by Marcus Loane
Consciousness should not be off limits or defined as outside of scientific investigation. Some seemingly insurmountable puzzles in the past have been explained. Disease used to be considered utterly mysterious and caused by evil spirits or God. The existence of life used to be utterly mysterious and was explained by a variety of creation myths. The increase in scientific knowledge recently is fantastic and it is not slowing down. There has been progress in understanding how parts of the brain function. The answers won't come all at once but are put together piece by piece. The process is (rightly) slowed down by the ethics of dissecting live human brains for reasons other than curing disease.
You can't trust intuition
When thinking about consciousness, we need to abandon our intuitions. Our intuitions are there because they helped our ancestors to survive (or they are a side effect). That does not necessarily mean they are correct. Science has repeatedly shown that intuitions can be wrong. Intuition leads us to imagine there is some inner observer sitting in the brain watching a little screen (the Cartesian theatre) where everything is "presented". We know that is not the case but that is how it sometimes feels. How does the "inner you" process the contents on the screen? By an even smaller inner you inside of it? This leads to an infinite regress. Rather than explaining, it just pushes the same question further back. Some other explanation is needed. [The infinite regress problem also occurs when a Universe Creator is proposed. Who created the Creator? Who created the creator of the Creator? Theories that lead to an infinite regress don't explain. They just move the question elsewhere.]
A theory of consciousness is likely to be unintuitive and also unattractive because it challenges some of our most cherished notions about individuality and free will.
No soul stuff
Consciousness appears to be closely correlated with physical structures and processes. There is no need to postulate a magic ingredient. No need for souls. No inner observer. It is likely that the substrate is not important and this conclusion is based on the the thought experiment "biological prejudice" described later. Neurones don't have consciousness-stuff within them. They are made of atoms like everything else. I think the information processing, the organisation of the neurones, the wiring is what matters. Look inside a computer and it is difficult to imagine how the collection of plastic and silicon leads to complex data structures, metaphors and pictures. It is the same with brain tissue. Grey matter containing bundles of neurones does not look like it can contain meaningful content. In a computer, series of 1's and 0's are built up into structures that mean something even though they are still just 1's and 0's. In brains, neural connections are built up into structures that mean something. The structures symbolise something without resembling what they symbolise. How can this be? It is how the symbols interact with each other within the brain or computer processes that matters.
Virtual reality
Using data from the senses, the symbols (patterns of neural connections) are built up into a model of the outside world plus a model of the body (shape, position etc.) plus a self with beliefs, opinions, and memories. These continually changing virtual reality models are played out in your brain. They interact and modify each other depending on data from the five senses. The world seems to be "out there" but what you are experiencing is a virtual reality reconstruction of what the brain estimates is out there by using sense data combined with previous knowledge. This has been confirmed by the symptoms of patients who have had the neural connections between visual modules and memory modules damaged. They can describe what they see but not find the words that most people would use. For example they would describe a mountain scene in front of them as "a greenish bit poking up into a bluish area". Their brain cannot put together a proper model of what is out there without the connection to the memory modules.
There has been a lot of progress lately in understanding how vision works. It is now known that there are at least thirty areas in the brain performing some visual task. The processing of images is divided up and "labels" are attached, for example there would be circuitry for labeling edges, another for corners, another for finding vertical lines. These labels are then used to attach object labels such as tree or table or daughter, based on memory. These tasks are distributed throughout the brain and also distributed in time. They are not brought together into some final presentation.
What we see with our eyes is a reconstruction based on light patterns on our retinas and previous knowledge of what we expect to see. It is just an approximation. This works most of the time but is easily fooled by optical illusions. On a dark night it is easy to see a person lurking in the shadows and on closer inspection it turns out to be a bush. With some optical illusions the brain is unable to make up its mind and will flip between two guesses periodically, for example the duck-rabbit illusion or the wire-frame cube picture.
The models of the world and the body are hopefully, at least most of the time, accurate enough representations of the real body and real outside world. The self, however, while very complex and utterly fascinating could be a kind of story, a sequence of internal neural information processing events through time. Bodies containing selves use words like "I", "me", "thought", "belief", "aspiration" and "dream" to describe internal events. This works because other people experience similar neural information processing events and have learnt to emit the same sounds when performing speech acts to describe them. The model of the outside world also includes other people's bodies and their selves. The model of other people's selves is based partly on their observed behaviour and partly on how this self feels - "putting oneself in another's shoes".
The self - benign user illusion or source of suffering?
We enjoy our sense of self and interacting with other selves. Our thought processes seem important to us. Attempting to explain the self does not detract from that. The most wonderful thing is that the experiences can be shared with others.
There are psychiatric conditions where the distinction between the "world model" and "body model" breaks down so that the body no longer feels distinct from the surroundings. There are also conditions where the self no longer exists - senile dementia or schizophrenia can lead to loss of beliefs, memories and personality traits. In Buddhism, meditation is used to deliberately achieve a feeling of no self as the self is regarded as a source of suffering. Meditation is also used to achieve a feeling of oneness with the surroundings. This could be explained by the world model and body model merging temporarily in the brain of the meditator. For some evidence on these ideas see this MSNBC article on the neural causes of mystical experiences. Here is another article from Feedmag which describes recent advances in understanding the physical basis of the self.
How can you say that we are just a computer or bag of chemicals?
I have been asked that question. We are made of chemicals but we are not just a bag of chemicals. A car is not just some metal and rubber and plastic. It is more than the sum of its parts. It is how the parts are arranged that matters. We never hear people protesting that a car cannot be made only of metal, rubber and plastic, but that it must have special magic car-stuff or automotive essence permeating it. Folks used to insist that life must have a life-force - the discredited notion of vitalism. Now we know that the properties of life are a result of a lot of working parts. Proposing a life force was a way of not explaining life. Most successful explanations explain complicated phenomena in terms of simpler ones. I suspect that consciousness is similar and will be explained as a property (or set of properties) of certain information processing systems.
Consciousness is not binary
Consciousness is not a switch that is either on or off. There are levels of consciousness. Any doctor who assesses a patient with a head injury will agree. They do a test to determine the level of consciousness (on a scale of 0 to10 if I remember correctly). Consciousness can be "dimmed" with anaesthetics as well as switched off entirely.
Our early ancestors were unconscious, as a single celled organism is presumably not conscious. There is a direct line of ancestors from unconscious single cell to conscious human being. Each organism in that line is only very slightly different from its parent, this would suggest that no consciousness "switch" was turned on but that consciousness increased in degree very gradually. However this process may have been recent (in geological time). If the process started much earlier then it has implications for the consciousness of other animals.
In a similar vein, you were once an unconscious fertilised ovum in your mother's womb and you eventually became a conscious human. You have taken the journey from unconscious to conscious. What a pity that we forget those early stages. Was there a time as an infant when you first suddenly became conscious? It is more likely that it was a gradual process. How early was your first memory? Does it mean that you never consciously experienced your first few weeks/years of life? It may be that new born babies are not conscious, and that the process of learning and exposure to stimuli is necessary for consciousness in humans. Exposure to stimuli causes rewiring of the brain in as real a sense as a surgeon going into it and rewiring it. Any mother will reject the idea that their newborn is not conscious, but that makes evolutionary sense. Of course, the mother needs to regard their infant as conscious so that she cares more about its well being. However no one ever remembers being conscious when very young. It is very possible that the necessary brain wiring just is not there in the early stages and as a result consciousness is absent.
Do other species have an inner experience?
Chimpanzees can recognise their own reflection and preen themselves in front of a mirror. This is supposed to be a test of self awareness. Only a few other species pass it. Chimpanzees can learn sign language and invent imaginary friends (an analogue of god-belief?). Cunning experiments have shown that they appear to be able to hold the concept of someone else having a mind - they know that Sheila knows were the food is but Tom doesn't. If they have a concept of someone else's mind they should be aware of their own mind. We don't yet know precisely which information handling capabilities collectively constitute consciousness, so we can only guess about other species. It is unlikely that a spider has any conscious experience at all - its brain is too simple. I have mentioned chimpanzees as possible candidates for membership of the privileged group. Many people, myself included desperately want to believe that dogs have some kind of inner experience, but we just cannot say yet if that is the case. Your much loved canine friend could be a complete automaton with noone inside. Unpleasant as that thought is, we have to keep it open as a possibility. Intellect takes precedence over emotion and intuition.
Zombies
If you ever read any philosophy you eventually will come across a curious species. It is the philosopher's "zombie". This is a hypothetical being which is supposed to lead us to new insights. A zombie is a human being that is defined as having no inner experience but it behaves exactly as if it does - it will even insist that it is conscious. You would not be able to distinguish it from a person who really was conscious. A zombie is generally regarded to have the exact same internal physical structure (including brain structure) as its conscious twin. A zombie by being defined as indistinguishable from conscious humans, would be able to enter into deep philosophical discussions about the nature of its own inner experience (which according to the definition it does not have).
Now why is such a thing considered? It is supposed to show that consciousness is something extra, the intuition (perhaps a false one) that even if we explain every single detail about how the brain does what it does, we are still leaving something out. My problem with this is that we cannot know if an unconscious zombie is possible even in principle. I could argue that if we built a zombie which did everything that a human did, it's brain would have to include internal physical representations of propensities, emotions and beliefs including the belief that it is having an inner experience. It is more consistent to say that it is these internal representations and information processing structures, and their resulting activity, that is human consciousness. Try to build a zombie and you will end up building consciousness.
Could we have conscious machines?
Yes, there are 6 billion humans walking the planet already. The existence of us shows that if matter is arranged in a particular way, consciousness can result. This means it would not be breaking some physical principle to create a conscious non human machine. While at present technically formidable, it should be theoretically possible to create an inorganic conscious entity. Human intuition revolts at the idea. However intellect has a better track record than intuition. I can imagine the debates when we are confronted with a machine or network (the brain is a network) that exhibits intelligence and creativity. There will be a camp who confidently assert that no machine made of mere silicon could ever be conscious. The machine will reply that it knows it is conscious and who are they to tell it what it can or cannot feel. Quick, reach for the Off switch! In fact this is one possible test for consciousness. If an entity claims it is conscious and it was never programmed to do so then we should believe it. Unfortunately this leaves out entities that do not speak our language. Machine intelligence that can learn from experience (learning algorithms/some neural networks) has already been created. After exposure to environmental stimuli these systems restructure themselves just like human brains do and the end result may not be understood by their creators. It is these sort of systems that have the most chance of becoming aware and frustratingly it may not be possible to pinpoint exactly what it is that causes it because they are changed from their original (known) configuration. Actually, learning algorithms may not be enough and some kind of evolutionary algorithms may be needed (see my later article) to make such a machine/software entity. Similarly, computer based evolutionary algorithms can produce complexity that is impossible to reverse engineer so the programmer may have very little idea how his "creation" works.
Strange thought experiments.
Solipsism variation: Brain in a jar
You could be a brain in a jar. Your every experience could be artificially induced by the jar-masters. If this were true you would have no way of knowing. Try not to have nightmares - hypotheses which will always be impossible to disprove are best ignored.
Where is consciousness located? Remote controlled body
Most people feel like their consciousness is "located" behind their eyes. We know that our consciousness is dependent on a functioning brain. These two statements seem to be in agreement. However, try this thought experiment: imagine we remove your brain and place it in a jar (again!) and arrange a radio link between it and your body including all its senses. Your body would be able to walk around as normal, controlled by the brain in the jar. You would not feel any different. You would still feel as if your consciousness was located behind the eyes of your body. What does this tell us? The sensation of where consciousness is, is not connected to the location of the processes that cause it. The sensation of location appears to depend on the location of the senses (eg. the eyes). That is not the whole story. The brain in the jar is building a real time symbolic representation of the outside world plus the state of the body, using data from the eyes and other senses on the remote body. There is a model of the world with respect to the body, and this virtual reality model resides in the brain in the jar. Therefore the sensation of location really only depends on the state of the model in the brain. This explains why people can have out of body experiences and other strange phenomena which seem real to them. The brain's usual model of the world, the body and the self has been disrupted by chemicals or disease processes.
Here is another one: Last Thursdayism
The universe was created last Thursday. A deceiver god arranged everything including your memories to look like the universe is older. This is impossible to disprove but remember that is not a good reason to believe it!
Recently a monkey's head was transplanted unto another monkey's body. The face of the transplanted head showed every sign of being aware and interested in its surroundings. The following thought experiments are technically impossible with our current state of technology, but they are possible in principle I think.
A thought experiment about biological
prejudice
This experiment implies that computer consciousness is possible
in principle.
Neurones take a number of electrical inputs and produce an output. Like the logic gates that are built up into computers, they are quite simple devices from a functional point of view. Electronic simulations of neurones, either physically realised or in software have been around for several decades. Imagine we put you under anaesthetic and replace one of the neurones in your brain with an electronic device which does everything the biological version did. It responds to certain input signals with certain output signals. When you wake up you should feel just the same - you would not notice any difference. Everything functions as before. Using this process, if we replace every neurone one by one, eventually your entire brain will be an electronic device. At no point will you have noticed any change. Your identity is now inside an electronic device! Would you still be you?
A thought experiment about the unity of self:
Imagine we have a scanner that can scan a 3d object and make an exact replica. We don't tell you what we are going to do and we put you under anaesthetic and make a replica of your brain. We transplant your original brain into another body and we put the replica brain in your own body. When the anaesthetic wears off, which body do you wake up in? Who is the person in the other one?
Both brains have identical memories. They both remember their lives up to being put under anaesthetic. From my 3rd person point of view as the evil scientist, I would observe both bodies claiming to be the real you and accusing the other of being the replica, but which body would you feel you woke up in? How could you inform me? Both would insist to me that they are the original. One would feel no different but one would claim that their body had changed. Which one is the old you? The way I would explain this apparent paradox is that neither is the old you because "you" (the inner you) do not exist as a continuous entity through time. It only seems that way.
The Self
In the "unity of self" thought experiment, the closest to the original would be the one with the original body, because the body is an important part of consciousness. Chemicals from the rest of the body can affect the nature of consciousness by altering the electrical signaling, so the brain, while being the seat of consciousness is not the whole story. The body and the surrounding environment are important as well. I could change the thought experiment so that your entire body is replicated by our wonder machine. However that may lead readers to the easy (and false) conclusion that the real you is the original body and the replicant is somehow less real.
I say, there is no difference. You and the replicant are equivalent at the point of waking up. You both have the same memories and are both utterly convinced that you are the real you. From that point on though, you have different experiences so you become different people. So a self can be split in two. This is not surprising if you realise that the continuous self is an illusion. When you wake up every morning (and when the subjects in our experiment wake up from the anaesthetic) the brain processes start retrieving memories and construct a narrative, a mini history about a Self, a self consistent story of a person. Consciousness (the type with a Self) could be a series of these stories from moment to moment. Each story is self consistent but stories may not always be consistent with each other. Memories fade and Self history reinvents itself. Motivations are often deliberately attributed after actions have already occurred. I think that this is also going on all the time as an unconscious process. Multiple parts of the brain each doing their own simple task can have the net result of causing a behaviour, and other parts then concoct a story about an "I" who had a motive for the behaviour. The motive is then time stamped with a time before the behaviour, and the illusion is complete. The conscious part of the process firmly believes that it originated the behaviour. There have been experiments that suggest that this time stamping does indeed occur.
What would be the point from an evolutionary perspective? It is a result of being a social animal. It is a self perpetuating way of predicting others behaviour. We can concoct stories about the brain processes in other humans and reach better than chance predictions about their behaviour (also known as folk psychology or just plain psychology). If everybody does it, then it works.
Progress
I think the most progress in understanding consciousness will be made by abandoning deep set intuitions. Consciousness in humans appears to emerge gradually in infants as their brains increase in their complexity and power of processing environmental stimuli. This suggests that consciousness is related to complexity of information processing. Further developments in neuroscience and computer science may shed more light on the matter.
--
Marcus Loane