Holism, Reductionism And Consciousness – A Dialogue

The Gliders, Eaters and Abstractions

Neo said, – Conway’s Game of Life is a monochrome world with a dead-simple transition rule to determine which squares, in the next phase of the game, will be “on” or “off” (black or white, say). Yet, it is mathematically provable (this has been done) that one can build a Universal Turing Machine in such a world. It would be massive, requiring an assemblage of “gliders” and “eaters” (denizens of the flashing, 2D monochrome world the rule creates) – but it is possible in principle… I believe a complete (if unmanageably long and computationally intractable) description of mental function in terms of low-level cellular interactions is possible. It wouldn’t help us understand mentality, but that would be a failure of our imagination, not the model.

Zeo said, – Neo, it is clear that in that system patterns appear, but the issue is – if we remove the developed consciousness which recognizes those patterns qua patterns, will they have some ontological existence by themselves?  In the Conway’s Game of Life, we do recognize different types of “things”, but doesn’t that delineation or recognition of something as a thing belongs to us. If we limit ourselves to thinking merely in terms of the game itself, there is nothing but the on/off pixels, and the rule of transition.

Neo said, – Zeo, I disagree. The gliders and eaters are “real” – as real as the transition rule etc.

Zeo said, – In what way are gliders and eaters real?

Neo said, – We can arbitrarily but consistently label them as such. There is no other kind of “real”, except for essentialists.

Zeo said, – Neo, we can label them as such, and see them as separate from the whole thing. But there is where I don’t agree – as much as we understand the gliders and eaters in terms of abstractions (rules of the game), it will be because we will have in the comprehension an element that isn’t neither implicitly, nor explicitly present in the abstractions of the game.

Neo said, – Zeo, but then if you are talking about necessity of adding some other element, you’re talking the position of dualism. That doesn’t solve the problem, but merely cloaks it in mystery. Believing mentality (meaning, intentionality etc.) is something “added in” is the very definition of dualism.

Zeo said, – Neo, I’m not saying that we should add something in addition to the abstractions present in the computationalist account in order to understand consciousness. I’m saying that those people who think that they can understand consciousness in terms of computational models are bringing in notions, e.g. a whole-ness in the model, which are nowhere in the abstractions which they claim are “enough” for the explanation. (I would say that they do some kind of Hegelian sublation of those abstractions into richer concepts, but without neither explicating it, nor noticing it).

Neo said, – Zeo, I’m afraid I find your sense of “abstraction” impenetrably obscure. What do you mean by the term?

Zeo said, – Sorry about that… say we have apple, we can abstract color, mass, etc… In that sense, the color, the mass and so on, are abstractions. Or in the case of movement, we can abstract time the movement took, the path of the movement, etc… Or say, you have a quantum wave equation, we can abstract position, or moment, or energy of a particle, etc…

Neo said, – Zeo, Then a machine could “abstract” in your sense.

Zeo said, – I’m not claiming that a machine can’t abstract, but that in considering what can “come out” of the artifact whose model is based on some abstractions, we should limit ourselves to thinking in those abstractions, and be careful not to introduce some other “external” notions.

Can talk about holism vs. reductionism go without mentioning QM? I guess not.

Neo said, – Zeo, just so you understand, I’m a hard-line determinist. I think that Zeo’s behavior can be explained entirely in terms of low-level causation, with no “gaps” or question-begging. Introducing some additional property of “whole-ness” merely introduces causal gaps – skyhooks! – which do no explanatory work. It’s mysticism, gibberish.

Zeo said, – Neo, I am also determinist in a sense, namely that there is reason why things are as they are.

Neo said, – But what sort of reason is non-reductionist in a sense that the properties of parts cannot completely explain the behavior of “wholes” (as you put it)? Where does the extra causation (or whatever) come from?

It seems to me, you misunderstand the nature of reduction. At issue is whether the behavior of the system can be predicted (in principle, that is) and explained in terms of the properties and interactions between parts. This is always the case. “Emergent” properties are properties of our models, of our ascending to higher levels of abstraction in order to render salient important causation without computing all the grubby details.

Zeo said, – Neo, I note really that we use the term “abstraction” differently; in your case “abstraction” is something to which we ascend to, while I use abstraction to refer to what is more simple (so, probably to what we descend to), but anyway here is a quote from Stanford encyclopedia on quantum mechanics:

What this means, or at least what it appears to mean, is that there are, according to quantum mechanics, facts about composite systems (and not just facts about their spatial configuration) that don’t supervene on facts about their components; it means that there are facts about systems as wholes that don’t supervene on facts about their parts and the way those parts are arranged in space. The significance of this feature of the theory cannot be overplayed; it is, in one way or another, implicated in most of its most difficult problems.

– So contrary to what you said was always the case, this paper claims that the behavior of a QM system can’t be predicted nor can be explained in terms of the properties and interactions between parts.

Zeo continued, – BTW, you shouldn’t find anything weird with the idea that there is more to the whole than the aggregate/dynamics of its parts. In the basic subject/predicate proposition, you have that possibility; When we say A is B, it doesn’t have to mean that A is nothing but B. (and if “A is B” is to be cognitively significant A better be something more than B.)
– Neo, and using the same logic “’that whole’ is ‘such and such configuration of parts’” doesn’t have to be taken as equation between the whole and configuration of parts. So “whole is more than sum of its parts” is not something illogical, and as the case in QM shows it is not even something inexpressible by math! It may just look as “mysticism” in the context of classical physics.

Neo said, – It can’t (be predicted and understood) IN PRACTICE, but that’s because of computational intractability, combinatorial explosion etc. It arises because all computation has a physical basis, which means we need an exponentially large machine to compute low-level physics in complex systems. – I argue that science is about unifying and showing how various levels of explanation compliment and support each other. It is impossible to explain how a mind works in terms of cellular interactions, a computationally intractable problem. We’d need computers the size of galaxies! So we need emergent perspectives, higher-level abstractions, concepts of representation, function, information and the like.

Zeo said, – Neo, it is you … a conscious being that can understand all those “higher-level abstractions” (putting aside our different usage of “abstraction”), and for whom those are meaningful, and it is you who are adding those notions to the explanation. If you don’t bring them there, all you have are atoms causally affecting each other and nothing else… you can’t have anything else but the abstractions you have put in it.., namely causal relations in terms of space and time.

Does this make Zeo a dualist?

Neo said, – Eh? But I’m a collection of atoms causally affecting each other, too. You’re just reifying mental categories and creating insoluble linguistic riddles. You’re also moving towards dualism for reasons I outlined earlier.

Zeo said, – I’m not dualist, and I’m not reifying mental categories! As far as terms like mind, consciousness, qualia etc… go, I’m against reifying them as some kind of separate essence, and for sure I’m not talking consciousness (or anything else) to be some “magical substance” you add to your average zombie.

Neo said, – Zeo, I don’t think you think you’re a dualist, any more than I think I’m denying free will and consciousness. But some might say that I am, because they might feel my views imply that (but I just can’t see it). Even I consider those people wrong, it is true that it isn’t necessary for their case that I regard myself as denying those things. So, I’m saying your views lead to dualism, whether you know it or not.

Zeo said, – Neo, but why would you think that holism leads to dualism?

Neo said, – IMO, is a the “holistic” perspective is just a higher-level explanatory level, which should be fully complimentary and unifiable with lower-level explanatory levels. All perspectives can be valid, if they have predictive power and explain how things work. If you want to argue, say, that psychological descriptions of mind are as valid as neuronal ones, or that cognitivists need to ascend to higher levels to regard brains as carriers of cultural signs, symbols and meaning, then fine – count me in. But if you want to argue that those perspectives form self-existent, irreducible categories, that’s dualism.

Neo continued, – Zeo, I just don’t see anything you can say about the world as being anything other than metaphor and description. Either we can unify various levels of description, or one particular level must be regarded as primary and “literal” in a way the others are not. If something is “irreducible”, then its description cannot be couched in lower-level terms, and no lower-level description will capture its essences, in which case you have dualism.

Maybe some common ground is possible?

Neo said, – …Zeo, what’s the difference between “holism”, and systems thinking?

Zeo said, – Neo, by holism I mean the position that the whole can’t be reduced to its parts. Not in the terms of unpredictability (as in the case of chaotic systems), but in principle. In the kind of holistic thinking I have on mind, parts are abstractions… or one can say… it is not that a thing consist of atoms, but that it is divisible into atoms.

Neo said, – alright, let’s say I was to argue that for many phenomena, we cannot ever, ever understand them without regarding them as systems with emergent features, and relying on those higher-level perspectives completely in our explanation. Is that enough “holism”?

Zeo said, – No, as long as it just a matter of understanding, it is not holism, at least not of the kind I’m talking about. It would be holism only if it is in principle irreducible.

Neo said, – Zeo, systems theory regards systems as systems, within an environment. It is “holistic” in every sense you outline, yet nobody regards it as non-reductionist in the special metaphysical sense of being incompatible with reductionism. It is just another perspective.

Zeo said, – Different kind of holism I guess.

No, seems no common ground. And little more on the differences…

Zeo said, – Neo, btw, I don’t have problems by saying “whole is structure of parts”, I have problem with saying “whole=structure of parts”.

Neo said, – Verbal quibble. In some sense the whole equals the structure/sum of parts, but not in other sense.

Zeo said, – I don’t think it is mere verbal quibble, I think it is important distinction, between “A is B” and “A=B”. A=B is a relation of identity (relation which can be just between the thing and itself), A is B is subject/predicate relation in which the subject is more than the predicate. I noted that I accept “whole is aggregate of parts”, just to point out that I don’t have problems with physics, which treats wholes as aggregates of parts. For that, what is enough is to accept that “whole is aggregate/dynamics of parts”, identity isn’t needed… Neo,I think that the problems appear in questions where instead of treatment of A as B, the issue involves acceptance of A=B.

Zeo continued, – same goes, I think, for the question of consciousness… I don’t have problem with treating consciousness as functionalistic or computational system. I have problems with making identity between consciousness and such system, where consciousness is nothing but that model. Or said differently functionalism is good for understanding the relations of abstractions, and physics is also good at that. What I think is wrong is taking those abstractions and their relations as constitutive, and mistake artifacts made on base of those abstractions, for the real things which were merely analyzed in terms of those abstractions.

If you need to take sides who would you agree with – Neo or Zeo?