To put it mildly, I am not a fan of Russell--I recall that the verifiability theory collapsed in the 1960s--cf the work of philosopher Brand Blanshard--

@Bernd and Edwin: thank you for your answers. I will check for Brand Blanshard. Best wishes

I think this is still a very relevant issue. It is also clear from Savage that the definition of sense data and the shift in Russell's thought need very careful handling. And thanks for flagging up those two texts.

My own view is that something like sense data are necessary to provide a grounding for physics that one might call a metaphysics or in a slightly different context an epistemology for physics. Physics is about defining patterns of causal relation and those patterns can be considered instances of operation of dynamic dispositions. If disposition is defined by calibration with rulers and clocks then physics can be done without probing deeper. But if we want to know what the dispositions are dispositions to, I think Russell's initial intuition was right in that they are dispositions to determine changes in 'sense data' of some sort. If we want a metaphysics or if we want to understand meaning and how our minds work and how we 'know' then I do not think we can do without something of this kind.

To my mind Russell's big mistake, from the outset, is to think of sense data as objects rather than relations, which he termed sensibilia when considered on their own. He then moves further away from relation and abandons the subject. I think to do this is to abandon anything like physics, because, as Russell himself admits, all physics is about relations. (I think Whitehead does better. Russell sees that physics is essentially dynamist like Leibniz but can perhaps never quite abandon the 'objects' of logic and language.) He ends up wanting 'qualia' to be intrinsic features of matter, on the basis that physics can only deal with the relational. But he does not explain how intrinsic features could be 'known' or talked about without them too being relational. My suggestion to get out of this is that 'qualia' or indeed 'sense data' are unique in that they are always maximally proximal relations. There can be no non-relational 'objects' we could know about and discuss. Everything must be relational, but the difference in the case of sense data are that they are the most proximal relations to 'me' or the subject, whatever that might be. I assume that the subject is in fact another relational unit in line with physics, so there is no 'non-physical' dualism involved.

My impression is that Russell gets quite close to something like this some of the time but then nearly always backs off. As a result Russellian monism is in the end incoherent. But I think a lot of the arguments he mustered at the beginning are more cogent than a lot of contemporary work on the subject.

All relationships are relationships between something and something else-there are no disembodied relationships--at the sense level objects are the primary--you cannot go anywhere without them--that is what is the given in perception--

I disagree Edwin. All relationships can be between relational entities. So nothing is disembodied but nothing in non-relational. In other words all entities are happenings or processes as in Whitehead or Leibniz. Objects are merely convenient abstractions. Inasmuch as they seem to have intrinsic natures, like 'bodies' these always turn out to be illusory 'visualisations' based on the relations of perception itself. Kant's unknowable thing in itself was a retrogression from Leibniz's potentially fully knowable dynamic unit - now known as an element of modern field theory.

I fear I am on a different planet than Jonathan--so let me try once more--objects are given directly in perception (whether the naked eye or thru telescopes or microscopes etc.). Abstractions are formed by integrating objects in a certain way (they are mental integrations)--all relationships have to be relationships of something--you cannot have a relationship between nothings-of course objects exist even when we don't see them at any given time--but we have to discover them at some point--atoms were discovered first by inferences from chemistry by observing the way particles interacted but the chemists had to observe things in reality--later we could see them thru advanced technology--

I think you are on the same planet as Aristotle, Edwin, together with a lot of contemporary academic philosophers. I am on the planet favoured by Heraclitus, Plato, Leibniz, Shoemaker, and I think maybe Schmeikal. As Russell pointed out, modern physics never visits Aristotle's planet, and we are talking about the epistemology of modern physics. It has no objects. As James Ladyman says 'Every Thing Must Go'. The idea that objects are given directly in perception has no meaning to a scientist.

As I understand it, although our physics is called Newtonian, the fabric of ideas, and in fact the notation, owes as much if not more to Leibniz and the tidying up of the conservation laws in the 1670s-80s. Leibniz saw Aristotle as backtracking on Plato's insights, although he could see that Plato's understanding of the illusoriness of reality needed reformulating. As Shoemaker and Ladyman, amongst the contemporary philosophers, show, physics can only be knowledge of causal relations because they are the only things that can cause us to have any knowledge of the world. Any conjectures about 'intrinsic' aspects of junctures between these relations must, as Bernd says, be mental abstractions derived from our inferences about causation. That is not to say that there is no external reality, just that any non-relational aspects of it are invented by us for comfort purposes, since physics has no need of them.

Dear Victor,

I am inclined to say "Yes" to your question. As physicists such as Schrodinger, Penrose and Davies may have alluded to in their scientific and popular writings. The vein of reasoning some have referred to as: quantum consciousness or the quantum mind. The abiding contention (sense versus empirics) seems to stem from rigid delineations between: science and philosophy; mathematics and physics. At the same time there is always what I call the "imperialism of disciplines" one particular discipline trying to dominate other areas of human knowledge. Such is the case with economics and its incursion into almost every main branch of knowledge. Economics itself being accused of being Physics-envy. Yet isn't this what physics is attempting to achieve in its search for a universal theory of everything? Another contention raised: Grand Unified Theory (GUT) vs. Theory of Everything (TOE).

There is definitely a role for sense-data in modern physics. Bertrand Russell's evolution on the subject of sense-data was very significant. Initially he was puzzled about the subjective nature of the experience of a penny, for example, which projects an elliptical aspect from most vantage points, a circular one face on, a rectangular one edge-on, and every variation in between, all shrinking with distance from the penny. This entire assembly of aspects was considered the sense-data of the penny, different observers experiencing different aspects of the whole assembly. Paradoxically, the volume of the penny itself being unobservable beyond its surface, was not part of its sense-data.

After his epiphany Russell realized the distinction between physical and experienced space, and that all the sense-data were internal to the mind, the objective coin itself being external, beyond direct observtion, corresponding to Kant's "Ding an sich". What finally convinced Russell was consideration of the causal chain of vision. Light from an object in the world enters the eye, where it is transduced to a neural signal in the optic nerve, from whence it is eventually transformed into a pattern of activation in the visual cortex. Russell (1927 p. 137-143) observed that a potent source of confusion in this matter is a confusion of physical space with perceptual space. For although our percept of the external world appears external to our head, it is not external to our true physical head, but only to our perceptual head in perceptual space. All of our perceptual space, including the externally perceived world, is inside our physical head in physical space. (Russell 1927).

"Perhaps there is nothing so difficult for the imagination as to teach it to feel about space as modern science compels us to think ... This question is very important, and must be understood if metaphysics is ever to be got straight. The traditional dualism of mind and matter, which I regard as mistaken, is intimately connected with confusions on this point. So long as we adhere to the conventional notions of mind and matter, we are condemned to a view of perception which is miraculous. We suppose that a physical process starts from a visible object, travels to the eye, there changes into another physical process, causes yet another physical process in the optic nerve, finally produces some effect in the brain, simultaneously with which we see the object from which the process started, the seeing being something 'mental', totally different in character from the physical processes which precede and accompany it. This view is so queer that metaphysicians have invented all sorts of theories designed to substitute something less incredible. But nobody notices an elementary confusion". (Russell 1927, p. 137 - 143)

An elaborate tretise on the fascinating history of this epistemological debate can be found here...

http://cns-alumni.bu.edu/~slehar/webstuff/consc1/consc1a.html#hist

A brief illustrated summary of the two contrasting epistemologies can be found here...

http://cns-alumni.bu.edu/~slehar/epist/epist.html

A Cartoon Epistemology presents the ongoing debate between  "direct perception" and "representationalism".

http://cns-alumni.bu.edu/~slehar/cartoonepist/cartoonepist.html

This is indeed highly relevant to modern physics because it reveals the indirectness of perception and conception, and the fact that not only our mental cogitation, but also our "external" perceptions (sense-data) are actually internal to our mind, and thus necessarily confined inside our brain.

Dear Victor,

Looking at the second paper you cite for us I am struck by something I had not thought of before, although I would expect others to have done so. Russell says:

'If we are to know anything about the table, it must be by means of the sense data - brown colour, oblong share, smoothness etc.'

Moreover, the suggestion is that these are the data given so immediately inexperience that we consider their presence incorrigible - we cannot be wrong about sensing the brownness of the table.

My understanding would be that modern neuropsychology is quite clear that there is an error here. The brown that is incorrigibly present need not at all be the source of our knowledge of the table. In fact the brown will be the result of inferences during perception pathways about relative intensity of colours in object and ground etc etc and also of inferences involving evocation of a memory of tables. The brownness might very well not seem brown at all if it were not attributed to a table. If we thought that we were actually looking in a mirror then we would interpret the same input as indicating the pure silveriness of the mirror (silveriness being a colour sensed when something is associated with vicarious hues based appropriately on nearby  structures).

So the basic epistemological tenet that all our knowledge comes from these immediate data is quite wrong. Sense data considered in this functional framework have no place in current epistemology. What I have not got to yet is whether Russell resolves this later. It would seem to me that the interesting sort of sense data worth retaining are indeed incorrigible but have nothing much to do with knowledge - which will be based on inferences that occur earlier in perception, before the sense data are profferred up to the experienc ing subject.

Sorry but I reject Russell and Kant; Russell obviously was influenced by Kant who viewed the outside world as unknowable.. Without sense data we can know nothing at all. Consciousness has to be conscious of something. Cs is our means of awareness of the world. and our basis for forming concepts. Perceptions are not inferences but the directly given. Now what about sensory qualities like color? They are neither just in the mind nor just out there. They are the form in which we experience things--for example, blueness is the form in which we experience a certain wavelength of light. The wavelength of light is real. Color is the way we process it. Contrary to Kant the fact that our brain processes information does not invalidate knowledge; it is the basis for knowledge. Kant thought that because we have a means of awareness (the senses and a brain) that we could not know anything. The implication was that true knowledge would have to be by no specific means.This may have been the most disastrous error in the history of philosophy. It gave rise to post modernism: the end of philosophy.  The Ghate and Locke article attached may be of interest as an alternative view. Ghate has a separate article just on Kant but I am not sure I have it,

@all contributors: thank you for all interesting thoughts.

@Jonathan: thank you for your answers. Just a few thoughts: perhaps perception-data can be misleading depending on tools and lense that we use, for example a brown table may be perceived as black, especially if we use dark blue eyeglasses, or if we are colour-blind. This question seems important in the context of Quantum Theory, where there are a lot of different interpretations, all of them try to explain the same data. 

@Jonathan: to continue, i see that the problem with quantum theory is not with many worlds interpretation, but the too many interpretations problem. Some of these interpretations do not support common sense realism. Moreover, quantum theory seems to have deep problem: the quantum measurement problem. For example, in his Nobel lecture, Louis De Broglie gave title: the wave nature of electron, so it seems to me he disagreed with wave-particle duality of electron. What is observed in experiments is wave nature of electron, but duality is a philosophy created by Copenhagen interpretqtion proponents. This seems to suggest that quantum theory is at best incomplete, and at worse it may be possibly wrong. So what do you think?

http://philsci-archive.pitt.edu/5325/1/QTwrong.pdf

Dear Eric,

I absolutely agree that Kant was wrong. But that seems to be another issue. 

I am not convinced that qualities are the 'form' in which we experience 'things'. 

Blue is actually a brain code for a wide range of dynamic dispositions - transmitting mostly short wavelength light, reflecting mostly SW light light, emitting mostly SW light and quite differently, depolarising our 'blue cones' - a disposition of light itself. But these are all dispositions, not things. 

The world is knowable if we accept that all we are knowing are dispositions and instances of operation thereof. That is where Leibniz was light years ahead of Kant. Russell is a minor figure in comparison but at least it was the fact that he was asked to do somebody's lecture on Leibniz that led to Russell contributing significantly to the re-emergence of Leibniz as a central figure in Western thought.

I do not understand what a brain code is--blue is not a thing--it is an attribute of things, just like size, shape, weight--entities are the integrated sum of their attributes-I am no expert on Leibnitz but I understand he was a rationalist (reason divorced from experience) and believed in fictional entities called monads along with religion--I simply cannot go there--

Sorry I meant Edwin, not Eric, my senior brain does not seem to retain names for more than a second.

The evidence of the senses is the base of all knowledge including all science--skipping that step and going to concepts is an example of rationalism--cutting off concepts from reality--the problem with positivism was not reliance on the senses as such--it was using the senses divorced from concepts--the best refutation I know of can be found in Brand Blanshard's Reason and Analysis, 1962, but I think positivism had collapsed long before that date--note that two types of (opposite) mistakes: reason without the senses and the senses without reason (rationalism and empiricism). The article I attached before based on Ayn Rand argues that you need both together and how one builds on the other--

Dear Victor,

I agree that data received in perception may mislead us about the nature of outside dispositional patterns - so we may think a horse is a cow on a dark night or we may think the table is black when we have just come in from bright sunshine, but I was thinking of a much deeper issue.

The assumption in certain philosophical positions seems to be that blue of the sky or brown of the table, or maybe by mistake black, are what our minds have to start with in order to infer the existence of objects and their behaviour - i.e. all our knowledge of the world. But modern neuroscience confirms what was evident to a few astute people like WIlliam James long before. Our knowledge of the world is derived by complex inferential processes starting right at the retina, and greatly extended in the visual cortex and association areas like movement and face areas, before it ever gets to the regions where blue or brown are experienced - which are likely to be prefrontal, although there are considerable complexities to that story.

So the idea that maybe we associate particularly with Hume that all our knowledge comes from sensations is just plain wrong. Perhaps a good example is kinaesthetic sense. For most of the time we get kinaesthetic information about how our hands and feet are moving about in the world in co-ordination with external objects but we do not actually sense anything. No 'impression' or 'idea' ever occurs. The brain makes use of constant kinaesthetic inputs (most particularly oculomotor kinaesthetic input, without which vision collapses) to gain knowledge of the world without any 'experience'. When you become familiar with a new staircase in a new house you get to know the fine detail of its structure, so that you could climb it in the dark, without being aware that you have. 

Perhaps the most striking example is that movement is inferred in the retina by ganglion cells. These cells respond to changes in light intensity that may indicate movement. The precise nature of the movement is inferred in the MT region at the back of the cortex. Not until all this inference has gone on is there any possibility of an experience of a particular sort of movement. 

So I fear Edwin is wrong on empirically proven grounds. Nothing is 'given' in the way empiricism suggests. We have to distinguish the signals that are 'given' to the retina from what is 'given in experience' because the latter is derived from the former through a very complex chain of inferences, even for the simplest sensation of seeing blue.

Russell was very good on keeping up with physics but my impression is that he should also have done some homework in psychology. You would have thought he read James's Principles, but it doesn't look as if he digested them.

The fact that sensory information is processed is true--but this does not mean that what we experience is not there--this is a Kantian notion--the processing is unconscious but the result is knowledge of reality--the idea of knowledge that is not processed by any means is a contradiction in terms--it would imply that real perception would involve some type of magic--our knowledge of reality at the perceptual level is not an inference--it is an experience--

I agree, Edwin, but the quotes in the Savage paper definitely indicate that what is called a foundational view of knowledge equates the incorrigible elements of experience with the   only source material for knowledge. But I think we agree that what is incorrigibly present in experience, like the sense of some patch of blue, is not the source material, but something derived from source material that is processed unconsciously. When I 'see a word' there must have been a mass of processing before I see it because I see it just as the word, I do not see it gradually being built up. And all those priming studies show that even if I never notice the word processing has gone on that may affect my later knowledge of something. 

So I do no think Hume can have been right.

I think you are saying that percepts are sensations tied together--this was a very common view but I think it is probably wrong--I am told that infant research does not support it though that is not my field--note that a patch of blue, in fact, has to be a patch of something--a patch is something-- attributes do not exist independent of entities-

I have never heard of the idea of sensations being tied together, Edwin. I am not sure what it could mean. You say that a patch of blue has to be a patch of something, but what justifies this statement? I think if you asked the average reasonably intelligent twelve year old if they agreed they would say ' well, not exactly, it doesn't mean that'. I wonder if you are taking language on a holiday here. You might well find a room full of philosophers who agreed with you, but that was Wittgenstein's point.

If I look out of my window on this moonless night I see a patch of black - what something is it a patch of? If I see a patch of blue in a landscape that is in fact a reflection of sky (which is not a thing) off a lake but I cannot tell if it is a factory roof, what is the patch of blue a patch of (other than blue). It is not a patch of lake either. If I see this landscape in a hypnagogic lucid dream what is the patch of blue a patch of? The classic example of course is the patch of blue in a rainbow, which is not a patch of any thing because a rainbow is a manifestation of a triple dynamic relationship of sunlight, water and retina.

I would recommend Larry Hardin's Unweaving The Rainbow: Color For Philosophers or even better his update this year in Perception and it Modalities edited by Stokes, Matthen and Biggs, or Dale Purves and Beau Lotto on Why We See What We Do.

what do you see in a rainbow? a blue stripe! there are no errors in perception--perception is the given--mistakes can be made conceptually, that is, in conceptual identification- e.g., looks like a dog, ooops, it's a coyote--at the top of the empire state building: wow, people on the street look like dolls--in perception things look smaller in the distance--but to say they were dolls would be a conceptual error-a stick looks bent in water but it does not follow that it is bent--you can tell by taking it out of the water--conceptually you learn that straight sticks look bent in water--re sensations: it used to be a axiom of psychology that percepts were composed of integrated sensations but I think that view has been given up--what about a precept that seems amorphous? this would mean that you could not conceptually identify it perhaps due to poor visibility or you are coming out of a coma--you would say: I see something but am not sure what it is--

I answered this by saying that in some conditions you cannot identify what you see conceptually because of the conditions--so you might  not be able to name it--but to see it as something rather than nothing you would have to see at least one attribute, e.g., a black something until further information can be obtained--Of what relevance is this to philosophy? No more relevant than to say that brain damage can affect our cognitive capacities-- If our species had to live like this we would have become extinct eons ago--in reality we don't live in a nightmare universe perceptually--

An unconditioned state to the mind? What does that mean?

I think the role is existing because everyone who knows or grasps the object with his(her) brain thus he(she) is the role itself. 

I really cannot follow this thread--I don't know what conditioning the mind means--you are free to bring your mind to the conceptual level by your own choice--being free of thought, as a way of life, means self immolation--reason is your main means of survival-

Dear Jonathan,

Corrigibility, and its opposite, are value notions, indeed truth-value notions. Furthermore, while we might say that inferences take place in the retina, etc., this is a different use of the notion of inference than that used when speaking of justification of beliefs. We could say that the retina is thinking. But its "thoughts", and particularly its "inferences" would not be subject to the sorts of standards (of truth, validity, etc.) that ours are. Processes in the retina are matters of fact. Facts are not claims. Indeed, this sort of point was the basis for Sellars' criticism of the role of sense-data in epistemology. Sense-data too, as Russell characterized them. are matters of fact. They cannot have a role in the epistemology of modern physics because, whatever they are, they could not have epistemic significance. Granted they, like retinal "inferences," could not be false, but the price is they could not be true either. As I understand it, this is why they were discarded as having a potential role in justification of belief. 

In general I think your approach is laudably Leibnizian. Leibniz had strong reasons for conflating the epistemic and the causal, and it may be that you do too. But there would seem to be a price to pay in terms of thinking of ourselves as justifying our beliefs.

Dear Ian,

I have tried on several occasions to find the detailed justification underlying the more interesting statements from Sellars and Russell on perception and knowledge. Each time I get to the detailed texts I come to realize that they are constructing arguments without first grounding their premises in plausible biophysical dynamics.

Philosophers are often very good at fashioning valid arguments but the impression I got from moving from studying in a department of medicine to studying in a department of philosophy is that the philosophers’ conclusions are often unsound because their premises are falsely grounded. And no surprise that is often because terms have been taken on holidays by being used without reference to a specific dynamic context.

Russell bemoaned the lack of knowledge of science in the philosophers of his day, and reasonably so if considering his good knowledge of basic physics. But he was just as guilty in terms of neuropsychology, I think. He should have thought harder before constructing his particular definition of sense data, which has no proper grounding and deserved to be thrown out. Descartes, Berkeley and Leibniz did much better at this level, despite having almost no neuroanatomy to go on. You can deduce what sorts of processes must be going on from very gross structure and exploring experience in different contexts, as Berkeley showed.

The depolarization of a retinal cell is a fact, but so is whatever biophysical event underpins an instantiation of a belief further downstream. It may be useful to say that no inference has been drawn until there is an instantiation of a belief about the inference further down but the point I think I was making earlier is that the deductive process that leads to that inference is occurring as early as the retina, so there is no direct interface between signals that have no truth values and a subject.

Maybe Russell is committing Ryle’s crime here – insisting on an analysis based on a concept of a person when we are discussing processes at a finer grain of structure. Descartes was right - you have to postulate relations something like knowledge for structures operating dynamically somewhere inside the brain, not for a 'person'.

So although I find your post civil, interesting and well constructed I find it hard to catch hold of the point you are trying to make because it seems to be couched in the sort of philosophical language that floats above dynamic grounding and is therefore liable to have no implications in the real physiological world. Specifically, I do not understand what you mean by paying a price? I am also unsure what conflating the epistemic and the causal would be since knowledge must have a causal basis. (Although in my own writings I have come to the conclusion that there is no valid concept of knowledge corresponding to the intuitive one. It all gets rather complicated.)

Maybe I am missing the point - please let me know if I am.

Dear Jonathan,

It would be uncharacteristic of you to miss a point, and this case is no exception. I suppose whether or not one can justify committing some version of  "Ryle's crime" gets to the heart of the matter. (In this vein, it is interesting that you recruit Descartes, who in the second Meditation is keen to stress that not only does the 'I think' accompany each 'idea,' but that it must be the same 'I think' so far as those ideas constitute the content of a mind (if not a person)). 

I guess I must grant that there is something like a floating above the (mere) physiological account at play in my suggestion. It is true that "whatever biophysical event underpins an instantiation of a belief further downstream" has the same causal/matter of fact status as the biophysical events in the retina. But notice that here you have used the description "underpins an instantiation of." One thing that the readers of Frege, post-Russell, (Wittgenstein, Austin, Sellars, Gareth Evans, McDowell, et. al.) have been impressed by is the Fregean observation that it is propositions that are true or false, whereas physical states of affairs are not. A slogan was: causes are not reasons. 

Descartes seems quite convinced, given what he says in the sixth Meditation, that biophysical events underpin an instantiation of perception and belief. But, and perhaps this is the sort of thing Edwin was pointing to, I don't use them in forming my beliefs. (And here I suppose is the Rylean crime.) My reasons for believing I saw a rainbow over Malvern (a belief I could be mistaken about) seem to have a grounding in a different kind of description than the causal explanation for my seeing a rainbow (or not) over Malvern. The difference in the kind of description can be captured by observing that the causes are biophysical facts, but it is what we experience, think or say that may be true or false. The game, as it were, of justifying beliefs is played not at the level of depolarization of the retinoid cell, etc., but rather at the level of rational relations of the beliefs that are underpinned by such events.

Would it be meaningful to say we find the inferences taking place in the retina, or biophysically farther downstream, convincing or unconvincing? And here we don't mean that the "output" may be misleading or not, but rather that the inference itself may be subject to such a standard?

That seems a very fair analysis, Ian.

I agree that there seems to be a mismatch between properties we judge to apply to propositions and those we judge to apply to causal biophysical chains. And Ryle was of course right to say that we should not analyse at the wrong scale. However, being a Leibnizian, I would like to think that we can make sense of how all this fits together if we are careful about apportioning roles inside the brain. As I indicated, I think we may need to have some new words to replace 'knowledge' when we get to fine grain and maybe we should abandon 'knowledge' at any grain, even if it puts epistemology lecturers out of the job of endless debates about what it is. I have tried to make some sense of how truth relates to fine grain in brains in some essays on my UCL home page (link on my RG page, essay on neurons and language and a much longer text of Reality, Meaning and Knowledge.) but I am very much exploring.

Descartes had problems knowing how much to attribute to a single soul unit. He was aware that he might be asking too much of an indivisible soul and in places admits that some of the thinking would be sub-contracted to nerves elsewhere. But in Passions of the Soul he comes up with some pretty plausible ideas about division of labour in brains, where I think he does a better job than Leibniz in some ways. His model is clearly not one to be followed closely but he is at least trying to make sense of internal dynamics and the place of thinking and language in that.

And I would not want in any way to deny the insight that Descartes has and which Leibniz extends, that the dynamics of thoughts and propositions seem to be incompatible with at least a mechanical dynamics of a Cartesian/Newtonian sort. I agree with Leibniz that indivisible dynamic units must be telic, for instance. But, also like Leibniz, I think the gulf ought to be bridgeable.

To my mind a true proposition is some sort of act of comparison of consonant patterns. So a bit like late Ramsay I think truth is a property of some utterance or thought. I don't go for Frege's third realm propositions. In Leibnizian terms a truth is a comparison between a predicate and a subject that contains that predicate. It is a matching. In a sense a statement is true in the way a gun barrel is true if all parts of the tube line up. But it is crucial that a truth be considered in the context of the relevant dynamic process. And since the relevant dynamic processes for propositions are unbelievably complex brain routines there is no general sense in which a comparison of any two physical structures or events can be true. A comparison of two channel marks on the way up a river is only 'true' if your have them in line  with your ship's course. The proposition 'I aime Mary.' is not true because it conflates English and French brain routines. Truth is matching by some very special set of rules tacitly accepted as relevant to the context.

So my thought is that we should be able to find a causal dynamic account of the truth of propositions but we have to be very careful which steps we attribute to which roles within the brain. 

I think we can evaluate inferences taking places in the retina, or in other early visual pathways. Convincingness may be a  more complex issue but I see no problem in principle. There is an illusion you get with signs on London Underground trains. Lettering formed in a matrix a bright dots appears to move left across a black screen. From six feet it is impossible not to see the bright dots moving left. However from two feet you see that the dots are produced by fixed LEDs in a grid and never move. Go back to six feet and the smooth movement of the dots is impossible to deny. Until I worked out how this worked I asked myself if the inference made by my early visual pathways at six feet was convincing since I doubted  it. By getting nearer I was able to see that it was founded on false premises. OK, we do not conventionally refer to the signals in early visual pathways as having propositional meaning but maybe we should. I think there is a tendency to confuse propositional meaning with propositional structure and inside brains I do not think that holds (for rather complicated reasons indicated in the essay on neurons and language).

Maybe we would all be happy to agree that the conception of sense data that Russell had in mind is no good. I guess my view is that there is something more appropriate but as yet very hard to pin down in precise dynamic terms that should replace it, rather than giving up on the idea of an internal locus that experiences by receiving signals in the way that Russell later seemed to.

I am going to have to end participation in this discussion because it has descended into extreme rationalism: concepts which are not tied to reality--this is Platonism come to full fruition--as an Aristotelian, who argued that knowledge starts with  information from the senses, I find discussions which do not make this connection to be incomprehensible. Ayn Rand's Introduction to Objectivist Epistemology is the only approach to knowledge that makes any sense to me.

Jonathan,

So I find your latest post really quite helpful as a window onto your view. And even if we agree that the notion of sense data, at least as originally conceived, may not have a role such as Victor is wondering about, we may also agree that there is something going on in the brain (that admits of dynamic explanation) which links what is perceived to the experience of it. We may even agree that this means there is an internal locus that experiences. 

I note, however, that you use the phrase 'act of comparison.' I don't know how much weight you put on this locution. But if an act of comparison is required, then locating where experiences occur, and identifying the dynamics of their occurrence, may leave an aspect of the experience unexplained, namely, and now I really do show my hand, the act which makes it possible to "bring the unity of experience to consciousness," as it were. (Now you'll be recognizing Kant.) Really both Descartes and Leibniz seem on board with this. (The historical figure in most stark opposition is, of course, Hume.) 

Experiences, they seem to think, cannot be merely given. They must be taken. So perhaps I lose you here. Nonetheless, I repeat I have found your comments quite helpful.

This where things get complicated, Ian, as you realise. We may not be so far apart and there is a difference between what one feels is the range of plausible analyses and the one of one's personal choice as a working model. We may differ more on the latter.

I would see the 'act of comparison' as being in one sense an act of a system that shunts signals around - a brain - and shunts two sets of signals into some event for comparison. But then we have the 'event of comparison' which needs to be an 'act', in a different sense, of some unit substance (of which there may be many) that relates to the two sets of signals in some way that leads to a 'weighing of one against the other'. The options for biophysical dynamics look to be very wide here. I have recently been discussing some of these things with Richard Arthur, who shares an interest in Leibniz and he has noted an intriguing passage in Leibniz's New Essays (which I quote in the discussion section of my 21st Century Monadology, googlable under that name) where L talks of images as if projected on to a highly folded membrane that retains prior information in its folds. The result will depend on the relations between new and old inputs.

I am not quite sure what to make of bringing the unity of experience to consciousness. I would see the unity of experience as a characteristic of consciousness itself - a feature of an event in the 'string of pearls' that is our conscious experience. That to me requires no particular explanation because we now know that all totally direct or fundamental interactions in physics have a unity of relation of a dynamic 'receiving' unit to the entire field of potentials that 'causally informs' that unit. We would expect any direct relation in nature to have the sort of unity we find in conscious experience.

I think you are right that Hume, despite being a very pleasing writer, missed the point in the end. He was good on miracles. I also think that Descartes and Leibniz had a faith in an enduring subject that looks unconvincing in comparison to Locke's more pragmatic view of continuing identity. But  I am still ensure that I really understand what Kant had to add. He seems to re-insert a non-parsimonious step of some sort. I continue to favour Leibniz even if he is not all that brilliant at taking apart the knowledge process in some respects.

@Dr. Jonathan Edwards and all contributors : thank you so much for all your answers. This thread has come to a point which is beyond my reach. My initial question was intended to see whether sense-data still has role in epistemology, or something else has replaced it. But thank you so much for all enrichments that you brought with. Best wishes

Victor,

I think, when it comes down to it, you've asked a question that can lead quickly to some of the more difficult questions in philosophy of mind. That, of course, makes it an interesting question. While Russell's classical notion of sense-data seemed to be pretty much abandoned after the criticism it received from Wilfrid Sellars and others in the middle of the century, it seems that on all sides theorists seek something to play the role of interface between all the relevant physical events that lead up to experience and the experience itself. The variety of views is daunting. Many seem not to really face this question at all. You are to be commended for doing so.

Here's a paper in the history of the idea of qualia, which discusses the idea of sense-data at some length. You may or may not find it useful. I found it useful.

http://www.timcrane.com/uploads/2/5/2/4/25243881/the_origins_of_qualia.pdf

@Ian: thank you for the link, i will read it soon. Best wishes

Yes, I always find Tim good at lucid exposition of a problem and its origins. I am less sure that he is good at a radical solution, but that's more difficult!

Dear Victor,

I am sorry if we have gone at a tangent but I am interested to know quite what your original puzzle was. We have been discussing whether we should consider 'sense data' a useful concept either in Russell's form or another. Can you give an example of the sort of epistemological issue that you feel might or might not need sense data?

I am a bit sceptical about the principle of verifiability in fact. Leibniz deduced the law of conservation of energy by reductio ad absurdum rather than by experiment and I suspect that applies to a very large number of basic scientific hypotheses. As Popper said, nothing is ever verified by experiment. All you can do is refute certain alternatives in certain idealised forms. You have to arrive at a positive conclusion through a vast number of reductio steps rejecting possibilities like Descartes's demon on the ground of a priori improbability. I am not sure Russell got his premises right.

I am interested to now what motivates your original query?

Dear Dr. Jonathan Edwards, thank you for your question. Perhaps i was asking not the correct question, but my initial motivation was from quantum mechanics. We all know from historians, that Niels Bohr, Heisenberh and all the founders of Copenhagen interpretation of QM tend to reject thet we can know anything about reality of electron. To summarize, according to them what we can know is only what the measurement tells us. But of course, realists like Einstein wanted some real description of electron. As for now, i got an impression that there are two descriptions of electron: a. The electromagnetic field model, and b. the probabilistic model of Quantum Mechanics. My concern is that whether a realistic description of microworld can be expected? Or all are just our perceptions based on sense data? Perhaps you can resolve some of my questions. Thanks

Thanks Victor. Those questions are maybe more interesting than technical issues about Russell. I would first ask what you are wanting in terms of 'reality' for an electron. As I see it most people, including Einstein, mean by that how one could imagine it to be in an envisageable form sitting in 'time' and 'space' as we think of them in terms of the appearance of the world. So in a sense it is to ask what is the 'appearance' of an electron. But since appearances are due to things like reflection of light an electron cannot have an appearance without specifying how you would observe that appearance.

I think Bohr comes into the story at a rather peculiar time. It is about the only time in the history of physics when all aspects of the world could be considered as 'particles'. With Eintein's photons even force could be particles. For 200 years prior to that people had to accept that at least forces were unenvisageable - something Newton himself hated.

Bohr's theory made it clear that there cannot be any 'particles' in the envisageable sense. People have hung on to 'wave-particle' duality for as long as they could but contemporary quantum field theory makes the idea of any particle unworkable (strings are no better I suspect). QFT indicates that the universe is made up of modes of dynamic connection of a completely unenvisageable sort. Our brains paint pictures of aggregates of modes for us, including a synthesised 'space' and 'time' but this is completely unsuitable for envisaging individual modes. 

That might seem to mean there is no reality, but there is a perfectly satisfactory reality as long as we see it in terms of dynamic, or causal, connections, which we can define in terms of the sorts of experiences our brains synthesise for us when they find such modes of connection operating.

In fact the 'oddness' of Bohr's theory should never have been considered odd because it had been predicted by Leibniz 200 years earlier. Even before Newton wrote the final equations of motion Leibniz had realised that the units of the universe were unenvisageable dynamic connections (which he called monads). He realised that 'matter' and also 'space' and 'time' as we sense them are all pictures painted for us by our brains and that the real reality can only be described in terms of the mathematical structure of the causal connections, with the experiences of observations as checking devices. Plato and Locke had almost got there but Leibniz realised that even our sense of shape and size and duration is just a painted picture inside the head. Kant agreed but then got muddled.

There are two reasons why Newton got his name on physics rather than Leibniz. The first is that Newton focused on getting the equations right based on detailed astronomical observations, whereas Leibniz was probably more concerned with logical principles. The second is that Leibniz's lesson about unenvisageability is very hard for a lot of people to accept. People preferred the naive realism of what they thought was Newton's physics. Einstein preferred it too. But modern physics has shown it is no good.

People talk as if electrons were particles that get a bit fuzzy and wavelike between observations but then behave properly when observed. But this is simply not so. A particle has to have a size and therefore a right side and a left side that can be engaged separately in interactions (maybe two photons, one bounced off each side). But electrons never ever show a right and a left side. They have a place of departure and a place of arrival maybe, but that does not make them particles.

So my simple answer to your question is that a reality of appearance or envisageability for fundamental dynamic units is very definitely not to be expected and Leibniz explained why 300 years ago. In fact Parmenides said the same thing 2500 years ago. It is easy enough to deduce by logic. It just upsets people who like things to behave like brain pictures.

Dear Dr. Jonathan Edwards: thank you so much from addressing my question. I think the problem is not just to choose between particle view and dynamic view, in your words. But perhaps, there are tensions among the followings (at least):

a. Between corpuscular and wave model

b. Between probabilistic and deterministic view

C. Between observer-dependent reality and objective reality

So much problems are there, that Bacciagaluppi and Valentini concludes that interpretations problems were not settled at the Solvay Conference at 1927. 

http://arxiv.org/abs/quant-ph/0609184

This is where I disagree, Victor. The 'interpretation debate' is dead and was never worth having because interpretation is simply a way of trying to rescue naive realism. No serious practical quantum physicist has any interest in it and no serious philosopher should either.

There is no corpuscular model - as I indicated above, there never were any particles, except in Democritus. Descartes and Leibniz realised that the concept of a particle is self-contradictory. That is how modern physics came into being. It was dumbed down by Newton for practical purposes but even newton knew there were no particles or corpuscles. Moreover, the dynamic aspect of a quantised mode is not even a wave. The wave function describes an ensemble or type of mode, not an individual dynamic unit. 

The probabilistic/deterministic issue is settled by the math. And Leibniz could see that the math had to be probabilistic at the ultimate level. If you have continuous dynamic rules and discrete dynamic units (quanta or monads) you can only have a theory that is probabilistic at fine grain. it is built in to the logic. Einstein's idea of an underlying determinism or Bohm's hidden variables create an infinite logical regress that helps nobody - as Bohm eventually realised.

Reality is both observer dependent and independent in different senses. No dynamic connection can exist with connecting to something that 'perceives' it in operational terms. But there is no need for that to be a human being or even a life form.

These are all false problems, I am quite sure. They all derive from a desire to envisage something we have no reason to think we should want to envisage - all we need to do it write down the mathematical structure, say it is instantiated and that it connects directly or indirectly to experience in operational terms.

Dear Dr. Jonathan Edwards: thank you for your answer. Ok i see that we have different views here, but i am willing to learn more from you. Would you please give some enlightenment on these problems: what is the exact mechanism of collapsing wave function, as reality is observer dependent as you say? And also why don,t we have an electromagnetic picture of wavefunction?

As of myself, sometime in the past i prove that Dirac equation can be transformed into Maxwell equations using quaternion unit. The meaning is that mathematically there should be neat connection between quantum picture and electromagnetic picture. My wish is to describe electron in a electromagnetic picture, i read elsewhere that Winston Bostick proposed his ring toroidal electron model in 50s. What do you think? Thanks

Dear Victor,

My understanding is that wave functions describe types (or ensembles) of indivisible dynamic modes. Since the mode is a dynamic indivisible it does not involve any 'mechanisms' since these would be accounts in terms of dynamic subcomponents and there are none. An instance of operation of a type of mode will include the type aspect of its dynamics (the wavefunction) and a token aspect that is stochastic. These have no separable realities associated with them. I think von Neumann was very unhelpful in proposing two processes. 

My idea of a wavefunction is that it is a bit like a descriptor of a particular gauge and pitch of corkscrew wire or a wallpaper with a repeating pattern. If you have a connection between observed events A and B then the wavefunction tells you that the connections has maybe 7 and 1/2 turns or repeats - it gives you a phase relation. But it should not be interpreted as suggesting that anything 'moves' or 'spins' or 'travels like a wave' from A to B because that would imply separate component realities for different points in space time within the connection and dynamic subcomponents connecting these. Quantum theories deny that. 

As I see it quantum theory takes us away from the intuitive concept of cause by, as James Ladyman puts it, 'microbanging' of billiard balls. In fact it takes us to the proposal of Leibniz - that everything progresses in a co-contingent harmony. All attempts to envisage a 'mechanism' for this should be abandoned. 

Dear Dr. Jonathan Edwards: thanks for your answer, that is very interesting. But are you sure that you do not overread the experimental data, and invoke your own favorite philosophy? I think we should also remember Occam Razor. Best wishes

Yes, but Occam's razor was what I was using! If we postulate that there is a quantised or indivisible or monadic level (all much the same thing) to dynamics then we have no need for further mechanisms. They lead to an unnecessary infinite regress. 

I originally come at this not from philosophy but from the physiology of perception (I am a medical scientist). If one considers perception carefully it becomes apparent that our sense of space and time cannot be directly related to either space or time - they are codes for something that we cannot envisage other than through such mental codes. We see things 'moving' at a particular place despite this being impossible if motion is a change of place. Our perception of movement must be based on a code signifying differentials, not 'what is there'. We see a cloth as red even in a nightclub with almost entirely blue light when almost all the light coming from the cloth is blue - but our brain differentiates against the background and we just see red.

I don't think over-reading data comes in to it, in fact. Leibniz shows why it only makes sense on a priori, reductio terms, to consider the world as based on monadic dynamic indivisibles (unenvisageable quanta). We now live in a time when the data confirms that very nicely. I think the data we do tend to overinterpret are the 'sense data', if you like, of our experience. We think they are 'veridical' when this is not something they could be, because they are only codes.

Dear Dr. Jonathan Edwards: thank you for your answer. May be Leibniz is right after all, but can you point to a paper showing how calculations prove that monadology explain experimental data? Without calculatiob, i am afraid we just replace a philosophy with another complicated philosophy. Best wishes

I don't think this is an issue of calculations, Victor. It is an issue of basic tenets. Leibniz's three most significant contributions were probably (1) the tenet that there is a quantity mv2 which we now call energy (the 1/2 cancels out), which is conserved in all dynamic interactions. (2) the tenet that all energy comes in discrete indivisible units or monads - presaging Einstein's photons and quantum theory in general and (3) the tenet that energy is not some add on property you can 'give' to a particle of matter or take away but the very nature of the dynamic units that constitute the aggregates we call matter - which leads directly to Bohr's idea of orbitals with a prescribed energy level rather than orbits of particles going at whatever speed they like. He also points out that we should not expect to find any billiard balls in trajectories at the fundamental level, presaging the Copenhagen story, but that idea was not new.

So Leibniz's contribution, like that of Einstein and Bohr was chiefly about theory structure rather than specific calculations but at least he was right about mv2, or at least on the right track.

@Dr. Jonathan Edwards: thank you for your answer. Btw i just found an interesting paper discussing quantum measuremen problem. Perhaps you would like to make comment: http://arxiv.org/pdf/quant-ph/0509042.pdf

http://arxiv.org/pdf/quant-ph/0509042.pdf

My understanding is that the measurement problem only arises if you want to think of quanta of action or energy as 'particles'. My understanding also is that the concept of particle has finally been replaced by that of a mode of excitation of a field. To be honest I do not think the author of that paper understands what he is proposing or the fact that by definition he cannot test his theory.

@Dr. Jonathan Edwards: ok, i see. Yes i agree that a good theory should be made testable. Yours

I recommend reading Aristotle. Sense data are the base of all conceptual knowledge--skipping that step has a dirty name; rationalism (reasoning divorced from the senses and thus from reality). We can thank Plato for this. It has been said that the history of the world is a duel between Plato and Aristotle. I agree. I am reading an interesting new book on this: The Cave and the Light by Arthur Herman which is focused on this issue..

@Edward: thank you for your answer. Best wishes