B E R K E L E Y ' S    M E T A P H Y S I C S

EXCERPTS

The mind-body problem

What are the mental and physical worlds? How do they stand in relation to each another? How do they interact? This book offers unconventional answers to these questions.

The mind-body problem is the usual name given to the philosophical controversy that has been raging for several centuries over the disputed and foggy territory that lies at the border between the two worlds - the mental and the physical. It has been painstakingly difficult to get and articulate the right answer to the above questions, as is evidenced by the extensiveness, and polemic intensity, of the writing in this area. Nonetheless I believe the effort of understanding and solving the mind-body problem will almost certainly be worth while.

The mind-body problem is worth studying for its own sake. It can plausibly be regarded as the most important problem in philosophy and the one that impinges most directly on the idea we hold of ourselves. In addition to its own worth, though, an inquiry into how the mind relates to the body may throw some light on a range of phenomena that are widely believed to exist, yet still stand outside the realm that has been recognised by systematic science: the paranormal. The present volume addresses only the philosophical mind-body problem: the next volume in this series applies the philosophy to a proposed solution of the problem of paranormal phenomena.

The conscious mind and the physical brain have sometimes been conceived of as real things of fundamentally different kinds. Nowadays, however, there is a consensus amongst natural scientists that the mental world is reducible in some sense to the physical world; and that physical things are the only sort of thing that really exist. Indeed, it could be described as a powerful orthodoxy rather than as merely a consensus. Claims for the existence of non-physical entities or beings are now normally dismissed out of hand by the scientific community. (Notwithstanding that belief in the supernatural is burgeoning outside the scientific community.)

I shall advance arguments against this conventional belief. And I shall offer further arguments for the opposite thesis that the mental world is the only one that really exists. The physical world, I shall claim, is a useful fiction that has been concocted out of the raw material of our mental life.

There are three main camps in this debate, and I will use the following broad terms to refer to them. Physical monism (or physicalism for short) is the theory that only physical things exist; mental monism (or mentalism) is the theory that only mental things exist; and mind-body dualism (or just dualism) is the hybrid theory that both mental and physical things exist and that neither can be reduced to the other. There are variants and vigorously contended interpretations of these positions, and I shall refine the definition of these positions later.

Where is your mind?

Does a conscious mind lie anywhere in physical space - or is it, in some sense, 'outside space'? The two lines of argument that I shall propose in this chapter are intended chiefly as an introduction to the mind-body problem. I hope that they will, at least, persuade the reader that physicalism is far from self-evident, and that the physicalists need to argue for their orthodoxy - they cannot take it as read. I would, of course, like to do more than this, as I am aiming to defend the opposite theory of mentalism. Realistically, though, I cannot hope to win converts to a view of the world so radically at odds with the prevailing view.

I am certainly not questioning whether the physical correlates of the mind can be located in space. Clearly they can. We know that a lot, and conceivably all, of what happens in someone's mind is strongly correlated with electrical and chemical activity in the nervous tissue in her head. That is to say, the physical correlates of the mind are obviously beneath the cranium. Neuropsychologists have made much progress in identifying which bits of the brain support which elements of the mind; and we may reasonably assume that they will make much further progress. Nevertheless, there remains the question of whether the mind itself - as opposed to its physical correlates - can be said to be in the brain. (Clearly, if the mind could be localised at all, then it would be in the brain. It would certainly not be anywhere else.) This, however, is the question that I am addressing: can the mind itself be in any place?

If the answer to this is no, then there are two alternative conclusions that might be drawn: either the mind does not exist at all, or it does exist but lies outside the entire system of space and spatial coordinates. It is, I believe, indubitable that the mind exists, as we each receive proof of this in every moment that we are awake; so, the conclusion that I am driving at is that the mind exists outside the realm of physical space. (I shall, however, address the question of whether the mind exists at all more fully below, because some philosophers like to make the perverse claim that their minds do not exist.)

Lest this sound rather mystical, I would at once draw attention to more mundane categories of things that lie outside the realm of physical space.

• Numbers
  Numbers do not exist in any particular place - although each number can be embodied in particular things, for instance the number 3 can be embodied in the three pencils on my desk. But it makes no sense to say that the number 3 itself resides in some special corner of the world.
• Algorithms
  A computer algorithm does not reside in any particular place. Though, of course, embodiments of algorithms can be 'localised' here and there - say, in the form of a program written in IBM's Fortran code on a sheet of paper; or the same sequence of characters displayed on a screen, or held on a magnetic disc; or encoded in an executable form in the processing unit of a computer.
• Institutions
  Institutions do not reside in space. The University of Oxford, as Gilbert Ryle famously remarked, does not lie in any definable volume of space, although its physical embodiments (its members and buildings) are certainly in particular places.

We must make an analogous conceptual split between the mind on one hand, and its correlate in the tangible brain tissue on the other. And, having made that conceptual distinction, I wish to argue that the mind itself is not localised in space - even though its tangible substratum (the brain) is.

I would emphasise that the analogies I have just made are only partial. Too much has been made by some authors of the similarity between the 'software-hardware' relationship in computers and the 'mind-brain' relationship in people. To be sure, there is a useful comparison to be made, but it is not a complete analogy. And I shall argue later that it is, in some important regards, radically wrong to suppose the mind to be the same kind of thing as a computer algorithm.

Nevertheless, I think these analogies are useful for focusing on the conceptual split that needs to be made between mind and brain. I have found that many people automatically fuse together their concepts of mind and brain. For them, the proposition that the mind cannot be localised - even though the brain obviously can be - simply does not make sense. I hope that reading this chapter might make the distinction clear.

In this chapter, I shall offer two broad lines of argument: an 'offensive' line that makes the positive assertion that the mind cannot be localised; and a 'defensive' line that attempts to repel the orthodox claim to identify the mind with something in the brain. On each line, I shall address what seem to be the principal counter-arguments.

Creeping androidisation

In his book, Chalmers refers to his arguments as 'absent qualia', 'fading qualia', and 'dancing qualia', but I do not think that either that division or those names are especially helpful. The thought experiment of what I shall call creeping androidisation is mostly described by Chalmers as having to do with 'fading qualia', but it really involves all three.

In this experiment, we throw ethics to the wind and embark upon irreversible brain surgery on you. By means of key-hole surgical techniques and micro-robotics, we make an opening in your cranium, and proceed with an automated plan of removing each of your brain cells in turn and replacing it with an artificial device that performs the same function as the neuron it replaces. The argument does not hang on the nature of the device: it might be a silicon chip, for instance. For the purposes of the argument, we stipulate only the functional properties of the device, its input-output characteristics.

If the neuron's behaviour were completely deterministic, that is, given a particular pattern of input signals, a certain output signal is produced, then the replacement device would have an identical input-output relationship. Since, however, the nerve cell's behaviour will, to some extent, be stochastic (that is, involving a certain element of chance), the artificial device that replaces it will use an internal random-number generator to yield an output with the same stochastic properties - the same probability distributions. As far as the argument for functionalism is concerned, those random numbers may be pre-coded into a table of random numbers (strictly speaking, 'pseudo-random' numbers) or they may be generated in real-time by some quantum noise device. By definition, functionalism stipulates only the functional behaviour, not the internal mechanism. So, as far as functionalism is concerned, it makes no difference how the random numbers are generated. But, as we shall see later, the choice of internal mechanism will actually have a crucial impact on androidisation.

Of course, if the glial cells also perform a functional role, then they too are replaced by part of the new implanted device.

Changing a few brain cells will have negligible effect. It is uncontentious to say that you are unlikely to notice any difference at all after the first one or two nerve cells have been replaced. But what is going to happen over the next hour, as the automated surgical machine goes to work on your whole brain? Will you notice any difference in your conscious mental experience? When the procedure is complete, when all your brain tissue has been cut out and flushed down the drain, and you have a complete android brain, will you still be conscious?

One possible answer, the one that Chalmers offers, is that you will notice nothing untoward at all. You will see the colour of the sky, smell and taste your lunch, feel angry about missing the train, just as you did with your original brain. If this is correct, then functionalism is established: we will have shown that consciousness is present purely by virtue of the physical execution of a function. This is because all we stipulated about the replacement devices was their functional characteristics. Therefore, if consciousness persists during and after androidisation, then it follows that the consciousness is present just because of that functionality.

Some people will feel quite happy with this answer, but most people, I believe, will feel instinctively that this cannot be right. One feels that if one's brain were to be replaced by a piece of machinery, then one could not possibly have the same mental experiences.

As Chalmers argues, however intuitively appealing that answer may be, it is a hard position to hold on to rationally. What if, for instance, we were to show you colour samples every five minutes during the hour that the androidisation takes place, and ask you whether they still look the same. Since the question enters your brain (be it the original brain, or the electronic replacement), as electrochemical signals on the auditory nerve; and since the answer comes out of your brain as similar signals along nerves (going to your hand if you are giving a written answer, or your mouth and throat if you are giving a spoken answer); then the electronic brain will give the same response as the original biological brain. For, recall that each replacement device is so constructed that it gives the same output for any given input. So, during this process, you will consistently report that your conscious experience is continuing unchanged.

So, if we accept the starting assumptions, then the only position left open for anyone who is trying to resist Chalmers' conclusions would be to claim that the conscious experience does indeed change, but that you cannot report it. This, as Chalmers convincingly argues, is massively implausible. It is, moreover, a basic and necessary assumption for any investigation of consciousness that when you report your mental experiences, you will be able and willing to do so honestly and accurately. It is introspectively hard to imagine experiencing, say, the annihilation of your visual field (as your visual cortex is androidised) while continued to assert that you can see things every bit as clearly and colourfully as before.

It is true that John Searle, for instance, wants to say that the conscious experience will indeed be degraded and eventually vanish, even though this cannot be reported. He envisages a bizarre and nightmarish situation in which you will watch your consciousness vanish while your body keeps on saying that everything is fine. That, however, raises the question of where your brain would implement and embody the (unreportable) belief that your conscious experience was vanishing. We have already established, by hypothesis, that all the information processing is going on just as before. The only way in which you could possess the belief that your consciousness was vanishing is by having a non-physical mental substance in which to implement the belief.

Chalmers thus takes us down a very strong chain of thought leading, seemingly inescapably, to functionalism. It seems to me the only way out is to assert that the replacement devices, although they have been designed to reproduce the behaviour of biological neurons, and tested in vitro to show that they do so, nonetheless behave differently in vivo. This is not a conclusion that is acceptable to any strictly physics-minded property dualist such as Chalmers. Nonetheless, I think this solution is correct. I shall elaborate on this claim below. First, however, in the next two subsections, I shall argue that Chalmers' property-dualistic functionalism is inconsistent even on its own terms.

Summary of Berkeleianism

The main points of Berkeley's mental monist philosophy can be stated very briefly. Early in his "Principles" he says that the following is obvious

... that ... all those bodies which compose the mighty frame of the world, have not any subsistence without a mind, that their being is to be perceived or known; that consequently so long as they are not actually perceived by me, or do not exist in my mind or that of any other created spirit, they must either have no existence at all, or else subsist in the mind of some eternal spirit; (Principles, 6)

it being perfectly unintelligible ... to attribute to any single part of them an existence independent of a spirit. (Principles, 6)

Berkeley probably hoped that people would immediately see the truth of this claim. To help them along, though, he makes the following remarks in his next section:

The sensible qualities [of all objects] are colour, figure, motion, smell, taste, and such like, that is, the ideas perceived by sense. Now for an idea to exist in an unperceived thing, is a manifest contradiction; for to have an idea is all one as to perceive: that therefore wherein colour, figure, and the like qualities exist, must perceive them; hence it is clear there can be no unthinking substance or substratum of those ideas. (Principles, 7)

Collecting the steps of this argument together, we can summarise it as:

1. Physical objects appear to comprise sensory qualities inhering in a real, but not directly perceptible substratum.
2. Because the substratum is imperceptible, the only idea we could have of it is an abstract general one.
3. But abstract general ideas are impossible.
4. Therefore objects are nothing but assemblies of specific sensory ideas.
5. But sensory ideas can exist only in the mind.
6. Therefore objects can exist only in the mind.
7. Hence: immaterialism.

So far, we have considered only Berkeley's theory of objects, which he says are constructed from ideas. This theory is summarised in the Latin motto, esse est percipi, 'to be is to be perceived'. In addition to such objects, there are also active agents or 'spirits', for whom the converse motto applies, esse est percipere, 'to be is to perceive'. Of those agents, Berkeley writes:

A spirit is one simple, undivided, active being: as it perceives ideas, it is called understanding, and as it produces or otherwise operates about them, it is called the will. Hence there can be no idea formed of a soul or spirit: for all ideas whatever, being passive and inert ..., they cannot represent unto us, by way of image or likeness, that which acts. (Principles, 27)

Admittedly, Berkeley was not explicit about what a notion is. My reading of Berkeley is that the notion of a spirit,and hence one's knowledge of one's own self, is constituted by acts of perception and volition. This is not how we normally think of 'knowing' something, but if we dissect the concept of knowing, we will find that the difference is not that great. When we talk about knowing the self or spirit, we are concerned with the knowing of a subject, rather than the knowing of an object, so there will inevitably be some difference in kind between the two acts of knowing. That is to say, the having of a notion (in Berkeley's sense) is necessarily different in kind from the having of an idea. Having an idea involves two things, one phenomenal, and one epistemic. The phenomenal part involves entertaining a sensation or image comprising qualia; the epistemic part involves being in a state whereby can one can use the knowledge of the idea rationally in the course of one's subsequent actions. For example, if I see it is raining, then I may have the visual idea of the rain falling: in part, that is constituted by the qualial experience, and in part it is constituted by the knowledge of that idea. Using that knowledge, I can report the fact that it is raining, and I can take my umbrella when I go out. When I act, however, in the instant of my exercising my free will, there is no qualial component, nor do I enter a state that can be used subsequently in rational deliberation: rather, the act is an event in which I do use a perceptual content to influence rationally the free choice of an action. Performing that act is itself a kind of knowing: one might describe it as knowing through doing. (It is similar to practical knowledge: knowing how to ride a bike is not constituted by having certain ideas, but is manifested through the act of riding the bike; likewise, knowing that one exists (as a spirit) is manifested through acting volitionally in response to one's perceptions.)

Figuratively speaking, we could say that knowing 'that one exists' (as a self) is rather like knowing 'how to exist'.

One final point of terminology. Berkeley sometimes uses the word 'God' to refer to the powerful mind that creates the world around us (although he does occasionally refer to 'spirits' in the plural performing this role, and I assume he is then alluding to the angels). He also uses the traditionally reverential male pronoun 'He' for God. I am unhappy with both terms, but I shall stick with them in this section for ease of exegesis. In later chapters, I shall introduce the more generic and less religiously loaded term 'metamind' and the appropriate pronoun 'it' in their place.

Corollary: consciousness is not physical

David Chalmers has drawn a lot of fresh interest to what has historically been known as the mind-body problem. He has renamed it the 'Hard Problem', to separate the philosophical problem of accounting in principle for how consciousness can arise in a physical world, from the 'Soft Problem' of scientifically studying the detailed correlation of mental and neural activity.

Under the Berkeleian theory of mental monism, of course, the Hard Problem becomes an easy problem, as the physical world is no longer considered to exist.

Nevertheless, the reasoning that has been presented in this chapter as an argument for mental monism can also be used independently to argue that consciousness is not reducible to the physical world. So, if you feel too uncomfortable with the radical claim that reality is primarily mental, you might nonetheless settle for the intermediate claim that reality is not exclusively physical.

The world that is described by physics consists of entities and operations that are defined wholly by their logical relations with other entities and operations within that world. A number of fundamental terms are allowed, such as mass and space, and all others are derived from them by declaring formulaic relationships between them.

Terms which denote things that are defined by their intrinsic qualities, as opposed to being defined by logical relations, do not and cannot feature in the language of physics. The immediate conscious sensation of seeing the colour red, for instance, is not something that is defined by its relationship to other things. There is a qualitative aspect that cannot be captured by relations. The terms that denote things of this kind can be defined only ostensively, not formally. We can do no more than say, "There! Look at that - that's what I mean by 'red'!"

Hence qualitative mental experiences can, by definition, never exist in the physical world. This is a metaphysical claim, but it is also a claim about language - the deeper reason for this is that the physical world itself is a fiction, a verbal construct.

A more concrete way of looking at the situation is to consider that all the physical facts of the world could be printed in black and white in a huge library of books. Somebody who read and understood all these books still would not be informed of the experience of red. Likewise, consider Jackson's thought-experiment of the neuroscientist Mary: she somehow lives, studies, and works in an exclusively black-and-white world. There, she studies the neurophysiology of human vision, and acquires all physical facts about human colour vision. She knows all about different wavelengths of light, and how they are labelled as "red" and "green", and so on, by people with colour vision. Then, one day, she escapes and sees colours for the first time. Now, she has acquired some new knowledge that she did not previously possess: what the colour red looks like. This, as Chalmers argues, shows that the world of conscious experience contains facts other than those of the physical world.

As we have seen, the reason that the voluminous description of the physical world, or Mary's laboratory notebooks, can capture all relevant physical facts without mentioning conscious experience is that they need only mention things that have been defined within the closed verbal system of the physical sciences. They include no ostensively defined terms that refer to the contents of conscious experience.

This situation is more radical than some people think.

Some people, such as Chalmers himself, want to integrate consciousness into the natural sciences by making it out to be another fundamental element of reality, alongside mass and space. This misses the point. Consciousness is an ontologically different kind of thing from mass or space. We know this because words that denote things in the conscious world (including the word "consciousness" itself) perform a different kind of function from words that denote physical things. The former can be given only ostensive definitions, the latter only formal definitions. Consequently, as I have argued above, the physical terms are incapable of bearing any referential meaning: the only meaning they have is a formalistic meaning constituted by their use as tokens within a closed language-game. Hence, the things they denote cannot exist. Contrariwise, the terms that denote things within the world of consciousness can and do bear referential meaning, and what they denote can exist. So, for Chalmers to suggest that consciousness could be a basic part of reality alongside mass and other physical primitives is to make a category-mistake (in Ryle's celebrated phrase). The physical world is, necessarily, derived by construction from the conscious world.

© Peter B. Lloyd, 1999. Last modified 28th February 2000.

[ Ursa Software Home Page | Peter Lloyd Home Page | Ursa Publishing | Mind Detox ]