Don't worry, Church of the Churchless visitors who aren't as interested as I am in the hot new neuroscience theory of predictive processing by the brain, which is why I've been writing about Andy Clark's book The Experience Machine recently.
I'll be on to other topics soon. But not quite yet, since I want to share some of what I learned by watching Shamil Chandaria's talk on YouTube about "The Bayesian Brain and Meditation." I heard Chandaria and Sam Harris engage in a fascinating conversation on Harris' Waking Up app.
That led me to watch Chandaria's talk, since he's knowledgeable about the whole predictive processing thing, and also an avid meditator based on what he said to Harris. So I figured I might learn something about a topic that's been a subject of debate on this blog: whether it is possible for the human brain to function without filters or models.
I and a frequent commenter said no, it isn't. Another frequent commenter said yes, it is. So this is why I bring up that subject in my observations about some slides from Chandaria's presentation that I'm sharing below in the form of screenshots.
The theory of predictive processing by the brain fascinates me because it explains some things about the brain in a clearer fashion. For example, Harris (who has a Ph.D. in neuroscience) noted in his talk with Chandaria that there are ten times more top down connections between the brain and senses than bottom up connections.
Meaning, the brain is giving a lot more information to the senses than the senses are giving to the brain. This is at odds with the traditional way of looking at this: the senses communicate what's in the world to the brain; the brain then cognizes about that information; then the motor part of the brain decides on actions in accord with what the senses are saying.
Actually, the reverse is true, according to prediction processing theory. The brain, which has no direct connection to the world, is continually making predictions about what is out there based on prior experience and sense data. Errors caused by a mismatch between a prediction and sense data are what the brain pays most attention to, not sense data that matches a prediction.
The first part of Chandaria's talk was a bit drier and more technical than the part about meditation, because that initial part centered on what a Bayesian way of looking at things consists of. Key concepts include priors, or expectations, which basically is a prediction before new data is available. Sensory data then combine with the priors to produce a judgment about what is happening in the world.
(I've probably garbled what Bayesian statistics is all about, since I'm no expert on it. Watch the first part of the video for an expert explanation.)
Focused attention meditation is what most people think of when the subject of meditation arises. The dim objects in the thought bubble on the left represent the variety of things in the world that attention could alight on.
If a meditator focuses attention on a single object, such as the breath (the nose image), then other objects receive less attention. Or in the jargon of predictive processing, a decreased precision weighting -- which probably should just be called "weighting."
Deconstruction meditation such as some forms of mindfulness, analytical meditation, and koan practice is different in that the goal is to descend in the hierarchy used by the brain to model the world. For example, on the far left of the slide is a row where the top level is entire faces, the middle level is parts of faces, and the bottom level is parts of those parts.
The brain constructs a face by taking raw vision data and processing it upward from lines, curves, and such, to eyes, lips, nose, and such, ending up with the highest level of a complete face. Deconstruction moves in the opposite direction. Chandaria used the example of saying "dog" twenty times in a row rapidly until the word becomes a mere sound, not a symbol for a canine. (The slide does this with "flowers.")
Now, while it might seem that deconstruction is in line with the idea that the brain can function without filters or models, note in the slide above that Chandaria calls raw body sensations and raw sounds "low level models." So even raw sensory data is still minimally processed by models in the brain.
And of course it is difficult to imagine being able to function as a human with only raw sensory data. Even newborn infants quickly move away from the "blooming and buzzing confusion" William James termed their initial experience of the world as pure sensation. Who would want to live as a newborn for their entire life?
Probably the closest Chandaria came to talking about something that believers in the brain being able to function without filters or concepts would applaud is when he spoke of awareness as being the substrate on which the generative prediction processing model arises. Of course, it seems obvious that awareness or consciousness is the foundation of all experience, without exception.
This gets us into some murky philosophical and neuroscientific waters, since it is unclear what non-dual awareness actually consists of, or what awareness consists of. Is it possible for there to be awareness without any object of awareness? Some people believe in this. Me, I don't, but I could be wrong about this.
At any rate, non-dual awareness is much more of a spiritual or mystical notion, than a neuroscientific one.
Emptiness, in the Buddhist sense, is one of the spiritual notions (in the blue box) that Chandaria says are similar phenomenology (meaning, as experienced) yet with different metaphysical narratives. That's for sure. Brahman, God in Hinduism, is very different from Buddhist emptiness -- which means the interdependence of all things, none of which possess inherent existence in their own right.
The orange'ish note in the middle says: "NB Awareness itself is empty. It's not a thing. Tendency to reify it." [I"m confident this is a typo, since he is talking about ND, nondual, awareness.) So whatever non-dual awareness is, it isn't a thing. Chandaria also said "Everything is a construction," which supports my contention that the brain doesn't function without filters and models.
The thought cloud with a bunch of objects in it shows that non-dual awareness, as Zen teaches, is simply the union of subject and object where awareness isn't divided into (1) an experience of something by (2) an experiencer of that thing. There's just experience.
Chandaria presented an interesting three-dimensional model of meditation. (Not mediation.) Thr three dimensions are attention, non-dual awareness, and deconstruction. Note that under deconstruction he speaks of brain functioning becoming opaque rather than transparent. Meaning, the goal is to understand how the brain constructs reality instead of having this be automatic.
The peak meditation that combines all three dimensions to the greatest extent is in the upper top right corner: "Deep non-dual absorption state. Minimal Phenomenal Awareness (MPE). [Or MPA; Chandaria could use a proofreader for his slides.]
Chandaria said there are only a few people working on predictive processing models of meditation. He included references to two papers, one of which he was an author on. He didn't explain this slide in any detail. Guess you have to read the references.
Meditation, along with psychedelics, is a way of reconstructing ourself after deconstructing ourself. The discussion in blue also supports my contention that the brain always is going to use filters and models. But it's possible to use our knowledge of the brain to reshuffle our assumptions, expectations, predictions, and such about the world.
Chandaria calls this entering into a reprogramming mode. There's still a program running in the brain. The outputs are just more to our liking, hopefully.
This was one of his last slides. Again, Chandaria says that our experience of the world is a construction, which strongly implies filters and models always will be with us, since those are the brain's construction tools.
But we have the ability to make our experience as beautiful as possible.