Is there a conscious perception finish line?

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.

Image showing P3 wave presence for report and absence for non-report tests
Image from the study: https://www.biorxiv.org/content/10.1101/2020.01.15.908400v1.full

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?

A competition between integration and workspace

Back in March, I did a post on a proposed Templeton Foundation project to test major scientific theories of consciousness.  The idea was to start with a head to head competition between the integration information theory (IIT) and global workspace theory (GWT).  Apparently that project got funded and, according to a Science Magazine article, there are now active plans to move forward with it.

The first two contenders are the global workspace theory (GWT), championed by Stanislas Dehaene of the Collège de France in Paris, and the integrated information theory (IIT), proposed by Giulio Tononi of the University of Wisconsin in Madison. The GWT says the brain’s prefrontal cortex, which controls higher order cognitive processes like decision-making, acts as a central computer that collects and prioritizes information from sensory input. It then broadcasts the information to other parts of the brain that carry out tasks. Dehaene thinks this selection process is what we perceive as consciousness. By contrast, the IIT proposes that consciousness arises from the interconnectedness of brain networks. The more neurons interact with one another, the more a being feels conscious—even without sensory input. IIT proponents suspect this process occurs in the back of the brain, where neurons connect in a gridlike structure.

To test the schemes, six labs will run experiments with a total of more than 500 participants, costing the foundation $5 million. The labs, in the United States, Germany, the United Kingdom, and China, will use three techniques to record brain activity as volunteers perform consciousness-related tasks: functional magnetic resonance imaging, electroencephalography, and electrocorticography (a form of EEG done during brain surgery, in which electrodes are placed directly on the brain). In one experiment, researchers will measure the brain’s response when a person becomes aware of an image. The GWT predicts the front of the brain will suddenly become active, whereas the IIT says the back of the brain will be consistently active.

Tononi and Dehaene have agreed to parameters for the experiments and have registered their predictions. To avoid conflicts of interest, the scientists will neither collect nor interpret the data. If the results appear to disprove one theory, each has agreed to admit he was wrong—at least to some extent.

The whole thing has a bit of a publicity stunt feel to it.  As I noted back in March, both of these theories make differing philosophical assumptions about what consciousness fundamentally is, and the authors of both theories used empirical data, as it existed at the time, when formulating their theory.   So I’m not expecting the results to be overwhelmingly conclusive.  (Although it’d be good to be proven wrong on this.)

Lobes of the brain
Image credit: BruceBlaus via Wikipedia

What might be interesting is the front of the brain vs back of the brain thing.  I’ve noted this debate before.  Some scientists, notably people like Tononi and Christof Koch, see consciousness as concentrated in the back part of the brain, in the sensory processing regions including the temporal and parietal lobes.  Others, such as Dehaene, Joseph LeDoux, and Hakwan Lau, think we don’t become conscious of something until it reaches the prefrontal cortex.

This also has relevance on the distinction between first order and higher order theories, that is, between theories that hold that the representations and processing in sensory regions are conscious ones, versus theories that hold that further “higher order” processing in the prefrontal cortex is necessary for us to be conscious of them.

Part of the difficulty is that scientists depend on subject self report to know when those subjects are conscious of something.  However, self report requires the frontal lobes.  There are protocols to minimize the confounding role of self report, such as comparing brain scans of people who see something and report being conscious of it with people who see the same thing but without the requirement to report it.  But a first order advocate can always insist that any remaining frontal activations are superfluous, that all that’s needed for actual consciousness is the posterior activity.

My own money is that the frontal regions are important, perhaps crucial.  But this is complicated.  It’s possible for sensory information from the back part of the brain to trigger sub-cortical activity, such as habitual or reflexive action, without frontal lobe involvement.  It’s even possible to remember what happened during that behavior and consciously retrieve it later, giving us the impression we were conscious of the event during the event, even if we weren’t.

But if we insist that consciousness must include emotional feelings, then I think the frontal lobes become unavoidable.  The survival circuit activations that drive these feelings happen in subcortical regions in the front part of the brain, which have excitatory connections to the prefrontal cortex.  Of course, you could insist that the felt emotions lie in those subcortical circuits rather than the cortex, but severing the connections between those circuits and the prefrontal cortex (like what reportedly used to happen with lobotomies) typically results in deadened emotions.

And all of this is aside from the fact that the introspection machinery is in the very front part of the prefrontal cortex (the frontal poles).  Are we conscious of it if we can’t introspect it?

As I said, complicated.  A lot of this will depend on the assumptions and definitions the experimenters are using.

Still, I’m curious on exactly what they plan to do to test the back versus front paradigms.  If they do figure out a way to conclusively isolate conscious perception with one or the other, it might answer a lot of questions.  And if they do plan to eventually move on to testing theories like local recurrent processing or higher order thought theories, this work might provide a head start.

What do you think?  Am I being too pessimistic on whether these experiments will validate or falsify IIT or GWT?  Or are these theories all hopelessly underdetermined, and we’ll still be arguing over them months after the experimental results are published?

Dehaene’s global neuronal workspace theory

I just finished reading Stanislas Dehaene’s Consciousness and the Brain.  Dehaene is a French psychologist and cognitive neuroscientist who is bullish on the idea of consciousness being something that can be scientifically investigated.  It’s an interesting book, one that I recommend for anyone interested in the science of consciousness.

Dehaene accomplishes his scientific investigation by focusing on what he calls “conscious access”, which is roughly equivalent to Ned Block’s access consciousness and David Chalmers’ easy problems, that is, the ability of our minds to hold content available for reasoning, decision making, and verbal report.  He holds this distinct from the sense of self and metacognition, which he sees being built on top of it.

And as I mentioned in the last post, he is not concerned with phenomenal awareness, typically characterized as raw experience.  He actually barely mentions it in the book, being largely dismissive of it.  He characterizes the hard problem as ill defined, and the idea of qualia, pure mental experiences detached from any information processing role, as an idea that in time will go the way of vitalism.  (Those of you who know me may wonder if I’m projecting my own views here, but no, my views in this area just happen to match his pretty closely, right down to using similar language.)

After discussing the empirical tests able to identify what kinds of stimulus lead to conscious perception, as opposed to unconscious ones, he identifies four signatures of conscious access:

  1. The amplification of an early sensory signal leading it to an “ignition” of circuits in the parietal and prefrontal circuits
  2. The appearance of a slow P3 wave in an electroencephalogram, a slow massive wave throughout the parietal and prefrontal regions about 300 milliseconds after the stimulus
  3. A late and sudden burst of high frequency oscillations, gamma band power
  4. A synchronization of information exchanges across distant brain regions with oscillations in sync

All of this is referred to as “the conscious avalanche”, the widespread activation of neural activity in a network including the prefrontal and parietal regions whenever a perception makes it into consciousness.

Which leads to Dehaene’s theory of consciousness, the global neuronal workspace, a variation of Barnard Baars’ global workspace theory.  The main idea is summed up by Dehaene as:

When we say that we are aware of a certain piece of information, what we mean is just this: the information has entered into a specific storage area that makes it available to the rest of the brain. Among the millions of mental representations that constantly crisscross our brains in an unconscious manner, one is selected because of its relevance to our present goals. Consciousness makes it globally available to all our high-level decision systems. We possess a mental router, an evolved architecture for extracting relevant information and dispatching it.

…According to this theory, consciousness is just brain-wide information sharing. Whatever we become conscious of, we can hold it in our mind long after the corresponding stimulation has disappeared from the outside world. That’s because our brain has brought it into the workspace, which maintains it independently of the time and place at which we first perceived it. As a result, we may use it in whatever way we please. In particular, we can dispatch it to our language processors and name it; this is why the capacity to report is a key feature of a conscious state. But we can also store it in long-term memory or use it for our future plans, whatever they are. The flexible dissemination of information, I argue, is a characteristic property of the conscious state.

Dehaene, Stanislas. Consciousness and the Brain (p. 163-164). Penguin Publishing Group. Kindle Edition.

Credit: OpenStax College via Wikipedia

The workspace is held to be in a network of regions including the prefrontal cortex, the anterior cingulate cortex, and parietal regions.  I’ve discussed before that the middle parietal regions are the central integration of sensory association processing in the cortex, and the prefrontal cortex is the central integration of motor planning, so it makes sense that these regions would be highly interconnected and interactive, with ongoing recurrent loops of communication between the sensorium and the motorium.

I’ve also written about the central importance of the prefrontal cortex in imagination, the simulation and evaluation of action scenarios.  This fits with Dehaene’s view of the central importance of this region for the workplace.  Indeed, Dehaene implies that conscious access is centered on the prefrontal cortex, which fits with its role as the executive center of the brain.

So this theory appears to have a lot going for it.  And viewed from a purely instrumental perspective, it seems predictive of a lot of observations.  Unlike Tononi’s Integration Information Theory, it doesn’t aspire to recognize consciousness in systems outside of the brain, just in the brain itself, particularly in human or primate brains.  (Although Dehaene does think it could provide insights into possible architectures for consciousness in AI systems.)

And yet, like most grounded scientific theories of consciousness, I tend to think it captures aspects of the reality, but not the whole reality itself.  Even if we restrict ourselves to conscious access, the descriptions of the workplace feel a bit too simple to me.  It’s described like its one big thing, like a type of giant data bus.

This isn’t to say I think conscious access doesn’t involve wholesale activation of the regions that Dehaene discusses, but I’m not sure I buy the implicit description of it as one unified whole.  Based on all the reading I’ve done, it strikes me more as a complex web of disparate subsystems communicating with each other, with cross talk between the streams creating an emergent thing that may resemble the global workspace, but more messy, noisy, and less coherent than the theory implies.

But maybe I’m just quibbling here.  Dehaene might argue that it’s the final result that matters, and he’d be right.  And the global neuronal workspace seems general enough to be compatible with a lot of observations, as well as other theories such as HOT (Higher Order Theory).  I suspect before it’s over we’ll need a collection of theories to account for all observations.  But only time and more research will tell.