Saturday, April 01, 2006

Part 7: There are voices in my head

Suppose that consciousness exists in the brain. Clearly, the brain has many other functions, most of which have nothing to do with experience or decision-making. Then why is it that some neurons cause consciousness to appear, while the others are merely pattern recognizers? If you think of Jeff Hawkins' six-layer cortex hierarchy, what is it that sits at the top of the topmost layer and experiences the world around me through the streams of abstract concepts that propagate up from my eyes, ears and other senses? What is it that decides what I am about to say next? Whatever it is, it is probably inside my neurons, but which ones? Why those ones? There is the problem of deciding which neurons are conscious and which ones are merely information processors. How do these neurons communicate with each other?

Dr. Steven Sevush, an Associate Professor of Psychiatry and Neurology at the University of Miami has an interesting solution - the single-neuron theory of consciousness. Sevush proposes that there is no fundamental difference between the "conscious" and the "non-conscious" neurons. In fact, all neurons are conscious - every single one of them. However, some are lucky enough to be located in a special area of the brain, the left lateral prefrontal cortex (PFC), where the outputs they produce directly affect my motor functions. Sevush refers to several experiments and argues that if there is an area of the brain where consciousness sits it is in the PFC. There, things get strange.

First of all, the neurons in the PFC are quite complex. Some have tens of thousands of connections to other neurons. Second, there is a substantial degree of interconnection between the PFC and Broca's area, which is responsible for speech and comprehension. Sevush argues that each of the neurons in the PFC receives inputs and produces outputs that are complex enough to account for the whole of our conscious experience. Each neuron is capable of being conscious on its own.

If that is true, then perhaps the brain does not simply have a single consciousness. There could be hundreds or thousands of conscious neurons, each one experiencing and interacting with the world independently. Since each of them receive almost identical inputs, they each produce very similar outputs, and it is the amplification achieved through the strength in numbers that is responsible for that which we percieve as a single consciousness.

Sevush gives two analogies. Consider a crowd watching fireworks. Each member of the crowd sees the fireworks independently and reacts with an "aah" or an "ooh". At a distance, one person would be inaudible, but the collective response of the crowd as a whole can be heard as a distinctive "aah" or "ooh". The crowd is an entity that reacts to the fireworks display, but the crowd's consciousness is nothing more than the combined consciousness of the individuals in it.

As an example of a response, Sevush cites Garry Kasparov's 1999 chess match against the World, where the grandmaster played a single game against a large number of opponents on the Internet. The World's moves were decided by a majority vote.

The single-neuron theory claims to be an alternative solution to what is called the binding problem: how do the huge number of neurons work together to produce a unified conscious experience? Basically, there are three possible answers. First is the neural network approach, which postulates that neurons work together synchronously by exchanging complicated non-linear signals, and consciousness emerges as a property of the whole structure, not of any single neuron. Like any emergent theory, it has its opponents. Second is the Daniel Dennett approach, which states that consciousness is an illusion, and what in fact goes on in the brain is far from being unified or coherent. We will come back to Dennett's ideas later. Finally, here is the single-neuron theory, which side-steps the binding problem by saying that each neuron is fully conscious on its own, and that which we percieve as the consciousness of a human being is the collective sum of the conscious experiences of a large number of individual neurons. Most of those neurons receive identical inputs and produce identical outputs, and sproradic errors from stray neurons are corrected by adding together the signals produced by all of the neurons - a simple, linear operation.

Sevush then goes on to speculate about possible experiments that could prove or disprove the single-neuron theory. The presence of complex neurons in the PFC, each one capable of receiving very complicated inputs from all areas of the brain, would give evidence for this theory, as would the presence of multiple such neurons, each receiving identical inputs and producing identical outputs. There is no experiment, however, that could prove the single-neuron theory correct - only ones that would give evidence for or against it.

So how could a single neuron posess a degree of consciousness that is as complex as the consicous experiences of a whole human being? The rest is pure speculation. It could be that quantum effects, similar to those of Penrose and Hameroff are relevant here. In fact, the quantum connection becomes more believable on the scale of a single neuron. There is no longer a need to maintain particles in quantum entanglement across several neurons, only within a neuron. This is still, however, a huge distance for quantum entanglement.

On the Consciousness DVD, Sevush goes further to make a speculative connection with string theory. A beautiful result that comes out of string theory (if it is true), is that nothing can be smaller than a string. This claim is not to be confused with similar claims that have been made about molecules, atoms and quarks. It is inherent in the mathematics of string theory that if we consider lengths that are smaller than the length of a string, then the physical equations "flip" inside out, and we get identical physical laws that hold at those scales. It is as if each string had a whole world inside of it. If that were true, then, Sevush argues, each molecule in each neuron could be conscious, too, and so would each atom in a molecule. Finally, each string of each atom inside me would be conscious and have the same conscious experiences as me as a whole. If every string is as complex on the inside as the whole universe is, then it is certainly complex enough to be capable of my conscious experiences.

It is easy to get carried away in speculations, so let me summarise the key premise of Dr. Steven Sevush's single-neuron theory. Each neuron is capable of fully complex human conscious experience. There is a collection of specially (luckily) positioned neurons in the area of the brain called the PFT which are responsible for our consciousness. Those neurons receive identical (or nearly identical) sensory inputs and do not communicate with each other in producing (nearly identical) outputs. The only cooperation that happens is a big summation of their outputs, which serves to (1) minimise sporadic errors from stray neurons and (2) achive signal amplification levels sufficient to control our speech and motor functions. What we percieve as a single conscious person is instead a collection of multiple conscious entities (the neurons) that combine their outputs and "vote" on each decision.