And a fair amount of a human's interaction with its environment is reflexive and explained post hoc. So again, what we call perception has to be tuned to needs.

Could be made a better inference with a theory of content determination — fdrake

I’ve been wondering where “aboutness” is in these causal chains, and how @Isaac proposes to infer 9 is about 8 from 9 was caused by 8. Since the model is not supposed to be descriptive but predictive, you might think intentionality could be captured by saying that 9 points to 1 — which is around the corner from your point about priors. That makes it at least as tempting to say that 9 is about 1 as it is to say 9 is about 8 (or 7 or 6 or ...). Why look back in the chain for meaning, instead of forward? It might also make sense to think of the logging at 9 being something like “Cycle finishing, coming back to 1 now,” or “Now connecting 8 to 1,” or “Got from 1 to 8, headed back to 1.” Don’t we expect to find not a bare, descriptive report logged to consciousness but one with some kind of directedness embedded in it?

Thank you for those responses. I don't see anything here to criticise on philosophical, conceptual grounds. It is consistent with what I would call a realist position since it takes for granted that there is stuff that is independent of perception and action.

The concept of qualia is very usefull for pointing the way to material processes. — Goldyluck

But you did not make use of qualia in your description - you described what you saw in terms of circles and size and colour - all stuff that is part of our shared world.

So how are they supposedly useful?

Don't you think the world we share is made up from qualia? — Goldyluck

Of course not. It's made of chairs and cups and trees and people. And that some (you?) might suppose otherwise is why qualia are a bad idea.

But the last are secondary, not primary. — Goldyluck

https://thephilosophyforum.com/discussion/comment/629929

In the article cited, there is a discussion about the ambiguity of "qualia".

Treating colour as prior to the things that are coloured is presumptuous, as would be treating coloured things as prior to colour. The one is never without the other.

My reply to you, again, is that nothing is added to our account of red and loud things by introducing qualia. They do not have explanatory value beyond our existing language. You described of flashing lights before your closed eyes quite comprehensibly without reference to qualia.

So you will be able to address the criticisms listed directly.

https://thephilosophyforum.com/discussion/comment/629929

I'd just add that it doesn't then follow that we perceive images or, alternatively, respond to images. Instead we respond to things that we perceive, such as red flowers.
— Andrew M

But the latter doesn't follow either. I mean, I agree it doesn't necessarily follow that we respond to images (as in 'I have an image of my house on this USB stick'), simply from the fact that such a definition is plausible, but then it also doesn't necessarily follow that we respond to things we perceive from the the fact that such an definition is only plausible. Neither case has been made nor refuted. — Isaac

It's the way we ordinarily use our words. That thing there that I'm pointing to (cue, a red rose in a vase on the table) is what I understand a red flower to be. I can see it, take it out of the vase, and react if I drop it.

That a scientist could potentially put probes on my brain and view an image of what I'm looking at doesn't change anything about what I'm looking at.

I think a lot is made here of the status of an intermediary in the process of perception, which seems to be to be wrongly hinged on epistemological concerns when it's rightly more ontological.

That something causes us to respond when seeing (what we call) a red flower is not in dispute. That there are intermediary step between the flower an our conscious 'logging' of having seen it is also (I hope) not in dispute. — Isaac

I think its a logical/linguistic issue. Our (public) use of words derives from our interaction with things in the world that we find ourselves a part of.

There's no dispute that perception involves a physical process that can involve any number of steps. The dispute (if there is one here) is with the suggestion that we're not perceiving the things we ordinarily think we are - that we're instead subject to a "veil of perception" (i.e., that we're either perceiving images or else inferring things via an image).

The ontological commitment seems to be that the proper object is the first outside of our body. We could just as easily say it's the first outside of our conscious awareness. — Isaac

Yes, our perception starts with "medium-sized dry goods", so to speak - the everyday things we observe and interact with.

But perception is not limited to that. Brains aren't outside the body, yet we can open up a skull and see a brain. We can use a microscope to see neurons. And we learn that seeds (which we can also perceive) grow into flowers.

We can also use more abstract language - "image", "representation", "appearance" and so on. But those abstractions depend on the concrete things we perceive and interact with. The "veil of perception" is an alternative conception that breaks that logical dependency.

I don't know if noises are even uniquely associated with environmental or internal state variables in Friston's work. — fdrake

No, as far as I know they're not. That was the point I was trying to make. I was giving an example of an input where the external/internal boundary made no difference in terms of being Markov separated. I could perhaps have used a hidden physiological state instead (might have been less confusing).

the noises aren't kept track of with state variables in the same way as environmental and internal states? They're instead kept track of with their distributional summary characteristic (the precision matrix of their joint distribution). — fdrake

Not necessarily. It depends on the prior priming of backward acting signal suppression as to whether noise is aggregated by some sort of precision matrix into the signal or if it is 'clipped'. We're interested here only in stochastic resonance effects rather than background noise (which can be aggregated without influencing priors. But... that's not really relevant to the point I was making, which could have chosen an example other than neuronal noise.

Any signal which has a threshold potential to meet can be modelled as it's own Markov blanket as all but the proximate contributory signals will be Markov separated. There's no modelling impact of this on a per neuron basis, but with neuronal clusters, their immediate sources have modelling implication, but the fact that those sources too have threshold potentials to meet, themselves from multiple sources, gives us Markov separated internal states.

if you wanted to look at 'the Markov blanket of neuronal noise and environmental + internal states' it seems to me either to use more than one concept of 'state' (one for errors, one for environmental and internals) or... — fdrake

Yeah, this is correct - as far as I know, but with stochastic resonance, it's a reasonable modelling assumption, I think.

Incidentally, this is (according to Friston himself, in a paper I did understand) Friston's preferred model of neuronal noise (in a paper I didn't understand!)

Putting it in less jargony terms, whatever errors we make in perception act upon the synthesis of sensory data rather than acting as their own sensory data. Errors are formed by the coincidence of discrepancies between emerging features of our perceptual landscape, rather than stored as their own form of sensory or environmental data (state). Part of the model are assumptions about how this error behaves in the aggregate. — fdrake

Except for stochastic resonance in non-suppressed signals from forward feeding neural clusters with specifically modelled 'meanings'. assumptions about input states can be detached from external states by the effect of neuronal noise.

If the transition from 9 to 1 could be thought of as a reset, I'd agree with the emphasis, but isn't it more that 9 provides a very strong prior for the next 1? So unless the prior effect's gone away by the time you get to the next 9, it seems to me too artificial to abstract from the process that 9 is the real content of 1. — fdrake

Maybe (see reservations below), but my main dispute with this analysis is that there's no need for us to use the 'real' content at all. My argument is entirely about the pragmatic content of 9 for various models. I've little time for the 'real' anything...only the useful.

That said, I think the fact the 9 informs 1 is not more special a part of the system (it's what I was implying by having my example continue ...8>9>1>2...), as you say...

Maybe a philosophical way of phrasing it, if you've got a chain like that, you can read the arrow as something like '1>2 = 1 informs the content of 2", if 1>2 and 2>3, you'd still have that 1 informed the content of 3 if that relationship is transitive. — fdrake

Yeah. If I could actually draw in these posts I would have gone for a network with several 1s leading to 2 and several 2s leading to 3 etc. but that seemed beyond my mathjax capability. I'm not sure that the introduction of ambiguity affects how we talk about the content of 9 though. I mean 9 itself will be a lose aggregate with fuzzy boundaries no matter how we divide it up. 'Flower' is too (where does the flower end and the atmosphere begin - is the CO2 inside a stomata part of the air or the flower? So I don't necessarily see a problem with us similarly loosely referring to 'all the 8s' informing any given 9, or 'all the 9s' informing any given 1.

In the real world it probably depends upon the weighting of steps — fdrake

Yeah, this is basically the point I'm trying to make. We weigh steps differently. No-one even has a non technical name fo the activity of the retinal ganglia, but the external hidden states we call 'the world' or 'objects' or 'a flower'... we have names for that stage, it's of huge significance to us. What I'm arguing is that the most proximate stage which we weigh heavily enough to name it, conceptualise it, is what we refer to as 'mental image', 'memory', 'concept', 'motive' etc.

Regardless, it doesn't seem to me a valid inference to go from: "X predominantly determined the content of Y's perceptions" to "Y perceived X". Could be made a better inference with a theory of content determination - eg, what makes that inference true? — fdrake

Possibly, but what I'm arguing is that the inference is useful, not that it's true. I'm saying that anything could be substituted for X that's in our model of a causal chain (with the caveat above, that each stage be treated as a loose affiliation of signal, not a deterministic single route), but that not everything should be. Not every X is as useful as every other X, but any X is a valid as any other.

And a fair amount of a human's interaction with its environment is reflexive and explained post hoc. So again, what we call perception has to be tuned to needs. — frank

As above, yes.

Isaac proposes to infer 9 is about 8 from 9 was caused by 8. — Srap Tasmaner

Not infer, declare. It's a choice we make, not a fact we discover.

Thank you for those responses. I don't see anything here to criticise on philosophical, conceptual grounds. It is consistent with what I would call a realist position since it takes for granted that there is stuff that is independent of perception and action. — Banno

Cool. 'Realist' is what I was aiming for!

'Realist' is what I was aiming for! — Isaac

That's what I thought from our previous chats. :up:

It's the way we ordinarily use our words. That thing there that I'm pointing to (cue, a red rose in a vase on the table) is what I understand a red flower to be. I can see it, take it out of the vase, and react if I drop it.

That a scientist could potentially put probes on my brain and view an image of what I'm looking at doesn't change anything about what I'm looking at. — Andrew M

That seems like an odd position to take. It implies that science is a pointless exercise, forever subsumed by whatever it was we 'reckoned' was the case prior to its discoveries. We used to talk as if the sun went around the earth, we talk of sunrise, the 'movement' of the stars. Should we then say that cosmology needs to change how it talks because we had a prior linguistic convention which assumed a geocentric universe?

If we start out with a very direct form of realism (because we have no idea how the brain works), and create linguistic conventions around that, I don't see why it has primacy over any new linguistic conventions arising from knowledge we've gained as to just how much internal states affect our perception and understanding of external objects. Edit - Actually, I'm not even concerned about 'primacy' here. The counter-claim being made is that it's not even 'allowed', let alone, secondary. Secondary I'd accept, even support.

I think its a logical/linguistic issue. Our (public) use of words derives from our interaction with things in the world that we find ourselves a part of. — Andrew M

I don't think they do, at least not exclusively. Our public use of words is derived as much from social beliefs, dynamics and feedback (often chaotic), as it is from the properties of objects.

those abstractions depend on the concrete things we perceive and interact with. — Andrew M

I agree.

The "veil of perception" is an alternative conception that breaks that logical dependency. — Andrew M

I don't see how. They just seem like two models to me. Why does the fact that one of them governs everyday interaction (including interaction with brains, fMRI scans, EEG etc) and one of them govern talk about how minds work mean that one breaks a logical dependency on the other?

If I use an instrument which relies on electricity to investigate electro-magnetism my results are thus constrained. I'm not told "you can discover anything you like, but you cannot change how we think electro-magnetism works because the machine you're using relies on electricity"

Cool. 'Realist' is what I was aiming for! — Isaac

Indirect realist, right?

No, as far as I know they're not. That was the point I was trying to make. I was giving an example of an input where the external/internal boundary made no difference in terms of being Markov separated. I could perhaps have used a hidden physiological state instead (might have been less confusing). — Isaac

I guess I don't see what the point you made is for then. It seems the following are true:
( 1 ) There are states.
( 2 ) Some stuff counts as a state, some stuff doesn't.
( 3 ) It's ambiguous when to count something as a state sometimes.
( 4 ) Neuronal noise doesn't have a state associated with its values (but it does have states associated with its precision matrix).

( 5 ) If you perceive environmental object X, X has external hidden states with it.
( 6 ) Those external hidden states are in direct contact with some internal states.
( 7 ) Direct contact means that the value of one state (external) influences the value of another state (internal) without passing through another state. A->B rather than A->B->C.
( 8 ) That means the Markov blanket of some internal states includes some external states.
( 9 ) A corollary of ( 8 ) is that the Markov blanket of all internal states includes some external states.

I think I can believe those things and draw a direct realist conclusion from Friston's work. Direct in the sense of the contact in ( 8 ), rather than perceiving reality 'unfiltered'. I don't want to conclude that we 'perceive the objects of the external world as they are' from that, I want to conclude that the values of external states actually saturate the perception process. (Box 1 here).

I think I've been playing fast and loose/using the vocabulary a bit wrongly in construing there being internal and external states, since 'internal states' are in the brain, and external states are outside of the brain - but sensory states and action states have connections to both internal and external states. I've been lumping in sensory states with internal states and action states with external states. Hopefully that hasn't done too much damage to what I've said.

Anyway, I don't think it follows from what I said above that "I perceive the state of the external world exactly as it is", just that "I perceive the states of the external world (using some model process)" and "That modelling process is in direct contact with the external world".

Yeah, this is basically the point I'm trying to make. We weigh steps differently. No-one even has a non technical name fo the activity of the retinal ganglia, but the external hidden states we call 'the world' or 'objects' or 'a flower'... we have names for that stage, it's of huge significance to us. What I'm arguing is that the most proximate stage which we weigh heavily enough to name it, conceptualise it, is what we refer to as 'mental image', 'memory', 'concept', 'motive' etc. — Isaac

Can you flesh out why from that it follows that we perceive mental images, memories, concepts etc? I don't see the connection.

Can't help you with that, Harry. Unlike you (seem to be), I'm neither a subjectivist nor a introspection illusionist. — 180 Proof

"The introspection illusion is a cognitive bias in which people wrongly think they have direct insight into the origins of their mental states, while treating others' introspections as unreliable."
-Wikipedia

If you're claiming that brains are the origin of experiences and asserting that all others who disagree are unreliable, then you're as much a introspective illusionist as anyone.

who [...] has had a clearer understanding of what mind is and its relationship to brains [...] why what someone else thinks about this relationship could be better than mine or anyone else's? — Harry Hindu

Call it what you will, Harry, but your "informationalist" position as expressed here suggests introspective illusionism (i.e. naive platonism) to me. — 180 Proof

Don't know how you interpreted skepticism of other people's introspective illusions as me being an introspective illusionist myself. Wouldn't that mean you're one too?

Is your experience of brains an illusion? How else do you or anyone know anything about brains if not by experience?

If you're claiming that brains are the origin of experiences and asserting that all others who disagree are unreliable, — Harry Hindu

Well, i'm doing neother ... :roll:

Don't know how you interpreted skepticism of other people's introspective illusions ...

I didn't, so ... strawman ergo another evasive non sequitur.

The differences of opinion concerning naive realism, direct realism, indirect realism and so on gain traction from failure to adequately set out the various claims.

I'd like to again draw attention to the intentionalist approach to the problem of perception. intentionalist approaches draw on the intentional character of perception: seeing things as a cup or as a tree or as a person. The interesting thing here is that this is quite similar to the Baysian modelling described by @Isaac and others. So for example if what you see fits in with what in previous encounters you have treated as a cup; treat it as a red cup but modify your model if needed. "Treat it as a red cup" is adopting an intentional attitude towards the cup.

Intentinalist explanations potentially show how a neurological account and an intentional folk-account can both be true.

Intentionalism is a form of direct realism. While other direct realists might say one sees a cup, an intentionalist would more accurately say that one sees it as a cup.

More duck-rabbits, of course. And this needs filling out. But it fits fairly neatly with the neuroscience, avoids the silliness of qualia and shows that we refer to flowers and not perceptions-of-flowers.

seeing things as a cup or as a tree or as a person. The interesting thing here is that this is quite similar to the Baysian modelling described by Isaac and others. So for example if what you see fits in with what in previous encounters you have treated as a cup; treat it as a red cup but modify your model if needed. "Treat it as a red cup" is adopting an intentional attitude towards the cup. — Banno

The concern some might associate with different kinds of realism is whether the things we encounter have mind-independent status. If one calls the encountered thing an apple, it would appear you're saying the apple is not mind-independent. What we call this view is really of secondary importance philosophically speaking.

The interesting thing here is that this is quite similar to the Baysian modelling described by Isaac and others. — Banno

Science is interested in how a collection of firing neurons comes to be interpreted as a single thing. It may be that scientists need to look at inclinations to "treat" a stimulus as distinct objects. I don't know. It seems that you're gliding over the scientific question to talk about language use?

an intentionalist would more accurately say that one sees it as a cup. — Banno

This implies a dualism. "It" and "the cup". What means "one sees"? Already in saying "one sees" are qualia involved. What do you even mean by qualia and saying that they are silly? Do you live in an extra-quale reality of objective, solid material, existing in an extramundane world, seen from which they are silly stuff while you unconsciously use them too when looking at the world from a high tower made of some stuff inaccessible to qualia?

If one calls the encountered thing an apple, it would appear you're saying the apple is not mind-independent. — frank

Why do you think it follows that if we call something an apple, it follows that we would thereby be committed to saying it is mind-dependent?

What we call this view is really of secondary importance philosophically speaking. — frank

I quite agree. I would add that what we are engaged in, at least to begin with, is finding an effective grammar, a way of talking.

The concern some might associate with different kinds of realism is whether the things we encounter have mind-independent status. — frank

All realisms take it that the stuff around us is mind-independent, don't you agree?

If one calls the encountered thing an apple, it would appear you're saying the apple is not mind-independent. — frank

I don't see why. Calling it an apple is mind-dependent. But the apple isn't. It doesn't need a mind to be what it is.

The stuff around us is always already interpreted; that is, we always take an intentional attitude towards it; we see this as an apple, that as a dish, that as a knife. But if you call the apple an orange, or see it as an orange,that makes no difference to the apple.

What do you even mean by qualia and saying that they are silly? — AgentTangarine

Read the Dennett article - even if only the first few paragraphs. Get back to me.

I don't see why. — Banno

I don't follow your reasoning either, but we both agree with Isaac: the objects we respond and react to are internal.

the objects we respond and react to are internal. — frank

I think that a mischaracterisation*. But @Isaac will speak for himself. I've found that he is unusually careful with his choice of words and found that on examination his view is a pretty straight forward realism.

As for you and I, we might usefully finesse the discussion. You said

The concern some might associate with different kinds of realism is whether the things we encounter have mind-independent status. — frank

Can you confirm that realism for you is the view expressed in SEP, that:

a, b, and c and so on exist, and the fact that they exist and have properties such as F-ness, G-ness, and H-ness is (apart from mundane empirical dependencies of the sort sometimes encountered in everyday life) independent of anyone’s beliefs, linguistic practices, conceptual schemes, and so on.

Just for the sake of checking that we agree on at least this basic use.


Then:

If one calls the encountered thing an apple, it would appear you're saying the apple is not mind-independent. — frank

I agree with you that calling the apple "an apple" requires a mind. Can you explain how this is incompatible with the use of "realism" agreed above?

* Edit: That mischaracterisation is that Isaac might well agree that the objects we respond and react to are internal but be puzzled why others might think that it did not involve flowers.

A more straightforward question relating to the OP: Do you see a place for some notion like "qualia" in your work?

Putting you on the spot...

Just bring honest here: I don't see you as coming to the discussion in good faith, so I'm not interested in clarifying any statements made by either of us.

Yeah, you've avoided any criticism using that tactic a few times. If you are not going to engage productively, don't bother to respond to my posts, and I'll go back to ignoring you.