I want to thank Alfredo (and Nathan Munn) for co-editing this blog, and creating a Forum. I will link to the Forum as soon as possible.
Dr. Pereira makes the case for astrocytes in the brain as substrates of consciousness, and specifically for a “neuronal global workspace”. (See the Glossary on Scientific Consciousness Terminology, also soon to be posted here.) Astrocytes are very interesting little critters. Until recently it was thought that astrocytes (literally, ‘star-shaped’ cells, but often taken to be synonymous with neuroglia) were merely neuronal support cells. They are that, but they also conduct information processing, as it now appears. However, I do not believe that the evidence currently supports a DISTINCTIVE role for astrocytes in supporting conscious (as opposed to unconscious or less conscious) brain events.
Here’s why.
1. Astrocytes are too widely distributed. Conscious contents (at least in humans) are primarily a function of cortex - obviously in interaction with subcortical structures. We know that because (a) specific cortical lesions lead to specific deficits in conscious, but not necessarily unconscious, cognitions. In contrast, (b) subcortical lesions, like bilateral cerebellar lesions, may not affect conscious contents or even the conscious state at all; © stimulation of cortex, but not subcortex, leads to specific conscious events in waking subjects (classically in Wilder Penfield’s neurosurgical patients); (d) loss of the conscious STATE involves, not a drop in neuronal firing, but a drop in fast and long-distance interactions among neurons and neuronal populations. There’s a lot more evidence for DISTINCTIVE conscious, but not unconscious brain events, but those points are basic.
2. Note that any adequate hypothesis about the brain basis of consciousness must account for the DIFFERENTIAL properties of conscious, but not unconscious, brain events.
3. Astrocytes do not differentially account for conscious, but not unconscious brain events.
Astrocytes are widely distributed in the brain, including the cerebellum. Yet a bilateral loss of the cerebellum does not lead to a loss of conscious contents or state. I believe (but would like a reference) that there are no brain regions that lack astrocytes - they are really basic.
(Astrocytes are the reason for the famous “white matter” of the brain, the huge body of white tissue that can be seen with the naked eye when a mammalian brain is dissected. The white coloring is due to high lipid content of astrocytes, which wrap around all the nerve tracts, both supporting and protecting neuronal signaling. Because neuronal cell bodies in the surface layers of cortex are not surrounded by astrocytes, they appear greyish to the naked eye. There is not question that astrocytes are vital. The only question is whether they distinctively support conscious brain events.)
4. Dr. Pereira points out that astrocytes are especially plentiful in Layer I of the cortex, which is often called a dendritic “feltwork,” so that it is the dendrites (input branches of neurons) that make up Layer I, along with glial cells. That is fascinating, of course, but it does not prove the claim that astrocytes are differentially associated with conscious experiences. Layer I contains numerous dendro-dendritic connections, so much so that it is believed that Layer I may itself be a major information processing highway. (Normally we would look to axonal-somatic synapses, as in the major white matter pathways.) But there is no evidence that I know of that Layer I is DIFFERENTIALLY associated with conscious, rather than unconscious, processes.
5. To test these hypotheses properly one would have to study knockout mice, for example, if one could vary the number of astrocytes without impairing vital functions.
6. Assessing conscious experiences in animals, using closely analogous measures.
In humans we assess conscious experiences behaviorally by means of accurate, voluntary report, the evidentiary basis for all of psychophysics for the last two centuries. That is an excellent method, and in animals the analogous method is “match to sample”, essentially a way of asking other animals whether some stimulus S1 matches some previous or simultaneous stimulus S2. Using operant conditioning one can obtain voluntary reports that are methodologically equivalent to human voluntary reports. Animal experiments are therefore directly relevant to the question of human conscious experience.
6. We start from humans.
Arguably, humans are our “criterial species” for studying consciousness, just as pigeons might be the criterion species for studying bird flight. Once we understand such basic biological phenomena in one species we can carefully look for homologies in others. That line of standard biological reasoning strongly suggests that mammals (characterized by rich, thalamocortical systems in the core brain) are conscious. Comparative biologists are exploring the possibility of wider homologies, among birds, reptiles, and recently, even cephalopods.
However, using the most conservative criterion of direct anatomical and physiological homologies, there is no current reason to doubt that non-human mammals are at least perceptually conscious.
7. Consciousness appears to be widely conserved in the animal kingdom.
Dr. Pereira seems to suggest that humans are conscious, but that perhaps other mammals may not be. That is always possible, of course, because we only have “atheoretical evidence” at this point in the scientific study of consciousness. That is, we do not yet have a shared theoretical basis for understanding consciousness (though we have some well-developed hypotheses). Until chemistry provided an understanding of O2/CO2 exchange via alveoli in the lungs, there was no theoretical basis for understanding that lungs (and gills) were vital for animal metabolism. However, Aristotle was still able to make an atheoretical case for lungs and breathing as highly conserved animal features, with a lot of empirical evidence.
8. Theoretically-based and atheoretical arguments by homology.
Dr. Pereira’s hypothesis is therefore not disprovable on theoretical grounds. But the available evidence, as far as I can tell, does not support the hypothesis.
Coming Soon: Current candidates for a brain basis for conscious experience.
I will write an entry soon on what appear to be some plausible brain substrates for conscious experiences. Again, astrocytes are surely involved in some way - they are just not DIFFERENTIALLY involved in conscious, compared to non-conscious brain events.
BJB