Restatement of hypothesis

I thought it might be an idea to set down again where I am trying to go with these ‘sra’ posts.

The quest is for the secret of consciousness, a facility enjoyed by most humans and by some animals.

The first part of the hypothesis is that consciousness comes in discrete units which, borrowing from the world of film, I call frames. Frames, which might endure for a second or so, are grouped into takes, which might endure for several seconds. Takes, in turn, are grouped into scenes. The frames of a take share a lot of content, the takes of a scene rather less. Scenes are more-or-less freestanding and successive scenes may share very little content at all – frames, takes and scenes all drawing on the same underlying knowledge and memories notwithstanding. We also allow the decomposition of frames into threads. See, for example, reference 4.

The second part is that consciousness arises from the activity of the neurons in a small patch of cortical sheet, somewhere in the brain, an essentially two dimensional structure occupying a couple of square centimetres or so. In crude terms, with an information content equivalent to one of the pictures which I can take on my telephone, perhaps 10 million bytes.

The third and last part is that, for present purposes, the activity of all the millions of neurons in this patch can be integrated, summarised into a signal which can then be analysed in the two-dimensional space of our patch, and, over the interval of a frame, into frequency bands. One might say a three-dimensional grid with two long sides – a thousand elements or so - and one short side – ten or so elements, taking non-negative real values for each of the resultant cells, values which might be binned into suitable integer ranges. This is described in a little more detail at the beginning of reference 2, which goes on to make some guesses, some suggestions about how this signal might be organised. We talk of  data occupying the lower frequency elements of the short dimension and process occupying the higher frequency elements. Future posts will say more about this.

A corollary is that while we suppose that for the duration of a frame this population of some millions of neurons can be considered as fixed and that their synaptic interconnections, many more millions of synapses, can also be considered as fixed, we do not need this fixing because it is the aggregate signal from the firing of neurons which is of interest, not the details about individuals. We do not say anything about the life-support, the chemical soup in which these neurons and synapses live – fluctuations in the content of which have been shown to have important consequences for the activity of nearby neurons.

We suppose that there is a compilation process – comparable to that whereby a program written in the Fortran of the 1970’s was readied for action – which builds the data part of this structure, pulling in stuff from all over the brain. A pulling in which may become difficult if the brain has been damaged or if supplies of energy and other consumables have been compromised, but which is not itself consciousness. Compilation also puts in place the connections which will support the activation process running in the process part of this structure (see below).

Compilation is revisited, updated for each successive frame and is renewed completely at the start of each successive scene.

Consciousness itself arises from the much smaller and simpler activation process, a process which pulses across the neurons which express our data structure, a structure which is self-contained and more or less unchanged for the duration of the frame. Everything that consciousness needs has been put there by the compilation process and there is no need to go anywhere else in order to generate the phenomenon that we seek, to bring the data to life. Whatever it is that makes red look red is there and there is no need to go scurrying off to some other part of the brain. Future posts will say more about this activation process.

A corollary is that we are not much interested in all the goings on between the brain and the periphery, in particular the sense organs, the eyes and the ears amongst others. In all the goings on between the input/afferent side of things and the output/efferent side of things. In all the large scale networks being uncovered at, for example, reference 8. The idea is that we are able to split the core business of consciousness away from all that, just leaving a compilation process to provide liaison, a process which we hypothesise but do not consider in any detail. The classic device of divide and rule.

Other people

Turning to what others think about all this, some people think that the answer to the problem of consciousness lies in the field generated by the electrical activity of the brain, that consciousness arises from a fancy field, the sort of field, that is, described at reference 3. This hypothesis leans in this direction to the extent that it supposes that the activity of the neurons on our patch of cortex can be integrated, summarised into a signal which can be analysed in two-dimensional space, and, over the interval of a frame, into frequencies.

Rather more people think that the answer lies in large scale integration across the brain, taking in parts of both the cerebral cortex and the brain stem, with plenty of discussion among them about how large is large. With plenty of both feed-forward and feedback. With plenty of re-entrance, that is to say circuits looping back on themselves. With damage to the wrong part of the brain stem being fatal to consciousness. The present hypothesis is very different, at least to the extent that it isolates and localises the very end of the chain of processing, of the neural activation which results in consciousness.

I should add, that I am not suggesting here that consciousness is for anything or even at the top of the pyramid. But it is there, and it does seem to be associated, if not more, with important human-only functions like choice and program execution. See reference 5.

Other matters

There are many way of extracting information from, of describing the noisy firing of large numbers – millions – of neurons. The wikipedia entry at reference 6 gives the idea. But I think it is fair to say that, moving away from relatively straightforward firing rates, a lot of people have become interested in the information carried by the timing and the relative timing of spikes. Here, however, as indicated above, I have opted for integrating and then analysing those spikes over space and time to give us numerically coded frequency bands.

I am still thinking about why one cannot have more than one patch of conscious at any one time. This being a more or less observable fact, but not popping out of the foregoing. Somewhere along the line there must be a process, a thread, of which there can only be one – unlike a lot of neural processes which can take place in the left hand side of the brain, the right hand side or both. An example of such a thread, about which I have posted in the past is the business of thinking in words, a business which requires access to parts of the vocal apparatus. See reference 7.

As an aside, I note that in this context it does not matter whether consciousness is right or wrong, a sensible reflection of the outer or inner world or not. Here we are just interested in the fact that it is and how that might come to be. Deranged or diseased, consciousness is usually still there.

Then there is the question of frame refresh. In the case of a film, it might be that a frame endures for about a second while the business of advancing the roll of film from one frame to the next takes about a tenth of a second, with the whole blurring into a seamless whole of moving picture. Now the part of compilation which is drawing in data from all over the brain can go on all the time, in the background as it were. But, on our frame hypothesis, at some point we need to build the network of neurons which delivers the data and processes which together do consciousness. One scenario is that our patch, our population of neurons is divided into two, with one being compiled while the other is being displayed. The switch from one to the other could be very fast. Another scenario would be to compile between displays, between frames. But this sounds too slow, involving as it does tinkering with large numbers of synapses – because while the population of neurons might be fixed, that of synapses is not – at least the population of synapses might be fixed, but their strengths, their weights are not.

Unlike a film, we do not have the frames to come ready and waiting. At the most, the next frame might be in preparation during the life of the current frame. While past frames is another matter, it being probable that they get written to or consolidated into permanent memory in some form or another. Which is not to say that they are accessible, but they are there, at least for a while.

With, in the foregoing, the distinction between takes and frames having been blurred over.

References

Reference 1: http://psmv3.blogspot.co.uk/2016/12/from-grids-to-objects.html. Introducing patterns on a grid. A grid which contains millions of points and which is based on a patch of cortex. Patterns which are based on repetition of some sort of the patterns of values on rectangles, rectangles which provide, as it were, a window onto the grid.

Reference 2: http://psmv3.blogspot.co.uk/2017/01/layers-and-columns.html. A post which develops the structure of layers and columns which we are proposing to host the experience of consciousness. A rather long post with no less than six diagrams and ten references.

Reference 3: https://en.wikipedia.org/wiki/Field_(physics).

Reference 4: http://psmv3.blogspot.co.uk/2016/11/an-assembly-for-consciousness.html.

Reference 5: http://psmv3.blogspot.co.uk/2016/09/what-is-consciousness-for.html.

Reference 6: https://en.wikipedia.org/wiki/Neural_coding.

Reference 7: http://psmv3.blogspot.co.uk/2016/11/on-saying-cat.html.

Reference 8: https://www.alleninstitute.org/.

Group search key: sra.