Neurons and logic gates
from the webmaster
The idea that the brain is a computer has been popular since the invention of the latter, and this remains a mainstream consensus. The feeling of many others that there is an essential difference between computers and humans has tended to be pushed aside as an uneducated delusion.
However, although much brain processing can be accepted as being analagous to computing, it may be that there is a fairly easy to detect structural difference between brains and computers. Classical computers are based on seven main logic gates using the binary code of ‘0’s and ‘1’s.
1.) With the AND gate, there are two inputs and one output. Two ‘1’ inputs are required to produce a single output of ‘1’. The function is so called because it is analogous to the word ‘and’.
2.) With the OR gate, if either or both of the two inputs are ‘1’, then the output will be a single ‘1’. The function is analogous to ‘or’.
3.) With the XOR gate, if one but not both inputs are a ‘1’, the single output will be ‘1’. The function is analogous to either … or’.
4.) With the NAND gate, there are two inputs, but a single output of ‘1’, unless both the inputs are ‘1’. This is analogous to the word ‘and’ followed by the word ‘not’.
5.) With the NOR gate, there are two inputs and if the inputs are both ‘0’, the single output is ‘1’. This is analogous to ‘or …. not.’
6.) With the XNOR gate, there are two inputs, and if the inputs are both ‘0’, the single output is ‘1’. This is analogous to ‘either …or’ followed by ‘not’.
7.) The NOT gate is the only gate with a single intput, and the single output is always the opposite of the input. This is analogous to the word ‘not’.
The material point in all this is that the logic gate has to produce a single output that is wholly determined by the sequence of inputs. If we can read the inputs into a logic gate, we already know the output. If we knew the inputs into 10,000 logic gates, we could, given time, work out their outputs without actually seeing them, and so on into the millions or billions of logic gates in particular sequences of brain processing. There may be such millions or billions producing the read out ‘remove hand from hot stove’ from our conscious neural processing.
Where this starts to break down is with the minority of brain processing, in which there is not one but two or more possibilities. Computer systems can deal with this sort of ambiguity only up to a point. A spell-check can produce a number of suggestions, each of which is the result of single outputs, on the basis of ‘if that then this is a possibility’. Similarly a search system for stations with trains to Edinburgh will initially have a number of suggestions which can be narrowed down by further single output logic gates.
What appears less easy to derive from single output systems is the choice between things that have no common currency, such as apples or pears, wine or beer, or even a small reward now against a larger reward in the future. These are not simple gradients between good and bad, but a complex landscape between good and good or between the lesser of evils. The computer systems are compelled to give a single output, even if this is the culmination of many such ‘0’ or ‘1’ single-output logic gates. Where there is a choice of such outputs, the computer can only resolve them by a narrowing down process using further single output logic gates. It does not have a way of ending the process where there are two of more logically processed but conflicting solutions. In the case of the brain, such conflicting outputs may arise from separate processing within areas such as the inferior temporal and the somatosensory.
The ability to resolve this may point to the evolutionary usefulness of consciousness, which conveys the ‘something it is like’ aspect of consciousness, such as eating apples, or of the future scenarios of benefits resulting from having adopted a healthier diet or exercise regime. It is possible this choosing is the trigger that pushes neurons into consciousness.
It has been a long-run puzzle that only neurons in certain parts of the brain, and those only at certain times, appear to be involved with consciousness. There is no evidence of special consciousness neurons or of a special consciousness brain area. The pyramidal neurons that are the excitatory neurons of the cortex tend to be more associated with conscious processing, but again only some pyrmaidal neurons, some of the time.
However, what recent studies have suggested is that conscious images arise in relatively small minorities of single neurons. The sensory neurons such as those in the temporal or somatosensory cortex have a conscious output that appears to be initially value or choice-neutral, prior to processing by prefrontal and limbic areas. However, it is possible that the reason for bringing images and sensations into consciousness in the first place is as a ‘service’ for the subsequent choice or evaluation processing. Other studies showing that actions that require perception but not evaluation, such as removing a hand from a hot stove, or putting a letter in a letter-box can be performed without the involvement of consciousness.
The need for choice and the need to have the images/sensations that underlie such choosing could be the crunch-time for the neuron, pushing it from algorithmic processing into something that accesses a fundamental level. This is where the decoherence of extended systems of phonons, pi electrons or other quanta could kick in. If we ask why this happens in brains, but perhaps not in rocks, the difference could be that rocks can rely on the algorithmic rules inherent in electromagnetism and gravity, and are never forced into the difficult choices arising in brains.