Global workspace theory (GWT) is the proposition that consciousness is composed of contents broadcast throughout the brain. Various specialty processes compete for the limited capacity of the broadcasting mechanisms, to have their content broadcast to the all the other specialty processes.
Global neuronal workspace (GNW) is a variant of that theory, popularly promoted by Stanislas Dehaene, which I’ve covered before. GNW is more specific than generic GWT on the physical mechanisms involved. It relies on empirical work done over the years demonstrating that conscious reportability involves wide scale activation of the cortex.
One of the observed stages is a massive surge about 300 milliseconds after a stimulus, called the P3b wave. Previous work seemed to establish that the P3b wave is a neural correlate of consciousness. Dehaene theorized that it represents the stage where one of the signals achieves a threshold and wins domination, with all the other signals being inhibited. Indeed, the distinguishing mark of the P3b is that it is massively negative in amplitude, indicating that most of it comes from inhibitory action.
The P3b has been replicated extensively and been seen as a pretty established phenomenon associated with attention and consciousness. But this is science, and any result is always provisional. Michael Cohen and colleagues have put out a preprint of a study that may demonstrate that the P3b wave is not associated with conscious perception, but with post perceptual processing.
The study tests the perception of subjects, showing various images while measuring their brain waves via EEG. Using a no-report protocol, in half of the tests, the subjects were asked to report on whether they saw something, but in the other half they were not asked to report. Crucially, the P3b wave only manifested in the reported cases, never in the non-report ones, even when the non-report image were exactly the same as the ones that did generate affirmative reports.
To control for the possibility that the subjects weren’t actually conscious of the image in the non-report cases, the subjects were given a memory test after a batch of non-report events, checking to see what they remember perceiving. Their memories of the perception correlated with the results in the report versions.
So, the P3b wave, a major piller of GNW, may be knocked down. The study authors are careful to make clear that this does not invalidate GWT or other cognitive theories of consciousness. They didn’t test for all the other ways the information may have propagated throughout the cortex. Strictly speaking, it doesn’t even invalidate GNW itself, but it does seem to knock out a major piece of evidence for it.
However, this is a more interesting discussion if we ask, what would it mean if all cortical communication beyond the sensory regions were ruled out, that the ability to acquire a memory of a sight only required the local sensory cortices? It might seem like a validation of views like Victor Lamme’s local recurrent processing theory, which holds that local processing in the sensory cortices is sufficient for conscious perception.
But would it be? Dehaene, when discussing his theory, is clear that it’s a theory of conscious access. For him, something isn’t conscious until it becomes accessible by the rest of the brain. Content in sensory cortices may form, but it isn’t conscious until it’s accessible. Dehaene refers to this content as preconscious. It isn’t yet conscious, but it has the potential to become so.
In that view, the content of what the subjects perceived in the non-report tests may have been preconscious, unless and until their memories were probed, at which point it became conscious.
This may be another case where the concept of consciousness is causing people to argue about nothing. If we describe the situation without reference to it, the facts seem clear.
Sensory representations form in the local sensory cortex. A temporary memory of that representation may persist in that region, so if probed soon enough afterward, a report about the representation can be extracted from it. But until there is a need for a report or other usage, it is not available to the rest of the system, and none of the activity, including the P3b, normally associated with that kind of access is evident.
This reminds me of Daniel Dennett’s multiple drafts theory (MDT) of consciousness. MDT is a variant of GWT, but minus the idea that there is any one event where content becomes conscious. It’s only when the system is probed in certain ways that one of the streams, one of the drafts, become selected, generally one of the ones that has managed to leave its effects throughout the brain, that has achieved “fame in the brain.”
In other words, Dennett denies that there is any one finish line where content that was previously unconscious becomes conscious. In his view, the search for that line is meaningless. In that sense, the P3b wave may be a measure of availability, but calling it a measure of consciousness is probably not accurate. And it’s not accurate to say that the Lamme’s local recurrent processing is conscious, although it’s also not accurate to relegate it completely to the unconscious. What we can say is that it’s at a particular point in the stream where it may become relevant for behavior, including report.
Maybe this view is too ecumenical and I’m papering over important differences. But it seems like giving up the idea of one finish line for consciousness turns a lot of theories that look incompatible into models of different aspects of the same overall system.
None of this is to say that GWT or any of its variants might not be invalidated at some point. These are scientific theories and are always subject to falsification on new data. But if or when that happens, we should be clear about exactly what is being invalidated.
Unless of course I’m missing something?
55 thoughts on “Is there a conscious perception finish line?”
I’m mostly fine with “there s no one finish line,” with one caveat. For moral purposes – a big drivier of discussions of “consciousness” – the finish line is, does this create an affective response? For a signal that only spreads around the local sensory cortex, for example, is that region sending signals to the prefrontal cortex (or other decision control regions) that effectively mean “I don’t like this, please make it stop”? Then it’s conscious enough to matter.
LikeLiked by 2 people
I agree. But according to GWT, a signal can get around in small peripheral connections that never trigger a broadcast. That might lead to physiological changes, changes that eventually cause the anterior cingulate and/or prefrontal cortex to push an affect into the workspace. It’d be a situation where we might suddenly feel anxious without being sure why.
So we might end up having an affect experience without the visual experience that caused it. Although if the visual imagery is still in the early sensory region, or the memory of it is still there, then it could likely make it into the workspace at that point. But if it’s been masked away, we might be looking at something unrelated and wondering what is going on. (Which, of course, happens.)
Paul, I think I kinda sorta maybe agree mostly. I think there may not be “one” finish line. I think there may be several. However, my best guess as to the situation you describe would look more like this: signals get activated in sensory (or other ?!) cortex regions, but do not get enough activation to get to the global workspace. Nevertheless, they send signals to sub-cortical regions (esp. amygdala, maybe midbrain/brainstem?) which induce hormonal responses, so increase (or decrease?) in heart rate, blood pressure, etc. Interoception of these signals may again lead to cortical activations which don’t get to the workspace but but do activate a mechanism somewhere in the cortex that recognizes those interoceptive signals as “bad”. This last activation may get into the workspace, which makes it globally available, potentially leading to “I’ve got a bad feeling about this.”
So the above describes 3 finish lines, unless you require affect, which would make it 2, or unless you require reportability, which would make it 1.
I think once you decide what counts, there is a finish line.
LikeLiked by 1 person
That’s a good way of putting it. In that sense, local recurrent theory is using the formation of a memory, even if transitory, as its finish line. While GWT is using global availability, including to reporting systems, as its.
Although technically, they’re both using report as the finish line. It’s just that some experiments find a way to buffer the information and extend the duration between stimuli and report.
The question is what counts for the system itself? For stimuli that represent opportunities or threats, I would think as wide availability as possible.
LikeLiked by 1 person
I think what you choose as the “system itself” is somewhat arbitrary. It’s just that some choices are more useful than others.
But when you talk about “wide availability”, I think you are choosing to look at the usefulness of the system and possibly missing the essence of what makes something an experience. [I brought this up to someone recently, but I forget where, I think Twitter] The size of the potential audience shouldn’t matter. You can talk about the audience of the global workspace, but what if you pare each audience member one at a time? What happens when you get down to exactly one? Has anything essential changed? What’s left?
I don’t think, biologically, the system is arbitrary. It’s whatever is needed to plan the movement of the animal.
What’s changed if we pare the audience down to one? Well, remember that the content for the workspace comes from the audience. So with one member, you have no workspace, just a specialized neural network, probably similar in scope to the ANNs in the device you’re using right now, albeit far more sophisticated.
Likewise if you only have one sensory network and one movement one, you have a pretty simple stimulus response situation, with again, no real workspace.
So, how many member processes are required? I think a better question is, which ones are required? It seems pretty hard to have what we think of as conscious experience without image producing processes. As well as affect producing ones. Which in turn implies enough sophistication for basic cause and effect prediction. (Otherwise the affect isn’t affecting anything.) Which also implies a rudimentary episodic memory system.
This gets us at some basic version of primary consciousness. But still missing a lot from the human level variety. Whether it amounts to consciousness is a matter of definition.
Does any of that get at your question?
“Well, remember that the content for the workspace comes from the audience.”
I’m not sure that GWT says that, and it would be easy to imagine that you can get content into the workspace from something that doesn’t respond to it, but we can adjust the thought experiment. What if there is only one audience member, besides the representing member, that does something functional with the representation? Is that enough? And if not, why? Does two possible responders make it a “workspace”?
On further thought I think I’m just getting back to an old idea, namely, that I want to identify the molecule (psychule) associated with wetness, whereas you mostly want to explain wetness, which you can’t do with just one molecule.
GWT does have most of the processes able to contribute to the audience. But it is possible to have audience members that never contribute, or contributers that never receive. But one way flow is generally not how the thalamo-cortical system does things.
It does seem like you’re looking for some line between an unconscious and a conscious system. I don’t think you’ll find it. If your computer screen is dark except for one pixel, do you have a picture? Not really. If we then add one pixel at a time, when do we get the picture? How many pixels are necessary? The exact same question applies on the distinction between a light source detector that a primitive worm might have, and an actual image capturing eye. How many photoreceptors are necessary for us to cross from light detection with directionality, to actual imagery?
Like I said above, I think it’s not the number of processes, but which ones you have. Which leads me back to that (admittedly oversimplified) hierarchy:
1. Reflexes and fixed action patterns
2. Sensory images
3. Affect driven action selection
4. Imaginative deliberation
Each of these require additional processes, audience members and contributors, to the workspace. I can’t see that there’s a fact of the matter on when consciousness begins. I personally incline toward 3, but if someone insists on one of the others, it’s not an argument we can settle with empirical data.
LikeLiked by 1 person
My take-away is that yet another specific idea about how the brain works turns out to probably not be the case, and the mystery continues.
My other take-away is that it re-emphasizes the need for a definition regarding the topic of study. Or a set of definitions. I quite agree with the “bundle” view of consciousness. The cargo carrying analogy is apt; here’s another: different kinds of engines, big, small, simple, complex. They all output revolutions of some speed and torque, but in myriad ways.
Seems like we can define cargo-carrying and engines okay. Maybe at the next Con-Con heads can get together and come up with something. 🙂
LikeLiked by 1 person
I wouldn’t hold my breath. I’m sure there will be new definitions presented at the next big conference. Probably dozens. For some, people will nod and say, “Interesting, yes, very interesting, but here’s how that supports my definition.”
It’s like I said before, all consciousness science is the science of correlating brain details with report. (Even no-report or behavioral protocols are ultimately dependent on previous reports.)
LikeLiked by 1 person
I was looking at a review of research on gamma waves (nobody is sure yet whether there is anything of significance regarding NCC to them ) and the author made the following comment:
“I think there’s an unwritten rule that whenever someone mentions consciousness, they’ll be made to look foolish by a subsequent paper.”
LikeLiked by 1 person
Gamma waves to me represent fast processing, so I would think they are a correlate.
I don’t think things are quite that bleak. Not quite.
The interesting thing to me here is that evidence came in that the P3b waves detected seem associated with conscious processing rather than conscious input, and GNT looses evidence that it once thought it had. Personally I just play along with this “global” business given it’s popularity, not because I have any reason to believe that phenomenal experience is created by means of the widespread brain transmission of information. But finding brain wave evidence associated with conscious processing? To me that seems pretty significant. Whatever is happening in my brain for me to figure out what to do, I consider this just as “conscious” as input information to this processor. It’s simply a different element of the game.
Does GWT theorize that every random sensation that I have, like a faint butt itch, needs to go “global” for me to have it? To me that would seem costly. But I can instead understand how thinking about what to do about a given situation might warrant a good deal of brain function, and perhaps top to bottom.
My own theory suggests that GWTers should find it best to say that their theory concerns phenomenal experience first, with informational dynamics (or “access”) brought in through that created phenomenal entity. I suspect this will be a simple transition.
Next I suspect they’ll find that it’s not global information itself that creates phenomena, but rather brain waves associated with that global information coming together. Why? Because in a causal world all physics based output should require associated mechanisms, whether EM fields or something else.
And in the end I fear that a random butt itch requires less than global broadcasting for it to occur, and so such evidence should retire the first letter of their title. But I can play along for now.
“Does GWT theorize that every random sensation that I have, like a faint butt itch, needs to go “global” for me to have it? To me that would seem costly.”
Not for an itch you don’t notice, or one you absent mindedly scratch at, but for you to become conscious of it, it does need to be global, at least for whatever period you’re thinking bout it, which might be very brief. The contents of the workspace can switch very fast.
“My own theory suggests that GWTers should find it best to say that their theory concerns phenomenal experience first, with informational dynamics (or “access”) brought in through that created phenomenal entity. ”
Most GWT proponents focus on access consciousness, regarding phenomenal consciousness as something that will just come along for the ride. Carruthers is an outlier in that regard, explicitly making the case that phenomenal content is nonceptual (irreducible) sensory or affect content that is globally broadcast. In any case, in GWT, the actual experience involves the receiving processes reacting to the content.
“Next I suspect they’ll find that it’s not global information itself that creates phenomena, but rather brain waves associated with that global information coming together. ”
Arguably, the propagation of the global information is the brain waves. (Well, in the specific case of the P3b, it’s the inhibition of the competing signals, for attention if not for conscious content itself.) In other words, they’re really the same thing at different levels of organization. And it’s all mechanism.
LikeLiked by 2 people
If an itch causes an itchiness input motivation for the thought processor, though this conscious processor disregards that itch given that it’s scratched automatically, I guess you’re saying that it doesn’t go global. Or perhaps it’s ignored given various stronger inputs, with the same result. And indeed, apparently the P3b evidence as processing fits in here pretty well. So maybe GWT could concern the processing element of consciousness rather than the input? Or at least for the human?
Surely you don’t mean to imply that most proponents of GWT, consider affect epiphenomenal? I realize that you don’t, but that’s not what most believe is it?
So he’s at least closer to me regarding the importance of phenomena? That’s good. Anyway I wonder if there’s any good reason in the GWT framework for access to be considered fundamental while phenomena remains secondary? I can imagine this being switched around without GWT losing a step, which I think would help the theory as evidence comes in.
As I see it the access side is essentially just information, though information which exists by means of the phenomena medium. For example the affect of toe pain is required for there to be access information about where this problem happens to be. It could be that even the colors that we see provide affect based phenomena somewhat, though information is provided through this medium as well. It’s somewhat like the way paper and pencil can be a medium for something drawn. So phenomena would exist as such a medium, and access would come along for the ride through that medium.
Happy to hear that Mike! Of course we haven’t pinned down the physics of affect yet, but let’s presume that there are material based dynamics here. So in order to potentially build something which has phenomenal experience, we’ll also need to add those physics based mechanisms.
LikeLiked by 1 person
“Or perhaps it’s ignored given various stronger inputs, with the same result. ”
That’s the thing. Achieving broadcast dominance is competitive. If you’re just sitting around, an itch might well win the competition. But if you’re engaged in something that requires a lot of mental effort, the itch might not break through unless it became particularly intense.
“So maybe GWT could concern the processing element of consciousness rather than the input?”
GWT is all about the processing. Of course, one process’ output becomes another’s input. And an output that is broadcast becomes the input of processes throughout the brain, which is what GWT is all about.
“Surely you don’t mean to imply that most proponents of GWT, consider affect epiphenomenal?”
No. My “carried along” phrase was was epistemic, not ontological. What I mean to imply is that phenomenal experience and access are the same thing, just seen from different perspectives: the subjective vs the objective.
“That’s good. Anyway I wonder if there’s any good reason in the GWT framework for access to be considered fundamental while phenomena remains secondary?”
It’s not secondary. For phenomenality to be secondary would imply that it’s something separate and apart from access. Even in Carruthers’ case, it’s a subset of access, not something separate.
“Of course we haven’t pinned down the physics of affect yet,”
I think it’s been pretty definitely pinned down to neural processes. Maybe we can argue whether those neural processes amount to computation (although I haven’t heard a concrete alternative yet), and a lot of things feed into those processes (glia, hormones, interoceptive signalling of body states, etc), but they’re at the core. We can always say, “maybe there’s something more”, but without evidence, or at least logical necessity, it’s just wishful speculation.
LikeLiked by 1 person
“the propagation of the global information is the brain waves”
Are you agreeing with EM field theory? This sounds like something I might have written.
No. Neural oscillations only require neural (wired) spikes, although perhaps they get some perturbation from local field potentials.
I’m not following the distinction you are making. Neural oscillations are brain
I’m not following the distinction you are making. Neural oscillations are brain waves measured by EEG. Are you saying there are electrical currents with no magnetic component? Of course, MEG can measure the magnetic element by itself but these originate from the same processes.
I didn’t mean to imply any distinction. I was using “neural osccillation” as a synonym for “brain wave”. I used it because it more clearly indicates what the phenomenon is.
Um, sorry, that comment came off with too much certitude. I haven’t read anything indicating that more than neural spikes are required. It’s always possible that data might come in tomorrow.
Ok. As EM field proponents point out spikes result in increased EM fields. Actually it is a key part of the argument.
“Synchronous firing of distributed neurones phase-locks induced em field fluctuations to increase their magnitude and influence. Synchronous firing has previously been demonstrated to correlate with awareness and perception, indicating that perturbations to the brain’s em field also correlate with awareness.”
Of course, one can argue the EM field is incidental.
LikeLiked by 1 person
Ok, I haven’t read anything in mainstream neuroscience. I should have remembered that somewhere there is a paper for every assertion. 🙂
If I may make an analogy:
Consider an elaborate domino set-up which is hidden by a large cloth suspended over the top. We’re interested in what the dominoes are doing, but we can’t see them. We can, however, hear them, and we can place microphones in various places around the cloth. Suppose something knocks down a domino at one end, which starts a wave of dominos falling in some pattern. Each falling domino event creates a sound wave which can be detected by the microphones.
So we’re talking about two different kinds of wave. In theory, it’s possible that one or more dominos in the system is perched so precariously that a sound wave, or many combined sound waves, could actually influence whether it gets knocked over. But such a claim would be extraordinary, requiring extraordinary evidence.
[waiting for the paper]
It seems more reasonable when you consider there are billions of dominoes. We really don’t have reliable intuitions when it comes to “billions” of things.
(Or, for that matter, really tiny things. As article I was reading made passing reference to the discovery that fruit flies can leverage electrical fields to assist in taking off for flight. The world is a very different place when you’re that small. At the neuron scale it’s even weirder. I’ve long thought EM is likely to play some role in what’s going on in our brains.)
LikeLiked by 1 person
Read some of the ephaptic coupling links on this page.
James C, I don’t have time to read all of those articles. I read McFadden 2002, possibly too quickly, but I did not see anything that would change my position.
Does the EM activity of the brain correlate with Consciousness? Yes. The EM activity of the brain correlates with whatever the neurons are doing, because neurons (and neurones) are generating EM fields in their normal operation. The question is whether that EM activity is epiphenomenal or causal with respect to Consciousness.
Can EM fields have causal effects on neurons? Yes. The mechanism of signal propagation in the standard model of a firing neuron works by responding to the EM field, so an outside source of that EM field could trigger that response.
Therefor the EM theory of consciousness is possible. But is it the best theory out there? I don’t think so. Why? Partly because I can’t see how it could work. Given the amount of noise in the system, and the fact that neurons are grown and not placed precisely, I don’t see how the EM fields can contain information that can be processed into precise functional responses from neurons. I have not seen any suggestions in the literature.
In contrast, we have proof of concept neural systems that perform all of the information processing functions attempted so far in the complete absence of EM effects. I refer to computer simulated neural nets.
So I await the explanation, or computer simulation, of how the total brain EM field can cause the specific report of a type like “I see a cat” distinct from a report like “I see a dog”.
You said you were waiting on the paper and I point you to a link to five papers.
The brain’s electromagnetic field is naturally suited for binding
because its information is pooled as a continuous, global whole — in
contrast to the discrete information in neurons (McFadden, 2002a, p.
43; 2002b, p. 25). While information in neurons is digital, discrete,
and slowly processed, information in fields is analogue, continuous,
and processed at light speed. The latter are features of minds that traditional, neuron-based theories have trouble explaining (McFadden,
2002b, p. 31, 2006, §12.5; cf. John, 2001, p. 208).
‘Light speed’ above doesn’t refer to the speed of thought, but to the
speed that synchronized activity propagates across brains. John’s
experiments (2001, pp. 185ff.) show that different brain areas synchronize their firing with zero time lag. So synchrony is mediated by
the brain’s global field which propagates at light speed, not by slowly
propagating neuronal firing patterns.
LikeLiked by 1 person
That might be like asking how a car’s electrical system helps it go forward compared to backwards. EM is, for obvious reasons, not likely to have a primary effect, but it may have a contributing or participating effect to its operation.
With so much of the signaling in the nervous system being analog, and much of it being right at the minimal signal / noise ratio, it does have “design” principle that compensate for environmental noise (redundancy, repetition, etc). But those mechanisms aren’t perfect (perfection costs too much metabolically), and I don’t doubt the EM fields are part of that environmental perturbation. So I think you’re right that it’s unlikely to be primary.
Yeah, if it was primary, we would have noticed by now.
It occurs to me the brain evolved in its own EM environment, which implies it evolved to either ignore or incorporate it. The former seems wasteful, and nature seems more generally inclined to leverage what’s available than ignore it.
LikeLiked by 1 person
James C, I very much appreciate the papers you linked. I read one and looked at the intro to another. But I’m looking for one model that, even hypothetical, could explain how it could possibly work. Synchronization of the neurons in the brain is a likely part of the human mechanism of consciousness, but 1., nothing in the papers suggested how you could get the EM field to reliably influence exactly the right set of neurons while not affecting nearby neurons, and 2., it’s extremely possible that simple neural mechanisms are responsible for the synchronization. Personally I suspect the claustrum is involved in that, being approximately equidistant from all the neurons needing synchronisation. That’s just a guess, but it seems to me to be designable, whereas with an EM field I wouldn’t know where to start. Also, if the neurons could interact via EM, what are all those dendrites and axons for? Also, it’s certain that neurons did not evolve to interact via EM. It’s possible the EM functionality “emerged”, but it would have to be without any basic change in the original functionality, since there is no significant change in the neurons themselves.
So again, if one of those papers describes a mechanism by which this could work, let me know.
LikeLiked by 1 person
On assessments of EM field theories of consciousness, it should help to grasp what is and isn’t being proposed here. Let’s try this.
We can agree that consciousness exists by means of brain function. Furthermore the brain to computer analogy seems productive to most or all of us. So when my toe gets whacked it should be productive to say that my brain processes this input, somewhat like pressing the “W” key on my keyboard may be processed by my computer. My brain seems to process input by means of neuron firing (at least), while my computer should use a microprocessor to do so. We know that my computer’s processed information effectively animates its screen by causing it to do associated things, such as display the letter. But once processed, what mechanism does whacked toe information animate so that I feel associated pain?
This is the role that EM fields are proposed to play — the consciousness mechanism. This is to say that toe pain, a visual image, feeling cold, and all phenomena, are proposed to exist by means of associated EM fields. Here there should be no consciousness without these fields, just as there should be no screen image without a screen to display them. (Note that this is an analogy which I developed many months before I’d heard of EM field consciousness theories.)
While we’ll need evidence that EM radiation does serve this roll, it seems to me that we should presume that at least some mechanism must exist as a consciousness medium. Computer processing alone should not be sufficient, since we know of no other computer output that has no associated output mechanism. It’s when we propose computer processing itself as mechanism, that we run into funky notions such as any computer which processes what a brain does, will thus produce phenomenal experience — even symbol encoded sheets of paper which are properly processed into other symbol encoded sheets of paper!
Let’s also take this speculation back to the post at hand, and so get into the tiny “second computer” which I consider to do less than 1000th of 1% as many calculations as the brain itself. Apparently some GWT advocates have been hoping that evidence would suggest that the P3b brain wave (which apparently is considered “global”), correlate with phenomenal experience. According to new evidence that Mike discussed in this post, apparently not. But it could be that once there are non-global brain waves which cause an experiencer to exist, when this brain wave based experiencer thinks about what to do (or consciously processes its experience), this demands whole brain function such as the P3b, as found. So it may turn out that GWT will end up being validated as an element of consciousness, or the processing part, though conscious input and output components might not display global neuron function.
(Here some may wonder how I can consider the teleological computer to do so much less processing, even if it turns out to require “global” brain processing. This is because what facilitates “the words in your head”, for example, might require whole brain function, even if those words don’t amount to much as processing themselves. While I’m not sure about GWT yet, I’m quite sure that whatever processing that the words in my own head may be said to do, this should be infinitesimal when compared against the amount of processing that my brain should do to create something that thinks them.)
LikeLiked by 1 person
There are variety of EM theories with various nuances and the link to the quote is a great compare and contrast of the various theories along with some historical background. The history of the idea goes back a ways and even includes Karl Popper as some one in its history.
The quote itself makes the point that electrochemical transmission methods are too slow to account for the synchronicity between different regions of the brain. They also don’t explain how different types of processing done by different regions gets merged into an integrated conscious experience.
Regarding dendrites you may have not read my first post on this topic but I linked there to a theory regarding the apical dendrites of the L5 pyramidal neurons. I would be curious to how you describe the function of the apical dendrite. The L5P neurons are the most prevalent in the cortex and are present everywhere in the cortex as well as in other brain areas. The neuron body is in layer 5 (near the bottom). The apical dendrite extends from layer 5 to the top of the cortex where they dead end in an apical tuft. Billions of these neurons with dendrites lined up in parallel from near the bottom to the top of the cortex.
LikeLiked by 1 person
Yes all those apical dendrites ending in one place seems interesting. I feel that I should qualify this by saying that I know very little about the biology of brain function however. It’s really only the big picture stuff that I feel qualified to intelligently address. I’m an “architect” rather than “engineer”. It’s my hope that you’ll find my architecture effective for your engineering purposes though. I have quite a lot to say in that regarding.
Given that psychology remains such a primitive science, how might neuroscience explore issues related to consciousness? Unfortunately by making up its own psychology. Thus the exploration of consciousness by means of neuroscience. has become quite a joke. I’d like to help fix this problem by means of psychology alone. To me this seems like the proper approach since psychology supervenes upon neuroscience.
Shoot me an email sometime. I do enjoy speaking in the public domain, though not everything between us should be there.
LikeLiked by 1 person
[James C, see my response to your latest in a new comment thread below]
Looking at the diagrams it seems like even with multiple electrodes, and certainly with the 1-dimensional signal level, there is just too much information being summarised as a crude signal to separate out the different effects of perception, action and attention. It would be interesting to see what the effect is of removing the ‘no report’ signal levels from the ‘report’ signal levels, to try to separate out what is the separate effect of reporting. Maybe high inhibition in the ‘report’ case reflects selecting a single, very focused action set, whereas the ‘no report’ case just reflect incorporation of new data into the captured information without taking any action. Also that taking action takes longer (out to 300ms+) whereas ongoing perception can cycle about every 200ms.
LikeLiked by 1 person
That would be good to see. It’s noteworthy that, even in the no-report case, activity is still fairly widespread in the posterior regions (albeit briefly), including what looks like penetration into the posterior parietal and temporal lobes. Which fits with what the participants would have needed for their reported performance on the memory test. (The authors claimed that performance required episodic memory, but I can’t see why.)
[This is a response to a recent comment by James Cross, which thread was getting too long]
James C, you said “The quote itself makes the point that electrochemical transmission methods are too slow to account for the synchronicity between different regions of the brain.”
The speed or slowness of a signal has no bearing on whether it can synchronize others. Imagine there is a series of sticks of dynamite in a line, with one hundred feet between each stick. I’m standing at the end of the line, but I want to synchronize the explosions. Here are some options:
I have a fuse running to each stick that will burn 100 feet in one second. I start the longest fuse, and every second after that I start the next-longest fuse. Assuming I’m one hundred feet from the closest stick, one second after I light the last fuse all the sticks blow up.
Let’s do that again, except, having run out of 100ft/sec fuse I’m stuck with 1ft/sec fuse, much slower. This time, after I light the longest fuse I wait one hundred seconds before I light the next-longest, and so on. One hundred seconds after I light the last one, they all go off.
Now let’s say I have no idea how far away each stick is, nor how fast the fuse to that stick runs. But I can try different timings, and I can try them over and over, and the timing between lighting the fuse for a given stick and the time that stick blows up never changes. With proper feedback and an ability to learn, I can figure out which fuse I need to light, and with what timing, such that they all go off at once.
Neurons can be synchronized by the normal (non-EM) actions of other neurons.
James Cross, I would once again like to thank you for your links. That last one provided a nice overview of L5p (which I assume stands for Layer 5 Pyrimidal) neurons.
In your most recent comment you said “I would be curious to how you describe the function of the apical dendrite.”
Dendrites are usually considered the loci of inputs to a neuron, so the apical dendrites would be one of the sources of input to the L5p neuron. Specifically, I believe those dendrites are in layer 1, and I think the inputs are coming from other cortical regions. But it may be that most/all of those layer 1 inputs are actually suppressive. Mike, do you know?
So the dendrites of the pyramidal neurons go up and down the layers, forming a kind of column. I know there appear to be what is called cortical columns, and/or mini-columns, but I don’t know how many L5p neurons there are in one column. For my own theory, I see these columns as the units of unitrackers, possibly one column per unitracker, but also possibly some few columns per unitracker. When the column[s] is[are] active, that unitracker is “on”.
So my next question is, if those apical dendrites (and all the other dendrites) are the inputs, what about the axons? Where do the outputs go? Apparently, at least some go to the thalamus. Usually a neuron is depicted as having just one axon, but I think some can branch. I’d like to know if the Lp5 axons branch. Another thing I would like to know: how many axons from one column go to the thalamus? I’m going to hypothesize exactly one, because that would be interesting, but I could see a few being the answer.]
[Warning: wild, yet I think plausible speculation follows]
So to finish the answer to your question, I see the function of the L5p neurons as coordinating the processing of the unit of consciousness, the unitracker. When the unitracker/column is activated, perhaps by being revved up to firing at 40 Hz or so, the signal from the L5p axon goes to the thalamus where it contributes to a pattern of firing in thalamic neurons. This pattern could represent the target of only this unitracker, or the pattern could essentially combine this unitracker with other unitrackers. A mechanism for exactly this combining has been demonstrated by Chris Eliasmith. This set of neurons in the thalamus would constitute the famous global workspace. These thalamus neurons can then broadcast back to the cortex.
What would the role of the apical dendrites be? They could be the site of the suppressive effects coming in from the prefrontal cortex. They could also be the site of incoming synchronization signals, although my current wild speculation would be that the synchronization signals come from the claustrum. Anyone know what layer signals from the claustrum go to?
[googled claustrum axons. Apparently they go mostly to layers 2,3, but also some to 5,6. Note, the layer 2,3 neurons have the XOR functionality]
“But it may be that most/all of those layer 1 inputs are actually suppressive. Mike, do you know?”
Can’t say I do. I can’t recall reading anything specifically characterizing Layer 1 that way, but it’s not something I’ve delved into. I do know most of the connections from the cerebrum to the midbrain are inhibitory.
On pyramidal axons and the thalamus, a lot do go to the thalamus, but there are an enormous number of cortical to cortical connections, as well as cortical connections to other subcortical regions. We can say that all sensory input, except for olfaction (smell in vertebrates has always been weird), goes through the thalamus before reaching the cortex, but a lot of the inter-regional activity bypasses it.
I am not sure anybody really understands so I can’t judge where your ideas are any better or worse than mine with the present state of knowledge.
What’s strikes me, however, is the fact the dendrite is supposed to be an input.
The apical dendrite runs from layer 5 to layer 1 where it terminates in an apical tuft which is apparently connected in some sort of network with tufts from other L5 neurons.. Wikipedia states ” Layer I consists largely of extensions of apical dendritic tufts of pyramidal neurons and horizontally oriented axons.”
So what’s wrong with the picture?
Where is the input to layer 1 if the apical dendrite is supposed to be where the output goes? Layer 1 looks like an area of a city where all the roads are one way going out of it.
My guess is that the apical dendrite on the L5 neurons isn’t simply input. It looks more like an antenna running a relatively long distance through most of the cortex but I know appearances can be deceiving.
The theory in the paper I link to says: “The apical dendrite operates within the axis circuit to stabilize neural activity, which enables conscious impressions to be steady and to be sustained over long periods of time.”
The difference between my theory and EM theory is that my theory is plausible with all of the known functionality of neurons. The EM Theory would require the discovery of some mechanism by which neurons can be (exquisitely?) manipulated by EM fields.
There are known mechanisms for detecting EM fields in the animal kingdom. At least some birds detect the earth’s magnetic field, and I think some fish detect electric fields. Those are the first places I would look for a mechanism, expecting Nature to repurpose what already works. I’m going to guess that nothing like those mechanisms is in the human brain.
I wouldn’t expect a “detector” of EM fields to provide any clue to a mechanism for how EM fields can manipulate neurons.
Was your use of the word “exquisitely” was a deliberate reference to this quote from McFadden?
“However, in a busy brain there will be many neurons teetering on the brink of firing and these undecided neurons may be exquisitely sensitive to the em field.”
There is an extensive discussion of how em fields can influence neuron firing in McFadden’s paper.
My understanding is that this cortical structure of 6 layers is all over the cortex not just in the prefrontal areas. And pyramidal neurons are also in other parts of the brain.
James, I don’t know whether you missed this link I provided in another post about the claustrum where researchers could turn off consciousness in a woman with electrical stimulation of the claustrum.
Unfortunately, nobody has been able to replicate it and the single woman in the paper had part of her hippocampus removed, Still the paper seems suggestive of something.
On the claustrum:
“The rise and fall of the claustrum epitomizes the hunt for consciousness in the brain.”
We were wondering the other day why talk about it died down. Apparently there were active attempts to replicate the clautrum’s role in turning consciousness on or off, and none were successful.
I agree. I read the same article.But I still wonder what was going on in that one woman’s brain.
Unless the researchers were just making it all up.
I don’t see any reason to suspect they made it up. But there may have been equipment, methodology, conflating variables, or other issues. It’s why reproducibility is so important. Even the most competent scientists overlook things.