Thoughts on dualism and related matters.

Does reality contain something in addition to the bits of matter and energy (call them, collectively, Weltstoff) whose interactions the Standard Model deals with, something called ‘Mind’, of which the question ‘How does Mind interact with Weltstoff’ can be sensibly asked?

No. Mind is implemented in Weltstoff. Interaction comes for free.

To have a complete understanding of reality, do we need to talk about entities other than Weltstoff?

Of course. And before we even get to Mind, let’s consider Life for a moment. Life represents a colonization of Weltstoff by information, in the form of DNA, and the dance of genotype and phenotype makes the world a very much more interesting place than it would be without life. Organic chemistry is a big field already, and we go from there.

And Mind represents a further level of colonization by information. When I Google ‘brain storage capacity’ the highlighted result is 2.5 million gigabytes. Very difficult to quantify, but it’s ‘really, really big’.

Finding the sentence ‘Why would any collection of matter in the universe be conscious?’ in the first paragraph of Annaka Harris’s book, Conscious, I am rather bemused. We are not talking about a ‘collection of matter’, we are talking about a living information-processing system.

Two predicates there: (1) living, and (2) an information-processing system (IPS). For everything living, we know that the ultimate answer to ‘why’ questions is, ‘because it worked for that organism’s ancestors’. As for IPSs, now that we can make comparisons to a member of the genus with a completely different architecture—the digital computer—we have some inkling about the potential for relatively small ‘collections of matter’ to have enormous causal efficacy.

What does Mind do? Fundamentally, I would say, it makes the world a familiar world. Everything we encounter with our senses falls into place, more or less. The same is true of all organisms with a brain.

The evolution of language is yet a further colonization of the material world by information: words represent a move in the direction of digital processing, having a definite structure in terms of phonemes. And a fourth layer is the invention (finally we are in the realm of invention rather than evolution) of written language. (And digital technology is a kind of fifth layer, come to think of it; if true AI ever happens that would be a sixth layer.)

Mind is implemented in Weltstoff. But of the architecture of the human brain we can say very little; of course we know a good deal now about what the various areas of the brain ‘do’, but the fine details of, say, what neuron firings correspond to ‘driving east on Occidental Road while listening to Sugar Magnolia on the sound system’ or ‘working on this blog post’ are not something that we have access to, or would be able to form theories about even if we did.

It seems to me that the brain is ‘meaning all the way down’ in this sense: that the state of every synapse in this brain I’ve got here depends in a lawful way on (for all we know) everything that this organism has experienced in its 65 years of life, but that no finer-grained description can be had.

If we supplied a sufficiently powerful machine learning system with, on the one hand, a complete record of neuron firings in a subject brain over a year, and on the other hand, detailed information about what the organism housing that brain was doing at every moment of that year, it would presumably be able to detect patterns so that (after the training period) it could, from subsequent neuron firing data alone, assign correct truth-values to predicates like ‘Joe is in deep sleep | dreaming | talking | listening to music | doing calculations | eating popcorn | tying his shoelaces’ and on and on. But asking the machine learning system to produce an understandable theory covering how those patterns work and interact would be asking the impossible.