Are Attention and Consciousness the Same?

Not really, though what consciousness is in physical terms remains obscure.

A little like fusion power, the quest for a physical explanation of consciousness has been frustratingly unrewarding. The definition of consciousness is fraught to start with, and since it is by all reasonable hypotheses a chemical process well-hidden in the murky, messy, and mysterious processes of the brain, it is also maddeningly fraught in every technical sense. A couple of recent papers provide some views of just how far away the prospect of a solution is, based on analyses of the visual system, one in humans, the other in monkeys.

Vision provides both the most vivid form of consciousness, and a particularly well-analyzed system of neural processing, from retinal input through lower level computation at the back of the brain and onwards through two visual "streams" of processing to conscious perception (the ventral stream in the inferior temporal lobe) and action-oriented processing (in the posterior parietal lobe). It is at the top of this hierarchy that things get a bit vague. Consciousness has not yet been isolated, and how it could be remains unclear. Is attention the same as consciousness, or different? How can related activities like unconscious high-level vision processing, conscious reporting, pressing buttons, etc. be separated from pure consciousness? They all happen in the brain, after all. Or do those activities compose consciousness?

A few landmarks in the streams of visual processing.  V1 is the first level of visual processing, after pre-processing by the retina and lateral geniculate nucleus. Processing then divides into the two streams ending up in the inferotemporal lobe, where consciousness and memory seem to be fed, while the dorsal stream to the inferior parietal lobule and nearby areas feed action guidance in the vicinity of the motor cortex. 

In the first paper, the authors jammed a matrix of electrodes into the brains of macaques, near the "face cells" of the inferotemporal cortex of the ventral stream. The macaques were presented with a classic binocular rivalry test, with a face shown to one eye, and something else shown to the other eye. Nothing was changed on the screen, nor the head orientation of the macaque, but their conscious perception alternated (as would ours) between one image and the other. It is thought to be a clever way to isolate perceptual distinctions from lower level visual processing, which stay largely constant- each eye processes each scene fully, before higher levels make the choice of which one to focus on consciously. (But see here). It has been thought that by the time processing reaches the very high level of the face cells, they only activate when a face is being consciously perceived. But that was not the case here. The authors find that these cells, when tested more densely than has been possible before, show activity corresponding to both images. The face could be read using one filter on these neurons, but a large fraction (1/4 to 1/3) could be read by another filter to represent the non-face image. So by this work, this level of visual processing in the inferotemporal cortex is biased by conscious perception to concentrate on the conscious image, but that is not exclusive- the cells are not entirely representative of consciousness. This suggests that whatever consciousness is takes place somewhere else, or at a selective ensemble level of particular oscillations or other spike coding schemes.

"We trained a linear decoder to distinguish between trial types (A,B) and (A,C). Remarkably, the decoding accuracy for distinguishing the two trial types was 74%. For comparison, the decoding accuracy for distinguishing (A, B) versus (A, C) from the same cell population was 88%. Thus, while the conscious percept can be decoded better than the suppressed stimulus, face cells do encode significant information about the latter. ... This finding challenges the widely-held notion that in IT cortex almost all neurons respond only to the consciously perceived stimulus."

 

The second paper used EEG on human subjects to test their visual and perceptual response to disappearing images and filled-in zones. We have areas in our visual field where we are physically blind, (the fovea), and where higher levels of the visual system "fill in" parts of the visual scene to make our conscious perception seem smooth and continuous. The experimenters came up with a forbiddingly complex visual presentation system of calibrated dots and high-frequency snow whose purpose was to oppose visual attention against conscious perception. When attention is directed to the blind spot, that is precisely when the absence of an image there becomes apparent. This allowed the experimenters to ask whether the typical neural signatures of high-level visual processing (the steady-state visually evoked potential, or SSVEP) reflect conscious perception, as believed, or attention or other phenomena. They presented and removed image features all over the scene, including blind spot areas. What they found was that the EEG signal of SSVEP was heightened as attention was directed to the invisible areas, exactly the opposite of what they hypothesized if the signal was tied to actual visual conscious perception. This suggested that this particular signal is not a neural correlate of consciousness, but one of attention and perhaps surprise / contrast instead.

So where are the elusive neural correlates of consciousness? Papers like these refine what and where it might not be. It seems increasingly unlikely that "where" is the right question to ask. Consciousness is graded, episodic, extinguishable in sleep, heightened and lowered by various experiences and drugs. So it seems more like a dynamic but persistent pattern of activity than a locus, let alone an homunculus. And what exactly that activity is.. a Nobel prize surely awaits someone on that quest.

• Unions are not a good thing ... sometimes.
• Just another debt con.
• Incompetent hacks and bullies. An administration ends in character.
• Covid and the superspreader event.
• Outgoing Secretary of State is also a deluded and pathetic loser.
• But others are getting on board.
• Bill Mitchell on social capital, third-way-ism, "empowerment", dogs, bones, etc.
• Chart of the week: just how divided can we be?