Sunday, November 05, 2006

Consciousness

I subscribe to a physicalist view of consciousness. That is, I believe that consciousness is fully reducible to natural processes occurring within the material brain. Whilst we don’t yet have a full working theory of consciousness, I consider that this is most likely attainable in principle. However, even if such a complete naturalistic theory can never be formulated, I do not believe that this implies that consciousness requires anything other than matter/energy operating within space/time, and certainly do not accept that anything supernatural is required. In such a case, I think that our inability to explain consciousness (and such related concepts as qualia) by natural means alone would be due to its complexity and to limits of our imagination, ingenuity, language, or knowledge, and would not be because consciousness is dependent upon some non-material substance or additional properties for its existence.

The theories of consciousness that propose the mind to be dependent upon some non-material substance, or upon some additional mental properties, are known as dualist theories. These can be broadly classified as either Substance Dualism, in which the mind is thought to be a separate non-material substance, and Property Dualism, which proposes that when matter is organized in the appropriate way (i.e. in the way that living human bodies are organized), mental properties emerge. Substance Dualism was notably defended by Descartes, but is today associated primarily with theology – where this non-material substance is identified with the soul.

Substance dualism seems to me to be multiply fallacious, and perhaps impossible to verify or falsify. I would contend that it falls foul of Occam’s razor, since it posits some unknown supernatural substance to explain the mind, which we know to be closely associated with the material brain, and leaves more questions unanswered than it actually answers. To the proponent of substance dualism, I would submit the following questions:

1. On your hypothesis, what exactly is this other ‘stuff’ that mind consists of?
2. Since the material brain is intimately tied up with consciousness, what exactly is it that this substance does that the material brain does not do, and how does it do this?
3. How do the material brain and this other stuff interact to produce consciousness?
4. Why is the material brain required to play a part in consciousness at all? Why not just have ‘mind’ that is completely independent of any physical brain?
5. How do you explain the fact that brain damage can result in a (temporary or permanent) loss of consciousness?
6. How do you explain that stimulating the brain in certain ways repeatedly causes certain feelings, memories, and thoughts?
7. How does this immaterial substance cause a physical effect e.g. the firing of neurons and releasing of chemicals?
8. At what point in human evolution did we acquire this ‘mind’, how did that happen, and where did it come from?
9. Do other animals have any of this substance?
10. At what point during the evolution of the universe did this mind stuff come into existence, and how did that happen?
11. Are we born with this mind substance already present, or does it enter later? If the latter, when does it enter, and how?
12. How might we verify or falsify your theory of consciousness?

Property Dualism does not suffer from many of the problems of substance dualism, but it still needs to explain how these proposed emergent properties of the mind are able to cause physical effects within the material brain.

My own particular view is that some non-naïve variety of Identity Theory is the most compelling explanation of consciousness (although I might be convinced by some other materialist theory of mind, such as Functionalism). This is a materialist theory that offers a strongly reductive option by identifying conscious mental properties, states and processes with physical ones, most typically of a neural or neurophysiological nature. In other words, our conscious experience when we have a thought, or when we experience some phenomenal experience (such as seeing a colour, smelling a flower, feeling pain etc.) is identical with the specific brain state that accompanies this experience.

One might ask why physical processing within the brain should give rise to the rich inner life that we experience. However, on the Identity Theory, the concept that consciousness somehow ‘arises’ from something is a misnomer – and one that has led to much debate about what this extra something is, and how it arises from and interrelates to the physical brain. I would contend that consciousness is in fact identical to physical brain states, so a thought is nothing more than a particular configuration of firing neurons, chemicals, and other physical phenomena within the brain. Likewise, the experience of seeing a colour, or feeling pain is just our first-person experience of some brain state or other. These states are not necessarily the same from person to person, or even each time we experience something (token identity).

Now, this seems at first to be counterintuitive. How can a thought be nothing more than a brain state? Surely, our consciousness is more than this? In fact, the most intuitive idea would seem to be that our consciousness is something entirely separate from the physical brain – some other substance entirely. This is one of the appeals of Substance Dualism.

By contrast, Property Dualism, which is more modest in its claims, holds that there exist mental properties (i.e., characteristics or aspects of things) that are neither identical with nor reducible to physical properties. Conscious properties, such as the colour qualia involved in a conscious experience of a visual perception, cannot be explained in purely physical terms and, thus, are not themselves to be identified with any brain state or process.

However, on the Identity theory, both of these viewpoints are mistaken. There is no other ‘substance’ of which mind is constituted, and there are no mental properties that are not reducible to the physical. Although this may indeed be counterintuitive, perhaps we should instead ask exactly why our first-person experience of brain states would not be what we call thoughts, memories, feelings, qualia etc. How should these brain states manifest themselves, if not in these ways?

The brain states and our conscious experiences are just two sides of the same coin, two ways of seeing the same phenomenon – one from the first-person, and the other from the third-person. I am saying that the correlation between the brain states and our experience is more than just a correlation; it is an identity. Just as the presence of heat when molecules are excited is not just a correlation – they are different ways of looking at the same phenomenon. It is a (masked man) fallacy to say that we recognise heat, but do not recognise moving molecules, therefore heat is not moving molecules. In the same way, it is a fallacy to say that we know what thoughts and sensations are like, but do not know what brain states are like, therefore thoughts and sensations are not brain states.

Interestingly, when the brain is monitored during the process of seeing something, and the subject then being asked to remember this something, the neurons in the brain seem to fire in a similar way. This suggests that memories are just the brain recreating a simplified version of its internal state when the event first happened.

On the Identity theory, it is just a ‘brute fact’ that there are such identities, and the appearance of arbitrariness between brain properties and mental properties is just that – an apparent problem leading many to wonder about the alleged explanatory gap. Qualia would then just be identical to physical properties. There is therefore no real explanatory gap on this theory, and science will, in principle, be able to explain consciousness fully from a third-person point of view. Of course, as science explains things from a third-person point of view, it can never describe ‘what it feels like’ to smell a flower, since this it not the type of problem that it addresses, but I don’t think this implies that ‘what it feel like’ is some sort of knowledge that requires the mind to have additional non-physical properties.

This concept of qualia (properties of sensory experiences) is one that is often cited when arguing for dualism. However, I would suggest that the assumed connection between qualia and consciousness is actually a red herring. For example, how can we be so sure that a dog does not experience qualia? Why would it not experience some feeling of what it is like to scratch itself, or eat a bone, or see its owner return? Why would a simple creature like a butterfly not experience some feeling of ‘attractiveness’ in a flower that it is drawn towards? Colour vision is present in some lower animals, so why wouldn’t they have some experience of ‘redness’? Hence, I would say that lower animals likely have experience of qualia, and science cannot describe how this ‘feels’ to them, but that doesn’t imply that they possess some sort of non-physical mental property.

I would suggest that qualia are merely functions of perception, and are thus likely to be present in other animals too. We are conscious of our perception of qualia, because we are conscious, but that does not imply that consciousness is a prerequisite for experiencing qualia, or that qualia in themselves tell us something interesting about consciousness. Our experience of qualia will likely vary from that present in lower animals, as we have an additional layer of self-awareness that they don’t seem to possess. However, in my opinion, the experience of qualia alone cannot be used to justify property dualism.

The so-called Knowledge Argument is sometimes used against such physicalist theories. This is commonly expressed by recourse to two famous thought experiments. In the first, courtesy of Thomas Nagel, we are asked to imagine what it would be like to be a bat. We might know everything about the working of the bat’s brain from a scientific point of view, but we can still not imagine actually being a bat. In other words, whilst we might know all physical facts about a bat’s brain, there is still some knowledge missing – the bat’s first-person experience of being a bat. Therefore, so the argument goes, the materialist theories of consciousness are flawed.

However, this argument does not refute the theory that the bat’s conscious experience is not identical with its brain states. The fact that we can never know this experience from a first-person point of view does not mean that it is not a purely naturalistic process. After all, the only way in which we might experience being a bat would be for us to run some kind of ‘bat simulation’ within our own brain. Even if this were ever scientifically feasible, we would still lack the language to be able to express our experience - not to mention that the memories of this utterly alien event would need to be translated to some other form to be stored with our own brain. This would, by its very nature, distort and anthropomorphise the event.

The other well-known thought experiment (created by Frank Jackson) concerns a person called Mary, who is kept in a black and white room from birth. She becomes a brilliant neuroscientist, and an expert on colour perception. She knows all of the physical facts about human colour perception, but has never seen, for example, the colour red for herself. Upon leaving the room, she experiences red for the first time, and apparently learns some new fact that she didn’t know before. Since we presumed that she knew all physical facts about colour perception, this new fact must be non-physical. Therefore, the materialist viewpoint is shown to be fallacious.

However, on the Identity theory, Mary is not learning a new fact, she is merely learning the same fact, but from a different point of view – first-person rather than third-person. The thought experiment does not refute the identity between the brain state and Mary’s experience of seeing red, so the metaphysics of the physicalist perspective is not challenged.

Actually, I think that the Mary's Room thought experiment is an example of the Masked Man Fallacy.

Here is an example of this fallacy (courtesy of Stephen Law):

1. John Wayne is someone that Michael knows to have appeared in True Grit
2. Marion Morrison is not someone that Michael knows to have appeared in True Grit
3. Therefore, Marion Morrison is not John Wayne

But this conclusion is fallacious, as John Wayne was Marion Morrison's stage name. This illustrates that personal knowledge or belief is not a property of an object that can be used to disprove identity.The Mary argument can be rephrased in a way that follows the example above, and makes the fallacy much more evident:

1. Mary has all the physical information concerning human colour vision before her release.
2. Upon her release, Mary believes that she comes to know new information about human colour vision
3. Therefore, not all information is physical information.

Now we can see the problem. Mary might believe that she comes to know new information upon her release but, in reality, she is just seeing the same information from a different perspective. So, she comes to know no new facts about human colour vision, but just sees the same facts in a different way. Therefore, this is an example of the masked man fallacy.

Paul Churchland (in the essay 'On functionalism and materialism’) makes a similar point when discussing this argument. He suggests that it is making an error of equivocation with the word ‘know’. Knowing in the sense of knowing physical facts is not the same as knowing based upon first-hand experience of the phenomenon. Furthermore, we can allow for a duality or multiplicity of types of knowing, without committing to there being non-physical information. Rather, there are multiple mediums through which we can come to ‘know’ the same facts.

To summarise, I believe that thoughts, memories, sensory experiences etc. are just our first-person experience of brain states. That seems quite alien to us, but the fact that we don't recognise thoughts and sensations as being identical with brain states doesn't mean that they are not. Consciousness is just what you get when you have a material brain consisting of billions of neurons, and a feedback loop that gives it self-awareness (not located in any particular spot – this is a category error). There is no additional non-material substance required, nor are there any emergent mental properties. Science may never be able to describe what it ‘feels’ like to see the colour red, for example, but I think it is an unjustified leap to think that this implies that the brain possesses some non-physical properties. The limitation here is that we are trying to describe first-person experiences in third-person terms – and failing. In my opinion, the more parsimonious explanation is that our first-person experiences just ‘are’ these brain states, and the limitation is just one of description.

For an interesting discussion of qualia, see this. For an interesting paper about the explanatory gap, and why it may be illusory, see this and also the article by Thomas Clark here.

Wednesday, November 01, 2006

Considering a minimal god hypothesis

For this post, I will restrict my argument to a minimal god hypothesis that was proposed to me by somebody on the Butterflies and Wheels website.

This god is defined as follows:

1. Rational
2. Omnipresent in physical space, with the universe being part of God, rather than as a separate creation.
3. Neither omnipotent nor omniscient in that the future is undetermined, and human actions are free.

Now, at this point, I think that we need to answer a very pertinent question. Does this god have any observable effect upon the material universe? That is, does the existence of this god make our universe different in any predictable and observable way from how it would have been if this god didn’t exist?

Of course, the trite answer to this is yes, as without god the universe wouldn’t exist at all. However, that merely begs the question, and so gets us nowhere. So, are there any other predictions that we can test by observing the universe? If yes, then we are at liberty to test these predictions, and compare with reality. Most theists do in fact attribute all sorts of characteristics to their god. We can then make predictions about how such a god would be expected to manifest itself, order the universe, and to act. However, if you wish to declare that your god is transcendental and immaterial, is undetectable by any scientific means, and that we can make no predictions about the type of universe that we would expect to see if it exists, then I think that you have a problem. For, if the existence of your god entails exactly the same observations already entailed by naturalism, then there is no reasonable argument to be made for believing it - as I shall seek to demonstrate.

Although I can never hope to prove that such a transcendent and immaterial god doesn’t exist, the principle of Occam's razor dictates that we should reject such a hypothesis. For, if we have perfectly good explanations for how the universe behaves that are based upon what we can observe, and what we can infer from this, then why would we want to introduce into our theories a supposed entity of a type never encountered before in science, and whose existence we can never verify or falsify – even in principle? To be sure, our explanations are not complete yet, but I see no reason to believe it better to introduce such an ad-hoc entity as an explanation, rather than to just say that we are still working on it. Is it more rational to say that we have just a material universe (which is all that we can observe), and which is ordered in such a way that we can attempt to fathom the way that it operates; and to then attempt this fathoming by adductive, deductive, and inductive reasoning based upon what we know? Or is it more rational to say that, in addition to the material universe that we can observe, there exists some universe-creating transcendental, immaterial entity whose existence or effects we can never observe – and hence which we can never verify or falsify? I would say that the former is more rational.

This logic still applies if you are to contend that this god does, in fact, do all sorts of things that we can observe, but that it makes it appear as if these things happen purely by some non-intelligent mechanistic process.

Now, one way around this is to suggest that the god hypothesis is actually a more rational answer in the absence of any other evidence, as it actually explains the existence of the universe in the most parsimonious way. However, I would contest this, as there are already plausible scientific explanations for the existence of our universe. We are able to construct viable working scientific theories to explain how the universe might have come to exist from a prior state of ‘nothing’ (which I would define for this purpose as an absence of space, time, energy, and matter. See this, or this for more information), or might have arisen spontaneously as a quantum fluctuation in the quantum foam as part of an Eternal Inflation model, and then evolved to what we currently see. These theories are based upon our observations, and our scientific knowledge and understanding of the world, and are able to predict exactly the type of universe that we observe. They do this without the necessity of positing some ad-hoc creator, or introducing any other ad-hoc assumptions, and are therefore more plausible, parsimonious, and powerful explanations than the god hypothesis.

For my purposes here, I shall concentrate on the Eternal Inflation theory. Now, you may object that this theory doesn’t explain why there exists a multiverse in the first place, nor why there is some sort of process that allows universes to come into existence from it, so we still don’t have a complete explanation. That is indubitably true. However, your god hypothesis gets you no further, as you need to explain why your god exists at all; never mind why it has the characteristics that it does, and why it would choose to create the particular universe that we see (which is almost entirely lethal to our type of life), rather than one far more suited to us, and far more economical in scale and age. So, if you are proposing that there is no way, even in principle, that we can attempt to verify or falsify your god hypothesis by recourse to observation (as it is immaterial, transcendental, etc…), then I think that the most rational option is to disregard it. Contrary to what one might initially imagine, I would say that such a god hypothesis is not the most parsimonious or plausible one available.

However, you may choose to go another route. You might say that the apparent fine-tuning of the universe is evidence for the existence of your god. Such a statement does indeed constitute an explanation for the observations. However, if we are able to formulate a scientific hypothesis that also explains these observations, but does not require the introduction of this ad-hoc universe-creator, then it will be both more parsimonious and plausible than the god hypothesis. In fact, we do have such a hypothesis. The Eternal Inflation multiverse theory follows naturally from the combination of the Big Bang and Inflation theories – both of which are sound scientific theories. If we are to accept these two as our starting point, then the existence of a multiverse is the de facto conclusion. In fact, to deny the existence of a multiverse, we would need to propose additional ad-hoc laws that force the Big Bang (followed by Inflation) to have happened only once.

I should stress here that the Eternal Inflation theory (and other multiverse concepts) was not just pulled out of a hat in order to explain away the apparent fine-tuning found in the universe. Rather, it followed naturally from previously accepted theories. However, it does nevertheless explain the apparent fine-tuning as, in an infinite set of ‘pocket universes’, there is bound to be one with our particular characteristics. In such a universe, we are able to exist to have these thoughts, and to make the observations.

As a further observation on this, the proponent of the fine-tuning argument is also required to explain why a universe designed to ultimately produce us would be a huge and almost entirely radiation-filled vacuum, and why it would take billions of years before we would evolve. The scientist can explain this, as any apparent teleology is purely the result of natural processes. As a matter of interest, from a scientific point of view, teleology has been almost entirely expunged from our theories. Apparent teleology in the evolution of life (which, incidentally, is a fact. The explanation for this evolution is the theoretical part) has been explained by Darwin’s Theory of Natural Selection. In addition, from a scientific point of view, any apparent teleology in the ordering of the cosmos is now explained as the result of mindless natural laws.

Of course, there are some interesting loose ends here. An obvious one here is consciousness. I believe that consciousness is ultimately reducible to configurations of matter/energy within space/time. It is possible that we may never have a full explanation of consciousness, but I think that giving up on this venture now would be rather premature. Even if we are never able to explain it fully, we are not justified in inferring that consciousness requires some other ‘something’ that is not part of the material universe. I also tend to think that the problem of qualia is a bit of a red herring, as I suspect that it is not dependent upon consciousness at all – but merely upon perception. Ultimately, I feel that any inability to explain these things by natural means would be as a result of our lack of imagination, ingenuity, language, or knowledge, and is not because they are dependent some non-material stuff for their existence.

Another interesting case is that of mathematics. I tend to think that mathematics is in some sense a fundamental property of the structure of our universe, and is not a purely human construct (although, obviously, some aspects of it are a human construct). So, I do not subscribe to the Conventionalist position, but am inclined rather more towards the Realist stance. However, I think that whilst mathematics is in some sense 'out there', this is a result of the fundamental structure of space/time, and it is not in some way transcendent. Of course, we are still left having to explain why we find that mathematics is a part of our universe, but that is just one part of the overall problem of the existence of the universe - and for which the god hypothesis is not the best answer. Interestingly, Max Tegmark's Ulimate Ensemble does allow for worlds in which mathematics is radically different from ours (for example, in which 2 + 2 = 5).

Perhaps at this point I should attempt a summary of my points.

· We only know that the material universe exists (in as much as we can ever really ‘know’ anything).
· A theory that explains the existence of our universe by recourse to what we already know to exist, using and building upon existing theories that are logically sound and evidentially well supported, and making reasoned inferences from these theories, has to be more plausible than one that does not.
· In addition, a theory that requires the introduction of entirely ad-hoc elements has to be less parsimonious than one that does not.
· A theory that predicts the specific universe that we observe, rather than just some universe, has to be more powerful than one that does not.
· A theory that can be tested (and possibly falsified) has to be preferred over one that has no such options available.
· The god hypothesis as proposed by you requires the positing of some ad-hoc ineffable and immaterial universe-creating entity. The existence and characteristics of this entity cannot be explained in terms of the things we know to exist in the material universe, and by known physical laws. In addition, it cannot be observed as some distinct entity within the material universe. Further, its existence alone makes no predictions about the structure of the universe that we would expect to observe. However, by means of the introduction of a further set of completely ad-hoc assumptions, it can be made to fit with the evidence that we see around us. By such means, the existence of this entity becomes entirely untestable and unfalsifiable.
· By contrast, the Eternal Inflation multiverse theory explains the existence of our universe in terms of things that we know to exist, and of other well-founded and evidentially supported theories. It requires the introduction of no entirely ad-hoc elements. Further, in many specific ways it does predict the type of universe that we see around us. As yet, it cannot be tested or falsified directly. However, physicists are currently working on indirect methods by which it might be tested. Hence, the Eternal Inflation is a more parsimonious, plausible, and powerful hypothesis than the god hypothesis. It might not be true, but is at least a better explanation of what we observe to be the case than is the god hypothesis. Therefore, it is less rational to opt for the god hypothesis as one’s explanation of the existence of the universe, as better explanations exist.

At this point, the ‘sophisticated’ theist might be tempted to reach for one of the standard arguments for the existence of a god. However, this doesn’t really get them anywhere, as all of these arguments have well-known rebuttals.

For example, three perennial favourites are:

The Kalam Cosmological Argument takes the following form – 1) Everything that begins to exist has a cause. 2) The universe began to exist. 3) Therefore, the universe has a cause. 4) This cause is God. However, this argument fails, as the first and second premises are not demonstrably true. Current cosmological theories propose an eternal multiverse; and physicists have already proposed mechanisms whereby uncaused events can happen. Even apart from this, if we were to accept the conclusion of point 3, this does not imply that the cause is any particular god (or even a god at all: it could be a highly advanced alien civilisation), so the argument fails doubly.

The Cosmological Design Argument states that the precise fine-tuning for life of the fundamental constants appears more epistemically probable given theism than it does given naturalism. However, this fine-tuning for life arises as a natural consequence of cosmological multiverse theories, by means of the weak anthropic principle. That is, a universe in which we have evolved must necessarily have the appearance of fine-tuning for life. Within the infinity of possible ‘pocket universes’, some will allow for life to develop, and some others will not. We necessarily exist in a universe that does. As with the previous argument, this argument does not imply that the universe creator is any particular god, or even a god at all.

The classic Ontological Argument has the following form: 1. God is the most perfect ('the greatest') being conceivable. 2. It is more perfect ('greater') to exist than not to exist. 3. Therefore, God must exist. However, as Kant pointed out, existing is not a property of a thing at all i.e. existence is not a perfection. Therefore, the second premise is false. Even if we were to allow the second premise to stand, we could substitute things other than God in the first premise i.e. X is the most perfect Y conceivable, and thereby prove its existence – which is an absurdity. Even if we were to grant the validity of the argument, we would have to ask whether the entity whose existence we are proving bears any resemblance at all to the god commonly conceived of by the Abrahamic religions? The God depicted in the Bible, for example, is far from perfect by any normal definition of the word. In fact, I would go as far as to suggest that such a ‘perfect’ entity is not the god of any known religion.

So, all of these arguments fail, and, even if they didn’t, do not lead to the conclusion that the god in question is one of those commonly conceived of. It might instead be a committee of gods, a supremely evil god, or a highly advanced alien civilisation etc. Actually, this same reasoning applies to almost any argument in favour of the minimal god hypothesis.

Now, back to the previous discussion. There is one fundamental problem with the Eternal Inflation theory. We are left with no fundamental explanation of why there is some universe-creating mechanism at all, and why the proposed mechanism works in the way that it does. We are just forced to accept this as our ‘brute fact’ – our starting point from which to deduce everything else.

It is at this point that one might be tempted to think that the god hypothesis scores its knockdown win by positing god as the ultimate explanation. However, a little more thought shows that this victory is in fact illusory. For, why does god have the characteristics that it does, and why create the particular universe that we see? More to the point, why does this god exist at all? Why not just no god and no universe? In other words, the proponent of the god hypothesis is forced to accept the existence of such a god as their ‘brute fact’. This god does not come from anywhere, or exist for any reason (since this would imply some Meta rule of existence and attributes applicable to gods) - it just exists. So, the proponent of the god hypothesis has the same problem here as the proponent of the Eternal Inflation theory – with the addition that their hypothesis is less parsimonious, plausible, and powerful, so it loses again.

The proponent of the god hypothesis might try to insist that their god is somehow a ‘necessary being’, but no formulation of this argument has ever avoided rebuttal, so this line of reason is not at all persuasive. As Hume pointed out, insofar as this concept coheres at all, we might as well just posit the universe (or multiverse) as that thing which is essential.

So, to conclude, I contend that the minimal god hypothesis that I have been considering is not the argument to the best explanation, and so should be rejected on the grounds of rationality. Whilst the Eternal Inflation theory may not be the correct explanation of reality, or even the best one, it is nevertheless a far more reasonable one than the god hypothesis, and so is to be preferred.