On an aspect of attention and consciousness

Figure 1

Figure 2

Contents

• Introductory ramble
• Eye movements
• A process diagram
• The frames, layers and objects of LWS-N
• Conclusions
• References

Introductory ramble

Over the years I have given quality time to the business of counting things, like the number of floors of a tall building, or the number of courses of bricks in a wall, with just eyes and brain, without mechanical aids, like pointers or photographs. Some of this has been documented at references 1, 2 and 3. The present post is another episode in this story – which we continue to find both interesting and instructive.

We continue to think about how all this might work in LWS-N, introduced at references 8 and 9.
So this episode is about counting the length of a row of holes in a segment of an overhead shelf in a carriage of a Southern train running between Victoria and Epsom. Around 2245 in the evening and a small amount of drink – a couple of whiskies – had been taken a couple of hours previously. Outdoor spectacles, as was proper. Noticed in a preliminary way at reference 11.

At the time, the right hand end of the shelf did not fade away as much as it does in this snap; it was easier to count the holes there than might now appear.

The trick seems to be to focus on a hole, then slide the attention one hole to the right, then increment the count on arrival. Note that I always seem to do this task from left to right, perhaps a consequence of reading from left to right. Perhaps I should test whether it is harder the other way? While courses of bricks and floors of buildings are usually – if not always –counted from the bottom up.

One difficult bit is not to let attention – or the foveae – slip off the current hole, as once that happens there is no way to find one’s place again. The whole point is that one hole, one course of bricks or one floor is pretty much the same as another.

Another difficult bit is maintaining the count. It is all too easy, perhaps by losing the rhythm, to lose the count. Perhaps only by one or two – perhaps one is not sure whether one is at 23 or 24 – but plenty enough to spoil the count. Keeping the count going steadily, maintaining a rhythm, does seem to help. Counting in musical time – say three or four time – does not seem to help. I have not tried counting in groups and then doing mental arithmetic at the end: for example, 17 groups of three plus two left over makes 53.

On this occasion, despite the odd wobble, I got three consecutive counts of 48, a count subsequently confirmed with the help of Powerpoint. The only catch here being that, a couple of week previously, I had had a confirmed count on an apparently similar shelf of 46 holes, noticed towards the end of reference 4. Perhaps a different batch of shelves? Perhaps wishful thinking driving the checking? But this discrepancy does not bear on the present post, so I have not yet investigated.

The key ingredient of Powerpoint is that it allows for a snap of the thing to be counted to be marked up, in this case in groups of five. The groups can then be checked individually and counted without difficulty. A wheeze not available to unaided eyes and brain. A wheeze which allow the count to be made without needing to maintain attention for the duration. One can take breaks, for teas, fags or whatever. The count has, as it were, been de-skilled. Anyone can do it if they have Powerpoint – or something of that sort – at their fingertips.

Turning to what the brain is up to, we suppose that its natural tendency is to be scanning input – external or internal – all the time. A tendency derived from our savannah past in Africa, always on the lookout for a threat or an opportunity. And there is a standing instruction to the brain to raise an interrupt if there is a suggestion of either. An interrupt which shifts attention, in this case away from the count, to something or other else for further examination, examination which might be brief and result in almost reflexive action –  although not truly reflexive because central cortical resources have been deployed. Not something which has been triggered somewhere way down the spinal cord, as might be the case with, for example, touching something which is painfully hot.

Because we are doing the counting task, the brain has to be told, somehow or another, not to do this, not to raise interrupts. To concentrate on the nominated task and not to allow some extraneous, some extra, interpolated frame of consciousness. Not to allow the eyes to saccade off the current hole, as they will not be able to reliably saccade back again.

Figure 3

This seems to be a case of a decision which has been reported to consciousness being able to block eye movements – saccades – of a sort which are usually triggered unconsciously. Blocking which is the result, or is at least associated with, intense concentration on the current subject of attention, whatever is in the foveae. And which, we imagine, would be much more difficult if one was attempting to attend to something which was not in the foveae. Possible, but difficult. And much easier if one is looking straight ahead, in the way of a pointing dog: lots of visual things seem to be much simpler if body, head and eyes are all lined up.

This counting is related to the tea cup carrying task described at reference 5, in that in both cases one needs to maintain attention on the task in hand. One needs to be able to block the brain’s tendency for its attention to wander about. Failure in this second case resulting in spillage, instead of loss of count.

Then the are the instructions for perfecting your golf swing which involves, as we recall, staring intently at the ball you hope you are about to hit, while emptying your mind and trying to ignore any noise – as in the sequence in the film Goldfinger, when Auric rattles the change in his pocket to put James off his swing, an incident recently mentioned at reference 7, one of the warm up posts to the present one.

In all these tasks requiring attention and concentration, there is no room for self images in the LWS-N stack. These are all part of the baggage which has to be dumped so that as much as possible of the brain is focussed on the primary task: the idea is to swing the golf club, not to think about swinging the golf club, let alone about rattling change. And people who are trained to it, will be largely unconscious of what is going on around them, although some of this last will inevitably have penetrated some way into the brain, which cannot, for example, turn the ears off, however important the task in hand – but it can block or otherwise turn off the processing train, somewhere along the way.

Going further, one might argue that when such people are performing on form, they are barely conscious in the ordinary sense of the word at all. They are awake, alert and functioning, but entirely absorbed in the task at hand. If interrupted, they might even have trouble saying what exactly it was that was interrupted.

Eye movements

As one tracks along a line of holes such as the shelf of Figure 1, the subjective impression is that eye movements are in the lead, followed up by head movements. It would be interesting to put an eye and head tracker onto the case, to see exactly what is going on, but in the meantime what follows is not much better than a plausible guess.

We suppose mainly smooth movements of the eyes, movements during which vision is maintained, unlike the blanks during saccades.

Eyes hold the current hole.

Eyes slide from the current hole to the next hole. The task is at its most fragile during this slide; the least disturbance and position on the line of holes is lost, the cursor is lost.

Eyes hold the newly current hole. Brain increments the count. And so on.

The brain can relax slightly during these holds. But it does need to keep hold of the current hole and the current count.

Every now and again, regularly in the case that the task is being executed smoothly, the head adjusts its position, rotates slightly on a more or less vertical axis, assuming here that one is looking at the holes, rather than up or down at them, rotation which means that the eye balls can move back to pointing more or less straight ahead. One is aware of the head making these small movements every few seconds, but one is not aware of the compensating movement of the eyes.

These rotations are clockwise in the case that one is counting from left to right.

Figure 4

We illustrate all this with a diagram adapted from reference 3. The visual layer of LWS-N contains a chunk of perforated shelf, suggested in blue. Attention starts at the second hole from the left, marked with the red spot. Attention then starts to drift to the right, culminating at position 3 where attention has slipped off its first hole, but has not yet slipped onto the second. A fragile position at which concentration is needed and disturbance is likely to be fatal as attention does not lock onto the new hole until the fourth position and is only consolidated in the fifth. Attention is then firmly on the new hole and there are enough brain waves left over to increment the count.

It is easy to attend to one of these holes in the shelf, perhaps made even easier by the hole being circular, roughly the shape of the foveae. Indeed, it is easy to attend to any almost any feature of the visual scene. We have already mentioned the pointing dog, with is being easier to attend to something when the head is held straight and the something is straight in front. To that extent, the existence of the feature is irrelevant: the brain knows when it is looking straight ahead – but in practise this only seems to work if there is, in addition, something for the visual system to latch onto. Indeed, if one stares at a blank wall, the brain works quite hard to find features to attend to, and usually finds them. Perhaps jumping from feature to feature, but clearly preferring there to be features. Or at least, designed to look for and respond to features.

A process diagram

Figure 5

All this is summarised, roughly in the way of an SDM diagram, in the snap above.

The task as a whole is made up of four elements, with the three right hand elements repeating, more or less independently, in parallel. This suggested by the parallel task lines.

Each cycle, each hole as it were, is made up of four elements, executed one after the other, including in the four the ‘test for end of row’. We also allow an optional, slight relaxation during the hold. While the elements themselves are more or less conscious, but not self conscious, the decision to move from hold to slide is usually unconscious.

Loss of context is intended to suggest the error conditions of lost hold or of lost count, on either or which the task is abandoned.

The frames, layers and objects of LWS-N

In time, LWS-N is organised into frames of consciousness, each thought to be of the order of a second or so in duration, at least on average. With the idea being that when there is more change than can be conveniently accommodated by refurbishment, the brain moves to new build, that is say the brain compiles a new frame. There would, for example, be a new frame when the eye jumps to some entirely new position. A new frame, with new, or at least renewed, objects. This was recently reviewed at reference 6.

While the word frame evokes a static visual image, and there is something in that evocation, a frame is not completely static. It allows a certain amount of change, say predicted change or change within certain parameters. So the appearance during a second or so of a ball flying across the field of vision can be predicted and built into the frame, albeit at the expense of other details. But an ongoing frame is bound to an object of attention, and if that changes there is a new frame – subject to the trickery which follows.

Note that frames are organised serially, one after the other, and once the brain moves on, the previous frame is lost. One cannot go back to the previous frame, as might be possible if they were organised as a stack with pushing in and popping out. As is possible, for example, with the history preserved in the forward and backward arrows, top left in the Microsoft Edge browser.

Figure 6

In space, LWS-N is organised into layers and objects, with at least some of the layers being topically organised. And with one of those objects usually being the current object of attention.

The current proposition is that, in this counting task, a frame is held for the duration of a successful count. And the brain does the trick of having a single hole as the object of attention most of the time, while allowing that object to slip from one hole to the next.

So building on Figure 4, we get to Figure 6, purporting to show four successive stages of the count, amounting to one of the cycles of Figure 5. For the purposes of this task, the brain has reduced the image of the shelf at Figure 1 to a series of holes against a background, which each hole being a layer object with two parts, a (dark) frame and a (pale) interior. We call this layer 1. In real life, these holes would appear rather closer together than is convenient here.

With the additional blue dashed lines indicating the one or two layer 1 objects bound into a composite object by a layer 2 object denoted by red dashed lines – which lines also mark out the zone of high activation. This layer 2 object, the object of attention, can grow and shrink smoothly while sliding right on layer 1, or, equivalently, while layer 1 is sliding left on layer 2, while the layer 1 objects snap in and out of it. A blend of continuous and discrete activity which implements the notion that the current hole has a continuing existence, while a succession of particular holes are first attached to it, then detached from it. From which I associate to the Christian mystery of the Trinity. Hopefully this mystery is more tractable.

Part of the difficulty of this counting task is the difficulty of maintaining an object of attention which is not also a straightforward layer object on the straightforward visual layer.

Deciding exactly how any given situation would be modelled in LWS-N is, at present, not much better than guesswork. It also seems likely that, as with SDM modelling, introduced at reference 10, there are various ways in which one could express this task in the frames, layers and objects of LWS-N – and we are not yet in a position to know which way is the right way. A more familiar example might be Word or Excel from Microsoft: both are powerful, general purpose tools and there are usual many ways to achieve any one end. And we dare say the same is true of the programs, if we can call them that, expressed by genes in chromosomes. Perhaps, in the present case, it varies from person to person, from time to time?

But the two figures following summarise the current proposition.

Figure 7

Figure 8

Note the speculative use of layers 3 and 5 in Figure 7 for working data, for working memory. The second of which does not get into consciousness at all, except in so far as one loses the rhythm and the count collapses.

What we mean by survival in Figure 8 is that losing the current count, or finishing the count does not mean that the count is lost as well. One often loses one place, while still knowing what number one had got to.

Conclusions

We have sketched a mapping of a counting task into LWS-N, into a single frame of consciousness of LWS-N. An unusually long frame, with this unusual length correctly suggesting the difficulty of the task.

The layers foremost in consciousness are what are called above layer 1 and layer 2, the background, the holes and the current object of attention.

References

Reference 1: https://psmv3.blogspot.com/2018/09/supplementary.html. Counting courses of bricks.

Reference 2: https://psmv3.blogspot.com/2018/09/shopping-lists.html. The tall buildings section.

Reference 3: https://psmv4.blogspot.com/2019/04/a-speculation-moving-from-one-tooth-to.html. What is involved in stepping one’s attention along the teeth of a saw. A proposal involving saccades rather than tracking movements from one tooth to the next – with the teeth of a saw replacing the holes of a shelf we have here.

Reference 4: https://psmv4.blogspot.com/2019/12/shopping.html. Getting a different answer for the holes in a shelf. Towards the end of the post.

Reference 5: https://psmv4.blogspot.com/2019/02/a-special-sort-of-hand-eye-coordination.html. Climbing the stairs with cups of tea.

Reference 6: https://psmv4.blogspot.com/2020/01/another-look-at-frames.html.

Reference 7: http://psmv4.blogspot.com/2020/01/moving-cars.html.

Reference 8: http://psmv3.blogspot.co.uk/2018/01/an-introduction-to-lws-n.html.

Reference 9: https://psmv4.blogspot.com/2019/06/a-further-update-on-seeing-red.html.

Reference 10: http://psmv2.blogspot.com/2015/11/how-does-brain-do-sort-of-things-that.html. At the time of writing, the helpful reference at the end still works.

Reference 11: http://psmv4.blogspot.com/2019/12/belcea.html.

Reference 12: http://pointingdogblog.blogspot.com/. The source, via Bing, of Figure 3.

Group search key: sre.

Note

The section called ‘The frames, layers and objects of LWS-N’ was revised on Thursday, 6th February, 2020.