Dennett has his own schematic for sensory processing in the brain. Are you familiar with it? What's your assessment of it? — frank

I'm passing familiar, but only in that it's been brought up as being compatible with a Bayesian inference model (which is my preferred model). If you have any sources, they might be relevant here?

it might be simplest to call what System 1 gets up to "caused" but a lot of those habitual responses are caused in a way that will pass muster and reflect repeated earlier effortful examination of things by System 2, so we might as well call those causes "reasons" in the non-causal sense too, the sense in which "justification" is not a synonym for "rationalization".

If challenged, a person might engage System 2 and begin a process of assembling the evidence they are comfortable claiming underlies a belief; we can call that "rationalization" in a wide sense, allowing that we might approve or disapprove of their claims for evidentiary import, etc. (Sellars seems more or less to claim that saying "I know" just signals we are now playing a language-game that requires everyone to put on their System 2 hats -- it puts a claim "in the space of reasons".) Or they might refuse to engage System 2 (wait -- is that even possible??) with a bare "I just know" and philosophers tend to frown upon that. — Srap Tasmaner

This is very close to the way I think about knowledge (even 'truth', much to @banno's chagrin). As a signifying word which tells us we're playing a different game. In my experience, the game thus signified is more a social one than the private change of systems you suggest here (though I do like the fact that your model gives us the binomial distinction we're looking for).

What seems, again, in my experience, to be signified by a shift to the term 'knowledge' is that there will be agreement among others in one's social group. I 'know' the earth is round (anyone in my social group will agree). I 'believe' it's raining (someone who's actually outside might disagree). I know at a first read, most will baulk at this "I don't care what other people think" but the key is understanding that we have imaginary social groups to which we wish to belong, as well as actual groups. None of is is to undermine the processes we use to make this assessment.

The evidence for all this (as you so rightly were about to ask), is weak. It has a rich heritage though (Asch conformity), and a modern update (engagement of social status brain regions during both knowledge and evidence assessment), together with a host of studies inbetween, but you'll find plenty to the contrary as well. (Hey, my belief counts as knowledge, I've read most of the studies to the contrary!).

So there are two kinds of criticisms that can sometimes and sometimes not be made about the same beliefs:

formed with minimal "input" from the environment and considerable input from your other beliefs or "gut reactions";
"insulated" or "protected" from possible revision.


Philosophers don't like either of these but will let the first slide so long as you are open to revision; the second is more or less sinful. Are there good general-purpose ways of talking about these things? — Srap Tasmaner

I think this matters a lot, but again, as I mentioned earlier, I think it's important to distinguish simply being open to revision (our beliefs are in a constant state of being revised), and being open to revision via an approved method, much like the engagement of system 2, you mentioned. How much this relates to the ontological status of 'knowledge' as opposed to just what constitutes good habits (in the Ramseyan sense), I'm not sure.

It’s the “checking” part that makes the difference, between acting like the observations you happen to have made so are plenty (and possibly even being averse to makings further observations that might compel you to change your mind), and actively seeking out more observations to make sure that they continue the pattern. — Pfhorrest

You might instantly form a belief that the place is dangerous from that. But do you know it is dangerous? This is where a person has to actively and consciously examine their beliefs. — Philosophim

This all makes sense as far as identifying a process (and a worthy one at that) is concerned, but there's still this odd ontological leap, like there are two kinds of thought 'knowledge' and 'belief'. I don't see how a scalar process can result in a binomial distinction. Are you (either of you) perhaps suggesting the the moment a sub-conscious belief is consciously checked (and passes) it becomes knowledge? Or is some degree, and method, of checking required?

Either way, how do you avoid the problem I mentioned at the beginning that one cannot distinguish the presence/absence of evidence from unchecked belief?

We cannot say that our belief in A is justified by the deductive argument 'If B then A, B therefore A' because our belief that B might be what is at fault, or our belief that 'if B then A'. — Isaac

Basically the presence and decisiveness of evidence (the consequence of this 'checking procedure) is itself a belief.

So you have a belief that A. You check it and find what seems to you like evidence to the contrary (call that a belief that B). The belief that A and the belief that B are just two beliefs. One must obviously decide which counters which, but the ontological status of A hasn't changed, it's still just a belief, despite now having a contrary belief associated with it.

It seems as if you want to set up a hierarchy where initial beliefs are somehow considered suspicious, but subsequent beliefs about evidence for/against them can be treated as practically objective fact, and I don't see any reason why you'd want to do that.

That's not contrary to my views at all. (I've just been saying in the past few posts that a "belief" as I mean it is formed from a "perception", which in turn is exactly some interpretation of evidence). I think maybe you're reading in more than I intend to say. — Pfhorrest

Ah, then I've gotten confused. You said earlier that

I'm counting being uninterested in checking your beliefs against the senses as "ignoring empirical evidence".

Plenty of people ignore claims that there is empirical evidence to the contrary of their beliefs, rather than actually check if those claims pan out. That's being unresponsive to reality. — Pfhorrest

It's this that's caused the confusion. How can it be that people are "uninterested in checking your beliefs against the senses" and yet at the same time you acknowledge that all belief are the result of interpretation of input from the senses?

it's still in a feedback relationship with the lower level models - promoting certain actions, certain adjustments, the formation of certain perceptual features, the exploration of our environment to form new task relevant features etc. It doesn't look like there's a distinct "submission" operation to consciousness, it's more that the apportioning of conscious awareness is interweaved with a concentration of bodily effort and attention relative to a task. — fdrake

@Kenosha Kid. It looks like @fdrake has already said what I just answered to you - I should really read the whole thread before replying.

Another way of making the point: that conscious awareness "coming online as it is" isn't in a temporal order with perceptual feature formation (this, then that), it's part of the hierarchical order within perceptual feature formation (this is an upper part of that). If the time part is weird (since the higher order parts time lag the lower parts); the apportioning of conscious awareness is a procedural component (systemic part) of perceptual feature formation - rather than a distinct procedure which the results of perceptual feature formation output to. — fdrake

Yes, an important point to make in understanding this. The order in which our sensory signals are received and processed is not that in which we consciously report them as having done so (even to other systems in the brain). so , a classic experiment with this is the advancing circle. The subject is shown a series of flashing dots with circles around them, they flash in sequence to create the illusion of movement. The dot and the circle are in exactly the same place (one around the other) and the brain obviously see them at the same time, but depending on the direction of the induced movement you'll see the circle as being either slightly ahead or slightly behind the dot. You brain re-arranges the timing of the signals to suit what it thinks has most likely happened. a more mundane example is switching a light switch on. You see the light before you feel the switch move. Your conscious knowledge of how switches and lights work re-arranges the signals temporally so that they seem to line up with the order we model them as being in.

It would be as if the effectiveness of the hierarchical structure in place entitled subsystems operating under top-down constraints to take credit for being thus constrained and count the very constraints they work under as their value added, and all to set up entitling awareness to claim credit for much more than just playing its allotted role. — Srap Tasmaner

Yes, that's very much the way I see it. 'The way the apple tasted to me' is simply not something that one has access to at the same level of consciousness as the '...to me' part operates. It think intuition pump 7 illustrates this well. Chase has no more access to the 'taste' of coffee he's processing than the neuroscientist might. The cascade of mental events that his chemoreceptive system started when he sipped coffee has simply had more consequences than he can later gather up and report as being 'the taste of the coffee' in anything other than an arbitrary and constructed manner, and it's only later that ...to me' even enters the picture. For a start, the coffee would undoubtedly signal one response to maybe' sweetness' in one part of the olfactory system, and another to maybe 'sourness' in the occipital system in response to say labelling (the label 'Bitter Coffee' for example). so which one would be 'the way it tasted to me''? Yeah, it makes no sense at all at that level.

The best I think we've got is 'the way I later chose to report the coffee tasted to me'.

The thought experiment is to help you understand the abstract context of the above. If I have an unexamined belief (which has nothing to do with the technical neurological process of how that belief was formed) and it just so happens to be right, was my unexamined belief knowledge? — Philosophim

You're mixing up two different things and I'm getting confused. There's this issue of an 'unexamined' belief as in

The abstract point I'm making is we can have beliefs that are examined, and beliefs that are unexamined. Is an unexamined belief knowledge? — Philosophim

Then there's this issue of a belief without any basis, as in

the belief I mentioned did not tie to anything that would indicate Joe dated a woman last night. — Philosophim

The point I was making as counter to @Pfhorrest's argument here is that it is impossible to generate a belief which is not based on some interpretation of the evidence (input from the outside world). It just neurologically can't be done). So everyone already has a justification for all of their beliefs, the justification is whatever external inputs caused it.

The issue of whether a belief is 'examined' is another matter - the effort one puts in to gather even more external data relevant to the belief. Here the issue is scaler and the answer can be none, but in fifteen years of working on beliefs I've yet to see any evidence of a single belief which is 'unexamined' in this sense. People are quite keen on having their beliefs provide them with useful predictions and to do so there cannot be significant clashes with whatever objective states of the world there are.

I'm not being 'purposefully obtuse' here. It's the crux of the problem with these kinds of approaches. They disguise ideology as method. To create this categorisation of predictive though as 'beliefs/knowledge based on an ideologically defined method of checking them is fine when it's upfront about it (I personally use that method of checking my beliefs, I think it works best), but I oppose the suggestion that it is ontologically relevant, that it somehow describes the objective difference between two types of predictive thought. It doesn't There are only beliefs and there are various methods for checking these beliefs. We can say whether a belief has been more or less well-checked, but only after specifying the method of checking we're measuring that against, and the choice of method is underdetermined.

What measure of consciousness are you using then?


We were talking about definitions, not measures. — RogueAI

You said "Reporting on mental activity isn't even a necessary condition for consciousness, let alone a sufficient one.". To make such a claim requires (as far as I can tell) an empirical data-set which includes people being conscious but without any reporting activity going on. In order to acquire that data-set you'd need a measure of consciousness so that you can tell the people with no reporting activity are nonetheless conscious. I just wanted to get clear what that measure is you're using, otherwise I can't have any real understanding of what you're saying.

I'm not saying first-person subjective experience is a sufficient condition for a definition of consciousness, but it is a necessary condition. Agreed? — RogueAI

I don't really know what 'first-person subjective experience' is in this context.

What action best describes scrolling down this page and coming across the picture of the two cars and identifying them as cars? I suppose, in a page of mostly text, an image is surprising and, when we become conscious of something surprising, the instinct is to identify it? — Kenosha Kid

Most of the time it seems to be planning. We can recognise planning quite well as systems specific to imagining a scenario seem to be associated with the same neural pathways as would be used in the actual scenario. Not that this is relevant here, but this happens to be my personal favourite 'consciousness' explainer. The reason why we feel like we're living some story is because we're constantly scenario-planning and to do that we have to integrate our current environment into the 'the story so far...' section of the film. What seems to us as 'being instantly aware' of the scene we're in is really a set of imagined actions taken within that scene. So, to use your example of seeing a picture of a car on a page of text, what you later report as 'instant identification' is really instant simultaneous planning of potential scenarios - 'What word would I reach for if I had to communicate that object, what body positions would I take up if I had to interact with that object, what endocrine response would I need if that object came closer, what other systems would I need to bring online if it moved, made a sound...and most importantly of all...what sensory input am I likely to get next as a result of the shapes in this one.

There's quite a lot of evidence for this theory, in the form of seeing these areas light up in response. Interesting aside, there's some consideration that Schizophrenia might be caused by something as simple as getting the post hoc story of these mental events the wrong way round (so that the 'what if...' preparations seem to precede the sensory cause, and therapy aimed at this has has some success.

Anyway, I'm rambling. The point is that there's no reason to think that 'identifying the object' is an event in any singular manner. At best it's a collection of predictions about what might happen next and how you might interact with it in a whole bunch of different brain areas. We only really bring these together for the purpose of the next set of predictions the 'story so far...'

It think this is the main point of attack on qualia, even in the most public, effable, and extrinsic use. There really doesn't seem to be a single act of 'recognition' at all. I do see what you're trying to get at (I think) by saying that, the way it seems to you (recognition) has to still be something and we can use that, explain it etc just like any other mental phenomena. I think that's true, but then 'qualia' would mean something more like 'plot device' and I really don't think that's where qualists wanted to go with it.

I can imagine that, as infants, we might have suffered a period of time in which the brain had to learn how to do this. — Kenosha Kid

Another fascinating (to me) aside, but I must stop getting sidetracked. Have a look at this paper, if you fancy, it's really interesting.

So there's a sense, then, in which our conscious perceptions are being assessed by our unconscious brains in order to fire/learn correction processes. Is it your assessment that this is done for the purpose of improving our awareness, or is that just a nice side effect? — Kenosha Kid

The theory goes that surprise in costly, so brains have evolved to minimise it. There's even a theory that surprise opposes self-constitution and so life itself evolved to minimise it!

Not that this thread or Dennett's article is about judicious use of labels. Ultimately it doesn't matter whether we call them qualia or something else, so long as it's clear that 'qualia do not exist' means 'ineffable, intrinsic, private and immediate properties of consciousness do not exist' and not 'properties of consciousness do not exist'. — Kenosha Kid

I could almost get behind that but would have to add that consciousness may not be the kind of thing that has the kind of properties we're talking about. Consciousness seems to be a set of processes, story building. If that were true, and we were to say "one of the properties of consciousness is...", I'm not sure how to complete that sentence without just offering the definition itself. Surely it should be "one of the results of consciousness is..."

the belief I mentioned did not tie to anything that would indicate Joe dated a woman last night. — Philosophim

In your thought experiment, maybe not, but what I'm saying is that such a situation is neurologically impossible. No matter how much you insist you did, there is no know (or even plausible) mechanism by which a belief can be formed without the sensory, or interoceptive inputs to form it.

That's an incomplete definition of consciousness. Reporting on mental activity isn't even a necessary condition for consciousness, let alone a sufficient one. — RogueAI

What measure of consciousness are you using then?

There is therefore something objective and operational about colours as we perceive them. — Olivier5

I think this conversation is on the wrong thread, but briefly - there's a substantial difference between "something objective and operational about colours as we perceive them" and claiming there's such a thing as the subjective experience of 'blueness'.

Consciousness is just the tendency to be able to report on mental activity and it's caused by the neurons which produce language, movement and other awareness-mediated responses being stimulated by the neurons constituting the processing of sensory inputs to which that awareness relates. — Isaac


As the subject, this is my tendency to be capable of reporting -- but not just on any mental activity, on my mental activity. I'm wondering if there are pre-utterance steps where some subsystem perhaps tags the analysis and speech prep being done as "me related", or if there aren't, and why we need or don't need such steps. — Srap Tasmaner

Self-identity is a whole other massive topic, although it's related here, but like saying "I wonder what's in that wardrobe, Edmund". Broadly the idea of a me/other divide is important at a somatosensory level in the assessment of sensory inputs (that's my hand waiving about out there), but not at the reporting level (we can really easily insert ourselves into false memories, have out of body experiences, experience the emotions of others) here tagging stuff as me-related doesn't seem nearly so 'sticky' and is almost certain done post hoc (when we try to analyse past responses from memory). Where it's important again is in somatic responses, we need to know I moved my arm. Somatoparaphrenia is the condition studied to give insight here (where people think part of their body does not belong to them) and Alien Hand Syndrome (where people seem to assign the movements of their limbs to another identity). These seem at first glance to be perhaps similar, but the differences are revealing (I know I keep banging this drum, but these things do get misrepresented so I'm going to say it gain - very small sample size, very large pinch of salt). The former condition seems to be the result of deeper neural processing of sensory inputs not being sent on to neocortical areas - ie we don't get to assess the 'meaning' of the signals. This could be taken as evidence that the identification of 'self' takes place at higher levels than the basic somatosensory system. The latter, however - more on topic here- seems related to occipital cortex damage as simple as -"It doesn't look like my arm, so maybe it isn't may arm".

Interestingly (to me anyway), and this relates to what I was talking to @Kenosha Kid about, patients with somatoparaphrenia and Alien Hand will come up with and have deep beliefs about, all sorts of plausible real-world stories to explain their situation. It's something very common in most anosognosias, they're accompanied by an absolute conviction in the story explaining it. We need, it seems, some coherent story at a very high level (meaning it combines lots of input data) which explains the sum total of our sensory inputs.

I only skimmed instead of rereading, but it seems to me Dennett might have added here that I am generally expected to know non-inferentially, and perhaps infallibly, whose qualia are rattling around in my consciousness, and to know that they are mine rather than yours. — Srap Tasmaner

As with the allocation of responses to objects, I think any allocation of ownership to responses would be mixed, and mostly post hoc (with the possible exception of immediate bodily responses - like catching a a ball) so yeah, Dennet could well have added that to the list of things to be thrown out.

If I may summarise, then, the conscious perception of my field may include something caused by a car without the car 'tag' (recognition of car object with or without dorsal data), then moments later updated with that tag. So I consciously see the light caused by the car before I see the car. — Kenosha Kid

Yep, that's about it.

to clarify, it's not a conscious decision to identify a car, right? Whenever the car recognition output is presented for conscious consideration, it's not doing so because I'm studying a patch of light and trying to figure out what it is. This is all going on in the background. — Kenosha Kid

I see what you're getting at, and generally, yes, but I'm going to be careful unless I get called out on it later - we need to take care when talking about conscious awarenes. It's not a binomial thing, you are more or less aware of signals depending on your level of attention which varies. Here, I think, is an important issue which might touch on some of the stuff you're trying to work out with regards to the 'presentation' of the image as you put it. The 'identification of 'car' (to the extent that it happens that way - remember it's just one model with a very small sample size study to back it up) happens subconsciously. What you're typically aware of depends on what you need that identification for - Are you about to say the word that goes with the object, are you choosing the right object from others, are about to interact with it...Whichever following action requires you to identify it as a 'car' will determine how the fact of that identification reaches your awareness, if it does at all.

Am I right in saying that, as you describe it, data from our conscious perception is fed back into these myriad cascades and may affect (or indeed effect) some of these unconscious processes — Kenosha Kid

Yes, cortices have backward acting neural connections whose job it is to suppress non-matching signals and they're often (but not exclusively) informed by the consequences of conscious recognition. A classic example is sensory priming where you are exposed to a distorted sound/picture/smell, you're then exposed to the undistorted version (which you interpret the meaning of at least partly consciously), then when you next are exposed to the distorted version it seems much clearer. Your higher models of what might be being said are suppressing the signals from your primary sensory areas which have all the 'noise' based on your conscious awareness of what the similar experience just resulted in.

All this is then adjusted post hoc to make a coherent story.

Do these processes rely on this, or can we recognise objects just based on pre-processed data? This is again going back to the idea of sensory data categorised as unimportant, such as the sound of a car engine on a busy Manhattan block. — Kenosha Kid

I'm not sure what you mean by 'pre-processed'. I'm going to take a stab at it assuming you mean to ask if conscious awareness is a necessary feature of bacwkard acting signals. No, but it is usually involved at the higher model levels which, of course, will have an influence over all the models below them. Dream study is probably the best case study for this (again huge pinch of salt required here, very difficult to study), but it seems likely, because of the way memory storage works, that basic object recognition, including limited function, must be able to take place without conscious awareness because they do so during deep sleep dreaming, but appropriate response, form detail and complex function do require some conscious awareness because they only seem to enter stage during REM sleep where you're semi-conscious.

What I meant about not throwing the baby out with the bathwater is that there still remain objects of subjective experience, such that I can see a car without consciously determining it to be a car (even if 1 ms ago I didn't see a car), and that this object is private (internal processing from my raw sensory input to instantaneous apprehension by me) and immediate (I see car as car object is presented to me, which may be some while after I see light from car), but not intrinsic or ineffable, and that these objects and the processes that yield them (e.g. neuron that recognises car) underlie our pre-theoretical conceptions of what theorists call qualia. — Kenosha Kid

Yeah, I think I agree with you there - but as I think @fdrake is getting at (though I've only skimmed the other responses - not much time this morning, sorry) that there could meaningfully be identified an object of conscious experience leaves absolutely nothing for Qualists. I don't want to be pedantic, just to clarify my position, but I don't think it would be be a case of throwing the baby out with the bathwater to say that Qualia are completely useless and have no place in studies of perception. The objects of perception to which qulaists would like to attach subjective properties, I think are not in danger should we reject qualia wholesale. That said, if we want to reserve the possibility of studying, say, the taste of coffee, We do need some fuzzy-edged set of responses we might associate most with it. I'm not opposed to that kind of grouping. Maybe these could be the new 'qualia', but I think, given the sullied history of the term, we'd better reach for something else.

And yet we all can agree than certain cars are blue, and others not. — Olivier5

Indeed. So we have a choice. Discard all neurological evidence and pretend things are the way the seem at first blush to be...or...dive in with curiosity to find out how things might actually be, even if the prevalent theories are counter-intuitive.

What seems to happen with consciousness, perception, free-will..basically anywhere where neuroscience might have some input, is that the response is to vehemently assume our first blush reckoning about it must be right and then filter all the data through that. Can't see much point in that approach myself, but each to their own I suppose.

there still exists an object of subjective experience as tagged by its outline. — Kenosha Kid

I agree here insofar as you're saying that 'whatever we say of the realised car we can say of the unrealised grey blob', but that wasn't my intended target (sorry if my writing's not clear - in my defense I'm recently writing on a phone on the train and so I don't do as much overarching editing as I should). I brought up the {could-be-a-car} just to make it easier to imagine the conscious results cascading out of a sensory processing event. I've probably just ended up confusing matters. If you understood what I was saying about the conscious results which cascade out from the other firings extrinsic to the ventral pathway, then we needn't even worry about this blob/car distinction. The idea is that what we can say of one is what we can say of another - which, in terms of the subjective experience associated with either, is nothing much.

Here the 'car' quale means nothing more than that the image is presented for conscious appraisal with the 'car' tag, i.e. the 'car' neuron having fired with some degree of certainty. — Kenosha Kid

I may be reading more into 'with' than you intended, but to be clear about what's going on neurologically, the image is being 'presented' without the car tag. Much of what's going on, even consciously, is going on before the car tag. You later (perhaps even seconds later) re-tell the story as happening in a better order (saw a thing->worked out it was a car->thought 'I could drive that'). Expermenting on this is really difficult because of the time lag in fMRI and the non-specificity of EEG and the like, so take this with the very large pinch of salt attached to small sample sizes (neurosurgery patients). Nonetheless it seems as if it's very difficult to match your subjective report with the actual process, it seems we all lie about what happened.

I'm not sure what responses you mean. — Kenosha Kid

Again, not sure what all these goings on means. — Kenosha Kid

OK. Maybe I'm attacking the wrong target. To me qualists want to say that there is something it's like to experience red, there's the 'redness' experience, or the 'car' experience. To do this, they invoke the 'way it feels' in response to sensing 'red', or 'a car'. What I'm trying to show is that we cannot, even in principle, distinguish the 'way it feels' in response to red, or cars, from 'the way it feels' just right now in general. The cascade of neural responses is continuous, there's no break in higher level backward suppression at the point of seeing red, so the conscious 'feelings' are unattached. We attach them later in retrospect.

So however it feels to see red is however I decide it feels. Some responses I might attach to seeing red, others I might decide had nothing to do with the colour but were a response to something else, or something I just happened to be feeling at the time. And...what's crucial, picking up on Dennet's point in intuition pump 8, is that I might be wrong in a way that a third party could theoretically determine. I don't have privileged access here, there's no difference between first-person and third-person guesses.

This is really just by way of explaining what I mean, I'm not necessarily assuming you're suggesting otherwise.

Yes, I agree, this is Dennett's rejection of the intrinsic nature of qualia. I'm with you, and Dennett, on this, and I agree that this is what philosophers usually think of as qualia, incorrectly. — Kenosha Kid

Cool. I think we're much on the same page then, which makes it interesting to tease out any differences.

Not sure of the extent to which my reply here helps, but when you've had a chance to have a read of that...

It's really difficult to stick completely to exegesis when so much of the question of what Dennet might have been getting at requires some external 'rounding out' of what the issues are, so I sympathise with your posting dilemma. I'm happy to see if we can continue the subject matter here and if everyone gets annoyed about that then we'll just have to nurse our slapped wrists.

if the error arises ultimately from treating experience as an entity which bears properties, how is it any better to treat it as an entity which bears relations (or other high order predicates)? — fdrake

Indeed. I hadn't noticed until you quoted it, but I think there is a tension there which I'd certainly not side with Dennet on, if that's what he's implying, but I'm not sure. I read that quote as saying that the one thing we can be sure of is that the expected response to sipping coffee (action, not experience of) has changed. This is not memory reliant because it's not internal - "Didn't I used to enthusiastically ask for a second cup?". So something's changed. We can drop 'sweetness' out of it altogether if we like. Event A used to lead to event B, now it doesn't so some A-B causal link is different. Without getting into Humean induction issues, I think we can say that much.

The salient point in the devil's advocate is that the "fundamental error" seems to be claiming that or acting as if we experience experiential entities (which have or may be experiential properties), rather than experience itself being a mode of our interaction with entities.

That looks to me one way of fleshing out it being okay to say "The coffee tasted sweet today" but not "My subjective experience of today's coffee was partially constituted by a quale of sweetness". — fdrake

Yeah. My memory of reading the text before looking at it again for this thread was actually just something like - "we can't tell if our changed responses to an event are because the interpretation of the sense data have changed or because our memory-associated judgement of them has changed. As such we don't have any privileged access to these 'qualia' than the neuroscientists do, let's drop the whole term" I'd actually forgotten all the more nuanced arguments leading up to and stemming from that, so it's certainly what I see as the central argument. If there's nothing privileged about our access, then there's nothing first-person and the whole matter might as well be discussed in the third-person terms we already have.

it's always good to hear an expert expunge. — Kenosha Kid

I should then perhaps clarify my expertise is actually in belief, decision-making, uncertainty etc, this stuff about perception is something of a sideline I got into when working with a colleague in cognitive psychology so it's a) second hand, and b) all one preferred interpretation, I'm sure there are others. That said, it'll be interesting to try and flesh out some of the issues.

I'm not sure whether you're saying that the recognition of the car is part of the experience I am conscious of, which is also what I'm saying, or whether we consciously recognise the car, which flies in the face of my experience, and also seems to contradict the idea that the brain is adept at filtering out irrelevant sensory data that we are, consequently, unaware of (e.g. the sound of a car engine at night after living a month in Manhatten, versus the sound of a gunshot). — Kenosha Kid

The former. Although it is important to note the cases you alluded to where we're conscious of our processing ("is that a car over there?") because in these cases we're still having an experience - so what is it an experience of? We haven't identified the object yet. Are we experiencing the quale of {some vague grey shape in the distance that we can't quite make out}? Possibly, but 1) that rather detaches quale form the object of experience, and 2) deciding what the thing is is definitely part of the experience, so what happens when we realise it's a car?

Neurologically (according to hierarchy theory) we have a single neuron which will eventually light up (and start it's chain of responses) to the recognition of a particular object. You have a neuron for me (which is delightful) and it initiates a chain of responses every time it recognises me (whether by name, or prose style, or the blue square that is my avatar). Proof of this is quite surprising (summary here). Anyway. The point is that this neuron is triggered after a long chain of neurons all of which have in turn triggered a number of other neurons. (Imagine each neuron in the chain has, say, five exit pathways, only one of which goes on toward the 'car' neuron). The cascade of effects triggered by your interaction with that picture (and the environment you're in at the time, and any other neural processes which were half-way complete when they were interrupted by seeing that picture) will have, by now, had consequences, other than the triggering of your 'car' neuron many of which you could be consciously aware of.

This process is only noticeable when you're not sure if it's a car or not (hence the introduction of cases where you're conscious of that uncertainty), but it's happening, lightening fast, even in cases where your post hoc story is "I saw the car then all these other responses followed". That, provably, is not what really happened.

Now I agree you could say "All of this goings on were associated with what I finally decided was a car", but they weren't really. As I said earlier, some of them were associated with other sensory inputs, but some of then (and I think this is the most important bit) were associated with neurological processes which hadn't yet finished which had nothing to do with the 'scene' at all. It just seems unreflective of what's really going on to say anything about your subjective experience 'of the car'.

The left car is blue. I am conscious of it being blue. I am not conscious of figuring out that it's blue: it's blueness is presented to my consciousness. — Kenosha Kid

Basically, as far as we can tell from the studies that have been done on various forms of achromatopsia, it never is. At no point in time can we trace anything like the 'blueness' of the car being presented as an property to your conscious awareness. In fact all the evidence seems to point to there being nothing but a series of responses to the final object, of which it's colour is only one aspect. So, for example one type of achromatopsia might present with an inability to name colours, another with an inability to distinguish them, another with an inability to respond to them, (Dennet goes through some of these in the Qualia article, but not in much depth) all separate and functioning in other aspects. someone has even been reported to have difficulty with yellow objects of certain shape, but not yellow 2d images. Basically, what seems like a recognition of 'blueness' is actually an object specific tendency to respond. If someone asks, what colour is that car, you'll have a tendency to seek the word 'blue', if someone asks you to pick the blue car, you'll arrange the necessary spatio-motor response, but these are separate systems and can (in lesion studies) be switched off separately. There's no 'awareness of blueness' - at least as far as neuroscience can tell.

I assume you mean that the timescales involved in consciously working stuff out is much slower than the timescales of photons-hitting-retina to conscious-of-image. It can't be too much later. I have present experience for a reason: present problems require present solutions. — Kenosha Kid

That's tight, but the timescales change all the time. When you first see the car you might have one post hoc story which is almost coincident with the process of the retinal signal to the object recognition and sensori-motor responses. But seconds later you'll have a slightly different post hoc story, minutes later another one, until (as will be familiar to us all) years later you have a totally different story of how you felt), again, we could call these stories qualia, but since they are in a constant state of flux, it seems incredibly difficult to get any useful function from doing so - "Which qualia ar you referring to? The one just now...or now...or now..."

I hope I haven't missed the point you're getting at completely here!

I've formed beliefs that just sprang up out of nowhere. — Philosophim

The question was how do you imagine that these beliefs formed. A belief, neurologically, is a very complex set of neural connections which lead to a tendency to act a certain way. Normally these get set by repeated triggering from sensory inputs and responses. What I'm struggling to see on your account, is how you see such a complex array of network connections as a 'belief' just randomly occurring without being prompted to do so by stimulus from sensory inputs?

OK, what is your explanation for how non-conscious stuff, when assembled the right way, can produce consciousness? Because that seems like magic to me — RogueAI

Consciousness is just the tendency to be able to report on mental activity and it's caused by the neurons which produce language, movement and other awareness-mediated responses being stimulated by the neurons constituting the processing of sensory inputs to which that awareness relates. I'm genuinely dumbfounded as to how or why anyone finds this in the least bit difficult to imagine.

If those things were impactful on the coffee experience (and we know they can be), the sensation of sweetness could not be modelled accurately as a unary predicate/property. There's just no place in a logical property for more than one term. That is to say, it's a higher order predicate of those things - at least a relation. — fdrake

Exactly. This is what I was trying to get at in my reply to @Kenosha Kid earlier from a neurological perspective. 'Sweetness' to the extent we can even identity the sensation is way down the line from the coffee hitting your taste buds. A whole ton of stuff has got involved before then including, crucially, stuff that isn't even part of the event right now. Which prevents the qualist from saying the 'sweetness' is a property of the whole event. A property of your entire life and the environment you've interacted with up to this point, maybe, but that's definitely not on table.

I think it's more likely to mean that intellectual act I did when talking about "the sweetness of the coffee I had today", fixing some aspect of a memory using introspection, can all too readily produce unrealistic accounts of the thing in question. The error being that there was some sort of experiential entity which bore that property, contrasted to the fact that the coffee tasted sweet to me. — fdrake

I agree. I don't know but I think this is where Dennet I leading with...

he pretends to be able to divorce his apprehension (or recollection) of the quale--the taste, in ordinary parlance--from his different reactions to the taste. But this apprehension or recollection is itself a reaction to the presumed quale, so some sleight-of-hand is being perpetrated--innocently no doubt--by Chase — Dennet

Some people might say Dennet's denying there's anything to 'experience, it just goes from input to response, but I think he's saying, and I agree, that the recollection divorced from a response is incoherent. We always 'experience' events post hoc, never in real time. The experience is a constructed story told later (sometimes much later).

A flattening of standards between that which concerns people's self reports of experiences and that which concerns all else. — fdrake

Yes. I like the conclusion of the neural wiring pumps on this...

If there are qualia, they are even less accessible to our ken than we had thought. Not only are the classical intersubjective comparisons impossible (as the Brainstorm machine shows), but we cannot tell in our own cases whether our qualia have been inverted--at least not by introspection.

Actual psychopathologies can even confirm this.


Train's pulling in to my station....perhaps more later

I think they're wrong, but at least they're addressing the problem. — RogueAI

The problem being that you're incredulous?

My incredulity is that you find it at all difficult to believe that 80 billion neurons firing at a rate of up to 1000 per second could produce something as relatively simple as experiencing a phenomena. How many neurons did you imagine it would take? Another few billion? Should I contract some philosophers to investigate that for me, do you think?

I had zero justification for believing my friend dated a woman the night before. It was just a random belief. And I think that's the problem with your proposal. If you are to state knowledge is something we simply haven't had refuted yet, you allow beliefs without justification to be declared as knowledge. — Philosophim

Can you really think of a scenario where you'd have zero justification though? How would such a belief even get formed neurologically? Wouldn't it make more sense to say that the justifications were of one sort at t1 and of a qualitatively different sort at t2?

Following that, 'knowledge' would be beliefs whose justification was of a certain qualitative sort, which I think is more in keeping with his we actually use the word.

I think whether the "first order properties" or "second order properties" are called into question depends on which intuition pump we're talking about. Intuition pump (1) looks to me to be about first order properties and how they are ascribed. First order being eg. "the taste of this cauliflower to me now" and second order being eg. "(the taste of this cauliflower to me now) is private and subjective" — fdrake

...which I just noticed. Yes, I think that's true. Intuition pump 1 never really seemed like an intuition pump at all, as Dennet uses them, but more a definition of qualia raw, as it were. An honest attempt to at least start with " I see what you guys mean...but..."

What are the first order properties? — Luke

My understanding is that they are, for Dennet, simply the facts of the experience as we would relate it in, say, a story or a report. No different to the properties of an event.

Also see below...

OK, so in brief (with apologies if any of this is stuff you already know, or misses the point, I just want to be sure we're in the same frame)

Consider that we've no idea what caused the signals which just arrived at our primary visual cortex. We're either expecting them (efficient) or not expecting them (inefficient) and the task is to minimise the surprise (processing requirements) associated with the very next signals. This is just basic Bayesian Brain stuff. This region (V1) delivers the V2 visual cortex with signals relating to basic stuff like outlines, texture etc. The V2 region is in the same boat, it wants to minimise surprise. One way of doing this is by suppressing surprising signals from the V1 cortex. If it did this randomly we'd be in trouble, so it does so according to a model of the signals it expects to receive. At this point the signal splits into two streams, dorsal and ventral, but all along these streams the process continues through layers of expectation models delivering their (modified) signals and being suppressed in turn by the model above.

At some point, several models in, the ventral stream reaches a region which models objects and it will feed forward to areas associated with the object 'car'. Meanwhile, the dorsal stream has been merrily progressing away on the question of how to interact with this hidden state, without the blindest idea what it is.

So.

Point 1 the recognition that it's a car is part of your conscious experience of the hidden state. There can be no quale of a car because modelling it as 'car' is part of the response, quite some way in, in fact. And what's more, plenty of conscious responses have already been initiated by this stage. The dorsal signal doesn't even know it's a car before it's deciding what to do with it (cue amusing but completely unethical experiment with monkeys who've had the connection between their ventral and dorsal signals cut and can pick up and peel a banana but have no idea what to then do with it).

Even if we were to call the sum total of our responses 'quale', we'd have to have 'car' as part of, not the source of, those qualia.

You could say that the responses were to do with what I later determined was a car, but...

Point 2 the models which determine the suppression of forward acting neural signals are themselves informed and updated by signals from other areas of the brain. So no more than a few steps in and whatever hidden states we might like to think started the whole 'car' cascade of signals have been utterly swamped with signals unrelated to that event trying to push them toward the most expected model.

1 and 2 together, I think, make it very hard to talk of the 'quale' of a car in any meaningful sense. If there are 'quale' they certainly can't be properties of any identifiable thing short of 'my entire brain at that point'. We could define them statistically - there are measureable functions of activity in correlated brain areas we could theoretically use to give the quale some very fuzzy-edged owner, some host for it to be the properties of, but I really think doing so would be an act of trying to fit the theory to the facts.

Either way, the private, accessible to introspection, but inaccessible to third party, qualia of 'red' is an absolute non-starter neurologically.

Dennett says: qualia are magical and thus do not exist — Olivier5

Dennet says absolutely no such thing.

Let's say there is a continent called SomewhereLand which many have traveled to, but none have returned from — Hippyhead

Then how do we know many have travelled to it?

Upon what basis should we exclude any speculation about SomewhereLand? — Hippyhead

That we have absolutely no information on which to base speculation - "I think it's made entirely of pink blancmange", "Really? I think it's the intestines of giant space alien" ... What a fascinating conversation.

Where does he say that they are additional properties? In addition to what? — Luke

In addition to simply being properties of conscious experience. He says

Traditional analyses suggest some fascinating second-order properties of these properties. — Dennet

I'm taking that to be equivalent to 'additional'?

And yet, you stated earlier:

One need not deny the existence of sense data properties to deny that calling them qualia is of any use, or to deny that they then exhibit any of the additional properties associated with qualia. — Isaac


Your position is that you don't need to deny sense data properties in order to deny qualia. But you deny sense data properties (without explaining the difference) anyway? Okay. — Luke

Yes. (Except, of course, I think I have explained the difference). If there's something about my explanation of the difference you're still unclear on, I'm happy to expand, but the thread's not about my approach. I just wanted to clarify (as there seemed some confusion) that I was talking about Dennet's position in the quote you referenced, not my own.

True story (the flag and the bus).

One need not deny the existence of sense data properties to deny that calling them qualia is of any use, or to deny that they then exhibit any of the additional properties associated with qualia. — Isaac


What distinction are you drawing between sense data properties and qualia? — Luke

When we take in sensory inputs it sets off a large set of reactions in the brain, like a cascade. Most of those reactions are immediate feedback loops with the sensory apparatus themselves, the majority of which take place without any conscious awareness. Those that do have conscious awareness are always in review, post hoc constructions to model what just happened and prepare a response aimed at minimising the errors in that model. So to say that the properties of sense data are anything like qualia is to say that one of the properties of flags is missed busses (sometimes a flag flaps in the wind, which sometimes catches a person's eye, which sometimes causes them to trip, which sometimes means they twist their ankle, which sometimes means they walk slower ,which sometimes means they miss their bus). We can draw a line from the flag to the missed bus, but we'd be considered insane to regularly talk of missed busses as being one of the properties of flags.

You say that there are "additional properties associated with qualia", so what are the (non-additional) properties that you appear to indicate are shared by both sense data and qualia — Luke

Here I was referencing Dennet's position that...

I grant moreover that each person's states of consciousness have properties in virtue of which those states have the experiential content that they do. That is to say, whenever someone experiences something as being one way rather than another, this is true in virtue of some property of something happening in them at the time — Dennet

The additional properties being...

qualia are supposed to be properties of a subject's mental states that are

(1) ineffable (2) intrinsic (3) private (4) directly or immediately apprehensible in consciousness — Dennet

In exegetical sense I think ti's clear that he is not denying conscious experience has properties, but only that these properties are not of the nature associated with qualia.

I would go further to say that conscious experience does not have usefully definable properties at all. There's nothing it is like to eat cauliflower. There is only the entire sum of your current model at the time you happen to be eating cauliflower (as well as doing and sensing a hundred other things). It is unhelpful to talk about this process of inference as an 'experience' of something. The something here is part of the experience, not the cause of it. We make up 'the something' as part of our guess as to the cause of all the hidden states (including our own interception) that we have messages from, and constantly refine that model as new errors are found.

To put it another way, there's no such thing as a cauliflower for it to have qualia associated with it.

Proof that the grass is shorter now then it was before implies no necessity for the term “lawnmower”.

Your

intuition remains philosophical — Mww


........is out of context. — Mww

Okaaay...Can't make head nor tail of that, but I suppose you don't have to be a Martian to wear lederhosen.

But is it really a shorthand for something? Could it be that the mind is a sieve that electrical and hormonal events flow through, leaving behind words as markers?


Or better: is there a theory out there (involving qualia) that pictures the mind this way when all that really exists is words? — frank

I think Dennet's 6th intuition pump touches on this. It's my favourite pump, and often overlooked for it's implications. You could not tell the difference between inverted messages or inverted memory links. You'd have no way of knowing which had happened. Dennet uses this pump only to show that the supposedly private nature of qualia must, if they exist, be so private as to be unavailable even to ourselves.

I think, however it has wider connotations from a broadly neurological perspective. Associations, like the memory association Dennet imagines having been inverted, are all there is going on. The reason why intuition pump 6 delivers the punch it does is because it shows that there's nothing to the 'quale' other than the chain of mental events (and physical/physiological responses) of which it is a part. A seamless process of two-way interaction between our minds and our environment. At no point in the process of inference and error reduction have we ever finally 'received' the input from the hidden states of the external world, they're constantly in a state of two-way interaction.

I figured that part of the exegesis would involve doing what he asks to see what happens. — frank

OK...might be getting somewhere, but still an awful lot of dots to join up...

Our mental (and physical) response to eating pumpkin is complex and varied - got that bit (smelling lilacs makes you think of childhood or whatever). So... that relates to Dennet's "surely it is quite OK to talk of the way the juice tastes to Dennett at time t, and ask whether it is just the same as or different from the way the juice tastes to Dennett at time t', or the way the juice tastes to Jones at time t." - there's some way pumpkin tastes to you at some point, right?

But then he goes on to say "right here we have already made the big mistake. The final step presumes that we can isolate the qualia from everything else that is going on--at least in principle or for the sake of argument. What counts as the way the juice tastes to x can be distinguished, one supposes, from what is a mere accompaniment, contributory cause, or byproduct of this "central" way." Which seems exactly the matter waived away. Just because tasting pumpkin initiates some pathway of responses, doesn't mean it's anything more to do with the pumpkin than the chair you're sat on.

If it should become established that humans, ... don’t necessarily appeal to intuitions,... physically speaking, from proof of empirical brain mechanisms in which intuition is irrelevant, — Mww

Just curious how you see it possible for this to ever happen if...

...it [is] sloppy for anyone to use intuition as a psychological term; intuition remains philosophical — Mww

How could proof from empirical brain science say anything about intuition if it cannot use the term?

What intuitions did you encounter re: taste of pumpkins? — frank

I'm not even sure that sentence makes sense, but if it does, I'm afraid I have no idea what you're asking me. The closest I could get to what might be an intuition about the taste of pumpkins is that they would taste like the last time I ate a pumpkin??

A bird is a bird. Tautology. Nothing is being said in fact. — TheMadFool

I didn't suggest it wasn't tautologous. What is missing, was the question. What fact are you expecting Dennet to be able to communicate?

Can Daniel Dennett describe to us what his supper and the wine he washed it down with, presuming that was/is his evening meal, tasted/tastes like? — TheMadFool

The wine tasted like wine and the supper tasted like fish and chips. What's missing that can't be put into words?

I have a different assessment of the degree of certainty of that claim. — Pfhorrest

Ha! I expect you do.