Being here now -
the past and future present 

This Page:

 Top of page   Site map

  Steven Lehar's Isomorphic Gestalt Bubble Theory

  Global workspace theory 

 CPI = careful phenomenological investigation

'A functionally isomorphic, volumetric, analogue model' - no less. 

noticing the way you see, hear and feel the world around you

  Anticipation 

   Link to my rough and ready abstract diagram
of impulse flow/connections involved in consciousness.
[Not on this page.]

 Work in progress -  This page edited with
Namo WebEditor v2.03.    

 Comparison of some different views of consciousness 
 

 Comments on Steven Lehar's Isomorphic Gestalt Bubble Theory

 Several things come to mind for me as I read Steven Lehar's Gestalt Isomophism paper.

He makes a very emphatic case for the volumetric representation of perceptual space. And yet this is incredibly challenging to think about because I feel the need to try and marry up what he is saying with what I think I understand about how the brain works. Steve Lehar's response to objections of this sort is that we do *not* understand how the brain works and therefor should concentrate on the phenomenal experience. I am currently not satisfied with this and I am trying below to tease out the threads which don't hold together. But at the same time I think there is a lot to be said for Steven Lehar's approach to the problems of understanding consciousness.  

Analogue computation - controlling the homunculus

He talks about 'fields' of attraction to or repulsion from things within the volumetric representation and how these can act on the homunculus - the body image - to make it move and thus operate the physical body like a marionette. This makes sense to me. It is a devastatingly simple underlying principle which explains adaptation of the body's position and motion to physical forces like gravity and other accelerations as well. It takes just a little extention to explain the gamut of physiological and emotional attractions and repulsions to and from things and people in the environment. I can see how emotions and physiological desires [NB: Damasio would say 'appetites'] can attach to percepts, indeed this must happen. [For me a recent example of a simple, explicit manifestation of fields such as Lehar describes occurred whilst I was riding my bike home from work. I had just crossed the main road up the hill from our place and was riding down the pavement to get to the small road leading to our house. The pavement there has several posts of various sizes sticking up - a power pole, a couple of street lamp poles, a couple of No Standing signs, and some signs about keeping dogs on leashes at the grassy edge of the park. As I navigated down the pavement wobbling and swerving to avoid these [in a very expert way mind you! :-], I could sort of 'feel' the influence of the posts as a kind of warning zone as I passed by each of them.]

The dynamic nature of the overall system also becomes apparent, including the way in which percepts can move around. In fact the whole phenomenal world moves in response to turning of the head and body. We interpret this change as us moving within our world and this marvelous fact shows how the model is of 'self in the world'.

I think this sort of thing requires that the dynamic logical structures which constitute percepts are made of patterns of variations in cell firing rates as opposed to patterns of individual impulses. What this means is that neurons have a normal rate of firing to which they will return after being disturbed, but this normal rate can be increased by excitation from other neurons or decreased by inhibition. How the brain selects amongst such dynamic logical structures for those which best fit the circumstances is a key question. [see below for more about this].

Making a 3D world model

Another thing is that Steve Lehar talks about the brain converting the two dimensional stimulation of the back of the retina into the three dimensional volumetric model of phenomenal space, and he puts this in the form:
"I propose to model the information in perception as a computational transformation from a two-dimensional colored image, (or two images in the binocular case) to a three-dimensional volumetric data structure in which every point can encode either the experience of transparency, or the experience of a perceived color at that location."

My problem with this is that from personal experience it is obvious that the system is built on the basis of binocular vision. Binocular vision is the norm, not single eye vision, and single eye vision is problematic because depth perception is greatly hampered. The fact is that the brain creates two renditions of the visual scene and then builds a whole lot more 3D experience out of analogically measuring the differences and similarities between them.

I have been running my own subjectivity experiment by going around and doing things with one eye shut. This simple change makes a huge difference in my competence at performing all sorts of tasks. Even typing seems much harder; presumably I should have been doing far more typing on the basis of the feel of where my fingers are on the keyboard rather than the surreptitious peeking that must be my normal style ;-) What I seem to be finding is that I rely more heavily on the memory of where things are, that is things that are now out of sight which would be visible with both eyes open, but also I rely heavily on what I *know* about things in the distance. [It would be instructive to hear from someone who has only ever had the use of one eye to see if they can identify situations in which they have to do things differently from the rest of us, although I can see how they might not be inclined to talk about it.]

Whilst using only one eye, I find myself noticing much more the process of parallax and occlusion resulting from the different relative speeds of progress across my visual field of objects at different distances. The main time I have available to perform this experiment is when riding my bike to and from work. Luckily I commute along a cycle path for most of my journey which makes this relatively safe; I would not recommend anybody trying it on a bike on a main road or behind the wheel of their car!

This is CPI#3 [see below for what that means]
With our visual experience we two eyed, cyclopean, 3D world viewers are fooled into belief that we see everything in high quality 3D because we have the ability to see, one at a time, all the things we want to concentrate on in 3D. Shut one eye though, and keep it shut, to catch your brain in the act of making up as it goes along the 3D depth of objects and their locations. You have to *do this* in order to see what I am talking about. BUT DON'T DO THIS WHILST DRIVING YOUR CAR!

You will soon find that simple tasks, like washing up for example or hunting for things in the fridge, are still easy to do but the visual experience is definitely not the same. By and large you can still reach for things, pick them up, wash, rinse, and stack them but they only look fully 3D while you are actually holding and moving them. It is as if the fullness of the belief about their 3D extension has to be confirmed by kinesthetic awareness from one's hands and arms or by continuous occlusion of the background when they move.

I infer from this that perception is as much an act of memory as of gathering information from the environment. Yes, the information from the environment is essential but it is only when these patterns of stimulation have evoked previously made constructs which best fit the stimuli that we can be aware of *what* it is we are looking at. It is an interesting question just how many things we can consciously see at any given moment. In particular, how many things at a time do we truly see in 3D? How many 'layers' of objects occluding others can we truly see when concentrating our gaze at a particular spot?

Change and simultaneity

It took me a long time to realise the full import of what I see with only one eye open. It is precisely the change, the movement of things in relation to their background or foreground. I was becoming really puzzled until it dawned on me that the brain has evolved to register coincidence and simultaneous movement. Thus there is part of the visual cortex endowed with sets of cells tuned to detect movement in particular directions. It will be a question of fact just how fine is the difference of angular direction between the different sets. I guess that, as the whole system is analogue not digital in function, actual direction of movement will probably be defined by the extent to which different sets of direction detectors are stimulated. The important point is that it is the simultaneous signalling of the same movement which identifies those points of the visual field as parts of the same object. Those points are going to be mapped onto other cortical projections of the same area of the visual field, thus associating colour, boundaries, etc., with the movement. There must be regions of cortex and/or sub cortical regions which register such coincidences and associate them with other more abstract representations of location in space, in story line, and in a resource versus danger assessment continuum.

This idea of simultaneity is critical for understanding how, for example, we can see through the many twigs of one bush or tree and see another tree behind it and distinguish the two [or more]. With two eyes, it is possible but not easy when the wind is still. With one eye it is next to impossible until the wind blows, but then it is strikingly obvious. The effect, and the difference between full versus half cyclops, is enhanced if using binoculars to view trees and so forth in the middle distance.

Connection between brain activity and consciousness

As far as I can seem this simple experiment shows also the deep, intimate and inextricable connection between brain activity and consciousness. The location of viewed objects in space is a fundamental aspect of visual consciousness. To put the connection succinctly, the rendition of 3D in one's view of the world is achieved by comparing two representations of the world in the brain - one derived from each eye - and using the similarities and differences to establish the distance [and shape, etc] of objects. The important point is that this process requires very precise feedback from the visual cortex to the musculature controlling the convergence of the eyeballs and accommodation of the lenses. The clarity and distance resolution of the visual scene is maintained by means of this binocular activity. Shut one eye and distance resolution is severely reduced even for objects close by. It becomes dependent on the occlusions of parallax motion caused by movement of the head and prior knowledge of the distance and nature of the objects being viewed. With one eye shut it becomes very much harder to navigate through new places.

If I am understanding Steven Lehar correctly, he asserts that our brains are modeling all of the [visual] perceptual space of our subjective world by a process which renders *each* 3D volumetric point as transparent or opaque. I don't think this is quite right. I think our brain's visual system explicitly models *each part of the field of view* - that is to say the inner surface of something less than half of a virtual sphere. It does this by rendering colour, texture, and surface orientation as features of objects and locating those objects at whatever distance from self is appropriate. This distance, it seems to me, is rendered analogically as an association with the degree of muscular contraction necessary to converge the eyes and compress the lenses just enough to produce coincident, focused images of the object[s] under observation. For distances within reach the rendering is also by association with the muscular activity in limb movements. The 3 dimensional extension, apparent mass and other distinguishing features of objects must be encoded by association with all the non-visual areas of cortex which encoded the original experiences of those qualities of objects.

The explicit, fine detail of our visual experience is rendered in the cortex on the basis of current updating from foveal vision. At any given moment there is only a small angle of the visual field which is in sharp focus and detailed to the maximum resolution, all the rest is peripheral and/or remembered from the last few saccades. The vast bulk of this information is not remembered permanently. Only features sign-posting or symbolising elements of our self story are remembered as well as things posing potential threats.

Other issues:

 How far is the sky?

Why the 'dome of the sky' is at the distance it is - because there is a certain maximum distance beyond which the brain cannot use parallax to gain information about distance from self and in the circumstances of our prelinguistic ancestors that distance was well beyond any normal region of interaction with the environment.

Stability and adequacy of mental models

How does the brain settle on an 'adequate' representation of any particular object or process? Steven Lehar suggests harmonic resonance, of something or other, as being the key ingredient. This may be right because if patterns of standing waves can represent objects this may allow us to explain both the creation and stability of mental objects and perceptions as being dynamic logical structures which constitute energy minima within a given context.

 The 'standing waves' cannot be all there is to it however because the vector fields operate on objects - dynamic logical entities - which come into existence then die off, ie

• they switch in and out of consciousness [according to my favoured hypothesis of linkage to self model] and
• switch in and out of existence [ie an established cell assembly is 'on' or 'off'] .

Furthermore, objects have afferent labels - which must be a type of vector - which characterise them, on one or more affective continua, as attactive actual or potential resources or repulsive actual or potential dangers. Objects can presumably be neutral also but then they will only be part of the background; transiently the focus of attention but only as an extension or sign of something/somewhere/someone which is of interest. My point is that the affective vector/s must have the ability to vary in correlation with perceived spacial-temporal and/or social-temporal distance from self.

For example an unleashed and apparently uncontrolled and vicious dog on the footpath up ahead as I walk towards it on the way to the bus stop. I have seen the dog and it has seen me, and I have had many experiences with such badly behaved dogs in the past. As I walk towards it my disquiet increases and also the feeling of needing to make a decision about whether to carry on up to it and pass by where it is now standing, or whether to cross the road and pass on the other side. Many thoughts would cross my mind with much weighing of the options. If the place was a long way from where I live and where I or any of my children would usually go, I would probably cross the road if there was noone about who I could talk to about the dog. If it was in any street near where I live or where any of my children might go I would feel compelled to go ahead and find out just what the creature was up to. [NB: This is not about me being brave, because I am not, it is about neutralising a threat. In Western Australia there are State laws and local by-laws compelling the control of dogs by their owners so it makes sense to invoke the action of the local ranger as soon as one encounters this kind of problem.] Once again, my point is that decisions in such circumstances are triggered by the subjective affect - such as fear or anger - reaching some threshold level. Prior to the emotion reaching that level it was building up, ie increases by degrees along some kind of continuum.

This kind of variation seems to sit better with the idea of increases in the firing rates of neurons in a cell assembly as well as the idea of a cell assembly co-opting more and more cells within each neighbourhood supporting its activity. It does not seem to sit so well with the idea of harmonic resonance patterns unless we are thinking about changes between harmonics well above the fundamental where, conceivably, the difference between standing wave patterns would not be great. I mean the tripping of a decision threshold could be the spreading of harmonic resonance pattern into new regions of the brain, thus inhibiting what was occurring in those places just before. But the build up to the threshold level has to be described

 Novelty versus stability

The brain seems driven to focus on novelty in the environment and incorporate it into the model of the world so it is no longer novelty. This makes sense for several reasons:

• something new can be either a danger, a resource, or neutral.
• from a natural selection point of view, it is more likely that a creature could be ambushed and killed by an unknown danger that moved into the locality [or which was chanced upon in a journey through unfamiliar territory] than by a danger which had been discovered and memorised as to its current home and probable whereabouts
• natural selection would have favoured a tendency to active but reasonably cautious patrolling of the tribal territory and investigation of new things in order to know how best to respond to them in future

An interesting question concerns the things which cannot be reasonably accommodated into the model of the world. This applies more to the abstract things of our social and cultural world than the every day world of physical objects. Many of these things become shadow contents: things deliberately avoided or denied existence altogether.

 Homunculus
This last speculation leads back to a big question about how the brain ascribes solidity, mass, permanence and importance to things and processes in the model of the world. Steve Lehar suggests these are different varieties of 'fields' embodied in the metric of the phenomenal model, with the field strength varying as a function of distances of the self representation from the representations of the objects within the world model. Thus decisions about what to do and how to move are processes of analogue computation. The relative strengths and directions of the various field vectors are summed such that the overall effect at any given moment acts on the body image/homunculus which induces the skeletal muscles to move limbs, torso, tongue, and so forth as required for the situation.

Generally speaking the strengths and orientations of vectors for any particular kind of field will be represented by activity of neuron groups - so called mini columns - within particular characteristic areas of the cortex. As cortical regions are basically 2 dimensional sheets, each particular cortical region might map two aspects only of any given environmental feature although the colour coding regions of the visual cortex may deal with 3 colour channels if, for example, the two dimensional cortical sheet is populated by sets of 3 minicolumns in which each member of the set represents the signal strength of a different colour sensitive retinal cone cell.

Global Workspace Theory [GW Theory]

 GW Theory is A Theory of Consciousness put forward by Bernard J. Baars and Katherine McGovern which they describe as:

"a simple cognitive architecture that has been developed to account qualitatively for a large set of matched pairs of conscious and unconscious processes (Baars, 1983, 1988, 1993, 1997). Such matched contrastive pairs of phenomena can be either psychological or neural. Psychological phenomena include subliminal priming, automaticity with practice, selective attention, and many others. Neural examples include coma and blindsight.

"GW theory may be seen in terms of a theater metaphor of mental functioning. Consciousness resembles a bright spot on the theater stage of Working Memory (WM), directed there by a spotlight of attention, under executive guidance.

"This architectural approach leads to specific neural hypotheses. For sensory consciousness the bright spot on stage is likely to require the corresponding sensory projection areas of the cortex. Sensory consciousness in different modalities may be mutually inhibitory, within approximately 100-ms time steps.

"Sensory cortex can be activated internally as well as externally, resulting in conscious inner speech and imagery. Once a conscious sensory content is established, it is broadcast widely to a distributed "audience" of expert networks sitting in the darkened theater, using corticocortical and corticothalamic fibers. Among the experts behind the scenes are "self-systems," viewed as contextual data structures that both shape and receive information from the bright spot; they include parts of prefrontal cortex, but may range posteriorly as far as parietal cortex for visual orientation.

"The primary functional role of consciousness is to allow a "blackboard" architecture to operate in the brain, in order to integrate, provide access, and coordinate the functioning of very large numbers of specialized networks that otherwise operate autonomously (Mountcastle, 1978). All the elements of GW theory have reasonable brain interpretations, allowing us to generate a set of specific, testable brain hypotheses about consciousness and its many roles in the brain."

 CPI = careful phenomenological investigation

 Phenomenology is the practice of observing the appearance of things in order to understand how one's brain, mind and perceptual systems work. This is a bit different from what we normally do, or tell ourselves we should do, which is to carefully observe things in order to better know what they are.

I have been trying out a few simple, do it yourself, experiments in observation of my vision.

 CPI #1 [I discovered this one while sitting on the throne today :-]

• If you look at a spot, i.e. some noticeable blemish or other easily discernible small marking, on an otherwise plain white or very light coloured wall, then shut one eye and just keep the open eye focussed on the spot, after a while something funny can happen. The darkness of the view 'seen' by the other eye comes to take over nearly all the visual field except for the small area of foveal view being observed by the open eye.
  
  I haven't tried this in lots of different situations yet but I have found that it there are other features in the periphery of view of the open eye then the 'darkness' stays on the inboard, i.e. nose, side of the visual field

 CPI #2

• One way of examining the differences between foveal and global world images is by using one's hand to occlude the view of first one eye and then the other, alternating in quick succession. Do this by concentrating on an object some metres away and using the hand on extended arm with the hand facing palm inwards to cut off the view of one eye and then the other. A little practice allows one to see that the object is 'moving' from left to right and return. If the object is on a vertical line, such as a doorknob next to the edge of the door, it is possible to see that doorknob so occluded appears to move from one side of the vertical line of the rest of the door edge to the other side. The line of the door edge above and below the foveal view, particularly beyond where the hand is obscuring the view, is a centred cyclops view and the line/edge does not move.
  
  I have confirmed for myself what I think is a demonstration of the foveal focus versus the global peripheral, view:
  Standing 3m from a pillar [of square plan cross section] in the basement car park at work concentrating my gaze on a spot on the edge of the pillar a bit more than 2m above the floor. This spot is in line with a pipe suspended close to the ceiling about 1.5m beyond the far side of the pillar. I am 166 cm tall so this means I am looking slightly upwards. By holding my forearm and hand vertically in front of me, palm flat and facing me, fingers straight and together and upper arm horizontal, I can move my arm and hand from side to side. This occludes, in turn, each eye's view of the whole vertical line, from floor to ceiling, made by the edge of the pillar. By doing this I can cause the point I am focusing on, on the edge of the pillar, which is in line with the distant pipe, to move from side to side like the doorknob described above. At the same time, I can move my attention down the edge of the pillar without also shifting my gaze downwards.
  
  *What I notice is that from a certain point downwards the edge of the pillar is not moving despite each eye's access to it being alternately occluded by my forearm.
  
  *In this particular situation the boundary between the moving foveal view and the static peripheral view happens to coincide with the upper reach of distinctive grey stripes painted on the pillars to make them visible to drivers. This boundary is about level with my chin and about 16 cm below my eyes’ fixation point.
  
  NB: The car park was deserted apart from me. I recommend doing this sort of thing in the presence of no one except informed friends or colleagues. :-) Otherwise, having a ready-made term like 'CPI' may come in handy.

 CPI #3

• If you shut one eye and keep that eye shut - assuming that you are normally a two-eyed sighted person - while you try to go about the things you normally do, you can start to notice things about the way the world looks. I have already written about this above. I tried to raise discussion of this on the 'Mind and Brain' group at Yahoo before Xmas 2003 and then again in June 2004 but was not successful. I put a complete copy of my 'Mind&Brain' messages in the footnotes of my diary #5.

Site map

Back to top    |   to end

10

20

30

Continuous Development

I am developing this page and this site. 
If you see something that needs improving and you want to tell me, please drop me a line or two at: 

   How to contact me   

Many pages on this site are unfinished, some are only just started.
They are none the less here because it is the ideas that are important.
You don't have to believe what I say about things, but if you come across a new way of seeing your world as a consequence of reading something here then your ability to think and act freely has been enhanced.
If that be so then my efforts have been worthwhile.