In philosophy there is space for 'the unconditioned' — Wayfarer

There is also the fundamental philosophical maxim, 'know thyself' with the concommitant emphasis on self-awareness — Wayfarer

in this discipline 'we are that which we seek to know'. — Wayfarer

I very much doubt scientists would happily admit that they have a blind spot. Do you find they do? I'll assume not.

So you merely claiming that philosophy doesn't isn't going to cut it. Science will claim it doesn't too. What can you give me by way of evidence?

what we perceive is experienced as being perceived immediately, or do you experience some time lag between turning to look at, say, a tree and seeing it? — Janus

I wasn't talking about the time lag, I was talking about the sense of directness. I don't get a sense that I see things 'directly'. Sometimes the form of something is unclear, but I don't believe the actual thing is unclear, I believe that my vision is not delivering me a clear image (in my naive everyday belief). I don't recognise this idea of directness that you're talking about. Vision seems quite clearly indirect to me, I imagine a world made of solid, clear object and yet many of the I can't see clearly. That's my day-to-day experience. Not a direct one at all.

As to how it "informs my art" it's the difference between accepting what you perceive just as it immediately appears to you, giving yourself over to it and becoming absorbed in it — Janus

I've heard artists talk this way. It's not something I really 'get' but it sounds almost integral to the artistic process. I suppose that's probably why all my drawings are shite.

Whenever a terminological framework has the purpose of explaining human consciousness(meaningful human experience) and/or other kinds of consciousness(such as non-human meaningful experience), and it is based upon either internal/external, or physical/non-physical, or even perhaps both, then those practices are doomed to fail as a result of not having the explanatory power to be able to take proper account of that which consists of both internal and external things, physical and non-physical things. — creativesoul

I can see how that might be the case, but I don't think dividing states into internal and external suffers from that problem as it still retains the possibility of modelling something which is both (a person in their environment for example). The division doesn't prevent both sides from being in the model.

Meaningful experience exists in its entirety, in simpler forms, prior to our knowledge. <-------That's the pivotal ontological consideration which ought inform the selection/creation of our terminological framework. — creativesoul

I think you're making a mistake in assuming that because something exists prior to our accounting for it, it must be that our accounting is wrong if it doesn't represent it fully. You're making tow unwarranted assumptions. Firstly that {that which exists in its entirety prior to our accounting practices} can be represented with only one 'true' model, that there's only one 'true' way to account. There may be many, hundreds. Secondly that our accounting practices must capture the entirety of the thing they're accounting for. I see no reason why they should.

1) the science of Markov blankets doesn't directly address the philosophical issue of subjective experience (as explained in the first paper) — Michael

Agreed. The idea of there being no coherent thing called a subjective experience is a different matter. As I said, we can discuss qualia.

2) colour terms like "red" don't (only) refer to some property held by some external world cause but (also) by something that happens "in the head" (even if you want to reduce qualia/first-person experiences to be something of the sort described in the second paper). — Michael

No. You've misunderstood what the second paper is about. The key is in the introduction, which you quoted from but didn't seem to take account of. It's quite clear...

We need to understand the meta-problem in a way that is (broadly speaking) behavioral rather than making essential reference to phenomenal experience itself. In practice, this means the goal is to explain the things we say and do, while bracketing the question of whether or not they reflect phenomenal experience.

They're asking the question why we think there's a hard problem at all, not whether there is one. Nothing in the paper is about perception, it's about the meta-model we have of perception in folk psychology and why we have it.

They even quite specifically repeat the point I've been making...

seeing red and feeling pain (just like seeing dogs, cats, vicars, and even (Letheby and Gerrans (2017)) having a sense of self) are themselves inferred causes.

...and...

Instead, our brains construct qualia as ‘latent variables’ – inferred causes in our best ‘generative model’ (more on that later) of embodied interactions with the world.

...and...

Qualia – just like dogs and cats – are part of the inferred suite of hidden causes

.

When they say...

in perception, we seem to become highly confident of something, where that something does not quite mandate high-level beliefs about the state of the distal world itself

...what they're getting at is a meta-model of how perception works (our own internal model of what's going on) which includes this high-certainty, low clarity 'feeling' of experience which works alongside our knowledge that what we see might not be what it seems.

They're arguing that qualia are a Bayesian model of the high certainty policy on any perception which acts as a base on which to build models of less probable possibilities. It's a model space in which possibilities can be explored. It's...

not some kind of raw datum on which to predicate inferences about the state of body and world.

In other words. We do not 'see' qualia. They are (in the paper) an inferred part of our internal model of how perception works.

the claim is that qualitative contents reflect mid-level sensory encodings apt for the selection of local action, and/or steeped in interoceptive information. These strikingly certain, sensorially-rich content states are then mistaken for something else (something ‘beyond content’) when we engage in certain kinds of imaginative exercise that hold them fixed while varying the distal realm

What is at question is the truth or falsity of the models, in the sense of correspondence, and that is whether the models are a fair representation of what is supposed to be being modeled. — Metaphysician Undercover

Oh. Well in that case I couldn't care less.

That is exactly the evidence I have been giving you — Metaphysician Undercover

You haven't provided a scrap of 'evidence' yet, you've just been asserting things so far.

I clearly indicated that overlapping does not prevent a system from being defined. I said it prevents a system from being defined as "discrete" — Metaphysician Undercover

It doesn't. A discrete system can still be defined despite being open. I gave the example of a cell.

do you see that the "cell membrane" in your example is a third thing? — Metaphysician Undercover

No. It's part of the cell, so part of the system.

the definitions employed by systems theorists are false premises — Metaphysician Undercover

Definitions are not premises. It just declares how the word will be used.

when the system acts in a way such that it is influenced (caused) to behave in a way which is neither the result of observable external causes, nor the system itself (2nd law), then we ought to conclude internal causes which are not part of the system itself. To conclude "hidden" external causes is a false conclusion, because properly designed experiments have the capacity to exclude the possibility of unobservable external causes. — Metaphysician Undercover

What experiments? I've never read of any experiments which reveal a system acting in a way which is neither the result of internal nor external states. Could you cite a paper?

by your description, it is not inside the system, it is the boundary. — Metaphysician Undercover

No. The boundary is part of the system.

you did say neither one nor the other of the two possible explanatory methodologies — Mww

I don't agree there are only two possible explanatory methodologies.

If it be agreeable that the domain of philosophy is rational thought in accordance with logical law, and the domain of science is empirical experiment in accordance with natural law, and furthermore that no human ever performed an experiment without first thinking how it should be done in order to facilitate an expected outcome.....we arrive at both a clear chronological succession and a clear methodological distinction — Mww

Ah, I see. Yes, fair enough. Philosophy, thus defined does come first. But...(you knew there was going to be one)...what is true of philosophy the practice is not true of any actual philosophy. Rational though comes first. Kant's theory of rational thought doesn't (necessarily).

If that were true, there would never be such a thing as a paradigm shift, whether in science, ethics, metaphysics or anything else. If there ever was that which is sufficient reason to cause the collapse of an antecedent condition, then that thing could not be contained in that which collapsed. — Mww

No, but cultural changes can occur for all sorts of reason (my favourite involves Lorenz attractors, it's really cool), which can then lead to changes in culturally mediated paradigms. The change needn't be rational.

Is it a far-fetched personal cognitive prejudice, or is it a case of the more things change, the more they stay the same? — Mww

Not all that far fetched...

https://www.frontiersin.org/articles/10.3389/fnsys.2016.00079/full

..but there are significant differences when it comes to active inference (which has elements of predictive processing, but is not restricted to it) to do with that troublesome second crossing of the Markov boundary which is so often forgotten. The active state. We move, interact with the world, harvest data, even change the world to fit our models better... and all this is part of the process of inference. Does Kant have an equivalent?

If you were arguing that all of us should abandon realism and avoid the tendency to use terms that suggest we believe there is an unshackled truth beyond our models I’m all for that. By all means challenge me whenever I let such vocabulary slip in. — Joshs

Well then we are on the same page it seems.

if I understand you correctly,
you believe such realist terms SHOULD be part of our scientific and philosophical claims — Joshs

Not so much 'should' as unproblematically are.

anything we say about such a cultural-independent realm is contingent on and relative to our practices, which are always changing. — Joshs

Agreed.

Any claim of an asymptotic movement of scientific knowledge toward representation of something independent of that movement itself is a claim within a practice that is itself changing — Joshs

Also agreed.

It is an invitation to see for yourself if what appears to be an internally generated representational model of an outside doesn’t qualtiatively alter the sense of that outside in the act of representing it. — Joshs

Absolutely, but there's consistency too, we couldn't think two straight thoughts in a row if every time we thought something it changed the model of the thing we're thinking.

Yes, but it seems that to you this is a bug, a contextual imposition of cultural bias and distortion on an autonomous scientific enter­prise from the “outside”. — Joshs

I'm not sure what I've said to give you that impression, but I don't feel that way. My beef, such as it is, is only with the equally culturally embedded philosophies claiming to be anything beyond optional. Optional narratives, I'm all for.

In other words. We do not 'see' qualia. They are (in the paper) an inferred part of our internal model of how perception works. — Isaac

Even if they don't say that we see qualia, the point still stands that colour terms like "red" (can) refer to this qualia, which according to them is a Bayesian model, and so colour terms like "red" don't (only) refer to some property of the external cause of the sensation.

The mistake you seem to be making is to conflate our external world model with the external world. But as they say, "despite holding all the phenomenal facts fixed, how the world really is might vary, even to the point of there being nothing at all bearing the properties so confidently represented as being present" which only makes sense if "the properties so confidently represented as being present" refers to something. And that something is "in the head". This is the sort of thing that colour terms like "red" refer to, even with this Bayesian interpretation of perception, and any further claim of this kind of redness being a property of some external cause of perception is a naive projection.

There is, perhaps undoubtably, some external world property of things that causes most humans to see red (i.e. infer these Bayesian models), but that model (and its properties) isn't that external world thing (or its properties). You can use the same word to refer to both if you want, but as I have said, and as you seem to be demonstrating, that leaves us susceptible to equivocation, and the erroneous claim that there is a "right" way to see things like colour. A human isn't wrong if he sees colours as birds do; his visual system is just atypical of humans.

Interesting paper. Between you and I get all kinds of nifty stuff to rock my epistemic water vessel, so sincere thanks for it.

“.....and he did not set out theories of learning....”
(From the link, under “A Note on the A Priori”

Why would he, when the theory on knowledge he did set out presupposes it? It is either tautologically true, or a useless exercise, to suggest we know things we haven’t the ability to learn. If the former, a theory of learning is unnecessary; if the latter theories of knowledge are all catestrophically irrational.
———

The active state. We move, interact with the world, harvest data, even change the world to fit our models better... and all this is part of the process of inference. Does Kant have an equivalent? — Isaac

Dunno about changing the world to fit our models; seems sorta backwards to me. Be that as it may......

The active state would be cognition. The process of inference would be the tripartite logical syllogistic functionality between understanding (major), judgement (minor(s)), and reason (conclusion). Now, as you’ve said, albeit in a different way, re: the talking is not the doing, this is how we talk about it, how we represent to ourselves a speculative methodology, but the internal operation in itself, functions under the condition of time alone, such that cognition is possible from that methodology.

Not sure that’s a very good answer, but best I can do with what I’m given, and considering my scant experience with Markov blankets.

Takes nothing away from the paper, though, don’t get me wrong.

Whenever a terminological framework has the purpose of explaining human consciousness(meaningful human experience) and/or other kinds of consciousness(such as non-human meaningful experience), and it is based upon either internal/external, or physical/non-physical, or even perhaps both, then those practices are doomed to fail as a result of not having the explanatory power to be able to take proper account of that which consists of both internal and external things, physical and non-physical things.
— creativesoul

I can see how that might be the case, but I don't think dividing states into internal and external suffers from that problem as it still retains the possibility of modelling something which is both (a person in their environment for example). The division doesn't prevent both sides from being in the model.

Meaningful experience exists in its entirety, in simpler forms, prior to our knowledge. <-------That's the pivotal ontological consideration which ought inform the selection/creation of our terminological framework.
— creativesoul

I think you're making a mistake in assuming that because something exists prior to our accounting for it, it must be that our accounting is wrong if it doesn't represent it fully. You're making tow unwarranted assumptions. Firstly that {that which exists in its entirety prior to our accounting practices} can be represented with only one 'true' model, that there's only one 'true' way to account. There may be many, hundreds. Secondly that our accounting practices must capture the entirety of the thing they're accounting for. I see no reason why they should. — Isaac

The mistakes you think I'm making are ones I'm not. We need not know everything. Our models need not be able to account for everything.

However, if we are to place confidence in a model of meaningful experience(consciousness), it ought be that the model is amenable to terms of evolutionary progression. It ought be simple enough to be able to account for the simplest meaningful experiences while having the richness of potential to be able to account for our own highly complex meaningful experience, as well as all other meaningful experiences in the meantime.

You missed the point of the ontological consideration, and neglected to address the elucidation of the issues raised that followed from my initial reply to you. All good though. No worries.

Multiple models can all be useful and contradict one another. They cannot all be true and contradict one another. Not sure why truth has been invoked here...

What can you give me by way of evidence? — Isaac

Only rational argument, which apparently doesn't cut it.

Notice that when psychologists play ‘gotcha!’ and talk about how our naive perception is fooled by illusions and tricks, that the ‘real’ truth of what we experience is hidden from us , they are referring to a level of analysis that first needs to be constructed by us as a fresh perspective. In other words, in order for some some phenomenon to be declared ‘hidden’, the conceptual framework within which its hiddenness is intelligible must first be invented as a fresh form of conceptualization. Could one not then follow the phenomenologists and say that both the ‘naive’ and the hiddenness-savvy frameworks are different varieties of direct perception, the second being an elaboration and transformation of the former? — Joshs

It sounds right that both are different varieties of direct perception, but the naive view is a (direct) conception of directness, whereas the "hiddenness-savvy framework" is a (direct) conception of indirectness. All our perception and conception seems, experientially speaking, to be direct and that doesn't change even if the conception is of indirectness. We can conceive of indirectness, of hidden process, of lack of immediacy, but we cannot perceive indirectness, we cannot perceive hidden process or lack of immediacy, because if we perceived it it would not be hidden; it would seem immediate to us. That's my take anyway.

Vision seems quite clearly indirect to me, I imagine a world made of solid, clear object and yet many of the I can't see clearly. That's my day-to-day experience. Not a direct one at all. — Isaac

So, let's say you are sitting in a room, and you imagine, say, that room full of "solid, clear objects" and yet what you see at any moment is only what you are directly focused on, attending to. You can turn to any of these objects and see them clearly (if the light is sufficient of course). Say you're now staring at your computer screen; it's right there immediately in front of you. That's the sense of direct realism; just that anything you focus your attention on appears right there directly visible, immediately present to you. Of course, the science of perception tells a very different story.

Note, I'm not arguing for the "truth" of one perspective over the other. I also understand that you believe that all views are culturally mediated, but I think it is arguable that the "naive" view is "native", just on account of the fact that perception seems immediate, and it is only when detached, "objective" investigation and analysis is carried out that it is possible to come to the conclusion that perception is not "really" immediate.

I found that Predictive Processing paper a couple of weeks ago, by chance. There's a scholar (might be mentioned therein) by the name of Andrew Brooks who has specialised in Kant on Cognitive Science, you can see one of his papers here and also various entries on the Stanford Encylopedia of Philosophy.

Interesting paper, I was particularly struck by this:

To see the contrast, we need to return to a work mentioned earlier, Anthropology from a Pragmatic Point of View. In this unjustly-neglected work, Kant tells us that anthropology is the study of human beings from the point of view of their behaviour, especially behaviour toward one another, and of the things revealed in behaviour. Anthropology in this sense contrasts with what Kant understood as empirical psychology, namely the introspective observations of our own mental states. Kant’s rejection of introspection and turn to behaviour have a very contemporary feel to them. (For more on Kant on introspection, see Brook
2004.)

The Anthropology is important for other reasons, too. In particular, it illuminates many things
in Kant’s picture of cognition. To make sense of behaviour, character, etc., Kant urges early in the
work, we must know something of the powers and faculties of the human mind: how it gains
knowledge and controls behaviour. Thus, before we can study character, etc., we must first study the
mind. In fact, this study of the mind (Anthropological Didactic, he calls it) ends up being three quarters of the book. In it, Kant discusses many topics more clearly than anywhere else. In one
amusing passage, Kant indicates that he was, if anything, even more hostile to the use of
introspection to understand the mind than I have indicated. Introspection, he tells us, can be a road to "mental illness" (Ak. VII:161). Strangely enough, Kant never seems to have asked whether
anthropology in his sense could be a science.

Did Kant become a proto-behaviorist?

Behavourism eschews any consideration of the mind whatever, so it's not likely!

Note I did say "proto-behaviorist", not "behaviorist" and certainly not "radical behaviorist" (viz. Skinner). In any case in that paper it is asserted that Kant rejects introspection, while saying that behavior can only be understood subsequent to "studying the mind". How would it be possible to study the mind other than via observing behavior, if introspection is ruled out?

Behaviorism is a systematic approach to understanding the behavior of humans and other animals.[1] It assumes that behavior is either a reflex evoked by the pairing of certain antecedent stimuli in the environment, or a consequence of that individual's history, including especially reinforcement and punishment contingencies, together with the individual's current motivational state and controlling stimuli. Although behaviorists generally accept the important role of heredity in determining behavior, they focus primarily on environmental events.

Behaviorism emerged in the early 1900s as a reaction to depth psychology and other traditional forms of psychology, which often had difficulty making predictions that could be tested experimentally, but derived from earlier research in the late nineteenth century, such as when Edward Thorndike pioneered the law of effect, a procedure that involved the use of consequences to strengthen or weaken behavior.

With a 1924 publication, John B. Watson devised methodological behaviorism, which rejected introspective methods and sought to understand behavior by only measuring observable behaviors and events. It was not until the 1930s that B. F. Skinner suggested that covert behavior—including cognition and emotions—is subject to the same controlling variables as observable behavior, which became the basis for his philosophy called radical behaviorism.[2][3] While Watson and Ivan Pavlov investigated how (conditioned) neutral stimuli elicit reflexes in respondent conditioning, Skinner assessed the reinforcement histories of the discriminative (antecedent) stimuli that emits behavior; the technique became known as operant conditioning.

The application of radical behaviorism—known as applied behavior analysis—is used in a variety of contexts, including, for example, applied animal behavior and organizational behavior management to treatment of mental disorders, such as autism and substance abuse.[4][5] In addition, while behaviorism and cognitive schools of psychological thought do not agree theoretically, they have complemented each other in the cognitive-behavior therapies, which have demonstrated utility in treating certain pathologies, including simple phobias, PTSD, and mood disorders.

From here

So, the mind is not so much "eschewed" as it is understood in terms of reflexes and conditioning, and it is just on account of this understanding that introspection is ruled out, because the introspectively generated story that individuals tells themselves about their motivations and states of minds are not understood to give the "real picture" as to what is going on.

I'm not saying I agree with this view, but just seeking to get clear about what it entails.

In any case in that paper it is asseted that Kant rejects introspection, while saying that behavior can only be understood subsequent to "studying the mind". How would it be possible to study the mind other than via observing behavior, if introspection is ruled out? — Janus

That's a very interesting question. One of the papers that @Joshs linked to is about that - Killing the Straw Man: Dennett and Phenomenology. It distinguishes between the 'straw man' depiction of introspection as the mere 'reporting of what comes to mind', and the discipline involved in phenomenological analysis illustrated with reference to Husserl's Logical Investigations.

Skinner was one of the two guys I loved to hate as an undergrad. (The other being Ayer.)

It distinguishes between the 'straw man' depiction of introspection as the mere 'reporting of what comes to mind', and the discipline involved in phenomenological analysis illustrated with reference to Husserl's Logical Investigations. — Wayfarer

I did read that paper at the time it was linked, and I'd have to go back to it to be sure, for which I don't have time right now, but from memory Zahavi points out that the practice of phenomenology does not consist in introspection as it is usually understood, but in reflection on the nature of experience just as it immediately seems to us,"back to the things themselves" (not the "things in themselves" :wink: ) . This latter would seem to consist in a kind of generalizing exercise of the memory. Maybe @Joshs will give his perspective here.

Absolutely, but there's consistency too, we couldn't think two straight thoughts in a row if every time we thought something it changed the model of the thing we're thinking. — Isaac

Yes, it depends on how much it changes. Our faces are changing all the time, but I might still recognize someone I haven't seen for twenty years.

No. It's part of the cell, so part of the system. — Isaac

Now you contradict yourself. You said very distinctly and explicitly: "Everything within the cell membrane is the system, everything outside of it is not". Now you say the boundary itself is inside the system. Or are you saying now, that the membrane is inside itself, being both inside the system, and the thing which everything within it is inside the system? The membrane is inside itself?

So let me get this straight. The membrane is not the boundary, as you said earlier, the membrane is inside the system, therefore part of the system, and inside any proq|aaposed boundary. What is the boundary then?

I suggest that you do not have a "boundary" at all, just principles whereby you judge some things as part of the system, and other things as not part of the system. And unless your principles are stated as spatial principles, your use of the spatial terms "internal" and "external" is misleading. The parts of "a system" may be scattered around the world, like a network of microwave communications, certain things being designated as part of the system, and other things as being not part of the system, and to use spatial terms like "internal" and "external" is rather misleading, because the things which are not a part of the system are intermingled with the things which are.

Would you agree with this characterization? If there is not a spatial boundary which circumscribes an area of inside, leaving another area as outside, then really what you have is a judgement as to which things are a part of the system, and which things are not. And the things which are not a part of the system could be physically inside the system, or they could be physically outside the system. So for example, the molecules which pass through the cell membrane in the process of osmosis, may or may not be part of the system both before and after they pass through the membrane, depending on how one models "the system", how things are included as part of the system. Being inside or outside the membrane is an arbitrary difference, if the membrane is assumed to be part of the system, because the membrane is not a boundary in this sense.

Nice.
————-

How would it be possible to study the mind other than via observing behavior, if introspection is ruled out? — Janus

The answer is in the paper. Simply put: we don’t.

So, we don't study the mind? Or we don't rule out introspection? Note the claim in the paper is that Kant did rule out introspection, And this relates to the Zahavi paper which asserts that introspection is not the way of phenomenology, but rather reflection of the general nature of experience is.

So, 1) the science of Markov blankets doesn't directly address the philosophical issue of subjective experience (as explained in the first paper) and 2) colour terms like "red" don't (only) refer to some property held by some external world cause but (also) by something that happens "in the head" (even if you want to reduce qualia/first-person experiences to be something of the sort described in the second paper). — Michael

The problem with this way of modeling is that it sets a boundary to the outside of the system, but it does not set a boundary to the inside. As Isaac describes, everything inside the boundary is designated as inside the system. This means that causal influences which change the system must come from outside the system. The only change caused from within could be the cause of the entropy of the second law of thermodynamics.

Consider a sphere, and all within the sphere is internal to the system. Now suppose there are changes to the system which cannot be accounted for by outside influence. A simple example could be the cause of existence of the system itself. We cannot say that it is the system which causes these changes because the system on its own can only follow the second law of thermodynamics. And in the simple example we'd have to conclude that the system caused itself.

But if we put a boundary to the inside of the system, suppose an infinitesimally small centre to the sphere, and we allow that this "internal" is not a part of the system itself, that problem is resolved. We can now have internal causation to the system, which is not a part of the system proper.

This is why systems theory is not very good for this type of modeling, it only imposes one boundary, between the system and everything else. In reality though, we need two boundaries, one between the system and the external, and one between the system and the internal, because the system is always going to exist as a medium between the two extremes, which are not part of the system itself.

the point still stands that colour terms like "red" (can) refer to this qualia, which according to them is a Bayesian model, — Michael

Here's a Bayesian model

P(S1,…,Sn)=∏i=1np(Si|parents(Si))

What colour is it?

Here's a Bayesian model

P(S1,…,Sn)=∏i=1np(Si|parents(Si))

What colour is it? — Isaac

If that’s what they mean by quaila being a Bayesian model then they’re demonstrably wrong because nothing like that happens when I see red, or indeed when I see anything. Perception/experience isn’t maths. So it’s as the first paper I referenced said: this Markov blanket theory says nothing about the philosophical hard problem of consciousness, it’s just a mathematical description of an organism’s functional response to stimuli.

Regardless, they’re the ones saying that redness is a Bayesian model, not me. My point is only that by their own account of perception redness isn’t a property of some external stimulus, contrary to your claims.

Between you and ↪Wayfarer
I get all kinds of nifty stuff to rock my epistemic water vessel, so sincere thanks for it. — Mww

Happy to help.

Dunno about changing the world to fit our models; seems sorta backwards to me. — Mww

It's just about direction of fit. If I have a model of my lounge and the chair's not actually in the place I was expecting it to be, I don't have to update my model, I can move the chair.

The active state would be cognition. The process of inference would be the tripartite logical syllogistic functionality between understanding (major), judgement (minor(s)), and reason (conclusion). Now, as you’ve said, albeit in a different way, re: the talking is not the doing, this is how we talk about it, how we represent to ourselves a speculative methodology, but the internal operation in itself, functions under the condition of time alone, such that cognition is possible from that methodology.

Not sure that’s a very good answer, but best I can do with what I’m given, and considering my scant experience with Markov blankets. — Mww

About as extensive as my understanding of Kant, so it's a wonder either of us can understand a word the other says. I'll give this some thought though. First impression is that cognition can't be an active state because it doesn't interact with the external states, but...

You missed the point of the ontological consideration — creativesoul

Could you perhaps repeat it for me?

Only rational argument, which apparently doesn't cut it. — Wayfarer

You made claims about what philosophy is able to do. I asked if it actually did. there's no rational argument can be brought to bear on that question. It's answered with examples.

you can see one of his papers here and also various entries on the Stanford Encylopedia of Philosophy. — Wayfarer

Thanks. I'll stick it on the reading list. Have you mentioned it to @Mww?

To study the mind presupposes it. So.....if mind is the unconditioned relative to human cognitive systems, what is there that can presuppose? To posit an antecedent to an unconditioned is a contradiction. Which relates to introspection, in that the mind ends up studying itself, which must be impossible. Now we got all kindsa metaphysical roadblocks, in that we are mistaking the replication of the doing of the deed, for the deed itself being done.

It just may be Kant’s greatest philosophical gift was not to try to explain stuff that didn’t need it.

As usual, I have no idea what you're talking about. The Markov boundary is a statistical feature of a network. It's not an object. It is at the membrane, not the membrane itself.

nothing like that happens when I see red, — Michael

How would you know. Do you have privileged and unfettered access to everything that happens in your brain?

There are several experiments which show that the mental processing of sensory inputs is, at least, extremely similar to Bayesian modelling, as subjects update expectations with almost exactly the same function as would a computer carrying out Bayesian modelling.

Not only that, but once you define a self-perpetuating system, then it's state, by definition can be defined by the Fokker-Plank equation (a probability function of the state of any system which has an attractor - a shape, or state it tries to maintain against random decay). When you write this equation from the position of the system's inferences about the external world (which it must have in order to resits random decay) they come out exactly identical to Bayesian model evidence functions.

So it's not only evidentially demonstrated that you carry out these functions, but it is mathematically demonstrated too.

they’re the ones saying that redness is a Bayesian model, not me. — Michael

They're not. They're saying that our experience of certainty during the modelling of red objects about the the colour is itself a meta-theory about perception which (like all our models) can be expressed in Bayesian terms.

My point is only that by their own account of perception redness isn’t a property of some external stimulus, contrary to your claims. — Michael

That's not anywhere in the text. I've already explained what the purpose of the paper is, it's written in the introduction to it. It is explaining the prevalence of certainty in our meta-theory about perception. It makes no reference whatsoever to what 'red' is actually a property of. The whole paper is about the role that the idea of qualia plays in our meta-theory of perception. It's not even about actual qualia as a part of the brain's process of perception.

The paper says:

"Our claim is that when the brain estimates that a suite of mid-level re-codings, couched in terms of features such as redness, roundness, loudness, pulsatingness etc. etc., as highly certain, it can simultaneously compute that this vivid set of (perhaps 100% agent-certain) contents is consistent with multiple ways the real world might be."

It quite clearly uses the words "redness", "roundness", "loudness", "pulsatingness", etc. to refer to "mid-level re-codings", not to any property of the external stimuli, and distinguishes this "vivid set of ... contents" from "[the] ways the real world might be."