise way that ensures the conservation laws is left unexplained. — boundless

It's unexplained either way imo. I just am not compelled to commit to the idea that its brute nature requires appeal to anything beyond local dynamics. I don't need to appeal to the whole universe (the only isolated system that exists) to observe examples of conserved quantities from interactions, as implied by conservation laws, in local systems. And I imagine you could say the same thing if the local system was further decomposable so one could focus on what is happening at a single component of it.

Well, this is an interpretation-dependent question. — boundless

Yes, it especialliy depends if you interpret the wave-function as a physical object I think.

Does physics ground mathematics? — Ludwig V

Do the smallest scales of existence ground our use of math?

Absolutely.

H'm what does "in some sense" mean? — Ludwig V

The kind of synchronication between internal and external states as described by active inference / free energy principle.

But I don't navigate that environment (under normal conditions); the environments I do navigate are all "external" to the body. — Ludwig V

Imo, the body is on equal footing to the rest of the environment in the sense that the brain synchronizes with the body by sensory inputs in the same way it would to any other sensory inputs from the external environment.

Obviously, I am in some sense equating the "I" and some states of the brain.

I know what a "plus" task is. Hence, I know that brains/neurons don't do the plus tasks that I do. — Ludwig V

At this point, I can only assume you are taking on some dualistic notion of the world that gives a profound ontic separation between you and your brain that I just don't agree with and find isn't evidenced by either science of philosophical arguments. I am not going to be able to make you understand what I am saying without you giving up this kind of dualism.

Are you by any chance saying that brains/neurons do plus tasks without knowing what they mean? — Ludwig V

A brain performs a plus task by sequences of membrane depolarization, not by looking up and applying meanings. That's how anyone acts blindly. That's partly why no one can give a non-circular definition of what 'plus' is that eliminates underdetermination / indeterminacy. We don't know anymore than our brains because our brains are exactly how we perform these tasks.

Can you explain in what sense you do mean "mechanistic"? — Ludwig V

I just mean mechanistic in the sense of one event causing the next event and the next event in a way that is divide of any kind of extra meaning. Like knocking down dominos where one falls causes the next and the next and the next in a mindless ways. But I am not assuming any limits on complexity or non-linearity or recurrence or anything like that.

Yeah but how would you answer the point that you don't need numbers to represent something?

But the information isn't numbers, the symbols are not numbers.

Did you communicate this message with numbers?

I didn't mean mechanistic in such a narrow sense as you do here.

I don't understand anything beyond that — Ludwig V

Its just the idea that mechanistic component cannot be inherently interpreted in terms of a semantic component. If you look at a brain performing a plus task, our description of 'plus' is not interpretable in terms of our description of how neurons arr actually performing the task.

I don't need to explain how mindless neurons do plus tasks, because they don't do plus tasks. — Ludwig V

Neurons are precisely what is performing a plus tasks for you. The biology and dynamics of neurons account for everything about your ability to do a plus task.

Saying that I act blindly when I do a plus task is saying that there is no need of and no room for an explanation how I do them. — Ludwig V

No, its saying that you don't know how you do them. But neurons explain how you do it and in principle they would even explain how you don't know how you do it.

I think you believe a philosophical thesis that physics is the ultimate grounding of everything — Ludwig V

In what sense do you mean that physics does not ground everything? Physics describes the smallest scales of existence which grounds everything else and upon which all higher scale behavior depends and emerges from.

and that you therefore infer that my neurons must be doing something relevant. My problems are first that I don't accept the philosophical thesis that there is/must be an ultimate grounding of everything and second that you don't seem able to explain what the relevance of my neurons is. — Ludwig V

I don't really understand what you could possibly mean by saying here other than rejecting mainstream science. In what sense are neurons not relevant? When you are performing a plus task it is due to the behavior of neurons.

I hope there's a typo there and you meant that know-that is a special case of know-how. — Ludwig V

Yes, thats what I meant.

Yes, when I was thinking about fundamentalness in a different way in terms of how physics doesn't seem to paint a picture where there is a constant, fundamental set of objects at the bottom of the universe which just change arrangement over time. And then thinking about whether this helps some aspects of the hard problem.

This perspective is clearly 'holistic': a property of the 'whole system' (the conservation law of the total momentum) 'dictates' how the properties of the subsystems (the particles) behave. — boundless

Not sure I agree. Its a property of the interaction so I wouldn't say it is necessarily holistic, though I would say the two different descriptions were equivalent.

Not only that: in the case of entangled quantum systems, there is a clear indication that what is 'more fundamental' is, in fact, the whole system of entangled objects, and this is not reducible to the subsystems. — boundless

I think this is interpretation-dependent imo but I know many people do believe something like this.

Just for curiosity, has been treated the 'Wigner friend' scenario in stochastic realistic models?

It hasn't but I talk about it in this post:
— boundless

Wigner — Apustimelogist

Essentially, there are always definite, objective outcomes but the statistics of the world are oberserver-dependent. This contextuality isn't specifically about observation but the statistical constraints when stochastic systems are coupled (e.g. a measurement device and system bring measured or any other kind of system-environment interaction perhaps).

do you think that 'physical realism' is undermined by the fact that the 'fundamental building blocks' of 'physical reality' are not spatially separable? — boundless

No, if there is a reasonable explanation. Obviously explanations may seem reasonable or unreasonable to different people.

I do believe that we have to reject the idea of 'fundamental building blocks' altogether BTW. — boundless

In what sense? I may agree in some sense and have thought about that, motivated by the hsrd problem of consciousness. But may not have been in the same sense you mean.

Why is your view restricting this to numbers? There is no reason you need to represent things with numbers.

I mean, there is no alternative. There are extreme nonlocal correlations in quantum mechanics; you cannot get rid of the strangeness.

Sorry, very late reply. Will be long post.

No, CFD (counterfactual definiteness) simply implies that physical quantities have definite values at all times. de Broglie Bohm's interpertation is a perfect example of an interpetation which has CFD. MWI violates CFD because it assigns multiple values to hypothetical measurements, so it can be realistic and 'local' (although in a weird sense... after all, what is more nonlocal than a 'universal wavefunction' split at each measurement?). — boundless

Alright, fair enough!

So, in your view, if the particle configuration is definite at all times, how can you describe non-local correlations without a non-local dynamics/kinematics which involves some notion of simultaneity? — boundless

I had trouble formulating a reply to this. I don't have enough insight inside these theories to make strong statements I would like; nor are these theories and surrounding literature really complete in a desired way. I will just offer different perspectives which are incomplete whether from lack of literature or my own ignorance.

Perspective 1:

The original stochastic mechanics by Nelson has an explicit non-locality issue where marginal probabilities of particles depend on velocity potentials related to other spatially separated particles. I believe this is thought to be similar to the Bohmian issue.

In the non-locality section of his book, quantum fluctuations, Nelson explicitly shows that in principle a non-Markovian as opposed to Markovian diffusions resolve this issue (pdf for book can be found on webpage below) :

https://web.math.princeton.edu/~nelson/papers.html

And there is at least one variation of stochastic mechanics where non-Markovianity is explicitly used and this eliminates that non-locality issue that was identified (clicking the link below is a direct download to the pdf of the paper: Stochastic mechanics of reciprocal diffusions by Krener and Levy)

https://math.ucdavis.edu/~krener/51-75/68.JMP96.pdf

Again, I don't have access to any real deeper insights into these theories here and their further implications within the theories. All I know is that Nelson saw this non-local problem and it seems to be solvable in principle, especially via dropping non-Markovianity.

I guess I might as well note that Nelsonian stochastic mechanic has two other major issues - incorrect multi-time correlations and something called the Wallstrom problem but I think both issues can be regarded as more or less resolved or resolvable based on recent formulations and papers.

Perspective 2:

This is not stochastic mechanics but still a stochastic interpretation based on showing mathematically a very general correspondence between unitary quantum systems and indivisible stochastic ones:

https://arxiv.org/abs/2302.10778
https://arxiv.org/abs/2309.03085

In the following paper:

https://arxiv.org/abs/2402.16935

They argue their theory is causally local - analyzing with Bayesian causal models they find that measurements of observers do not causally influence each other (Sections VII-VIII near end). Entangled stochastic systems do causally influence each other but this is because their non-factorizable transition matrices have encoded their initial local interaction. It is just the nature of these systems they will fail to factorize until a 'division event' because statistical information is encoded cumulatively in the transition matrix (in the words of the author). I have no idea how this perspective relates to the first because they are just different stochastic formulations of quantum mechanics. Perspective 2 is actually explicitly non-Markovian; but again, there is no explicit connection that can I can see that would relate it to the issues in the first perspective or vice versa.

Perspective 3:

This not specific to the stochastic interpretation but an attempt to explain away non-local correlations in a way I find appealing. Has roots in various authors (e.g. Pitowsky will be mentioned momentarily) but perhaps best exemplified by the 1982 papers by philosopher Arthur Fine:

https://scholar.google.co.uk/scholar?hl=en&as_sdt=0%2C5&q=arthur+fine+1982&btnG=

It establishes equivalence of Bell violations to the absence of a unique joint probability distribution. Recent generalization by Abramsky:

https://scholar.google.co.uk/scholar?cluster=12086196826892314859&hl=en&as_sdt=0,5

In following paper Abramsky talks about contributions of Pitowsky:

https://scholar.google.co.uk/scholar?cluster=17313080888273101986&hl=en&as_sdt=0,5&as_vis=1

Who noticed that Bell inequalities are actually a special case of Boole inequalities which have roots in the work of George Boole in the 1800s:

Boole’s problem is simple: we are given rational numbers which indicate the relative frequencies of certain events. If no logical relations obtain among the events, then the only constraints imposed on these numbers are that they each be non-negative and less than one. If however, the events are logically interconnected, there are further equalities or inequalities that obtain among the numbers. The problem thus is to determine the numerical relations among frequencies, in terms of equalities and inequalities, which are induced by a set of logical relations among the events. The equalities and inequalities are called “conditions of possible experience”. — Pitowsky

For certain families of events the theory stipulates that they are commeasurable. This means that, in every state, the relative frequencies of all these events can be measured on one single sample. For such families of events, the rules of classical probability — Boole’s conditions in particular — are valid. Other families of events are not commeasurable, so their frequencies must be measured in more than one sample. The events in such families nevertheless exhibit logical relations (given, usually, in terms of algebraic relations among observables). But for some states, the probabilities assigned to the events violate one or more of Boole’s conditions associated with those logical relations.

A violation of Boole’s conditions of possible experience cannot be encountered when all the frequencies concerned have been measured on a single sample. Such a violation simply entails a logical contradiction; ‘observing’ it would be like ‘observing’ a round square. We expect Boole’s conditions to hold even when the frequencies are measured on distinct large random samples. But they are systematically violated, and there is no easy way out (see below). We thus live ‘on the edge of a logical contradiction’. An interpretation of quantum mechanics, an attempt to answer the WHY question, is thus an effort to save logic. — Pitowsky

The force of this perspective basically is that what Bell violating correlations may have a formal cause not a physical one. The bizarre correlations could be formally entailed when certain statistical conditions are fulfilled, regardless of what system is being talked about. No information is actually being communicated across space between particles.

The question is then about what causes these joint probability absences? According to Fine, it is from non-commutativity.

Now there are many sources that attest to the fact that non-commutativity and associated uncertainty relations can be generically derived within generic stochastic systems, at least under certain conditions. In fact, this can be seen in the Path integral formulation where non-commutativity in that formulation comes from the non-differentiability (because of stochasticity) of the paths. Normally people see these paths as computational tools (purely out of incredulity). In the stochastic interpretation they represent actual definite trajectories particles may take.

Given that they are entailed formally, such correlations may occur in other areas with similar structures. Infact, it has been suggested that such non-local correlations are in principle possible in classical light: e.g.

https://arxiv.org/abs/2401.01615

Note, that classical entanglement is well-established in classical optics but it is usually only formulated in local "intrasystem" scenarios as opposed to the non-local "intersystem" scenario proposed by the paper. Given the setting is purely classical, the formal presence of non-commutativity or joint probability absences may be sufficient to provide the central mechanism for Bell violating correlations in that scenario or any other kind (e.g. social sciences they occur for probabilistic reasons albeit not as relevant because not about locality/nonlocality).

Whats most interesting is perhaps you don't need remarkably strange assumptions to get non-commutativity or virtually all quantum predictions our of stochastic systems.

For instance, the gist of the major Nelsonian stochastic mechanical assumptions are basically as follows - 1) particles follow paths by Newtons law had they been perturbed randomly; 2) the diffusion is time-reversible - which can be derived in kinds of equilibrium contexts where entropy regarding trajectories is maximized; and 3) the diffusion coefficient is inversely proportional to particle mass. And from that you can even reproduce the perfect spin (anti)correlations and Bell violations like in following dissertation and paper published from it (assumptions listed in dissertation).

https://scholar.google.co.uk/scholar?oi=bibs&hl=en&cluster=16239473886028239443
https://scholar.google.co.uk/scholar?cluster=15973777865898642687&hl=en&as_sdt=0,5

Despite the fact many would say it produces unphysical non-local correlations (obviously I have tried to argue via Fine's theorem that these may be in some sense a formal entailment that transcends physics), I think its definitely relevant to ask why it is even possible for virtually all quantum predictions to be derived from some very pedestrian assumptions in the first place. Why is it that indivisible stochastic systems with definite outcomes reproduce entanglement, decoherence and interference? Its kind of miraculous - if such non-locality should be impossible for particles in definite positions, why is this behavior even derivable?

Point 4:

My last point will be about your point about simultaneity of relativity and preferred reference frames. I think my point would be that such issues are no reason to discount a stochastic interpretation because these issues seem to be quite general. They occur in hydrodynamics, they occur for relativistic brownian motion, for thermodynamics. Markovian diffusions in general are known to not respect relativity and have superluminal propagation (mentioned in second link below too). It seems that when you start talking about things like probability and randomness, their relation to relativity just is never straightforward, and so areas outside of quantum mechanics have been or will be grappling with this same kind of issue also: e.g.

https://scholar.google.co.uk/scholar?cluster=17685845957935258058&hl=en&as_sdt=0,5&as_ylo=2023

Relativistic fluid dynamics [1] is an important tool in the description of vastly different physical systems, such as the quark-gluon plasma formed in ultrarelativistic heavy-ion collisions [2] and accretion disks surrounding supermassive black holes [3]. Early models of relativistic hydrodynamics were constructed in the mid twentieth century by Eckart [4] and Landau and Lifshitz [5], but these were later found to possess unphysical behavior signaled by causality violation [6] and the fact that in such theories the global equilibrium state is not stable with respect to small disturbances in all Lorentz frames [7]. These issues are not inherent to the formulation of viscous fluids in relativity.

https://scholar.google.co.uk/scholar?cluster=16512488009491179103&hl=en&as_sdt=0,5

Before outlining our approach, a general remark might be in order. Usually, a diffusion theory intends to provide a simplified phenomenological description for the complex stochastic motion of a particle in a background medium (e.g., on a substrate [5, 30, 33, 34, 35] or in a heat bath [20]). Thus, there exists a preferred frame, corresponding to the rest frame of the substrate (or, more generally, the center-of-mass frame of the interaction sources causing the stochastic motion). It is therefore not expedient to look for Lorentz or Poincare invariant spatial diffusion processes (cf. Sec. 5 of Montesinos and Rovelli [39]). Accordingly, we focus here on discussing simple diffusion models that comply with the basic requirements of special relativity in the rest frame of the substrate.

(a long time ago, I read of some versions of this interpretation which are Lorentz invariant. So, I guess that this kind of 'simultaneity' doesn't necessarily imply a rejection of special relativity. I don't remember however the details) — boundless

Hmm; just to be short, I feel like the issue is very up in the air and not simple. Skepticism isn't quite unwarranted imo. Certainly there seem to be stochastic field theories that can fulfil relativistic predictions but apparently have preferred frame plus some of the Markovian superluminality.

Edit: Just re-phrasing / clean up. Shouldn't change content but additional point:

Paper with interesting suggestion if non-locality appears in classical optics, it suggests that it should be compatible with Lorentz invariance:

https://scholar.google.co.uk/scholar?cluster=13776304742041840922&hl=en&as_sdt=0,5&as_vis=1

Can't comment on what math says at all but I am guessing the logic is that classical electrodynamics is already in some sense compatible with Lorentz invariance/covariance.

Really enjoyed reading this :up:

Yes. In a sense, the processes act blindly. But that implies that they follow rules, which they don't. They do not differentiate between following a rule and not following it. They don't recognize rules. So they don't explain them - any more than they explain why 2+2=4 and not 5. — Ludwig V

I don't really understand what you are trying to say here. These processes are not meant to explain the rules, they explain our behavior despite underdetermination.

If "the world" is not coherently accessible — Ludwig V

It should be If "the world" is not coherently accessible independently of perspective.

So we can make that inference from within perspectives. Obviously, this is a hope in some sense [i.e. problem of induction]; but, that is trivially the case for all claims. Obviously, such a claim is very much an abstraction that fills a role in our understanding of things. Our understanding of things includes the concept that things exist when we don't look at them, even if there is no fact of the matter in how someone could possibly objectively characterize things in a perspective independent manner.

How is that not reductionist? The bitter truth is the physics is just another way of conceptualizing the world, another lens through which to survey it. — Ludwig V

I'm not ruling out other explanatory frameworks, but there is a clear asymmetry in the sense that physics undergirds all behaviors in the world but not the other way round. Maybe a better way to talk about it is in terms of scales. We can describe how the world behaves on different scales. Behavior on larger scales obviously depends on behavior on smaller scales, regardless of the kinds of descriptions you use. Ultimately, physics is the only framework that describes the smallest scales of reality on which everything else emerges in some sense. That's not to exclude or say we don't need or want explanations on other scales [nor mean there is super hard explanatory reduction].

Causes are not correct or incorrect. They just are what they are. — Ludwig V

Yes, and they explain in a proximal sense all our rule-following behaviors in principle. I'm not really interested in some kind of objective sense of correctness. I just don't find it an interesting issue and in my conceptualization where all of our knowledge is basically idealizations regarding enaction in our unfolding flow of experiences, the idea of monolithic rules doesn't even seem well-founded to me except in a sense which is idealized, which is about what I think of as cognitive instrumentalism or pragmatism.

Forgive my ignorance, but I had this naive impression that an algorithm is a rule. — Ludwig V

Well, maybe in the loosest possible sense of a rule; but the point is that what neurons are doing in my brain are not related to the semantics of "plus" and you don't need the semantic notion of 'plus' to explain how mindless neurons do 'plus' tasks. Obviously, characterizing what artificial or biological neurons do is not exempt from underdetermination or indeterminacy either.

Edits: [ ]

Regarding tacit knowledge, I think all we do is essentially tacit at some level. Traditionally people often split up know-how and know-that but to me, know-that is a special case of know-that - or at least that is how it is implemented. Know-that is enacted.

However, this is not a simple matter of physical laws, but requires an intervening layer, like the software in a computer. — Ludwig V

Yes, I wasn't deliberately trying to exclude other things; after all, most of neuroscience does not appeal to physics. I was just trying to hit home that meaning behavior comes from processes which are independent of our own notions of meaning. Physics is the ultimate grounding since brain dynamics, computational behaviors are in principle implemented in the entities of physics. At the same time, I think I was trying to refer to something deeper than our perspective-dependent descriptions of the world since these are all in principle underdetermined and indeterminate whether in physics or neuroscience or machine learning, etc. By physical laws I just meant the way the world tends to behave independently of perspective; obviously this is not coherently accessible, but we infer that there id a world that exists and behaves consistently regardless of who is looking.

Those sub-routines will involve implementation of rules - otherwise they cannot possibly be successful or even unsuccessful. — Ludwig V

Well they would be the same processes which are not human-interpretable as described in the paper, the point being that if what they do can be explained in ways that is not fundamentally in terms of human-interpretable rules, we do not need to appeal to such kinds of rules to explain behavior, but instead to the kind of semantic-less descriptions of math, physics, statistics. Obviously, those descriptions themselves may be in principle underdetermined, but the point is that meanings are not the bottom wrung of explanation here - mindless algorithms are. Mindless algorithms drive our behavior despite underdetermination, which itself became apparent to us through the same kinds of mindless algorithms driving our categorization behaviors.

I like this Evan Thompson quote a lot!

But blindness resolves the infinite regress of interpretation and underdetermination, so it is a feature, not a bug. — Ludwig V

No it doesn't. It just explains how people can still act coherently under chronic underdetermination.

You think that the AI's hidden rules resolve the "problem" of blindness. I don't see how. If you are accepting that they are interpretable by humans, how do they not have the same problems as any other rules? To put the question another way, if the AIs rules cannot be understood by human beings (or even, if you insist, by other AIs, how would "correct" and "incorrect" have any meaning? To put the point yet another way, if the AIs rules really were uninterpretable by human beings, what meaning would "correct" and "incorrect" have? — Ludwig V

Because they don't resolve anything. All that it does is explain the fact that we act. Brain's are mechanistic systems which produce our behavior regardless of how we interpret it. The fact that I cannot interpret my behavior does not magically stop the physical chains of events that produce my coherent behavior. I don't need a determinate rule-interpretable understanding of what brains or A.I. do for them to perform the tasks they do.

Surely brains don't interact directly with other brains, but only via a chain that connects them - roughly, via the bodies they live in. — Ludwig V

I wasn't implying otherwise.

This is the idea that all science will, in the end, be unified into a single over-arching structure. That's an article of faith, or perhaps a programme of research. It certainly isn't a fact. What's worse, is that, by eradicating people from your causal chain, you seem to be reducing people to their brains. Perhaps unintentionally, but nonetheless, there's no conceptual space for them. — Ludwig V

What you are saying is directly opposite of what I had written.

Note I edited my original comment for clarification.

If the laws are underdetermined, how can they determine those mechanistic processes - except, perhaps, by some version of blind action? I do agree that there are complicated physical processes going on. But we do not know how to translate from the physical level of description to the human - it's called the hard problem. But if there were a translation how would it not be a matter of rules? — Ludwig V

Laws are descriptions or produce predictions, in our minds, about a world that exists independently of us - they don't determine the world, we pluck them out from our observations in perspective-dependent ways. Our descriptions can be underdetermined but that doesn't mean that things aren't happening in the world regardless of how we choose to describe them; nor does it mean that our descriptions are not useful to us. Those things cause our behavior even when I am not looking, even if I cannot characterize them in a determinate, perspective-independent way.

"people" for "brains" — Ludwig V

I mean, I don't understand how you could think this as some kind of over-reductive description when I literally said in the same paragraph the following:

And then, good understanding of whats happening here wants multiple levels of explanation spanning all fields from microbiology to evolution to linguistics, anthropology, social psychology to history and upward. No one field or level of explanation can do justice to everything. — Apustimelogist

[Brains] is one level of description, appeal, explanation - made necessary by the fact that it explains how people behave and think, at least in the proximal sense.

Of course brains interact with their environments. But they don't interact with other brains — Ludwig V

If brains are in their environment then ofcourse they can interact with other brains.

I have no idea what would persuade us to accept that any machine, biological or not, is not working from any human-interpretable rules. — Ludwig V

Its a well-established issue in machine learning and I already had posted a paper talking about it in the context of neuroscience this thread:

https://www.sciencedirect.com/science/article/pii/S089662731931044X

But in that case, we can certainly work out what's going on from the results. — Ludwig V

Which is always our interpretation of what is going on and falls to the same kinds of rule-following issues as initially described - which inevitably would result in another appeal to blindness.

in fact they are not acting at all in the sense that people act. — Ludwig V

Depends what you mean, I guess; but, not important.

But the idea that the laws of physics are not underdetermined is a big jump. — Ludwig V

Wasn't necessarily imply they weren't underdetermined; but the point was that rule behavior is not determined by rule abstractions floating about in a platonic dimension. It is determined by extremely complicated mechanistic processes in the world and our brains, as is the behavior which translates to our agreements about the applications of words and categorizations of behaviors.

Edit: [ ]

This is fair! A kind of dualist then?

I would just add that a thoroughgoing reflexivity between word and world implies that cognitive abilities and neuronal architectures are themselves responsive to, and continuously shaped by, the social world that they are exposed to and intertwined with. We can’t use biological concepts as the court of last appeal and legitimation for grounding conceptual meaning when they are not split off from the social milieu. — Joshs

Yes, definitely agree. Ofcourse, words and concepts must be inherently evolved, developed, learned, used in a social context. Brains in some sense synchronizing with other brains as well as other parts of the environments they navigate. And then, good understanding of whats happening here wants multiple levels of explanation spanning all fields from microbiology to evolution to linguistics, anthropology, social psychology to history and upward. No one field or level of explanation can do justice to everything.

To be clear, "metaphysical truth" isn't some vague term I've concocted. — Count Timothy von Icarus

Ofcourse, but I don't think it comes for free and imo it is neither necessary nor sufficient to explain people's beliefs about the world so it doesn't seem like a particularly good explanation.

On the deflationary view, "truth" has no explanatory or metaphysical import. — Count Timothy von Icarus

I would say there is a point here, since "metaphysical truth" is neither necessary nor sufficient for truth behaviors and people often either use the word cheaply or are wrong. If we get things right about correspondences to the world, then it must be mediated by messier, more elaborate mechanisms. At the same time; from my perspective, the notion of "correspondence" itself is rather cheap and thin, nor is it perspectiveless.

So my point isn't meant to be handwavey. It's a straightforward denial of the idea that "reason" should be thought of as simply the ability to follow the rules within the context of any specific language game — Count Timothy von Icarus

I think language games here would just reflect the activities and norms of reasoning which will naturally have emerged within a community. People don't reason well if they don't have the right information, if they haven't learned or been taught how to reason, if they are not attuned to the norms of what is deemed "reasonable" position to hold. And ultimately, what is doing the work is brains in their interactions with their environments and other brains.

But the assumption that this precludes access to a non-deflationary version of truth seems to need to assume that statements in language games are "what we know," not a means of knowing, or else that a short-lived positivist notion of correspondence truth is the only possible notion of metaphysical truth and that once it is defeated deflation must follow. — Count Timothy von Icarus

For me, "knowing" doesn't mean much more over the unfolding activities of our behaviors, thougjts, experiences. We predict experiences, enacting those predictions in various ways. I see language games are an exemplification, a subset of that process.

Well, better in virtue of what is the question, right? Better at approaching truth? Or better because they can be demonstrated from dominant hinge propositions in a given community? Is the goodness of an argument determine solely by the expectations of the people who are going to hear it? — Count Timothy von Icarus

He thing is that an explanatiom being deemed better doesn't necessarily mean it approaches truth. Someone can be mistaken about what they think is a good explanation. Something can be a good explanation only because we don't have all the relevant information to deduce a better one. At the same time, we likely do have to learn about good reasoning and good explanations, and such learning isn't going to be isolated from the community.

Ok, I see. The wave-function is interpreted as an 'useful' fiction but at the same time the theory also adopts Counterfactual definiteness. How is non-locality handled in this interpretation? — boundless

Counterfactual definiteness? Yes, I guess; but again, when people talk about counterfactual definiteness, they are usually talking about the wavefunction and perhaps things like collapse. Stochastic interpretation would be talking about definiteness in regard to something else, so the concept has arguably changed.

Regarding non-locality? In the most up to date formulations of the stochastic interpretation, it should be as non-local/local as orthodox quantum theory. There should be no statistical signalling statistically, there is no collapse and measurements do not have any causal influence on each other across space or time. Obviously, there must be non-local correlations and stochastic laws between separated particles but it doesn't seem to me that this implies some kind of causal signalling. At the end of the day, stochastic systems are capable of producing non-local correlations from some unremarkable assumptions regardless of whether the stochastic interpretation is correct or not.