© copyright ERA 2014
(ISBN 978-0-9551535-3-2)
A theoretical proposition for retinal detection of a field potential unfolding from the light array to form the basis of spatial and orientation awareness within the phenomenon of vision. Propagated through the colliculacortical pathway, the proposed perceptual structure (tectal vision) would form the basis of multi-‐sense spatial awareness and orientation through the integration of field potentials. Abstract: If, as we attempt to envisage here and Vision-‐Space (VS) media illustrates, there are two independent raw data potentials, an implicit field potential together with an explicit detail potential, are being collected and segmented at the retina where conventional theory considers there to be just one, we can assume that much of what is currently being attributed in terms of receptor function, retinal circuitry and the visual pathways would be in need of fundamental review. Our task here is: • to suggest the possible nature of the oversight • look at what has been established to-‐date with respect to the neural circuitry • to see if we can make better sense of observations by adopting the new hypothesis • to see if this model matches with the experiential reality of phenomenal field. Given the scope of this remit can only attempt to outline a proposition here but the argument should identify directions for research to either confirm or deny. Our top down assumption driven from experiential encounter suggests that decoherence is taking place at receptor level and that both resulting phase and particle potentials are segmented for independent propagation through the visual pathways. Ultimately the question is; do we understand what’s actually involved in an act of observation? Approach: If we are attempting to explore the subjective realm of consciousness and visual awareness in particular, it is essential to gather a reliable body of intuitive insights from those that work in the experiential realm to set along side, guide and even inform scientific experimentation that is essentially indirect.1 Vision as we experience it, is prior to science. In contrast intuitive record (art work such as paintings) can establish a direct connection to our relationship with the real. From direct intuitive exploration of phenomenal field it is apparent that its function and capabilities are dependent on two independently computed and composed data-‐sets each presenting different ‘takes’ on the actual setting under observation (real setting). These data structures together with their dynamic of information exchange have been artificially modelled in a unique system of visual presentation known as Vision-‐Space2 (as opposed to picture space –
1 Merleau-‐Ponty, The Primacy of Perception that; “Science manipulates things and gives up living
in them. It makes its own limited models of things; operating upon these indices or variables to effect whatever transformations are permitted by their definition, it comes face to face with the real world only at rare intervals.” 2 The Perceptual Awareness Centre – Perceptual Technologies http://www.pacentre.org/pt/index.php
1
© copyright ERA 2014
reliant on the fundamentals of optical projection). As such we maintain that VS models key aspects pertinent to visual awareness.
Fig 1. Intro to decoherence
Intuitive assessments3 of phenomenal field imply that there are two independent structures operational each dependent on individual data potentials derived from the light array incident upon the retina. This duality apears underpins perceptual structure containing specialisms that can be associated with the dual characteristics of the particle and phase potentials of light. This in turn implies that the retina functions as a ‘converting membrane’ decohering4 the light array thus mediating transference from a micro scale of operation through to macro (classical) level mechanisms, a process entailing the preservation of both data potentials within neural circuitry. While decoherence may indicate that quantum processes are not directly involved in signal cascade and cue development to visual awareness5, the consequences of decoherence with respect to brain function and the emergence of mind may ‘live’ within us facilitating their mediation within phenomenal field6. The phenomenon of vision and hence the processes of VS appear to be watermarked with the reality issues associated with particle physics but by proxy.7 The suggestion is that our intuitive understanding of the data structures and the dynamic of information exchange occurring within the phenomenon of vision, are highly pertinent to the understanding of visual process. While these concerns are not the only ones involved in visual awareness, these factors play out back from the percept through the visual system to the retinal receptors to our understanding of light.8 Understanding these processes are also likely to inform and
3 Paintings – the intuitive record of key visual artists such as Chardin, Turner, Degas, Cezanne, Bonnard and the authors work (http://www.pacentre.org/era/artworks.php) 4 Stanford Encyclopedia of Philosophy, The Role of Decoherence in QM. http://plato.stanford.edu/entries/qm-‐decoherence/ 5 We would expect a continuous analogue signal to be in operation along side the digital (binary) with a unique neural computational process involving synthesis and convergence. 6 We would suggest that the ‘field’ data-‐set underpins multi senses integration (including audition). This in turn suggests that decoherence is key to play a key role in biological function. 7 Penrose & Hameroff, Hameroff & Watt 1992 suggest quantum coherence processes do play an active role in consciousness through microtubules. Some others have also sought to link the semi-‐entangled state of indeterminacy with vision, consciousness and art-‐work, an approach with which we do not ascribe. 8 Dark light and dark noise are known to mediate receptor firing in both rods and cones.
2
© copyright ERA 2014
involve our understanding of biological systems, concerns that lie outside the scope of this paper.
Fig 2. Proposed duality associated with the visual pathways
It is understood that the dorsal and ventral streams that can be used to characterise our understanding of the visual pathways can be traced back to the retina and both are contained within optic flow serving the cortical and sub cortical pathways. Retinal processes from receptor level are involved in the development of these pathways and segment data through them. The main areas of the brain influence the setting up and calibration of retinal cell structures ensuring that vision is clearly not a one-‐way process at both the phenomenological level and neural level. It is also apparent from intuitive examination of phenomenal field that even the experiential ‘object’ in ‘space’ delineations between central and peripheral vision broadly align with the ‘what & ‘where’ characterizations understood to underpin the two visual pathways. As a visual artist by training I think that it is not just pertinent to overlay first hand experiential explorations into the phenomenon of vision with what has be learned by vision science but that it is essential to do so.
Fig. 3 Painting with screen by the author
3
© copyright ERA 2014
Overview: Indicators from the experiential suggest that retinal processing structures contain a specialised function capable of ‘segmenting’ the light array, of decohering light input and preserving both functions for subsequent propagation and analysis within the visual pathways. Where the current theoretical model typifies the sampling of a ‘retinal picture’ (retinotopical mapping) based data-‐set (one for each eye) together with supporting processes to enhance and clean the signal of contaminating ‘noise’ 9, the intuitively observed structure of phenomenal field would suggest that this ‘noise’ element contains a second order data potential requiring synthesis as opposed to differentiation.
Fig 4. Spikes and thresholds (source Wikipedia)
The suggestion is that this data potential would pass through the subcortical pathway via the superior colliculus (SC) and brain stem structures as the basis of retinotectal vision in support of the dorsal pathway to ultimately provision spatial awareness and orientation within peripheral vision10. The characteristic of phenomenal spatial awareness being established via a field structure generating proximity cues unrelated to ‘pictorial depth’ and ‘depth of field’ that accord with the fundamentals of optical projection. The noisy field structure having characteristics more closely aligned with those of phase space with fixation being in part akin to the operations of an attractor as opposed to optical focus. Retinal processing: Dark light and dark noise are encountered in receptor firing at scoptic levels mediating firing in both rods and cones.11 The cross correlation functions observed in ganglion cells are known to be linked to quantal noise events taking place in the rod and cone receptors. If this noise, paired away from the photon dependent data-‐ potential for optimal absorption, were just a waste product then why is it preserved through retinal processing to be observed playing a role in ganglion cell output? A question raised by Greschner et al 2011.
9 Noise deriving both from within the raw signal and from the noise produced by neural firing
through a process of redundancy 10 It is understood that linkage is also made from the LGN to the SC resulting in a convergence potential. 11 The primary work on dark light and dark noise was carried out by in the latter part of the 20C. Leading investigators being Baylor, Barlow, Donner 1970-‐90
4
© copyright ERA 2014
In general, correlated activity between retinal ganglion cells (RGCs) can be produced by a combination of two factors: shared noise arising in common circuitry such as shared photoreceptors, and shared signal arising from stimuli with spatial correlations. It is unclear whether these two sources of correlations combine independently in natural vision, or alternatively, whether the noise depends on the signal.
Debates naturally arise as to the origins of the noise element under investigation and whether or not it contributes in some way to the shaping of differentiated spikes or some type of independent data potential. The argument for there being a dual data potential within the light array put forward here and provoked from deductions made from the intuitive records of artists would require a fundamental shift in our interpretation of retinal processes. Processes the include the function of receptor cells, the nature of the segmentation taking place within retinal layers involving the various types of cell receptive field (especially ganglion cells) together with the function of retinal circuitry involving the mediating roles of gap junctions, horizontal and amacrine cells. When considering the functional requirements of the potential phase based data potential leading to the proposed spatial field structure we will need to reconsider the operational range of receptor types in differing lighting conditions. We will need to look afresh at the role that dark light and dark noise may play in mediating receptor firing and its derivation. Consideration will also need to be given to the functional significance of the impulse and passive processes of light absorption in receptor cells. For reasons that need not concern us at this point, the supposition is that the passive phase related process characteristically builds towards a processing system driven by synthesis and will be largely associated with the dorsal, ‘where, spatial’ pathway. The measured and differentiated impulse driven process (spikes) will be more closely associated with the development of the ventral stream and the detailed registration of ‘what and objective form’ experienced within macular vision. As ever, these delineations are but characteristics associated with the pathways that interrelate and interweave like entangled webs throughout visual process. Nevertheless, we will be making an outline theoretical case for the retina’s separation and propagation of these two independent data potentials enfolded within the incident light array that would subsequently discharge from ganglion cells the through geniculatecortical and colliculacortical pathways.
Fig 5. Visual pathways (source Wikipedia)
5
© copyright ERA 2014
Experimental considerations: Before we embark on this theoretical exercise involving reflection upon related scientific papers we should seriously consider the base line assumption under which that experimentation was designed. Experimentation is by default organised and calibrated to observe processes in support ‘differentiation’ and not ‘synthesis’. For example, it will be necessary to consider the spherical geometry of the retina and the degree to which laying out flat, small sections of retinal tissue are likely to adversely affect the outcome of experimentation? I have yet to find experimentation involving the retina that seeks to present it with a natural stimulus, a real setting under photopic daylight conditions. It is after all, in response to these conditions that the eye evolved enabling us as the sentient being to engage and operate within the environment. The assumption is that a photon is a photon is a photon and the retinal and visual system processes these stand alone standardized units. Anything lying outside this base assumption would constitute an anomaly, something to be seen in relation to the base assumption. We are all broadly familiar with the idea of photons arriving over time and being considered responsible for an impulse based data potential (spikes) from receptor through to ganglion cell output streamed through the optic chasm.12 We can watch photographic film develop in the chemical tray. There is, in fact at least one clear place where action at the single quantum level can have importance for neural activity, and this is at the retina. Roger Penrose: The Road to Reality P.516 If, as VS theory suggests, we develop our primary from of spatial awareness direct from the light array under photopic conditions, stimuli (including photographic media) that don't supply the proposed specific environmentally charged data formation carried within the light array to the retina will, as an atypical input, adversely affect receptor firing and limit vital aspects of signal segmentation. Results taken from experimentation based on the current approach are likely to be incomplete and/or difficult to interpret. The experimental set up would to some extent dictate the results obtained and the confusion that follows. We suggest that a full data collection signature will only become apparent in the assessment of responses generated from across the entire spherical geometry of the retina and therefore across multiple receptive field populations. Studying the individual firing of receptors or even small groups of receptive fields in isolation is likely to inhibit the detection, propagation and hence comprehension of the phase related data-‐set potential. The second important point is that this specialist potential is contained within the light array coming directly from a real setting, the real setting being the spatial arrangement of object forming the scene illuminated under photopic (daylight) lighting conditions. The data structure will either not be contained, or incompletely referenced in other forms of stimuli presented to the retina, (eg. photograph and film media or photons fired from an artificial light source including a diffuse light field). This prevailing situation suggests that in evaluating scientific results we must adopt a ‘floating’ position with respect to both its intensions and findings. We don’t see ‘pictures’ in either eye! Vision is not a matter of deconstructing rectilinear optical projection in order to transport it to some perceptual screen. Vision is closer to a controlled hallucination a process that involves the sentient being set within the environment. We must endeavour not to make assumptions with respect to the nature of the stimuli presented to the retina without first envisaging the potential implications such restrictions may impose upon the system under investigation. The visual artist
12 This pathway would be more closely associated with X cells dealing with high definition data fed through the Geniculocortical pathway.
6
© copyright ERA 2014
chooses ‘still life’ set-‐ups as the primary means of investigation into phenomenal field for a reason.
Fig. 6 Still life with Aloe Vera by author
Disorder and the field structure: In addition to the environmental factors we also need to consider the probable data formations incident on the retina together with the probable receptor cell function required for their detection. These considerations should enable us to optimise experimental setups in order to verify assumptions and to assist in our penetration of research conducted without taking this position of oversight. This is an entirely valid undertaking if we accept and adopt the experiential ontology. We can refer with good reason, to our intuitive investigations into phenomenal field13 as being representative of the output from the system under investigation.14
"Blur" is technically a convolution with a non-‐negative, localised kernel (like defocussing a camera or projector). In the limit o f infinite blur you end up with the average over the image, a uniform field. Fig 7. Examples of blur and disorder
!
!
"Disorder'' implies spatial shuffling. It destroys spatial resolution but leaves the histogram invariant. In the limit o f infinite disorder you obtain a texture w ith the same histogram as the image. Prof. Jan Koenderink
13 Visual art of the author and others such as Turner, Cezanne, Van Gogh, Degas, Bonnard, Monnet 14 Psychophysics, Neuropsychology, Neurophenomenology, Visual Art etc
7
© copyright ERA 2014
We are now also assisted by the artificial imaging output (perceptually structured stimuli) from the VS imaging software. The stimuli it produces are in the process of being psychophysically verified. The information structure ‘blur’ does not appear within phenomenal field, neither does motion blur or depth of field, these occur within the optical records produced by camera technology based on optical projection. The phenomenon of vision is entirely non-‐photographically rendered. The data structure ‘blur’ is rarely used by visual artists as they explore and develop strategies to portray the nature of experiential reality. Research undertaken by Prof Jan Koenderink and Andrea Van Doorn identify15 that the structure of data within peripheral vision should be thought of as disordered. This implies that we should be looking for receptors and receptive fields capable of realising and working with a spatial organised texture. Artist have to varying degrees applied the texture producing tools at their disposal in pursuit of this dark data potential; canvas and paint applied by brushes. “We argue that locally orderless images are ubiquitous in perception and the visual arts”
Fig 8. Still life with cup and saucer by the author
15 Koenderink. J. & van Doorn A 2000 Blur and Disorder Journal of Visual Communication and
Image Representation vol 11 pages 237-‐244. Koenderink. J. & van Doorn A 1999 The Structure of Locally Orderless Images. International Journal of Computer Vision vol 31 pages 159-‐163
8
© copyright ERA 2014
Fig 9. Pine tree near Aix en Provence, 1995-‐97, Paul Cézanne, Oil on canvas, the Hermitage, St Petersburg, Russia
Fig 10. Self portrait, 1889 Van Gogh, Oil on canvas, Musée d'Orsay, Paris
9
© copyright ERA 2014
Fig11. E. Degas Mademoiselle Malo, Pastel on paper, Barber Inst of Fine Art, Birmingham, UK,
Fig 12. JMW Turner Boat in a Storm, Oil on canvas, Tate Gallery, London, UK
VS software algorithms undertake the disordering of a photographically rendered data-‐ set by randomising pixels within a proscribed area. When applied systematically within a radial geometry utilising depth-‐map data and centred on a given fixation point it generates a field structure with incrementally increasing spatial texture outwards in space from the fixation point (Fig 6.). It establishes an implicit spatially salient medium with the proximal spatial arrangements between objects and surfaces directly evidenced. We instantaneously comprehend theses spatial arrangements once we alight on the fixation point within the depicted scene. This awareness is not the result of conceptual analysis compiled through multiple fixations accruing second order spatial
10
© copyright ERA 2014
cues such occlusion or perspective. Field derived spatial awareness is ‘implicit’ and distinct from the notion of ‘pictorial depth’ and the optical formation of ‘depth of field’.
Fig 13. Self-‐similar sunflower pattern and its possible articulation as a field potential set out from fixation.
We believe that the implicit proximity cues play a significant role within visual awareness identifying that ninety percent of phenomenal field is not an optically degraded form of macular vision where motion is ‘tracked’. It forms a simultaneously understood spatial field where movement is understood in ‘flow’ from within the structure into which we (the perceiving organism) are spatially factored. This ‘take’ on reality comes with its own unique form of attention.
Fig 14. Picture space, photograph (optical projection)
11
© copyright ERA 2014
Fig 15. Vision-‐Space, monocular phenomenal field (perceptual structure). There other perceptually significant transforms involved in VS media relating to perceptual structure but these are not dealt with in this paper. For moving image examples of VS media using post process software and access to relevant supporting information see www.pacentre.org
The articulation of “spatial disorder” from a phase derived data potential is likely to involve what has been considered to date as merely non-‐contributory noise and involve mediation, summation, simultaneous firing over a short time interval within a restricted area together with stochastic processing leading to synthesis and data convergence. In essence we are looking for the origins of a dark data-‐potential feeding what remains a largely covert processing stream. Retinal cell function: With these factors in mind we can return to the physiology of the eye, its biological processes and the known properties of retinal firing to trace the hypothetical dual data potential, from detection through development to their respective streaming from ganglion cells to specialised areas of the brain for visual cue development. While I am not an expert in this field, if we look at the various component cells that make up the retina it is clear that most of these structures could equally serve both proposed process functions through photopic, mesopic and scoptic luminance levels. The delineation of data-‐sets can’t be clearly assigned to independent regions of the retina or cell types or even to how they fire or are connected to one another. As stated above, these processes form entangled webs, however through intuitive experiential encounter I think it is possible to become sensitised to the different nature of each and thus start to render visible aspects of their signatures and general characteristics as we assess the detailed scientific investigations that have taken place.
12
© copyright ERA 2014
Fig 16: Cross section of the retina (source Wikipedia)
As we have intimated, the current model approaches, considers and traces retinal process largely in terms of differentiation and thresholds. A process whereby light detection is defined by spikes, its segmentation from noise and initial aspects of scene analysis are undertaken prior to latter stage assembly type processing in the main brain via the geniculocortical pathway. One critical aspect of the associated retinal process being how to segment the photon related spike event from associated but unwanted noise contained in the both the light signal and generated from internal neural processes. However, scientists are increasing aware that this unproductive noise is counter intuitively not ‘filtered from’ of the system once segmented. It’s retained and appreciated to influence data processing throughout retinal layers in multiple ways. Noise becomes a functioning component of the system. At receptor level ‘dark light’ and ‘dark noise’ are encountered with rod and cone firing at scoptic luminance levels mediating receptor firing. Largely due the difficulty in making measurements, debate is still playing out between whether or not it’s source derives entirely from internal retinal processes or a mixture including the external incoming signal. To this we have to add the unresolved functionality of gap junctions between receptors and horizontal cells linking receptors and receptor types located in the outer plexiform layer together with amacrine cells performing a similar function in the inner plexiform layer. These cells develop extensive lateral connections and feed back loops that further mediate output to higher layers. It is generally understood that a parallel processing system is in play from the receptors themselves with the cones not simply conforming to a binary mode of operation from their two types of synaps with bipolar cells. Large scale correlated firing of cell groups is also observed and this is associated with phase like information being streamed from retinal ganglion cells to main visual processing areas of the brain. Horizontal cells: Are under the influence of neuromodulatory factors from within the retina and even from the other direction involving input from the main visual areas of the brain. They introduce a form of lateral inhibition giving rise to a center-‐surround
13
© copyright ERA 2014
structure to retinal receptive fields. Most species have two16 or more kinds of horizontal cell and through type dependent gap junctions with neighboring cells make multiple long and short-‐range interconnections involving feedback potentials forming networks across the entire outer plexiform layer.17 Detecting a coherence-‐based signal from the light array would require significant temporal coordination in addition to a unique processing capability. Horizontal cells have an array of morphologies that can be specific to species. The A type horizontal cell has been associated with colour discrimination in fish but the B type connect to rods and have associations with perception of brightness. In mammalian species the B type also connects to cones but their physiology serves to electrically isolate one area of the cell from the other, thereby separating a cone-‐photoreceptor relationship within the cell from its rod-‐photoreceptor relationship. The S potential horizontal cell recordings (linking cones and rods) identify that cone receptors first spike and then show a ‘rod influenced after effect’.18 A combination of data-‐sets in one packet, one data potential carried or piggybacking on the other?19 Bipolar cells: They receive the synaptic input from either rods or cones, but not both, and are designated rod bipolar or cone bipolar cells respectively. However, as we have seen the inputs from either rods or cones contain influences from either the other cell type or contain a dual data potential capability. Again, they act, directly and indirectly with input from horizontal cells that even have the ability to influence their receptive fields,20 to then transmit data potentials to the ganglion cells. Amacrine cells: There are many types of amacrine cells that are thought to integrate, modulate and interpose a temporal domain to the visual message presented to the ganglion cell. Both amacrine and horizontal cells are also thought to segregate motion and colour potentials and these can be broadly aligned to specialisms within dorsal, peripheral (implicit) and ventral, central vision (explicit) respectively and hence have associations with the two principle visual pathways. It would seem plausible that the various types of each kind of cell would be broadly assigned to pathway specialisms. Ganglion cells: In total there are thought to be 11 types of ganglion cells21 with very different receptive fields (dendrite arboration) in terms of their circumference and their penetration within the inner plexiform layer.22 Much work has been conducted to confirm that these types perform specific physiological functions.23 Ganglion cells release directly to the optic tract and are largely of two types; midget (small receptive fields and linked to small groupings of cones and rods), forming part of the P-‐pathway
16 A type cells are axonless, B type cells have axons. Dendrites of A and B types connect to cones but only B type cells connect to rods. 17 Yamada and Ishikawa 1965, gap junctions were identified as “fused membrane structures” specialized for electrical transmission 18 Steinberg 1969 distinguished rod and cone signal separation in S-‐potentials cells in mammalian retina. 19 Given that the evolution of the eye saw the lens structure coming after the chamber and the potential ‘implicit’ field capability we should in fact think about the spike element to be piggy-‐ backing on the wave function? 20 Influences including spatial opponency. 21 This number could be stretched to 13 on a technicality. 18 types of morphological types in the human retina. 22 Levick et al 1964-‐1983. 23 Perry et al, 1984; Amthor et al, 1989b; Watanabe and Rodieck, 1989; Bloomfield, 1994; Dacey and Lee, 1994; Yang and Masland, 1994; He and Masland, 1998; Rodieck, 1998.
14
© copyright ERA 2014
(parvocelluar) sensitive to colour and shape, and parasol (larger overlapping receptive field linking to relatively more rods and cones), forming part of the M-‐pathway (magnocelluar) sensitive to depth but not colour. In the fovea a single ganglion cell will communicate with as few as five photoreceptors whereas in the extreme periphery, a single ganglion cell will receive information from many thousands of photoreceptors. It is actively considered that the variety of response experienced within the centre-‐ surround receptive fields of ganglion cells represent several distinct mechanisms the most significant probably being its ability to perform spatial tuning. We would suggest that a potential exists for the centre surround configuration to detect spatial texture in the form of disorder distributions.
Fig 17. Suggested possible role for OFF centre receptive fields in determining a phase related distribution pattern associated with the detection of a single differentiated impulse from an ON centre receptive field
Linear and non-‐linear receptive fields together with their sensitivity to light and dark phases (contrast) suggest an ability to tune optimally to phase based data-‐potentials. Differences in a cell’s ability to perfect spatial summation distinguish X (associated with detail/what) from Y cells (associated with space/where). The SC is fed with ganglion cells types X, Y and W. Y and W 24 cell types have been associated with spatio-‐temporal low frequency response profiles and extreme sensitivity to moving stimuli W. Waleszczyk et al 200425. This makes the W cells projecting to the SC a prime candidate for the transmission of a phase based data potential. It is understood that collicular cells receiving cortical stimulation from W cells do so at latencies consistent with convergence of afferents both direct from the retina and indirectly from Y cells from the LGN. With respect to the central hypothesis, it would appear that we are required to challenge every aspect of what is thought to be occurring within the membrane even in
24 In the primate the W pathway (cat) is analogous to the K (Koniocellular) pathway also
associated with modulation between layers in the LGN. 25 W. Waleszczyk et al (2004) Motion sensitivity in cat’s SC. W cells with heterogeneous receptive fields could be associated with ambient vision and perceptual space -‐ local movement in the environment (Rowe and Palmer 1995).
15
© copyright ERA 2014
relation to receptor function. For example it’s generally accepted that rods are saturated in daylight photopic levels rendering them inoperable or incapable of contribution to visual awareness processing. Rods are thought to be productive at only at scoptic (dark) or working alongside cones at, mesopic (semi dark) levels of luminance. As far as I can determine from this theoretical exercise of overlay, two outline possible scenarios suggest themselves. The first option accepts the duplicity between rods and cones relating to their performance at different luminance levels with the second being more radical than the first, namely that rods are actually ‘optimised’ in photopic conditions. What we have thought of as ‘saturation’ in terms of a cell’s ability to differentiate individual photon detection events being not representative of its principle function in photopic conditions, that being the passive transmission of noise. Noisy photopic level rod output would require subsequent synthesis among groups if not entire networks of receptive fields making use of the specialist array of amacrine and ganglion cells.
Theoretical scenario 1. • We accept that rods are indeed ‘saturated’ at photopic lighting levels and that saturation ensures that in these specific conditions they play no effective role. • We assume that decoherence is taking place at the retina and that the phase data potential is passively absorbed and ‘embedded’ within a noise function. • We consider the possibility that this noise function carries with it information relating to spatial arrangements within the immediate environment. • We acknowledge that both rods and cones mediate their firing indicating the presence of what is termed dark light or dark noise. However we must assume that only cones are involved with the segmentation of second data potential. Visual environment Starlight Moonlight Indoor Light Sunlight Photopic Luminance -‐6 -‐4 -‐2 0 2 4 6 Luminance category Scoptic Mesopic Photopic Receptor type Rods only? Rods and Cones Cones only? Rod threshold Cone threshold Rod saturation Visual function Little or poor colour definition Good colour definition Fig 18. Current understanding of luminance levels with respect to receptor functions
•
• •
While we are able to detect the said functional operation involving dark light at scoptic levels as we are dealing with individual light quanta, we would presume that this ‘process’ is also operational in cones at photopic and mesopic levels of illumination but impossible to individualise or isolate. Given that decoherence would take place within the receptor we would expect that a debate would result with respect to the origins of the mediating noise. The segmentation of the noise reliant data-‐set from the photon event data-‐set would occur via the dual ‘on’ & ‘off’ structure of cone synapses with bipolar cells working in conjunction with the feed back capability provided by the first stage circuitry of horizontal cells in the outer plexiform layer.
16
© copyright ERA 2014 •
•
This first scenario would require that under photopic conditions we are reliant on the cones in peripheral vision to provide the sparse spatial awareness function apparent in peripheral vision. That a second data potential is integral to the ‘where’ processing stream providing us with a specialist form of spatial awareness derived directly from the light array, ie, it’s not reliant on conceptualisation of the scene from other cues, eg occlusion.
While this scenario has to be considered as a potential hypothesis there are obvious issues with its capability to fulfil the brief. • At a common sense level, I think the distribution of cones in peripheral areas may be too space to generate this hypothetical second data potential without support. • The shear amount of horizontal connections between rods is suggestive of their primary function being closer to synthesis than differentiation and so potentially implicated in the proposed process. • The rods are clearly involved in the dark light mechanism at scoptic levels which suggests that they are involved in the production of the filed data potential operational at photopic levels. • Rod population levels also intuitively suggest to me some functional operation at photopic levels. • With their small receptive field I would see the role of cone receptors as being supportive to a principle mechanism. The cone output deriving the noise potential via its OFF centre synapse providing an element of further grain resolution within the field potential.
Fig 19. Population distribution of retinal receptor cells (source Wikipedia)
We are left with either a potentially significant numbers problem with respect to noise sensitive receptor cell population in peripheral areas or a very significant total oversight with respect to rod function? If VS theory is correct, we must assume to have simply missed an entire data structure vital to visual awareness so I am drawn to seriously consider this more extreme second scenario. When I structure this argument it actually make better sense of retinal circuitry but also total sense in terms of the experiential reality of phenomenal field. As a non-‐expert, I consider there to be an intuitive fit?
17
© copyright ERA 2014
While we have observed that noise processing within the retina is making a contribution to visual awareness we are struggling to understand what this contribution looks like, how it manifests. I understand from experts that the same consideration is true of the role that decoherence plays in visual awareness. I would suggest that these two factors are directly linked and that we have failed to make the connections as there’s a fundamental misunderstanding enshrined in the scientific approach. This situation has gone unnoticed largely because vital cross-‐referencing with intuitive records made of phenomenal field articulated by visual artists has not taken place?
Fig 20. Large Pine Tree and Red Earth, 1880-‐97, Cezanne, Oil on Canvas, The Hermitage Museum, St Petersburg
1. We have not understood the nature of spatial awareness or what’s actually involved in an act of observation in phenomenological terms 2. We have failed to seriously theorise about to the role that decoherence might play in retinal processing, brain function and visual awareness 3. The scientific ontology has limited our analysis methodology to differentiation overseen almost exclusively by a reductionist outlook 4. Partially in response to the above, we have failed to presents the visual system with appropriate stimuli These factors may have ‘blinded’ us to a fundamental truth that runs right through our understanding with respect to the nature of reality and our place within it. We are partners in the development of visual awareness. The noise from light array is used and there’s actually a case for suggestion that biological noise generated internally may also play a part in the formation of perceptual structure being the mechanism through which the spatial signals are realised? We generate vision. Vision is a relationship we form with the real. Are we are currently living in a hall of mirrors a situation that persists due to the dominance of the ‘remote observer’ position adopted by science? Theoretical scenario 2. We therefor theorise • We assume that decoherence is taking place at the retina and that the phase data potential is passively absorbed and ‘embedded’ within a noise function.
18
© copyright ERA 2014 • • •
• • • •
•
•
• • •
We consider the possibility that this noise function carries with it information relating to spatial arrangements within the immediate environment. That ‘measurement’ of discrete events (photon events delineated by individual spikes) is not made within this data processing stream. Realising the spatial data potential embedded in the noise signal is not a matter requiring differentiation, it will be reliant on convergence and synthesis requiring considerable amounts of horizontal connections, summations, correlations and ultimately convergence. That BOTH rods and cones are specialised in segmenting that data potential from the light array. We should be looking at the function of linear and non-‐linear ganglion cell receptive fields in this respect? That this ‘data potential’ from rods and via cones is available to us at photopic luminance levels. That a second data potential providing us with a specialist form of spatial awareness is derived directly from the light array, ie, it’s form of spatial awareness is not reliant on our conceptualisation of the scene from other cues, eg occlusion. Realisation of this data potential will be made via a perceptual structure generated by the sentient being and probably making use of internal biological noise factors (eg from receptors and synapses). There will be little point in analysing individual rod events beyond observing the skeletal processes involved in segmenting the data potential for streaming to the main brain areas. The retina is unlikely to be the location where the realisation or unfolding of this spatial data potential takes place. Related ganglion cell output will simply possess phase like characteristics. The retina will only be functioning optimally with respect to the ‘collection’ of this noise orientated data-‐set when exposed to a naturally illuminated real setting.
Given this set of criteria we should not be expecting to observe rods detecting from inappropriate stimuli presented to it at photopic levels, elements requiring processes involving differentiation. Differentiated responses with respect to exposure to flat coloured or tonal stimuli would not be compatible with the tasks rods are designed to perform at photopic luminance levels. They would be required to passively absorb noise, all of it, without differentiation or measurement as phase receptors. If this were indeed the case then circumstances aligning with ‘saturation’ would represent the ‘optimisation of function’ and not its cessation or suspension. Connection circuitry between cones and bipolar cells to individual ganglion cells would suggest that they carry the differentiated explicit data-‐set familiar to central vision. Noise segmentation through the OFF centre synapses of cones would suggest that horizontal cells (outer plexiform layer) share considerable characteristics with amacrine cells (inner plexiform layer). Horizontal cells and gap junctions would then connect cones and their OFF centre signal to adjacent rods facilitating the transmission of the segmented noise component to the main noise processing stream. Cone output (fast track) via the ON centre synapses with bipolar cells also carries a segmented slow rod like component along with the differentiated spike, the noise like slow element almost piggybacking on the main signal. If we consider the slow wave form of rods as being easily distinguishable from the fast wave forms of cones, the observed mixing of
19
© copyright ERA 2014
these responses would appear to be counter productive and counter intuitive. Perhaps we can suggest that This slow wave form ends up being streamed along with the spike component to the Lateral Gesticulate Nucleus (LGN) and from there posted to the Superior Colliculous to converge with the main implicit data-‐set? One consideration has been that the rods need to utilise the faster cone pathway to transmit spatiotemporal data in advance of its regular pathways? All of this reinforces the notional possibility that by limiting our approach methodology to the evaluation of differentiated data potentials, science has managed to entirely avoid a data-‐set essential to the formation of visual awareness? By understanding that rod saturation at photopic luminance levels effectively rules out differentiation tasks, rod function has been only partially illuminated featuring contributions restricted to mesopic and scoptic luminance levels. If so, then could this omission in our understanding be aligned with so called covert processing, blind sight and tectal vision delivered through the retinocollicular pathway? The functional transition of rods through luminance levels? Rod noise and its proposed associated function at photopic levels would start to fail when environmental illumination levels were unable to deliver a robust enough data potential for streaming. When this condition was reached we would assume that a reflex kicks in initiating the well-‐understood light adaption period. We would suggest that this adaption period is also a transformation process affecting the function of the rod receptors where they start to align their innate functional sensitivity to differentiation tasks mimicking aspects of cone function that struggle at these lower luminance levels. This would identify that while cones support rods in photopic conditions via their OFF centre synaps with bipolar and horizontal cells, they also play a supporting role for rods in scoptic conditions through their ON centre synapses as rods change their functions form phase or noise detection to discrete detection enabling them to perform differentiation tasks. This transformation may be linked to the observed hyperpolarisation function seen to take place in the receptive fields of most retinal cell types?
Fig 21. The adaption period (source Wikipedia)
20
© copyright ERA 2014
This transformation in rod function suggests that spatial vision at scoptic levels is impaired by a loss of the implicit spatial function and not just by an incremental deterioration of acuity and colour vision. This again squares with experience as we regularly knock into objects or misjudge distances even though we can identify objects within the scene. It also explains the transition in our experiential objective assessments of the moon through different luminance thresholds. At scoptic levels the moon looks like a flat shining dish, clear but entirely lacking a sense of its spherical form. In daylight it’s clearly an object with volume in space. At dawn and dusk it’s possible to make out two distinct monocular images of the object as the brain struggles to modulated the available data-‐sets to generate a meaningful spatial impression. Towards night and conditions of real darkness one data-‐set fails entirely, in daylight the brain becomes increasingly able to resolve the object via constructive modulation and the double impression ‘sets’ into an appreciation of its spherical form. This set of conditions is especially apparent with the moon in crescent phases.
21
© copyright ERA 2014
Fig 22&23 Perception of the moon in scoptic, mesopic and photopic conditions
Implications for visual system overview: We would suggest that the principle function behind retinal receptor cells is decoherence through which we as sentient beings are able to make sense of the world of real settings through the medium of light. Another, but very different way in which decoherence would play an active role in the appearance of the physical world! Interestingly, if decoherence does occur at the receptor cell level creating ‘preferred states’ then these would now be ‘robust’ and less likely to be further influenced by the environment, where the environment would be the perceptual processes/mechanisms of the sentient being. The resulting data potentials would remain ‘segmented’.
Dark light and dark noise at the level of light quanta have been observed mediating receptor firing. Multiple variations in firing patterns of receptor cells have been recorded from what was at first considered to be the simple centre-‐surround binary process of neural firing. Within these variables being increasingly understood by scientists it seems plausible to theorise that distinct data-‐sets are being segmented from the light array. For example, it has also been observed that adaptation of the light input (changing its nature) affects the rod-‐influenced aftereffect of S-‐potential horizontal cells suggesting that if the stimuli fails to contain such a variable that would otherwise have been present in normal circumstances, the firing patterns at the retina reflects this deficit. An Off centre and On surround could theoretically perform measurements with characteristics reliant on phase formations giving rise to a texture value such as the registration of disorder (distribution pattern). These values would first find articulation
22
© copyright ERA 2014
within a local field potential (LFP) 26 but then require integration within a pan-‐retinal response. This mediated and then collective response could form the basis for the proposed enfolded field potential. The use of, or contribution of biological internal noise lies beyond the scope of this paper but again I would consider it to be integral to the proposed field structure that we generate. It is likely to be a functioning component within perceptual structure.
Within the ganglion cells layer we also see linear and non-‐linear receptive fields, two distinctive forms of firing (correlated and uncorrelated) that are starting to be considered to provide distinct modes of visual signaling (Meister et al. 1995, Schnitzer & Meister, 2003; Schneidman et al. 2006; Greschner et al 2011). It is understood that the correlated firing serves the purpose of networking interactions from ganglion cells promoting a phase like data-‐set to the visual system. The cause/derivation of this synchronized activity is thought to be a combination of shared ‘noise’ from photoreceptors and the stimuli. While ganglion cells are largely considered to be ‘feature detectors’ requiring differentiation functionality, this form of approach may well apply to just one segmented ‘take’ from the light array, to the formation of just one aspect of a duality.
Preparing the second ‘take’ developing from the segmented noise potential into the proximity and orientation cues apparent in spatial awareness would be reliant on an aspect of correlated firing only coming together under convergence processes as an integral and holistic data-‐formation further down the visual pathway in specialized visual processing areas of the main brain. At that point all the relevant inputs from across the retina that collectively form the integrated implicit data-‐set would then be ‘realised’ within a neutrally generated field data formation (unrelated to optical projection)27. This is suggestive of a form of phase space coalescing around an attractor delineating fixation. This field potential would then be experienced largely subconsciously as tectal vision but also deployed to underpin and support a multi-‐sense integration.28
The suggestion being that the holistic field structure ‘unfolds’ through the magnocelluar/dorsal associated pathway within the subcortical mid-‐brain structures. This would entail the SC receiving not a 2D topographic and retinotopic map along side its gaze direction/ eye movement data, but a distinct mode of signaling from ipsilateral views generating a unique form of spatial field awareness. These data-‐potentials break out in terms of perception pathways broadly as follows;
26 LFP are a type of electrophysiological signal. A ‘summation’ given off by a volume of dendritic synaptic tissue. 27 Neural adaptation (modification of the raw data received) occurs at all levels of the visual systems (including receptors) confirming a considerable degree of neural plasticity (a basic biological function). Much of this activity is thought to be confined to just improving the ‘quality’ of the traditional optical picture and the efficiency of retinal signal or delivering ‘constancy’ in perception (eg. light and dark adaptation. eg. constancy of colour perception despite the effects of aging.) Adaptation mechanisms are appreciated to paly a role is removing spatially temporally redundant signal components (background signal). 28 An example of aggregation from multiple traces from ganglion being required further along the visual system in order to determine spatial awareness might be illustrated by the output from directionally selective ganglion cells?
23
© copyright ERA 2014
Firstly a holistic and implicit form of spatial awareness generating proximity cues, with motion comprehended in flow • Then a, form, colour and detail explicit take on reality, with space through depth perception and motion comprehension via tracking This would entail the retinal membrane segmenting both data potentials from the incoming light array via receptor specialisms and supporting the subsequent circuitry for their development to a point where they are independently streamed to individual areas of the brain. This data-‐potential now manifest as a neutrally generated field would not be on the familiar 1,2,3D curve prevalent in current visual media imaging techniques. If we have to ‘visualise’ this data-‐set it may not be unlike the vector field (required to produce VS images) illustrated in the lower left portion of Fig 10. •
Fig 24. Left to right: a) Optical space, b) Radial depth-‐map, c) Radial vector formation (representative of tecturn vision?), d) Radial disorder
This sort of structure would then account for the retinotectal vision identified by Sherman 1977(?) and possibly associated with the well known accounts of so called blind sight. The spatially configured field potential would then be aligned with the ecology of audio spatial field in the lower layers of the SC. The resulting neurologically generated structure would then be capable of supporting other facets of multi-‐sense integration29 in the form of a perceptual structure coming together within thalamic related structures. The thalamus has been linked (Engle 1999) to consciousness via synchronized sweeping gamma waves.30
Within such a ‘neurally generated field’, eye movements could be coordinated efficiently providing vital spatial data for the predetermining of the physical shape of the lens
29 Multisense integration in the SC has been well documented and shown to be lead by visual
input and subject to environmental influences. MA Merideth & BE Stein 1986; DK Sarko & D Ghose 1212; L Yu, BA Rowlans & BE Stein 2010; MT Wallace et all 2004; A King 2008; JC Alvarado et al 2009; J Xu et al 2012. 30 Andreas K. Engel et al. in the journal Consciousness and Cognition (1999). This is disputed by some and the debate is ongoing.
24
© copyright ERA 2014
required prior-‐to saccadic eye movements. The process of fixation would then involve a combination of drawing focus (optical lens) through cortically controlled functions and its alignment with the attractor function determining the setting out of the neurally generated spatial field. This implies independent systems of awareness (conscious and subconscious).31 While each form of awareness would support an independent form of attention, it would be the sub-‐conscious neural field structure that would form the base spatially integrated contextualized holistic world view within which local but detailed formation could take shape. Certain aspects of attention operational within the field structure would be in temporal advance those forming within macular vision. This phenomenon of certain aspects of decision-‐making taking place prior to conscious awareness of incident is well documented leading to debates on ‘free will’. Awareness would involve considerations made ‘in’ time and ‘over’ time.
As indicated above, the development of the notional neural field associated with an implicit and holistic form of awareness is likely to be associated with the early evolution of the brain and eye and therefor controlled by older brain structures. In addition, the degree to which the ‘perceiver’ controls these generative processes through active participation in the setting out of the data-‐set dependent on ‘perceiver’s intent’ in the environment suggests that in early development the eye’s receptors and retinal structures and circuitry will be dependent on instruction from the brain. Indeed, even before birth and the onset of sensory experience, neural activity plays an important role in shaping the vertebrate nervous system, (R. Wong et al 1998). The observed retinal wave phenomenon through early development is likely to be associated with this process of neural calibration.32 There is evidence for both an instructive role assisting with retinal development (the refinement of transient retinotopic maps and their elimination) driven by the brain and delivering the basic neural wiring and also for a permissive function extending beyond the set-‐up stage. The ‘collection’ of the proposed phase related data-‐potential would require neural ‘calibration’33 instructed from within (genetically sequenced) as well as from a supply function as part of the generation of vision from environmental light input. The potential for retinal waves to help define ganglion cell function and arrangement together with their respective dendritic layering in the inner plexiform layer has been examined with positive outcomes by Rachel Wong and her lab, 1990’s -‐2005. It would seem logical that retinal waves were pre-‐calibrating dendritic growth prior to the onset of visually driven input in order to prepare circuitry to receive an incoming phase orientated data potential from the light array. This initial circuitry would be extensively horizontal in nature involving gap junctions between cells and cell types across the spherical retina and the resulting circuitry would lead to the SC and well as to the LGN.34 In addition we should expect to see the subcortical areas affecting performance of ganglion cells and even further forwards within the retinal processing mechanisms to amacrine and horizontal cells. Following the set-‐up process spontaneously driven retinal waves would be expected to fall away allowing the pre-‐ wired circuitry to respond to the environmental light array. Research in this area is
31 We broadly associate these ‘takes’ on reality with left and right hemisphere realities – I.
McGilchrist 2010. 32 The creation of VS media involves processes that imperceptibly modulate data-‐sets over time. (including binocular stereo views). The processes are very similar in appearance to retinal waves. 33 The calibration of multisensory integration processes has been observed to develop from stimuli even in anesthetized cats. J. Xu et al (2012) 34 There is some ongoing debate with respect to retinal waves afferent signaling to lamina-‐ specific projections in the tectum. (Wong et al 1998)
25
© copyright ERA 2014
ongoing but restricted by the difficulties of experimental set-‐up.
Application implications: The implications with respect to there being two distinct and independent data potentials relating to the physical world encoded within the light array are significant. VS would need to be investigated as the setting out point for a new form of illusionary space based on perceptual structure in the same way picture space was paired with central perspective and optical projection. VS software applications would generate the stimuli that would pair with the developing theory of visual perception driven by an ‘experiential ontology’ determining aspects of our perceptual structure. The implicit implication is that VS would urgently need to replace picture space as over exposure to our current virtual environments would actively encourage the development of atypical perceptual structures. Over exposure to non-‐perceptually structured virtual environments (optically structured) especially at key phases such as early childhood and in old age could be detrimental to health.35 There are potential but as yet unresolved links of such visual deprivation to ASD related conditions.36 As vision is essentially a biological process, under certain conditions we would consider that over exposure to deficient stimuli could lead to irreversible neural redundancy. Large-‐scale neural redundancy within the visual system could well trigger an unbalance in our biological control mechanisms and be a factor in other mid-‐brain related conditions such as Alzheimer’s.37
Fig 25. Current instrumentation, painting by the author
35 Hyperactivity and deficits in problem Solving Following Superior Colliculus Lesions in the Rat. Smith C. & Weldon D. Physiology and Behavior 1976 Vol. 16 381-385. Multiple Sensitive Periods in Human Visual Development: Evidence from Visually Deprived Children Terri L. Lewis 2005 Wiley InterScience Periodicals. Sparing of sensitivity to biological motion but not of global motion after early visual deprivation Bat-‐Sheva Hadad et al, 2012, Developmental Science 15:4 pp 474-‐481. 36 Does Television Cause Autism. Waldman M. et al, 2006, Working paper 12632 National Bureau of Economoc Research. 37 Synaptic plasticity defect following visual deprivation in Alzheimer disease model transgenic mice. Christopher M. William et al J Neurosci . 2012 June 6; 32(23): 8004–8011. doi:10.1523/JNEUROSCI.5369-11.2012.
26
© copyright ERA 2014
Ultimately VS systems should allow us to determine what’s actually involved in us being objective. It holds out the potential to be a meaningful tool in processes capable of breaking into perceptual structure, the nature of consciousness and the development of advanced perceptual technologies. We would need to consider ‘the meaning’ or ‘relevance of’ all photographic media and data collection devises standing in for experiential encounter at all scales, including those of astronomical telescopes and particle accelerators. If VS did indeed constitute a new form of illusionary space based on perceptual structure as opposed to optical projection then we would need to consider it as representing something of a renaissance moment with all that would entail.
Fig 26. Portishead beach by the author
John Jupe: Visual Artist and researcher 2013: © copyright ERA 2013 Contact:
[email protected]
27
© copyright ERA 2014
Bibliography – in addition to footnotes
H.B. Barlow (1977) Retinal and central factors in human vision limited by noise. From:Photo reception in vertebrates D.A.Baylor, T.D.Lanb, K.-‐W Yau 1979. Responses to retinal rods to single photons, J.Phyaiol 288, pp. 613-‐634 D.A.Baylor, G. Matthews, K.-‐W Yau 1980. Two components of electrical dark noise in toad retinal rod outer segments J.Phyaiol 309, pp. 591-‐ 621 591 Kristian Donner 1991, Noise and the absolute thresholds of cone and rod vision, Vision Res. Vol 32. No 5, pp. 853-‐866 M. Tegmark (1999) The Importance of Quantum Decoherence in Brain Processes D. Copenhagen*, K. Donner and T. Reutert (1987), Ganglion cell performance at absolute threshold in toad retina: effects of dark events in rods T. Reuter, K. Donner, D. Copenhagen (1986), Does the Random Distribution of Descrete Photoreceptor Events Limit the Sensitivity of the Retina Receptive Field J. Demb, L. Haarsma, M. Freed, P Sterling (1999), Functional Circuitry of the retinal Ganglion Cells Non-‐linear Receptive Field Web vision-‐ The Organization of the Retina and Visual System, http://webvision.med.utah.edu
28