The Master of Königsberg famously differentiated between what is given in perception by the senses (the "phenomenon") and the thing-in-itself, the Ding an sich (the "noumenon") – cf. http://en.wikipedia.org/wiki/Noumenalism#Kant.27s_usage; on Kant's view, the noumenon is opaque to direct experience and can only be derived from the phenomenon by a process of induction or abduction. Noumenalism is commonly cited as an argument against "naïve realism" – the belief that the world is pretty much the way we perceive it (cf. http://en.wikipedia.org/wiki/Naive_realism). The noumenalist argument against naïve realism goes something like this: "We can have no direct access to reality, therefore we can never be certain that our phenomenal experiences correspond to some noumenal reality. This being the case, to hold that the putative objects of our phenomenal experiences – material objects, physical events, concrete properties and relations etc. – correspond to the 'actual nature of phenomenal reality' is philosophically naïve."

We can indeed criticise the "pretheoretical categories" of entity with which we populate the "physical world", but this is a matter of investigating ontological commitment and the limits to our reduction of theories (or of our capacities to translate one theory into another). However, there is no reason to suppose that we are "hard-wired" to make any distinctions between phenomena save by the limits of our perceptual apparatus – thus, we cannot have optic perception of light at wavelengths greater or lower than the visible range. In some cases, our perceptual apparatus allows that one of our senses can experience phenomena that are beyond the range of another - we cannot have direct optic perception of light at wavelengths lower than the "visible range", though we can feel heat; we can have no direct auditory experience of very low-frequency sounds, though we can feel vibrations. But, other than these "perceptual capacities", we are "tabulae rasae" – there is not even the innate Cartesian capacity to make logical distinctions. This view is largely due to Locke (cf. http://en.wikipedia.org/wiki/John_Locke#The_self), though it has been largely modified by recent advances in pragmatic behaviouralist philosophies and by the evidence from the neurological and cognitive sciences.

As to Kant: Kantian noumenalism is a direct result of Cartesian dualism, and as such has had a widespread influence on the debate between monists and dualists. Generally speaking, an anti-dualist monism must show that the rejected domain (be it "the mental" or "the physical") is in some way an epiphenomenon.

1. REALITY

For my money, the problem is rather one of determining what it means for something to be "real". If our "criterion of reality" is causation (as with Davidson before 1985), then "what is physically real" will be "that which is the cause of a physical event"; if our criterion of reality is spatiotemporal localisation (as with Quine), then "what is physically real" is "that which occupies some region of spacetime". We can adopt the view that the contents of our phenomenal experience is "real" ("given directly in experience"), though we must also accept that such first-person experience is private and therefore illegitimate as the foundation of any public definition of reality. We can adopt a "heterophenomenological" approach in which we accept that a subject Susan is authoritative about what ***seems to be*** the case for her but by which we evaluate her statements of "what seems for her to be the case" by appeal to "what seems to us (and to others) to be the case" and to supporting evidence (Susan's more general behaviour, correlated regularities between the putative object of Susan's impressions and our more general beliefs concerning that "kind" of object etc. etc.).

We should bear in mind that the definition of "reality" is entirely informal and depends on metaphysical assumptions – and we should be most careful to distinguish between "reality" and "existence". Existence is a formal notion – "to be is to be the value of a variable", to quote Quine – and depends on quantifying existentially over the entities in some well-defined universe of discourse (our ontology depends on our choice of language). "Reality" is more a case of overt ontological commitment – thus, when we say something like "Othello's jealousy was the direct cause of Desdemona's death, but Iago's treachery was the indirect cause", we could claim that among the "furniture of the physical world", there really are "jealousies" and "treacheries"; our task here is to relate such entities to the more usual physical categories of "material object", "physical event", and "concrete process, disposition, or state of affairs".

2. LOCKE vs. KANT (round 1)

The idea that "we can have no direct access to noumenal reality" – in the following I'll call this notion "noumenalism"- is dependent on the notion that reality could be "other than we perceive it"; in the following, I'll argue against noumenalism as a product of the "Cartesian error" and suggest that saying that we can have no direct access to reality is not at all the same thing as saying that our apprehension of 'reality' is determined by our modes of access to a putative mind-independent world.

Noumenalism makes the Cartesian error of supposing that our sensory impressions could somehow "cheat" us in their reconstitution and representation of reality. Setting aside the "why" of the cheating, the error here is not in the notion of "reconstitution" – we do indeed seem to experience the world as 'unified', despite the different modes of access given by our sensory organs. The error is in the notion of "representation": the assumption that "reality" is *some thing* which can be represented ALSO presupposes that "reality" is represented TO *some thing*. We have a thing A that is represented to B and a thing B to which A is represented. If, against Descartes, we question the assumption of "B", the whole house of cards of noumenalism collapses. If there is no "B", there is no relation of representation; if there is no relation of representation, then "A" – the "noumenal reality to which we can have no direct access" – certainly can't be defined in terms of B having or not having access to A.

If I may use a metaphor : the stage of the Cartesian theatre might be brilliantly lit, but apart from the actors, the theatre is empty. Nothing and no-one is "watching" or "receiving" the contents of experience - indeed, the very idea of distinguishing "experience" and "the contents of experience" is already theory-laden.

We can, I think, argue that we have no need of a "spectator". When we distinguish between 'the contents of experience' and 'experience', we are postulating 'experience' as the province of a Cartesian subject and its 'contents' as "coming from without" – while it is true that I experience the taste and texture of the steak, what I am experiencing is NOT the taste and texture of the stake, but a 'false content' fed directly into my nerve system. This is not the way experience works, I think. My experience and the contents of my experience are not subject and object, but – if there is a distinction at all – at best subject and complement ("I am experiencing steakwise"). I'm not sure we can even make this distinction – there's no "fundamental experience" that can be separated from, for example "the experience of eating steak". If I might cite the classics, Cypher doesn't believe that the contents of his experience is false (otherwise he would have no reason for choosing the illusion). He simply accepts a metaphysical theory – that the causal web underlying his experiences is not the causal web of intuitive realism. For Cypher, the "myth of physical objects" no longer serves to explain the world of phenomenal experience – the myth of The Matrix is, within his narrative world, more compelling to reason. But his choice is still determined by a preference for the phenomenal reality of the experience of eating steak and drinking Château Margaux over the phenomenal reality of eating white goo and drinking raw alcohol.

My view is that, once we get rid of the Cartesian error, "the problem of noumenalism" in philosophical naturalism (PN) reduces to a simple contradiction implicit in PN's metaphysical assumptions. Now, PN postulates that there is a mind-independent reality. This postulates the weak assumption (WA) that things would happen whether or not they are 'known' to some mind (and hence the big problems posed by some interpretations of QM) – in other words, that the physical universe gets on with its business regardless of whatever we might say about it. Whatever it is about the apple that makes me see red is there even if it's midnight. Now, however tenable or not such views might be, they are NOT the strong assumption (SA) that "we cannot know reality".

"Mind-independence" is the tricky term. WA postulates that the world is there whether we think about it or not; it does NOT imply that we cannot have "knowledge about" the mind-independent world .

The error of SA is to understand "mind-independent" as "mind-inaccessible". The basic tenet of noumenalism is that Mind cannot have direct access to "noumenal reality". But if we remove the spectator from the Cartesian theatre of experience, there is no "misrepresentation" possible – "Mind" is not a spectator to whom Reality is being represented in a "narrative" which reunites the disparate modes of sensory impression. Mind is the narrative itself.

If Mind were the spectator of the narrative – the thing ***for which*** a unified image of reality had been reconstituted – then we would have four elements: "reality" (R), the "process or reconstitution" (P) by which sensory impressions of R are integrated; the "image of the world" (I) which is the result of 'P'; and the 'Cartesian spectator' (S) to whom 'I' represents 'R'.

We can, however, allow that 'R' is ***no more than*** 'mind-independent', and that 'P' is the process by sensory impressions of 'R' are integrated in experience. The "unified image of the world" 'I' is thus no more than a reification of 'P' – and this is the very banal observation that 'Mind' is a process, not an object. On THIS view, the "noumenalist objection" is just the very banal observation that we cannot have mind-independent experience of mind-independent reality.

All in all, noumenalism is based on the assumption that, although we can have no direct experience of it, there IS some reality which in some ways determines our phenomenal experiences (otherwise, we'd have nothing but phenomena, and this way Berkeley lies). Yet if we reduce "noumenal reality" to mind-independent reality, then it is evident that we cannot have mind-independent experience of that reality: experience is a 'mental process', and CANNOT be 'mind-independent'.

Of course, a lot of people would say "ah yes, but you can have an experience without thinking about it". Of course, but 'mental process' doesn't reduce to 'metacognitive process'. "Being aware" is not metacognition, but is rather a state of excitation fundamental to "having an experience" – if the 'experience' has no impact on your basic sensory and cognitive processes, then it's hard to say that you've "had the experience" (a high-level burst of gamma rays would be undetected by our immediate sensory apparatus – the 'experience' would be indirect, and given by the resultant destructive effects on bodily function).

That "reality" is mind-independent doesn't exclude that its is ALSO mind-dependent. Quite evidently, the notion of "reality" is itself mind-dependent: if there were no "minds", then the questions of whether "reality" is ordered, and "how" it is ordered, would never arise. Nor would the question of "order". All these questions are part of the "narrative" that we're acting out in the Cartesian theatre of Mind. But there are no spectators – just our fellow actors. While experience is 'private", Mind is not – it's an intersubjective domain which requires the interaction of human agents (our 'view of the world' is a matter of triangulation, not direct description. A term designates rigidly only when it can pick out one particular entity throughout all possible descriptions of time and space and for all possible speakers).

Most of all, Mind requires language if it is to serve as a means of triangulating from different perspectives of "the world"; however, this also requires that language is flexible enough to allow for a difference in perspective. While the categories into which we regiment phenomenal experience are linguistic and therefore general, certain linguistic resources (deictics, tense…) allow each speaker to interact with her linguistic community from a perspective that is perhaps local, but is nonetheless unique. "Mind" is itself a linguistic notion – indeed, we can ask whether the term 'linguistic notion' is not tautological – and, as such, cannot be reduced to the characteristics of an individual human being taken in isolation from any social or cultural context. Such a person would have no language, and therefore no equivalent to OUR concept 'Mind', which is irreversible tainted with the linguistic.

So, mind-dependence is not just a matter of "awareness", but also of language. We can push this as a partial redefinition of the hypothesis that "reality is mind-independent": "reality" would be *as it is* HOWEVER we might describe it; what matters is how well our descriptions of one part of reality integrate with our descriptions of other parts.

Now, this redefinition of "mind-independence" effectively brackets out 'awareness'. The problem is not whether our sensory impressions "reconstitute reality" faithfully or not, but whether the observation sentences we use to talk about those impressions are coherent with the observation sentences we use to talk about similar impressions at other times and in other places. Such similarities allow us to talk of 'phenomena', and to postulate relations between such 'phenomena' (to construct "special theories"); these "special theories" allow us to generalise from kinds of phenomena to "underlying regularities"; and the more "general theories" can be tested against experience as they *should* generate a certain class of observation sentences under certain controlled conditions (the view is restated in behavioural rather than mentalistic terms). Now, 'building up' from observation sentences to theories might serve a certain view of 'the abductive process', but we should ALSO bear in mind that the observation sentences are themselves the product of a given language and of the general assumptions that language makes about whatever 'part of the world' the observation sentences are supposed to be about (in this way, theories "face the tribunal of experience as a whole)".

We can go as far as to ask whether we should rather be postulating the language-independence of the world than its 'mind-independence' - or whether the two notions can even be distinguished.

***

I remarked above that, even if "reality" IS mind-independent, this doesn't exclude that its is ALSO mind-dependent. We could understand this as suggesting that

(1) mind-independent reality (MIR) and mind-dependent reality (MDR) are co-extensive;

(2) that MDR supervenes on MIR; or

(3) that MIR subvenes on MDR.

In its most banal form, (3) is no more than the recognition that the notion of MIR depends itself on language; if the subvenience relation were given ontological force, it would simply state that reality is mind-dependent.

(2) covers the familiar Humean supervenience of physical reductionism but could allow for less drastic or less top-down accounts (thus, if we want to keep a bottom-up ontology, we can employ such notions as 'emergence'; on the other hand, we can employ a top-down ontology and talk of theories "carving the world at the joints").

(1) is perhaps the notion that we should be discussing, rather than placing bets on a wrestling match between the straw man of "naïve scientific realism" and the self-defeating spectre of noumenalism.

3. LOCKE vs. KANT (round 2)

On reading the above, our good friend Nicole Pernat made the following remark :

"I agree that there is no self for which the world is re-presented. However, just as a thermometer gauges temperature while the mercury is not identical with the temperature, so our sensations & perceptions are not reality themselves"

I remarked to Nicole that when I was younger I was much taken by the "semiotic" account derived from CS Peirce: according to Peirce (and with apologies to the "real" experts!), a sign represents its object in three ways: either as an ICON - by some formal or topological resemblance (like a map, a graph or a photograph); as an INDEX – by some relation of "contiguity", and usually a causal relation (a footprint can indicate the presence of a person, a rash can indicate the presence of a virus…); or as a SYMBOL – by some convention or norm (language, of course, but also pictorial conventions, gestures – even silences!). Now, if we apply this simple model to the question in hand, we can see that ***the observable state of the mercury in the thermometer*** is in an indexical relation with the ***temperature of the mercury*** (its mean molecular momentum or whatever), and that THIS is in an indexical relation with ***the ambient temperature***. Each of the ***starred*** terms represents a "real" characteristic of the world, and each is linked to the other by a 'causelike' relation – the ***observable state*** is caused by the ***immediate temperature*** and the ***immediate temperature*** is determined by the ***ambient temperature***. In each case, we're talking about "real" states of the world (and remember that "the world" INCLUDES our perceptions) – what changes is the "epistemic system" by which reality is constructed (and I say "constructed" rather than "demonstrated" as *reality* is a linguistic notion, not some mind-independent characteristic of the world).

The observable state of the thermometer-system depends on our beliefs concerning the causal relation between the height of the mercury and the ambient temperature. We perceive the thermometer-system as being in a state which corresponds to, say, 8°C. There is a real state of the world which corresponds to our perception of the thermometer-system (let's call it "TH"). Unless we're hallucinating or misreading, the real state "TH" corresponds to a real state of the world which is the "temperature of the mercury as a system" ("TM"). The possibility of hallucination or misreading does indeed "come between" TH and TM (this is the entire premise of "The Matrix" series), but this more a question of interpreting the "real" content of a perceptual state.

How is this possible? Well, we've seen that there is a real INDEXICAL relation between our perception TH and the state of the (mind-independent) world TM and, given the physical and physiological characteristics of our perceptual systems, we take it that any ***perception*** has some cause – that is, that the perception is an index to "some" state - in Davidsonian terms, the phenomenal contents supposedly "supervenes" on some state of the perceptual system. The phenomenal contents is logically related to the "real perceptual state" (I'd say that the relation is one of identity), and the "real perceptual state" is the causal result of some "real stimulus". This is as much the case in hallucination as in "real" perception. In the Matrix, the phenomenal contents of the sleepers' experience corresponded logically to certain neurophysiological stimuli – the experience itself is "real" as it is a state which is the causal result of these stimuli, and the stimuli are "real" insofar as they are real physical events taking place in a definite region of spacetime. The sleepers are only mistaken in their naïve intuition that there is an ICONIC relation between the contents of their perceptual states and the stimuli – that when they see "a dog", the causal origin of that perception is indeed a mind-independent entity having certain physical, biological, and genetic characteristics. Whereas an index MUST have an object (it is caused by its object), an icon can be a pseudosign – just as one can create a "map" of an imaginary territory, or a "picture" of an entirely fictitious character. If you'll allow me the analogy, the "dog" seen by the Matrix sleeper is "no more than pigment on canvas" – there is "real pigment", but no "real dog". But whether the "dog" is "real dog" or "real pigment", it is still REAL.

The relation between the perceptual state "TH" and the state of the mercury-system "TM" is a relation between two real states, though WE must rely on circumstantial evidence to decide whether TM actually obtains (and this rejoins my comments on "states of affairs" and "possible worlds" on the philo group – we have to decide "which" world we're living in. I can develop further if you wish). If we take the "Matrix" example again, we have to decide whether we're living in a world where TH is caused by or correlates to the actual state of a real system TM, or whether we're living in a world where TH is caused by some Demon Machine stimulating our perceptual system. TH is not in doubt – though it is only accessible to "first-person experience". Now, while I can't appeal to MY first-person experience to found any "universal reality" of TH, our language does allow a certain "heterophenomenology" (to borrow Dennett's rather unlovely term http://en.wikipedia.org/wiki/Heterophenomenology): I can formulate an "observation sentence" of the kind "the thermometer gives a temperature of 8°C" to which any other observer can either assent or dissent – and should they assent to my description, I can hold-as-true that they are in an equivalent phenomenal state to my own. Of course (cf. the various "zombie" thought experiments), I have no *guarantee* that they are in such a state – my attribution is a matter of my having a Theory of Mind (http://en.wikipedia.org/wiki/Theory_of_mind) or adopting some kind of "intentional stance". Each person who assents to the observation sentence reinforces my justification in believing that a perceptual state equivalent to TH should hold for any observer, though it is of course impossible to PROVE that TH actually holds for any observer.

My argument that TH is "real" is perhaps founded on my direct and private access to my own perceptual states; while Cartesian phenomenology would disallow this as a statement of how things ***are*** to me, I can - to use Dennett's terms - be "authoritative" about how things ***seem*** to me. The context of my observation sentence is evidently "IT SEEMS TO ME THAT the thermometer gives a temperature of 8°C" – and in fact, I am making a statement about my perception, not about the world. But, as we appear as a species to have evolved similar responses to similar stimuli, and as my "understanding" of my perceptual state is largely dependent on the linguistic conventions which determine the formulation of observation sentences, there is a pragmatic precedent for holding that the statement of my perception corresponds to some state of affairs accessible to "other minds" – that is, that in the context of a ToM, the "reality" of TM is at least not *solipsistically* mind-dependent.

***

The "Matrix" (and similar "brain in a vat" thought experiments - http://en.wikipedia.org/wiki/Brain_in_a_vat) are predicated on the idea that TH – the perceptual state – "represents" TM (the state-of-the-world). As you'll recall from my criticism of the Cartesian theatre, the error is perhaps in the notion of "representation", for which there is not only an "of", but also a "to". We can, I believe, make a distinction between "representation" in the sense of "modelisation" and in the more Cartesian sense of "the act of presenting something to someone" (Piece made some comment about a sign "representing" its object in the same way that a diplomat "represents" his government, which would imply that our perceptions are "synecdoches" of their putative objects). In the following, I'll use the term "presentation" to cover the Cartesian notion.

If we take TH to be "structured" (that is, as seeming to be disposed in a certain way in a perceptual "space" and as having certain "parts" which bear a topological relation to each other), then it would "appear to us" rather like a picture. Now, if we hold to the Cartesian notion of "presentation", this *picture* is being "shown to" something (the cogito). If we return to the notion of an "iconic sign", the Cartesian theatre is here a cinema – the perceptual state constitutes a "photograph" or "film" of reality which is replayed live to the thinking subject. Thus we have the idea that our perceptions somehow "represent to us" the topological and mereological structure of "external reality".

We can criticise this notion on two levels. First, we can criticise its semiotic accuracy: as the putative cause of the perceptual state, the "state-of-the-world" is in an indexical rather than an iconic relation with the perceptual state, and we cannot therefore hold that the "real structure" of the perceptual object is accurately reflected in the structure of the perceptual state. Imagine, for instance, that reductionism to fundamental particles were true. We could at best hold that the "real structure" of the perceptual object ***causes*** the apparent structure of the perceptual state, and seek regularities and covariances between one and the other (though this would rather be a scientific investigation into quantum-level interactions between highly complex emergent systems, and in no way a phenomenological investigation). This is similar to correlating, say, a rash, a fever, and a cough with the presence of a certain virus – the virus in no way "resembles" the rash, the fever, or the cough: they are merely indices to its presence.

The second criticism is more directly philosophical, and depends on our rejection of the Cartesian notion of "representation". If we refuse the notion of the perceptual state as a sign "presenting" a real state-of-the-world TO some other thing (the cogito or whatever), there is no longer any need to postulate an iconic relation between our perceptual state and the state-of-the-world, as there is no distinction between the topological and mereological structure of the perception and its informational content: the structure does not "transmit" information – it "is" information. Speaking more phenomenologically, we cannot separate the "reality" of experience from its "contents": there is no distinction between the contents of a perceptual state and its putative Cartesian subject (the experience of "seeing a thermometer" cannot be prescinded from its grammatical subject – there is no "I" to which "seeing a thermometer" is presented, there is just "an experience of seeing some thing which is recognised as a thermometer".

Our phenomenal experiences would seem to be structured in both space and time – the keyboard is "nearer" to me than the monitor screen, and the sound of footsteps upstairs is "more recent" than the sound of a passing ambulance. They seem to have certain parts which bear certain relations to other parts. In our everyday interaction with the world, we don't even need to question whether the structure of our experience corresponds to the structure of the putative objects of experience. If we take "the world" as a domain of entities in causal relations, our experience is part of the world – we interact with the world as it is because we are PART of that world; the puzzle lies in HOW phenomenal experience is "part of the world", not in how phenomenal experience REPRESENTS the world.

We have no reason to ask whether our everyday phenomenal experience of the world "differs" from the world – it is rather "part" of the world. However, we can still investigate the KIND of causal relations obtaining between the putative "object of experience" and the experience. I see the monitor screen as a source of light, and it makes no "sense" to me to formulate my experience otherwise. It is most probably the case that there is a "causal" relation obtaining between the physical state of the screen and the physical state of my perceptual apparatus, and that this causal relation could be described in terms of elementary-level interactions between complex systems. But it's meaningless to seek ***parts*** of "my experience of the screen" in particular fundamental interactions – the experience is an emergent phenomenon, and its "parts" are determined by relations obtaining at the emergent level.