Downward Causation

heterogeneous-ontologyIs downward causation the answer? Does it explain how consciousness can be a real and important part of the world without being reducible to physics? Sean Carroll had a sensible discussion of the subject recently.

What does ‘down’ even mean here? The idea rests on the observation that the world operates on many distinct levels of description. Fluidity is not a property of individual molecules but something that ’emerges’ when certain groups of them get together. Cells together make up organisms that in turn produce ecosystems. Often enough these levels of description deal with progressively larger or smaller entities, and we typically refer to the levels that deal with larger entities as higher, though we should be careful about assuming there is one coherent set of levels of description that fit into one another like Russian dolls.

Usually we think that reality lives on the lowest level, in physics. Somewhere down there is where the real motors of the universe are driving things. Let’s say this is the level of particles, though probably it is actually about some set of entities in quantum mechanics, string theory, or whatever set of ideas eventually proves to be correct. There’s something in this view because it’s down here at the bottom that the sums really work and give precise answers, while at higher levels of description the definitions are more approximate and things tend to be more messy and statistical.

Now consciousness is quite a high-level business. Particles make proteins that make cells that make brains that generate thoughts. So one reductionist point of view would be that really the truth is the story about particles: that’s where the course of events is really decided, and the mental experiences and decisions we think are going on in consciousness are delusions, or at best a kind of poetic approximation.

It’s not really true, however, that the entities dealt with at higher levels of description are not real. Fluidity is a perfectly real phenomenon, after all. For that matter the Olympics were real, and cannot be discussed in terms of elementary particles. What if our thoughts were real and also causally effective at lower levels of description? We find it easy to think that the motion of molecules ’caused’ the motion of the football they compose, but what if it also worked the other way? Then consciousness could be real and effectual within the framework of a sufficiently flexible version of physics.

Carroll doesn’t think that really washes, and I think he’s right. It’s a mistake to think that relations between different levels of description are causal. It isn’t that my putting the beef and potatoes on the table caused lunch to be served; they’re the same thing described differently. Now perhaps we might allow ourselves a sense in which things cause themselves, but that would be a strange and unusual sense, quite different from the normal sense in which cause and effect by definition operate over time.

So real downward causality, no: if by talk of downward causality people only mean that real effectual mental events can co-exist with the particle story but on a different level of description, that point is sound but misleadingly described.

The thing that continues to worry me slightly is the question of why the world is so messily heterogeneous in its ontology – why it needs such a profusion of levels of description in order to discuss all the entities of interest. I suppose one possibility is that we’re just not looking at things correctly. When we look for grand unifying theories we tend to look to ever lower levels of description and to the conjectured origins of the world. Perhaps that’s the wrong approach and we should instead be looking for the unimaginable mental perspective that reconciles all levels of description.

Or, and I think this might be closer to it, the fact that there are more things in heaven and earth than are dreamt of in anyone’s philosophy is actually connected with the obscure reason for there being anything. As the world gets larger it gets, ipso facto, more complex and reduction and backward extrapolation get ever more hopeless. Perhaps that is in some sense just as well.

(The picture is actually a children’s puzzle from 1921 – any solutions? You need to know it is called ‘Illustrated Central Acrostic’)  

 

77 thoughts on “Downward Causation

  1. Thanks Peter, quite the discussion over there! Carroll says:

    “The dynamical rules of the Core Theory aren’t just vague suggestions; they are absolutely precise statements about how the quantum fields making up you and me behave under any circumstances (within the “everyday life” domain of validity). And those rules say that the behavior of, say, an electron is determined by the local values of other quantum fields at the position of the electron — and by nothing else. (That’s “locality” or “microcausality” in quantum field theory.) In particular, as long as the quantum fields at the precise position of the electron are the same, the larger context in which it is embedded is utterly irrelevant.”

    This suggests we don’t need to know anything but microphysics to predict, say, the position of a particular electron in my arm 24 hours from now. Chemical, biological, behavioral and psychological regularities (“the larger context”) are simply higher level expressions of what the underlying microphysics entail. Higher level regularities don’t have any independent, ineliminable predictive contribution to make about the position of my arm and therefore don’t need to be taken into account to predict the position of the electron in question, even though they might be convenient short-cuts in such predictions. Is this plausible? And does anyone know of a proof of this conjecture?

    Of course such regularities will necessarily be *consistent* with underlying microphysical laws, but whether they are literally superfluous when it comes to predicting future states of affairs is a different question. Has it been settled?

    As for conscious experience (subjective qualitative states), Carroll thinks it can be accounted for by the fields and equations the Core Theory (“arising in a completely conventional way from the collective behavior of microscopic physical constituents of matter”) but neither he nor anyone else has provided such an account as far as I know. Which is not to say that either substance or property dualism is necessarily the case, but only that consciousness might not be a straightforward entailment of microphysical goings-on. So perhaps we shouldn’t assume that it is.

  2. “why it needs such a profusion of levels of description in order to discuss all the entities of interest.”

    I think the answer comes down to the limitations of our minds. Humans are apparently the only animals with symbolic thought. It’s a very powerful ability that allows us to conceptualize many things we’ll never have direct sensory access to, from things as small as elementary particles to intergalactic filaments. It allows an African hominid species to strike well beyond its original ecological niche.

    But it’s important to remember that symbolic thinking is a trick, a hack, a mechanism to put things in a manner that a primate species can understand. We use metaphors and analogies to understand things like protons, electrons, and black holes because we’ll never be able to appreciate them in their full scope. When we think about the solar system, we never think about it itself, but of the somewhat cartoonish models we’ve built from the diagrams and illustrations we’ve seen.

    While the hacks work very well, we can’t get away with one set of them for all of reality. The metaphors and analogies that work for physics start to become too complicated as we move up to chemistry and biology, and the metaphors and analogies we use for biology start to fray as we move up to psychology and sociology.

    All emergence is, is the fact that at certain points we have to switch models to understand what’s going on. I think people get confused about this because they tend to reify the metaphors and analogies, forgetting that they are only metaphors and analogies, not the underlying reality, which there’s no reason to suspect doesn’t operate smoothly from the lowest to the highest levels.

  3. Chemical, biological, behavioral and psychological regularities (“the larger context”) are simply higher level expressions of what the underlying microphysics entail.

    No, really not. What’s the microphysical definition of “Olympic Games”? Even the physical aspect of the games corresponds with an undefinable, indefinitely large set of microphysical arrangements. But the idea of the games has no microphysical interpretation. What about the Ulympics? They never existed, but I can describe them.

  4. I’m on board with pretty much everything SelfAwarePatterns says. We’re shallow information consumers, evolved to extract as much utility from as little signal as possible. Why solve systems for what they are when a ‘hack’ will do just as fine. The prostheses of science allow us to do this in a much more information intensive way.

    The real question, on this view, is why we should think things should be any other way.

  5. As we build levels of descriptions we are also building longer event horizons or objects. At the quantum level it is impossible to achieve distinction but moving up the ladder we can physically isolate with instruments individual atoms, then molecules and complex molecules and cells and organs, organisms etc. Qualia, then feelings to thoughts and concepts are actually formed metaphysical objects that we can pass between ourselves as SAP points out above as symbols.

  6. It might be that our limitations have something to do with it, but I think our understanding is a general capacity, not a collection of kludges fine-tuned to deal only with our anthropoid past. I don’t believe mathematics and logic are “a trick, a hack, a mechanism to put things in a manner that a primate species can understand”. If you don’t think these things are fundamental, I think you’re missing something.

  7. Carroll is arguing that “Chemical, biological, behavioral and psychological regularities (‘the larger context’) are simply higher level expressions of what the underlying microphysics entail.”

    Peter: “No, really not.”

    If not, then it seems you’d agree that we would need to take into account higher level regularities to predict the location of a particular electron in my arm in 24 hours. Knowing just the micro-conditions wouldn’t be sufficient, right?

  8. “but neither he nor anyone else has provided such an account as far as I know. Which is not to say that either substance or property dualism is necessarily the case, but only that consciousness might not be a straightforward entailment of microphysical goings-on. So perhaps we shouldn’t assume that it is.”

    Yeah, I agree. Carroll seems to use the same bluster for consciousness as he does for the multiverse.

    That said, I do think Carroll presents a clear view of what physicalism means – things have to be accounted for at the lowest level of whatever matter is. Additionally this matter can be defined, at the very least, as what it does not contain – any mental or proto-mental characteristics.

    This susses out physicalism from accounts that bring in “information processes” and/or “patterns”. If nothing else this leaves it clear what the goalpost for the physicalist actually is.

  9. Peter #6,
    I suppose it depends on what you mean by “fundamental”. I don’t think they’re fundamental in any Platonic sense, but I do think they’re founded on the most fundamental patterns out in the world (at least fundamental relative to us). In that sense, I’m an empiricist, although quasi-empiricist is probably a better description.

    I hasten to add that we do seem to have evolved instinctive intuitions about many of these foundations, giving us an innate appreciation of them. And I fully understand that most of what mathematicians and logicians do is not empirical in practice. Nonetheless, I think what they do starts with real world patterns.

    In my mind, none of this changes what I said above. We can work with concepts like 10100 years, but I’d be suspicious of anyone who thought they could actually imagine it in its full magnitude.

  10. Peter


    The thing that continues to worry me slightly is the question of why the world is so messily heterogeneous in its ontology – why it needs such a profusion of levels of description in order to discuss all the entities of interest.

    You sound like a physicist. Please o let the world be neat !


    Particles make proteins that make cells that make brains that generate thoughts. So one reductionist point of view would be that really the truth is the story about particles

    The truth of physics is about particles. But physics is a closed circle, and excludes consciousness. No amount of computation or slick mathematics goes from particles to consciousness.

    There is an understandable fixation with the reductionist issue as far as consciousness is concerned. Fluidity flows from physics (quite naturally and straightforwardly) but the Olympic Games .. is it theoretically computable to make a meaningful description of such an event in physics ? It’s possible to describe the physical components of the olymoic games, in general terms, but the Olympic Games has non-physical meaning. This is where physics is totally useless. If the description doesn’t involve kilograms, meters or seconds – and the notion of competition between humans of different nations at sport does not, nor is capable of such description – then physics cannot help. Reduction really is a myth.

    Reduction – which is based upon physics, let’s not forget – cannot cope with new semantic appearing at higher levels. It’s a phoney idea based upon (that old chestnut) Physics Is Everything. Physics Isn’t Human. Physics is a Gift of The Gods. Alas physics is all too human and that’s why reduction fails – and does so quickly, crashing in smithereens at the prospect of dealing with low level life forms- viruses and nematodes.

    JBD

  11. Carroll is arguing that “Chemical, biological, behavioral and psychological regularities (‘the larger context’) are simply higher level expressions of what the underlying microphysics entail.

    No, I believe he’s saying – and I agree – that we should stick to one level of description at a time. High level stuff is not merely the expression of micro stuff nor vice versa, so the position of an electron does not require consideration of “Olympic games” level facts.

  12. Peter, just to be clear, in his post Carroll says (and I quote):

    “And those rules say that the behavior of, say, an electron is determined by the local values of other quantum fields at the position of the electron — and by nothing else. (That’s “locality” or “microcausality” in quantum field theory.) In particular, as long as the quantum fields at the precise position of the electron are the same, the larger context in which it is embedded is utterly irrelevant.”

    This suggests to me that we don’t need to consult higher-level regularities (the “larger context” of chemical, biological, behavioral and psychological laws) in predicting the position of an electron in my arm 24 hours from now, since Carroll says (I think) that all the information needed for the prediction is in principle at the micro-level. I’m wondering if this conjecture has been proved, and if anyone knows of a proof.

  13. interesting post.great proposition. the “issue” with unification at the mental level of description is that if physicalism is the only answer then each human is a different arrangement of atoms at the physical level. differing biology leads to differing psychology. hence differing ways of “talking about” the same thing as evidenced by diverse artistic languages of music ,poetry,literature and schools of thought and in the sciences biology, physics, chemistry, and differing schools of thought like reductionism , synthesis etc. so finally we are left with the fundamental particle approach or you could call it an energy field whose probabilities are described by a mathematical function.

  14. I think Sean Carroll is right on and the problem is languages used between different levels of description. In that case, the mystery of consciousness has a solutution which needs details to fill it out. But this also means there may not be any one current science that can do that by itself and a whole new type of field may need to be started to relate these various levels of description.

  15. Peter: If you get a chance to check out Peter Hacking’s latest book on mathematics, you’ll be amazed that you saw it as anything but a bag of tricks–an exceptionally powerful one!

    The intuition of general cognition is something that my view predicts: short differentiations, we assume unities, identities, continuities, generalities. This is why your computer screen seems to glow continuously, why Aristotle assumed the celestial to be ‘pure’ in nature, ontologically distinct from the terrestrial, why so many Trump supporters assume Islam monolithic, and so on and so on. Short any metacognitive access to cognitive processes, short difference making differences, cognition has to seem undifferentiated, omni-applicable. How could it seem otherwise?

    But why do we need ‘general cognition’? If a myriad of tools and a workspace for knapping/experimenting with new tools or new tool/ecology combinations is all that’s required to explain humanity’s stumble forward, then why ask for anything more? What necessitates positing general cognition, which is to say, a form of cognition applicable to all possible problems? As with selves, it seems like exactly the kind of system evolution would not provide.

  16. I’ve read the blog post your refer to and commented on it (lost in the noise) and also read his recent book, The Big Picture. I’ve also watched a number of his YouTube lectures and follow his blog. None of what Carroll says about causation in that blog post seems sensible in light of his other statements on the subject.

    What kind of philosopher is Sean Carroll? He talks a lot about “ontologies”, plural, but if you follow closely he really only has one ontology, which is “The world is made of fields”. When he says that “stories” about reality work at different levels, he calls these “ontologies”, but he doesn’t really use the word correctly. He doesn’t believe in higher levels of reality only of *description*. So his “ontologies” are really *epistemes*.

    The higher level stories are expedients, useful because there are gaps in our knowledge and low level models become impractical to compute at higher levels of complexity. In other words the problem is epistemic. He’s an epistemological antireductionist and an ontological reductionist – a combination he calls “weak emergentism”. I think the strong/weak distinction is wholly misleading because in fact it turns on an ontic/epistemic distinction. He does usefully advise us not to cross the streams when using levels however.

    Crucially in TBP, he gives an example of an “emergentist” description which illustrates his attitude perfectly. He chooses a volume of gas. His fine grained description involves the velocities of gas molecules and so on. His coarse grained “emergent” description involves the temperature and pressure of the gas. What we call “temperature” simply *is* the average velocities of the gas molecules.

    What is special about this example is that the coarse grained description is a mere aggregate of the fine grained level. His coarse grained model is entirely reducible to his fine grained model. It perfectly illustrates his true philosophy – he doesn’t consider higher levels of description as real. He considers talking about them *as it* they were real a useful practice.

    In the book he argues that downward causation is a necessary feature of what I call ontological antireductionism. He’s quite wrong about this. Emergent properties are not required to be in causal relationships with their constituents by ontological antireductionism. They may well interact at their own level, but there is no requirement for downward causation that I can see. According to Richard H. Jones (Analysis and the Fullness of Reality), it’s a position argued for by a minority of philosophers in this area and Carroll may have simply read the wrong book on this.

    In TBP Carroll explains that there is *no such thing as causality*. It’s not required in quantum field theory or even in Newtonian physics. There is no equation for causality and no expression for it in any equation. Nor can causality be deduced from QFT. The universe evolves in a patterned way, but no causation is needed for this to happen. Why the insistence on *downward* causation from a man who doesn’t believe in causality anyway?

    OTOH in his book he says: “It’s not possible to specify the state of a system by listing the state of its subsystems individually. We have to look at the system as a whole, because different parts of the system can be entangled with one another.” (The Big Picture, p.100)

    In other words systems are irreducible!!! “We have to look at the system *as a whole*”, because entanglement is a relation that transcends individual particles and creates systems with irreducible emergent properties that cannot be understood or explained by simply aggregating the properties of those constituent particles. Fine graining the problem leaves out something essential, i.e. a structural feature. Carroll never returns to this example and does not seem to discuss it elsewhere.

    What Carroll shows us with this statement is that systems (or at least some systems) are irreducible. And so ontological structure antireductionism is a fundamental requirement of any philosophy based on reality. But Carroll himself does not go this far. The man doesn’t know whether he’s coming or going philosophically. But he’s a great guy, he writes and speaks well, and he knows a shed load more about physics than me.

    Downward causation is the least of his worries.

  17. Thanks Jayarava, you’ve saved me the trouble of reading Carroll’s book.

    Hopefully he can overcome the confusion you mention in the future.

  18. Guys,

    You might be interested to see my ‘summary of everything’. I tried to summarize ‘all knowledge’ by grouping wikipedia articles. It ended up separating into 27 different ‘levels of description’ (domains).

    I symbolize each level using a letter of the archaic Greek alphabet. Click on the name of each domain to take you through to the A-Z of wikipedia links covering all the main concepts for that domain. In total I needed about 1 900 wikipedia articles to summarize everything.

    It’s interesting that knowledge basically separates into 27 different vocabularies or ‘levels’. Here’s the link:

    http://www.zarzuelazen.com/CoreKnowledgeDomains2.html

  19. Mario Bunge [Matter and mind: A philosophical inquiry, 2010] gives many examples of downward causation eg purposeful behaviour leads to maintenance of hydration of an organism: the value of hydration to the organism, and caring about hydration, exist at a different level of organisation from the brute fact of osmotic pressure across a membrane. “Physics cannot…fully explain machines, as these embody ideas, such as those of value, goal and safety”. In terms of algorithmic complexity, the description of the homeostatic organism is shortest if includes a model of behaviour. I personally have a strong feeling that this is not just an artefact of human understanding, but that the minimum description length or “intrinsic computational capacity” [a la Crutchfield] is a physical property of the system, just as temperature is. Of course, this still hasn’t been fully operationalised.

  20. I find it exasperating that physicists jump into this discussion and make the assertion that we have to go down to the particle level to explain this. It is obvious that neurons (not blood or kidney cells) do this. Unfortunately not just the physicists with their particle conceit but math and computer scientists take over as well with their computational conceit. All the while most are clueless that the brain and CNS is another complex biological system composed of over thirty organs and sub organs. Oops I left out the philosophers and their conceit.

  21. If consciousness is a byproduct of the material that does no useful work nor has causal efficacy, its connection to behavior should be problematic. We can imagine endless versions of incoherent consciousness. Each would do as well as the actual coherent consciousness that is closely correlated with behavior. We can imagine, for example, a person laughing while having the gloomiest of thoughts or having no awareness at all. The laughter and associated behavior is physical and is unaffected by the epiphenomanal consciousness. So out of the infinitude of possible degrees of coherence, we wind up with the one that is coherent and perfectly correlated with behavior. For the materialist, it is all the same. For others it probably is not. But it defies belief, one out of an infinitude! But if consciousness does matter, it would have evolved along with other heritable traits. It would be coherent and correlated with behavior.

  22. David; Since they are dealing with a complex nervous system that they have no clue of how it can walk and chew gum at the same time, little wonder how they can concoct strange arguments. I think maybe what you are getting at is that the ability to be deceptive is an evolved trait for survival against predators or for politics in our world. Little wonder how it comes back to bite us.

  23. #19 “eg purposeful behaviour leads to maintenance of hydration of an organism”. Organisms that clearly are not concious also do this – plants, fungi, bacteria, and archaea all do it. So it’s not a very good example, since any conscious state that might be involved is clearly incidental to the feature. This is just “causation” on the same level.

  24. #21 David “If consciousness is a byproduct of the material that does no useful work nor has causal efficacy, its connection to behavior should be problematic.”

    Sure. This is why epiphenomenalism is not credible. It doesn’t explain anything.

    But worse, underlying epiphenomenalism, as in so much other philosophy of mind, is Cartesian Dualism – the idea that there is an ontological distinction between mental phenomena and physical phenomena.

    The abstraction from conscious states to consciousness introduces all kind of other philosophical problems. I’m starting to think that it is just a mistake. Abstractions like “consciousness” only exist as conscious states, they don’t exist in nature. So looking for the reality of an abstraction gets us nowhere.

  25. I really like Jayarava’s analysis in 16, and David Duffy’s remarks in 19 tend toward the same conclusion: to secure the structural and organizational reality of things above the micro-level. They challenge Carroll’s “poetic naturalism,” in which bottom level micro-physics is a description of what’s fundamentally real, and higher level phenomena are simply convenient descriptions (“ways of talking”). In principle, although definitely not in practice, Carroll says, we don’t need higher-level descriptions in making predictions about future states of affairs at any level, micro or macro.

    The empiricist’s criterion for what *really* exists is just that which is necessary for successful prediction and control in a given domain of explanation. For example, we ordinarily predict human behavior by attributing such things beliefs, desires, and other intentional and psychological states, and the reality of such states would be secured just by their having an ineliminable role in such predictions. Now, if we could in principle (if not in practice) predict human behavior without recourse to referring to psychological states, using only knowledge at the micro-level, then we could safely eliminate such states as fundamentally unreal. They are simply convenient explanatory fictions that help us navigate the social world (“ways of talking”). Carroll is, I think, saying the same thing about *all* purported phenomena above the micro-physical: in principle they don’t need to appear in a prediction about future states of affairs, so aren’t fundamentally real.

    But is this “in principle” claim true? If it isn’t, then according to the empiricist’s criterion of what exists, at least some higher level phenomena are just as real as micro-physical phenomena, since we can’t successfully predict or control events at a level or domain (e.g., that of human behavior) without referring to phenomena appearing in that domain (e.g., psychological states). To settle the question of the predictive completeness of the micro-physical level of description (still an open question, right?) would be to answer the question about the ontological status of higher-level phenomena.

    Btw, there’s an interesting review of The Big Picture which takes up the issue of Carroll’s “nominalism” (as opposed to realism) about the macro world over at Public Discourse, http://www.thepublicdiscourse.com/2016/09/17886/

    As for the causal role of consciousness (phenomenal experience), downward or otherwise, there’s no story on offer about how experience per se affects physical behavior-control mechanisms. Of course if it turns out that experience is the same thing as a sub-set of such mechanisms, then its causal role is secured. But it will be in terms of the physical, functional and representational properties of such mechanisms, not qualitative and subjective, that the causal story will be told.

  26. VicP, no I’m simply wondering whether humanity needs consciousness. Or is it a superfluity?

  27. Separating the elementary particle view and integrated macro views into “real” and “constructed” or “not real” categories is a false dichotomy. They are really just different views or aspects of the same thing. One is not more real than the other. Consider looking at all the quantum states of electrons making up a tree versus looking at the light reflections emanating from the tree. It’s still a tree. Another way of looking at it might be looking very closely at a photo so all you see is the pixels and then backing up until you see the whole picture. You are looking at the same thing, just processing the information differently.

  28. Thanks, Jayarava, for the comment. I did not make myself clear. Your quote of me is my effort to enunciate the mystery of the connection between behavior and consciousness. Why are the two perfectly correlated? So I ask why if consciousness were gibberish, it would change anything in a physical sense. Why would’t humans be, in all material ways, identical to the way they presently are given the elimination of consciousness? Or the making of consciousness nonsensical and uncorrelated with behavior; why would that matter for
    human behavior or neuron activity? I mean, why in the Dickens consciousness at all? Is it some kind of prank being played on humanity? Just say whether we would or could be the same, excepting awareness, in the absence of consciousness. This is a simple and straightforward question.

  29. What’s the microphysical definition of “Olympic Games”? Even the physical aspect of the games corresponds with an undefinable, indefinitely large set of microphysical arrangements. But the idea of the games has no microphysical interpretation.

    To me, this sort of makes the ‘Olympic Games’ something that ‘floats’. The Olympic games would have to have involved a series of physical events and with it micro physical events. But the idea of the games does not include those micro physical events, so it kind of seems to float above the micro physical – as if the games were a floating orb and the micro physical a flat plane below, with no connection between the two.

    I think SelfAwarePatterns aught to describe how it makes the ‘Olympic Games’ to him. It might be slightly different for him, but it might also include that the games is separate from the micro physical.

    I’m actually inclined toward Self’s explanation, but the thing is it rushes to explain away stuff without acknowledging personal perspective and inclination. As you might expect – we’re inclined to hide what we think are our mistakes. But even as he explains it away, for SelfAwarePatterns, the ‘Olympic Games’ probably still has some type of ‘float’ above the micro physical.

  30. Hi Callan S,
    If I’m understanding this part of the discussion correctly, I think there are two ways of looking at this: epistemic (what we can know) and ontological (what actually is).

    I like your float analogy. Epistemically, the idea of the Olympic Games does float, separate from the underlying, say, quantum physics, mainly because we have to shift our models so many times between those layers that attempting to relate them is pointless, first because there is no mind or computational system that could conceivably model it all together, and second because the current inaccuracies and blind spots within the models at each layer would likely snowball and make the resulting uber-model hopelessly wrong.

    That said, ontologically, I do think the idea of the games are patterns of patterns of patterns…ultimately down to elementary particles or whatever the ultimate reality is. There may be, in principal, a specific application of the Schrodinger equation that can describe the idea of the Olympic Games across the billions of brains and thousands of years of its history, but again, knowing it is forever out of reach.

    Of course, anytime someone makes a distinction between an epistemic and ontological view like I just did, they’re actually describing their theory, their model of that ontology, so that’s my theory. Like all theories, it should be subject to revision on new information. Who knows, maybe someday we’ll discover that there really are some things which emerge in the strong fact-of-the-matter sense.

  31. SelfAwarePatterns: “Epistemically, the idea of the Olympic Games does float, separate from the underlying, say, quantum physics, mainly because we have to shift our models so many times between those layers that attempting to relate them is pointless, first because there is no mind or computational system that could conceivably model it all together, and second because the current inaccuracies and blind spots within the models at each layer would likely snowball and make the resulting uber-model hopelessly wrong.”

    Dennett says a lot of things like this attempting to head off ‘greedy reductionists’ but I’m not convinced that mechanical idioms intrinsically lack the generality required to describe these systems in ‘middle contexts,’ forcing us to resort to homuncularisms. I think we’re looking more at a failure of imagination than anything ‘special’ about our traditional modes of solving systems like games.

    I also find the whole notion of ‘levels’ to be enormously loaded. If we spoke of ‘ways,’ instead, the ontic architectural possibilities are far more open (but then I think the epistemic/ontic is best left behind as well!). It’s a giant Wimsattian mess. Certain heuristic regimes are inter-applicable, while others are not. What advocates for mechanical regimes is simply the scope of applicability (which includes intra-applicability or ‘reduction’).

  32. The idea of the Olympic games, like any other non-physical notion, resides in an arrangement of neurons and their interconnections in you brain. If it didn’t, you would have no representation of it, no knowledge of it and wouldn’t even be able to discuss it. The notion can be communicated to others or observed independently and these others will develop their own representations of it, which might be quite different from yours, but will ultimately contain sufficient similar main constructs to make it the same “idea”.

    Once you realize that the idea must reside in a physical arrangement in your brain, then (for what it is worth) there must be an expression of it described by elementary particles. Nothing mystical to see here. Move on.

  33. Hi Scott,
    On failure of imagination, that’s always possible. But I’m reminded of something Sean Carroll wrote in an earlier piece, pointing out that we can’t use the standard model of particle physics to predict the periodic table of elements. If that’s true, attempting to model the Olympic Games with it seems similar to attempting to build a ladder to Pluto, in the sense that the tools that get us six feet in the air won’t be useful for much beyond that.

    “I also find the whole notion of ‘levels’ to be enormously loaded.”
    I don’t necessarily disagree. The levels are really just one pragmatic epistemic convenience. Other ways of looking at it might be more convenient depending on exactly what we’re doing.

    Incidentally, the convenience of abstraction levels exist even in paradigms where we thoroughly understand each level. It’s not productive to think in terms of transistor voltage states when working on an Excel spreadsheet, even though the workings of the hardware, device drivers, operating system, and application logic are all understood by various engineers and programmers.

  34. @ Stephen:

    “Once you realize that the idea must reside in a physical arrangement in your brain, then (for what it is worth) there must be an expression of it described by elementary particles. Nothing mystical to see here. Move on.”

    But then what you are saying is there is an intrinsic isomorphism between a particular arrangement of particles (bits of quantum foam, whatever is the bottom) and a particular idea.

    Given that’s a revival of Platonism, it seems rather mystical to me? The other alternative, as Rosenberg notes, is to accept thoughts are illusory:

    “A more general version of this question is this: How can one clump of stuff anywhere in the universe be about some other clump of stuff anywhere else in the universe—right next to it or 100 million light-years away?

    …Let’s suppose that the Paris neurons are about Paris the same way red octagons are about stopping. This is the first step down a slippery slope, a regress into total confusion. If the Paris neurons are about Paris the same way a red octagon is about stopping, then there has to be something in the brain that interprets the Paris neurons as being about Paris. After all, that’s how the stop sign is about stopping. It gets interpreted by us in a certain way. The difference is that in the case of the Paris neurons, the interpreter can only be another part of the brain…

    What we need to get off the regress is some set of neurons that is about some stuff outside the brain without being interpreted—by anyone or anything else (including any other part of the brain)—as being about that stuff outside the brain. What we need is a clump of matter, in this case the Paris neurons, that by the very arrangement of its synapses points at, indicates, singles out, picks out, identifies (and here we just start piling up more and more synonyms for “being about”) another clump of matter outside the brain. But there is no such physical stuff.

    Physics has ruled out the existence of clumps of matter of the required sort…
    …What you absolutely cannot be wrong about is that your conscious thought was about something. Even having a wildly wrong thought about something requires that the thought be about something.

    It’s this last notion that introspection conveys that science has to deny. Thinking about things can’t happen at all…When consciousness convinces you that you, or your mind, or your brain has thoughts about things, it is wrong.”

    Admittedly that doesn’t seem very satisfying either?

  35. SelfAwarePatterns: I was imprecise. ‘Levels’ as a handy dandy tool is well and fine, but ‘Levels’ in the ontological sense is very problematic. Their nebulous status seems to provide a lot of cover for a lot of intentionalists–I’m sure you agree! Especially when these levels become ‘autonomous.’

  36. I might have a simplistic view, but my views hold up as long they aren’t challenged.

    “What does ‘down’ even mean here?”

    I don’t agree much on the elaborate explanation about levels of emergence because there’s a much simpler and more intuitive way of describing this that I prefer.

    The fundamental aspect of downward causation is that it marks a distinction and opposition to the standard description of the physical world. The rules of physics as we know them. So it’s a concept that has its core meaning about being the specular opposite. It’s “down” because it opposes the “up”, as the standard description of physics.

    But this basic opposition means we are dealing with another context: a simple dualism. So, the way I see it, “downward causation” is just another descriptor standing for a dualistic model. Two contrasting directions (or vectors, this way, that way, upward and downward). Dualism.

    And I think dualism is born as a concept of another basic pattern, and basic epistemology: in order to perceive a “self” one must make a distinction between self and environment. This action of drawing distinction is what creates the whole paradox contained within dualism: it’s one substance (monism) that gets shaped (through self-reference) to have an inside and outside. A “me” and everything else: substance that belongs to me, internal, and substance that isn’t me, external.

    “they’re the same thing described differently.”

    But yes. If there’s one substance (monism), that gets shaped to form an inside and an outside (dualism), we simply have a model where the one thing (monism) is described in two different ways (dualism).

    (also: one way is “human”, the mind, meaning, the other way is “inhuman” at least in principle, Science, the system as it is described by being hypothetically outside it)

    Therefore: dualism exists while being contained into monism. Same as the downward causation exists as a concept (a description) while being encapsulated within a different model (that is too complex to use reliably = analysis paralysis).

    Downward causation seems to be merely simplification and pattern matching. Meaning written into a non-causal world. But in the end just a compatible description. Downward causation is at the same time compatible, inaccurate and incomplete, but pretty useful.

  37. Tom Clark:
    Higher level regularities don’t have any independent, ineliminable predictive contribution to make about the position of my arm and therefore don’t need to be taken into account to predict the position of the electron in question, even though they might be convenient short-cuts in such predictions. Is this plausible? And does anyone know of a proof of this conjecture?

    The problem is we cannot predict those categories of things. It’s just too complex and we would need a complete model of the world and track every transformation (Laplace’s demon). It’s not that prediction is impossible, it’s just not achievable with the limits imposed on us. But we shouldn’t mistake a limit imposed on us as the nature of the thing we are observing. We are just observing the nature of the observation (epistemology versus ontology).

    Usability (and practice) versus truth.

    Jayarava:
    OTOH in his book he says: “It’s not possible to specify the state of a system by listing the state of its subsystems individually. We have to look at the system as a whole, because different parts of the system can be entangled with one another.” (The Big Picture, p.100)

    In other words systems are irreducible!!!

    well, nope.

    He says that you cannot have a reliable model unless you model the whole thing without cutting just the slice you decided to use. It’s not “irreducible”, it’s, let’s say, holistic. You can reliably describe the system only if the description is complete. Otherwise you have to resort to the pattern matching of high level, emergent, heuristic descriptions. That are BUILT to recognize & match smaller pieces. The “hacks”.

    Tom Clark:
    Now, if we could in principle (if not in practice) predict human behavior without recourse to referring to psychological states, using only knowledge at the micro-level, then we could safely eliminate such states as fundamentally unreal. They are simply convenient explanatory fictions that help us navigate the social world (“ways of talking”).

    But is this “in principle” claim true? If it isn’t, then according to the empiricist’s criterion of what exists, at least some higher level phenomena are just as real as micro-physical phenomena, since we can’t successfully predict or control events at a level or domain (e.g., that of human behavior) without referring to phenomena appearing in that domain (e.g., psychological states).

    Yes, to an “empiricist” the emergent level is REAL. But because the empiricist decides that being trapped into the cave builds his ontology. Or: he puts epistemology before of ontology.

    The point is exactly this one. You won’t be able to obtain a complete model of reality without exiting it (it’s like Godel’s principle, you cannot describe a system while you are inside it, and so with a description that needs to recursively account for itself). Human condition is ontological for a being that exists within the system of reality. But it becomes false if you hypothetically imagine being transported outside the system and observing it (Laplace’s demon’s point of view).

    Downward causation (dualism) is the “ontology” of human beings, but still with the awareness that this “dualism” just describes our life as it is experienced, but not the actual truth of it (monism).

  38. SelfAwarePatterns,

    I agree I think, but my point is if we make a Venn diagram of how the float feels as one circle and ones own theory of the ontology as another circle, how far away is each circle from each other? They are kind of separate – without overlap? And to build upon the theory circle more and more just makes it a bigger circle – it doesn’t overlap the ‘Olympic Games’ feeling by just focusing on the theory. How do we drag the theory of ontology circle over to the Peter’s ‘Olympic Games’? And how could he potentially drag his circle?

  39. #23 Jayareva commented my example was not specific to consciousness: indeed it was a general comment about downward causation and the reality of levels all the way down. One may take this all with a grain of salt, but Crutchfield sees the emergent layers as “real” in that the minimum description of each layer of a dynamical physical system requires an “epsilon-machine” (FSA) of different mathematical complexity (they are all lower than a full Turing machine).

    As to tests of downward causation from consciousness, I just thought of a trial comparing a) meta-cognitive behaviour therapy, b) sham meta-cognitive behaviour therapy, and c) cognitive behaviour therapy for treatment of depression. The meta-analyses I have seen suggest a) and c) are at least roughly equivalent in terms of improvement of symptoms over time – which might be seen as favouring epiphenomenalism. Sham meta-cognitive behaviour therapy would I guess involve getting participants to carry out introspective exercises that we will all agree have no chance of working over the range of currently held models of human consciousness. This might be hard – I am already thinking of the procedure of “paradoxical intention” suggested by Maslow.

  40. Callan:
    I agree I think, but my point is if we make a Venn diagram of how the float feels as one circle and ones own theory of the ontology as another circle, how far away is each circle from each other?

    They aren’t separate. Ontology is the system itself, “float feels” is a circle inside the ontological one.

    One circle inside the other. The trick of this visualization is that it doesn’t explain the relativity of point of view. If you observe from inside the smaller circle, then the smaller circle is ontological.

    The complication is always because one confuses the relativity of the point of view as an irresolvable contradiction (two ontologies that contradict each other).

  41. Callan,
    Not quite sure I grasp what you’re asking, but I’m going to answer it as if you were asking how we derive the ontological from our subjective experience. I think the answer is that the ontological is always a theory, a subjective model of what we think is the reality. The only way we have to judge the model is by its efficacy. If using it as a guide proves effective, then it’s probably accurate to as least some approximation.

    Ideas like the one of the Olympic Games are cultural phenomena, existing in large numbers of minds, which can be studied with sociology. Each of us holds our own particular model of the idea, which is a psychological state. Psychological states are thought to be brain states. Brain states are neural firing patterns, which are molecular machinations, which is chemistry, which is composed of particle physics.

    Of course, there are lots of holes in our knowledge at all these levels, and particularly in between them. But we know enough for the overall model sketched above to be a highly plausible one. For example, many might object to the link between psychological and brain states, but the efficacy of that model for understanding mind altering drugs and other physical effects seems difficult to dismiss. That said, new information could be unearthed tomorrow that breaks or alters the link between any of the layers.

  42. @Sci (34)

    I think where you go wrong is assuming that there is an “interpreter” required in the brain to interpret representations. This is no more true than thinking there needs to be a homunculus to view the theater.

    There are indeed representations in the brain, and I believe the neuroscientists have well established this. I expect that interpreting them is caused by the associations between them and the processing the brain goes through to arrive at conclusions, but I don’t know if anyone understands this process fully enough to provide definitive answers.

    If you don’t believe that anything we think about has a neural representation, then you are truly believing in something mystical as the only other explanation seems to be some sort of dualism.

  43. SCI:
    I’m not sure I’m seeing the difference between holistic and irreducible?

    Oh, come on. What I mean is that to describe and predict the state of a system you need a COMPLETE description. You cannot just isolate a slice of it. But it’s still a reductionist description. It just needs to be complete.

    If you have to predict the movement of gases but THEN THERE’S A STRONG WIND, of course that wind is going to mess the prediction.

    If the description of a system isn’t complete of all its elements then it’s obvious the presence of an unaccounted element can mess the whole thing. But again, this doesn’t make this system non-reductionist.

  44. @Stephen:

    “If you don’t believe that anything we think about has a neural representation, then you are truly believing in something mystical as the only other explanation seems to be some sort of dualism.”

    I’m confused why you said this, since Rosenberg is a materialist whose argument is for eliminativism?

    But yes, neither Platonism nor Eliminativism seem very satisfying. Of course it’s not clear any of the “isms” are.

  45. @ Abalieno

    Glad you were able to clarify what you originally meant.

    I guess I will have to read Sean Carroll’s book, or at least that excerpt. When he mentioned the system being entangled it seemed to be suggesting something more than needing a complete model.

    Also not clear if you are giving your opinion or that of Carroll.

  46. My own. But I think it’s still valid even applied on that quote.

    It might be a slightly different interpretation, but I’d bet the context of that quote isn’t going to be much different. It only applies to the quirks of quantum physics instead of just normal physics that I have used.

    The quote simply said you cannot break down a system into smaller sub systems because they also influence each other. It still means that not counting ALL subsystems in the model makes the model itself unreliable.

  47. I didn’t want to oversimplify but we seem to be arguing the gap here. Particles are the most fundamental level I will assume but biology and then brains have to evolve under special conditions or the environment and paradoxically brains evolve to deal with the environment. The environment seems to be hidden in Carroll’s argument and the comments above.

    David #26; Consciousness is not superfluity when nervous systems are learning but eventually neurons engage in learned behavior without consciousness. We naturally acquire ‘behavioral conceit’ so we can do everything from walk to crack an egg with great dexterity. I think a greater question is can we have language whether spoken or the symbolic which we are presently engaged in on this blog without consciousness? The Jayne’s theory is that consciousness as we know it evolves with language. Many mammals especially primates re-adapt their nervous systems for game engagement right up to the Olympics. Philosophy and resultant sciences are an adaptation of our nervous systems and of course language. Maybe questioning consciousness is the highest game we can play?

  48. SAP in 41:

    “Ideas like the one of the Olympic Games are cultural phenomena, existing in large numbers of minds, which can be studied with sociology. Each of us holds our own particular model of the idea, which is a psychological state. Psychological states are thought to be brain states. Brain states are neural firing patterns, which are molecular machinations, which is chemistry, which is composed of particle physics.”

    On the materialist hypothesis, nothing exists but various combinations of physical phenomena, evolving over time. What I’m wondering about is whether, given only a micro-physical specification of the physical goings-on that, on the materialist hypothesis, constitute your current psychological state and your environment, a sufficiently smart system (a super-physicist, say) could predict your psychological state in 24 hours appealing only to how that micro-physical specification evolves according to physics. Or would it need to know about some higher level biological, behavioral and psychological regularities?

  49. Hi Tom,
    Before answering, let’s consider a slightly different example. Suppose I leave my house to go to work, but get delayed in traffic. If I call my boss en-route and tell her I’m stuck in traffic, she’ll instantly understand what I’m talking about. This appears to be a case of downward causation, since traffic is a phenomena that emerges from large numbers of cars on the road.

    But both me and my boss are getting by with a symbolic trick. We’re substituting the concept of “traffic” for the extremely complex web of causation that is actually happening. The immediate delay in my progress toward work is the slow or non-moving car immediately in front of me. The immediate cause of that car’s delay is the car immediately in front of it, and so on for every vehicle currently between me and my destination. The state of each car on the road actually fixes the current and short term future position of each car. (Well assuming there isn’t a plane crash or loose animal or something.)

    Along the same lines, all the low level micro-physical events fix all the future micro-physical states. We might speak of higher level concepts causing things, such as my fear of an out of control car causing certain neurons to fire, but when we do, we’re speaking imprecisely, just as I was when I called my boss. What is actually causing any one neuron to fire are its current state, the signals from other directly connected neurons, and other factors in its immediate vicinity.

    Of course, we could go even lower and talk in terms of molecular interactions, or atomic ones, or even of subatomic particles. At whatever level, the interactions at that level should fix the future states at that level.

    So, a super-physicist could, in principal, predict the future state of the atoms in my brain in 24 hours. (Quantum indeterminancy could conceivably limit their precision.) If they were also a super-chemist, super-neurobiologist, and super-psychologist, they could presumably predict my future psychological state. Obviously there are a lot of “supers” here, to the point that we’re approaching the territory of Laplace’s demon.

  50. “Reduction” is an illusion at best, nonsense at worst. Look at basic biology – “sexual reproduction” has a meaning above and beyond the motion of the constituent particles of which organisms are formed. The language of physics would never lead to a concept like “sexual reproduction”. Contrast this with – let’s say – the surface tension of a body of liquid, which is entirely reducible to particulate equations.

    In fact the failure of physics to cope with basic biology is fundamental. The notion of “organism” is not predictable from physics(hence – if it is to be meaningful in any way – not reducible to physics). “Organism” has semantic meaning to humans outside of mere physical metrics – but physics is a semantic closed circle, so any non-physical concept – gender, sexuality, society, post-industrial social chaos – is not reducible to physics.

    It’s another example of the cognitive limitation that humans placed upon physics when they created it. They can extract the basic physical parameters of a system but the interesting bits it can’t cope with.

  51. On the basis of Sean Carroll’s website posts (I haven’t read the book yet), I’m going to agree with Peter and especially Stephen (#27). And disagree with Tom Clark (#25) and Jayarava (#16).

    If “what *really* exists is just that which is necessary for successful prediction and control in a given domain of explanation” (#25), then psychological states are real, because there is no way to explain human behavior without them. That is a domain of explanation (which might be a better label than “levels”) we are fully invested in, so there’s no question of discarding it in favor of exclusively microphysical talk.

    That might be one reason to take talk of “ontologies” seriously, and not read it as “epistemes”. But an even better one is that it just works, and there’s nothing stopping us. Following Stephen #27, we can have our trees and our clouds-of-electrons-and-quarks too. Heterogeneous ontology worries Peter (slightly), but I think the solution is to just stop worrying. There are no gotchas lurking here.

  52. Paul Torek:
    If “what *really* exists is just that which is necessary for successful prediction and control in a given domain of explanation” (#25), then psychological states are real, because there is no way to explain human behavior without them.

    This is once again a false assumption.

    Just because there aren’t better ways available right now doesn’t mean this is some fundamental truth. There aren’t better ways because those ways are limited by human knowledge, but it would be stupid to think this is a proof.

    There aren’t better ways to predict the weather, so the thunderstorm must be a god who’s very upset with me. But today we do have better models than that. Science is the tool that opens those roads so that we can leave behind our rough cartoons. “Psychological states” are cartoons we use in absence of more detailed insight, but still cartoons, and still far, far away from a more fundamental truth.

    You have to rephrase it: accordingly to the computation accessible to a human being right now, “psychological states” are useful. They “exist” because we can’t access better ways. But this doesn’t exclude the EXISTENCE of better ways. It just draws the limit of the pragmatic knowledge we access at any given time.

  53. SAP in 52, thanks. You say “all the low level micro-physical events fix all the future micro-physical states…So, a super-physicist could, in principal, predict the future state of the atoms in my brain in 24 hours.”

    I wonder if this has been proven. I’m suggesting that it’s perhaps the case that higher level organization and structure, as described by laws above the micro-physical level, constrain the evolution of micro-physical events such that you’d need to know those laws in order to predict future states of affairs such as the state of your brain in 24 hours. Has this been categorically ruled out, and do you know of the proof?

    We can’t have a psychological difference without a physical difference, but whether in principle we only need to know micro-physical laws to predict future psychological states is another question. Do you know if there’s a definitive answer, and if so, could you point me to it?

    Paul in 54: “If ‘what *really* exists is just that which is necessary for successful prediction and control in a given domain of explanation’ (#25), then psychological states are real, because there is no way to explain human behavior without them.”

    I’m wondering if it’s been proven that in principle it’s impossible to predict my behavior in the next 24 hours without using laws above the micro-physical level. (Same question as I posed to SAP, just reversed.) If it’s impossible, that suggests that higher level ontologies are just as real as the micro-physical, not just “ways of talking” ala Carroll’s poetic naturalism.

  54. Tom,
    Obviously there is no such thing as super-scientists, just the plain old fallible ones with human biases and poor research budgets. No one is going to be using particle physics to predict my future psychological state, probably ever. For that matter, no one’s going to be using physics to explain why my iPhone hangs.

    But Sean Carroll has written many times that the particle physics of everyday matter and energy, such as the matter and energy in our brains, are completely understood. (Note that he’s talking about everyday physics, as opposed to the more exotic stuff like dark matter, dark energy, black holes, etc.) And that understanding doesn’t leave room for anything else to intervene and change the results. I’m not a physicist, so I have to take his word for it, but it fits with my layman understanding of the situation.

    But while physicists can say that, and maybe chemists to some degree, neurobiologists certainly can’t yet. Many of the molecular mechanisms in neurons and synapses are not well understood. And until someone can point to a particular neural firing pattern and identify it with a particular thought or idea, there will remain space for all kinds of unexpected discoveries.

    That said, I think the probability of downward causation, in the sense that we can’t predict the lower level physics without reference to the higher level concepts, is remote. I might feel differently if I knew of anywhere else in science where that had been demonstrated. (Again, we have to be careful of the imprecise language people, including scientists, often use. I might casually say that a hurricane knocked down my tree, but the precise immediate cause was the wind pushing against the tree for an extended period of time.)

    What’s much more likely is unexpected *upward* causation, such as a brain’s sensory image map processing changing in unexpected ways due to some poorly understood mechanism in synaptic vesicles or other molecular machinery. It’s already not controversial that mental states can be swayed by drugs, poor nutrition, or other factors mucking up these lower level processes.

  55. SAP 57: “Sean Carroll has written many times that the particle physics of everyday matter and energy, such as the matter and energy in our brains, are completely understood. (Note that he’s talking about everyday physics, as opposed to the more exotic stuff like dark matter, dark energy, black holes, etc.) And that understanding doesn’t leave room for anything else to intervene and change the results.”

    Agree that there’s nothing non-physical that could intervene, but again that’s a different question from whether we only need micro-physical laws to predict all future states of affairs.

    “That said, I think the probability of downward causation, in the sense that we can’t predict the lower level physics without reference to the higher level concepts, is remote. I might feel differently if I knew of anywhere else in science where that had been demonstrated.”

    Carroll writes in the blog post about this:

    “The dynamical rules of the Core Theory aren’t just vague suggestions; they are absolutely precise statements about how the quantum fields making up you and me behave under any circumstances (within the “everyday life” domain of validity). And those rules say that the behavior of, say, an electron is determined by the local values of other quantum fields at the position of the electron — and by nothing else. (That’s “locality” or “microcausality” in quantum field theory.) In particular, as long as the quantum fields at the precise position of the electron are the same, the larger context in which it is embedded is utterly irrelevant.

    “It’s possible that the real world is different, and there is such inter-level feedback. That’s an experimentally testable question! As I mentioned to Henrik, it would be the greatest scientific discovery of our lifetimes. And there’s basically no evidence that it’s true. But it’s possible.”

    So according to Carroll it’s an empirical, testable question as to whether the larger context of the electron as described by higher-level laws is necessary to take into account to predict its position in 24 hours. Which is to say there’s no armchair proof of the conjecture that we only need micro-physical laws to predict psychological states (which supervene on specific configurations of many trillions of electrons and other micro-physical phenomena, both in and outside the subject). As for evidence in favor of the conjecture, I don’t know of any examples of predicting psychological states from micro-physcial states. So I’m not sure where the confidence in the conjecture comes from.

    If you watched the great video series on the “Moving naturalism forward” meeting that Carroll hosted, you’ll know I’m taking Terry Deacon’s side in this debate, recap at http://www.preposterousuniverse.com/blog/2012/12/11/moving-naturalism-forward-videos-and-recap/

    Carroll says in the recap that “Most people in the room seemed to be willing to use a language of emergence and accept that higher-level descriptions had a kind of autonomy (you don’t need to know lower-level details to understand higher-level truths), but they also seemed to accept that higher-level happenings were in some sense entailed by lower-level happenings. Personally I prefer to think of parallel vocabularies rather than lower and higher levels, but it’s clear that some vocabularies are more comprehensive than others.”

    So maybe Carroll is an equal opportunity ontologist after all, with the obvious proviso that the micro-physical is the common denominator of all physically realized ontologies.

  56. Tom,
    Thanks for pulling all that together. I did watch some of that conference and read Carroll’s recap, although it’s been a while.

    “but again that’s a different question from whether we only need micro-physical laws to predict all future states of affairs.
    Actually, I think that’s exactly what Carroll meant in this snippet:
    “In particular, as long as the quantum fields at the precise position of the electron are the same, the larger context in which it is embedded is utterly irrelevant.”

    I don’t interpret Carroll being nearly as agnostic on this as you do. Notably:
    “it would be the greatest scientific discovery of our lifetimes. And there’s basically no evidence that it’s true. But it’s possible.”
    Which generally fits my own sentiment: skeptical but open to new evidence.

    If I recall correctly, Michael Gazzaniga in his book ‘Who’s in Charge?’ explored downward causation. I didn’t find it convincing, but it was a fascinating discussion.

  57. SelfAwarePatterns:
    Obviously there is no such thing as super-scientists, just the plain old fallible ones with human biases and poor research budgets. No one is going to be using particle physics to predict my future psychological state, probably ever. For that matter, no one’s going to be using physics to explain why my iPhone hangs.

    Well said.

    As Bakker uses to say, human beings are “shallow information consumers”. And the feature of being that is they have to do with a very limited access to information.

    So, again, the need for human beings to rely on non-reductionist models is exclusively because a full reductionist model is way, way too complex to handle (analysis paralysis, the opposite of being “functional”). Not because it’s not theoretically possible.

    We are BUILT to remain functional with very limited information, and built to NOT have a full, direct access. So it’s ontological that a full reductionist approach will never be possible (Laplace’s demon).

    Tom Clark:
    Carroll says in the recap that “Most people in the room seemed to be willing to use a language of emergence and accept that higher-level descriptions had a kind of autonomy (you don’t need to know lower-level details to understand higher-level truths), but they also seemed to accept that higher-level happenings were in some sense entailed by lower-level happenings. Personally I prefer to think of parallel vocabularies rather than lower and higher levels, but it’s clear that some vocabularies are more comprehensive than others.”

    This is essentially what Bakker calls “correlations”.

    We have two different descriptors, one high level, one low level. The high level one is imprecise because it’s just an heuristic using the limited information we have. A cartoon. A generalization. Whereas the other is a precise, complete description of the full physical state, that we can only theorize since it’s too complex and impossible to even remotely compute. Of course these are “parallel” vocabularies.

    They are the same thing. One is a shallow information description, the other is accurate.

    The high level one is an approximation that we are forced to use because we cannot deal with the “true”, high-density level. So we resort to use a shallow information level that is “correlated” or “parallel” to the lower one.

    It’s like looking at a map and call a thing “America”. Well knowing that the idea of America is built of a number of smaller parts. This is the skill human brains have, to reduce and wrap complexity into a simpler idea.

  58. Tom

    It’s definitely been predicted that your behaviour cannot be predicted. . Even in a newton in framework, never mind quantum.

    As you have such a humungous number of particles, it would be impossible to know the initial states if all of them simultaneously, as a matter of practicality. This would be essential to map all the future states of the system.

    As another matter of practicality, all particles are governed by the uncertainty principle, which states that the more accurately you can tell a particle’s position, the less accurately you can tell it”s momentum. Thus the current state of a system is undetermined as a matter of physics – never mind the one in 24 hrs.

    Describing these states as ‘quantum fields’ doesn’t liberate the uncertainty. It’s simply not possible to predict what of the many states a system in 24 hrs will be, although their likelihood is. I get the impression that Carroll is trying to insert some old-style newtonianism into quantum mechanics by stating that although the future of particles might not be definite, their wave functions are. Nice try i think but quantum fields are a mathematical exercise, you’ll never find them in nature.

    It’s certainly true I think that quantum mechanics is deterministic so I can’t disagree with him if that’s his intention. But if mental states are linked to particle states then quantum field theory is still no use because the only thing we can extract from quantum theory is meters, seconds and kilogrammes.

    It doesn’t matter if matter does ‘consist’ of particles : the problem is that physics ‘consists’ only of the physical measurables. If matter is doing something more than shuffling physical metrics about – which seems to be the case – then physics is of no help whatsoever.

    J

  59. Pretty sure reply 41 and 42 represent the circles that do not overlap, one right after the other 🙂 Where does one agree with the other?

    Abelieno in 41,

    They aren’t separate. Ontology is the system itself, “float feels” is a circle inside the ontological one.

    One circle inside the other. The trick of this visualization is that it doesn’t explain the relativity of point of view. If you observe from inside the smaller circle, then the smaller circle is ontological.

    The complication is always because one confuses the relativity of the point of view as an irresolvable contradiction (two ontologies that contradict each other).

    I kind of read this with the visualisation of a flat plane, viewed from it’s side, and then a mound rising from it that is actually speculation on the nature of the flat plane. To me your approach with the ‘irresolvable contradiction’ seems like the attempt in action to flatten the mound and make it lie flush with the plane. The attempt to have no contradiction between the plane of what is and the mound of what seem be. That’s why you say you have two ontologies when really ontology is defined as dealing with ‘the’ state of being, not states plural.

    And further complicating it, by that measure, it’s really three ontologies. One inside the other – and the conception of this is a third, via recursion. And every attempt to step back to see makes a further recursion and with it another ontology.

    So it’s ontological that a full reductionist approach will never be possible (Laplace’s demon).

    I think we’re probably inclined to feel that the smallest movements of our mind are the lowest resolution possible, when really they might be quite wide brushstrokes and reductionism can actually reduce to below the width of the brush stroke of a soul. In other words sufficiently crude behavior is indistinguishable from full reductionism – like a high resolution TV showing the big blocky pixels of a commodore 64 screen, the TVs own pixels might not reduce things down fully, but they are more than enough to reduce and reconstruct the commodore 64’s graphics for being smaller than the brushstrokes of the 64.

  60. Callan S.:
    I kind of read this with the visualisation of a flat plane, viewed from it’s side, and then a mound rising from it that is actually speculation on the nature of the flat plane. To me your approach with the ‘irresolvable contradiction’ seems like the attempt in action to flatten the mound and make it lie flush with the plane. The attempt to have no contradiction between the plane of what is and the mound of what seem be. That’s why you say you have two ontologies when really ontology is defined as dealing with ‘the’ state of being, not states plural.

    I’m not sure I understand this, but not really.

    If you imagine two circles, one inside the other, then it means you can define an inside and an outside. But it’s also obvious that for an observer within the larger circle the smaller circle represents the “outside”, whereas if the observer is within the smaller circle then the “outside” is the larger one.

    Relativity.

    The same can be done for the ontology. “Truth” of the system is relative to the observer. If the observer is bound by perspective to the smaller circle (the fictional rule), then the smaller circle becomes its truth. You can theorize there’s a foundational, deeper truth of the system outside, but being unable to go there, it remains speculation: Laplace’s demon.

    So ontology remains SUBJECTIVE. Relative to a point of view. Of human beings in this case.

    (though this line of thought is more important in the discussion of Free Will instead of “reductionism versus downward causation” that we’re discussing here)

    I think we’re probably inclined to feel that the smallest movements of our mind are the lowest resolution possible, when really they might be quite wide brushstrokes and reductionism can actually reduce to below the width of the brush stroke of a soul. In other words sufficiently crude behavior is indistinguishable from full reductionism – like a high resolution TV showing the big blocky pixels of a commodore 64 screen, the TVs own pixels might not reduce things down fully, but they are more than enough to reduce and reconstruct the commodore 64’s graphics for being smaller than the brushstrokes of the 64.

    We’ve briefly discussed the problem of reductionism above.

    Reductionism is ontologically opposite to (human) experience (the two circles discussed above). If human experience is filtered through language (Richard Rorty), then it means we “know” the world by dividing it into parts. But reductionism instead only works if we account for the whole, complete system. From the point of view of reductionism/foundationalism there isn’t a part of the system that is more important than another, so you cannot divide and focus of an area to circumscribe its description. It’s all or nothing. At the fundamental level a rock is like a brain, and there’s continuity between them until there’s no distinction between a brain and a rock. It’s all one system entangled together to form a complete description accounting for all laws of physics and particles playing in it.

    They are opposite forms of epistemology (and so suggest a dualistic model, or dualistic description of the same “monism”). One works by dividing, the other by using an holistic, complete and closed model.

    Also: you cannot use reductionism in non-reductionist ways. This is a very explicit contradiction in most comments here. You cannot predict the behavior of a single human being, or brain, through reductionism. Why? Because the reductionist level cannot be conveniently sliced into parts. A reductionist model predicts THE SYSTEM. Not an isolated part of the system. If you start dividing the system you are back into language, and stepping away from that foundational level.

    They are mutually incompatible descriptions that cannot be used together. If you do, you create a contradiction (the hard problem).

    If you use reductionism to try to predict a part of the system you end up using “upward causation” AT THE SAME TIME of “downward causation”. That’s what Carroll warned about, using at the same time conflicting descriptions, creating a paradox.

    But yes, as you say this dichotomy and dualism isn’t a binary position. Again, we are shallow information users, but this doesn’t negate us the possibility, through Science, to integrate more and more information. We cannot go to the bottom (foundational level), but we can go deeper.

    So, while we can never reach the ultimate reductionist level, we can at least increase the complexity of the language to better reflect the complexity of the fundamental reality (this is Richard Rorty again).

    Taken from the wikipedia:
    Postmodernists and post-structuralists such as Richard Rorty and Jacques Derrida have attacked foundationalism on the grounds that the truth of a statement or discourse is only verifiable in accordance with other statements and discourses. Rorty in particular elaborates further on this, claiming that the individual, the community, the human body as a whole have a ‘means by which they know the world’ (this entails language, culture, semiotic systems, mathematics, science etc.). In order to verify particular means, or particular statements belonging to certain means (e.g. the propositions of the natural sciences), a person would have to ‘step outside’ the means and critique them neutrally, in order to provide a foundation for adopting them. However, this is impossible. The only way in which one can know the world is through the means by which they know the world; a method cannot justify itself. This argument can be seen as directly related to Wittgenstein’s theory of language, drawing a parallel between postmodernism and late logical positivism that is united in critique of foundationalism.

    “Stepping outside” is exactly the formulation of reductionism we use here. The Laplace’s demon. And it’s true that it’s “impossible”. We cannot “breach” this formulation of dualism. Same as we are never “outside” the system we live in, so that we can account for all its deterministic representation. Once again Godel: since we are part of the system we cannot obtain a complete description that can close it.

    But this also doesn’t mean we cannot integrate more and more information to get closer to this fundamental level. Same as today we have means to predict the weather, approximately, without resorting to gods.

    (the only problem with your quote is that you cannot give a working description of a brain without also accounting for all the environment around it. the problem of “enactivism” that Bakker discussed a bit on his site)

  61. Pretty sure reply 41 and 42 represent the circles that do not overlap, one right after the other ? Where does one agree with the other?

    The contradiction between my comment and his is because of this:

    I think the answer is that the ontological is always a theory, a subjective model of what we think is the reality.

    He puts ontology (truth) after epistemology. Epistemology comes first.

    That means that he speaks from the point of view of the inner circle, where the “fundamental” truth is determined by experience (same as constructivists or phenomenology).

    It’s, if you want, the solipsistic perspective.

    I instead framed the problem by accounting for both, saying it’s relative to the observer (hence the perceived dualism).

  62. Very interesting discussion because we can deduce that biological cells even as one cell organisms are complex individual systems but resemble physical particles because they deal with an immediate environment that we call a higher environment. From a ‘panorganism’ perspective even physical particles are live organisms.

  63. Three thoughts:

    1- The subtitle of this blog. If we should be describing all of reality as interactions between particles, and anything else is a simplifying tool, who is doing the simplifying? It must be valid for me to reify some set of particles and their interactions into the Olympics, or else I am also rejecting the reification that is me. Cogito ergo valet reification. (Don’t blame me, blame Google translate.)

    2- Before reading the discussion, in the same sitting, I read a Notre Dame book review of Locke, Hume, and the Treacherous Logos of Atomism: The Eclipse of Democratic Values in the Early Modern Period by Robert Roecklein. (Roecklein wrote the book, not the review.) The idea is that the belief that the ultimate bits of the universe are imperceivable atoms leads to deprecating the experience of the common person which in turn leads to the political disenfranchisement of the common man. And, that this actually occurred among early modern philosophers. I found the whole thing a bit of a stretch, but then, I only read the review, not the author’s full argument.

    And then I got to some of the posters here, who are literally using atomism to dismiss the reality of the existence of those people altogether.

    3- Yes, it is true that a certain set of electrons experienced a voltage jump that allowed them to jump valence levels, causing a set of particles chemists would call a crystal in what I like to call my LCD to twist and thereby exposing…. Just as it is true that said pixel turned red because it is displaying part of a digit in a spreadsheet showing my projected back balance. (sigh)

    One needn’t be a Cartesian dualist to bequeath reality to the structures of things rather than the parts. Saying something is an emergent property isn’t dismissing it. It’s not just the parts (particles) it’s also the relationships between them.

    This dialectic between reductionism and holism was described quite well by Hofstadter back in the 80s, and I’m not sure we have that much to add.

  64. It’s not just the parts (particles) it’s also the relationships between them.

    Nope, it’s just the parts.

    The “relationships” are already high level concepts existing because of simplification that *we* need to understand the system on a broader level.

    The system itself (the particles) don’t know of belonging to a sub-system or another, or having an influence from here to over there. These are high level observations.

    So it’s either one (particles) or the other (relationships). Once again we should make sure to not mix different level descriptors.

    The idea is that the belief that the ultimate bits of the universe are imperceivable atoms leads to deprecating the experience of the common person

    The exact opposite.

    It’s because the ultimate bits are imperceivable that personal experience becomes ontologic (true).

    There’s no other way to be. If the ultimate bits were possible to achieve, then it would be seen as a journey from here to there. From illusion to truth. But since this isn’t possible, the illusion becomes factual truth.

  65. Abelieno: Are you actually saying that one can understand a system by knowing its parts without knowing how they are positioned and how they interact?

    Again, I point you to the subtitle of this blog: If any description about the ultimate reductionist one is an illusion, who is experiencing that illusion?

  66. Abelieno: Are you actually saying that one can understand a system by knowing its parts without knowing how they are positioned and how they interact?

    “Knowing its parts” includes their state and their laws. It’s all one thing. A distinction between laws, positions and particles are still high level descriptions, so close to a fundamental level, but not quite there.

    The subtitle of the blog plays too with the confusion of two incompatible perspectives. It’s always the same pattern.

    consciousness being illusion = reductionist description (consciousness is explained away)
    being fooled = high level concept

    Should I really keep repeating this over and over and over? We shouldn’t confuse descriptors because these descriptors are mutually incompatible.

    If you keep using them both you just reproduce that basic contradiction over and over.

  67. Abalieno,

    If you imagine two circles, one inside the other, then it means you can define an inside and an outside. But it’s also obvious that for an observer within the larger circle the smaller circle represents the “outside”, whereas if the observer is within the smaller circle then the “outside” is the larger one.

    Not sure why the inner circle becomes the outer circles ‘outside’? I mean, you treat the second circle as the outside to the inner circle it surrounds. All we’d have to do is add a third circle around the second circle and surely that third is the outside to the second circle – rather than the first circle being the outside to the second?

    Maybe something inner is always taken as being ‘the outside’? Even the inner circle takes something smaller inside it as ‘the outside’. That’d be more consistent, at the very least.

    From the point of view of reductionism/foundationalism there isn’t a part of the system that is more important than another, so you cannot divide and focus of an area to circumscribe its description. It’s all or nothing.

    Seems to deny our estimating/heuristic nature. We live in estimates, not absolutes. People will use reductionism in an estimate form rather than an absolute/all form. Saying ‘you can’t do that!’ is just going to be not being along for the ride. It’ll be a bulwark for ever decreasing specialist groups who have severed their connection to the greater community (the greater community who are using an estimate version of reduction). People can use a wrench as a hammer – saying they can’t wont stop them driving nails into coffins.

    That means that he speaks from the point of view of the inner circle, where the “fundamental” truth is determined by experience (same as constructivists or phenomenology).
    It’s, if you want, the solipsistic perspective.

    I’m not sure why you read SAP that way? You were saying he was talking in terms of fundamental truth?

  68. [a quick tear for posts lost to the internet ether][I restored it…. Peter 🙂 ]

    Abalieno,

    Just a post on one thing (as I lost my post on many things) just to check my posts are getting through (have had trouble with that before)

    But it’s also obvious that for an observer within the larger circle the smaller circle represents the “outside”, whereas if the observer is within the smaller circle then the “outside” is the larger one.

    That doesn’t seem terribly consistent? I mean, you have the second circle treating the first circle as the outside, but the inner circle treating the second circle as the outside? If we added a third circle that surrounds the second and first, wouldn’t it be more consistent for the second circle to look out toward the third as it’s outside?

    Or another form of consistency, the second circle treats the inner circle as the outside – and the inner circle treats something in it as the outside.

  69. Not sure why the inner circle becomes the outer circles ‘outside’? I mean, you treat the second circle as the outside to the inner circle it surrounds. All we’d have to do is add a third circle around the second circle and surely that third is the outside to the second circle – rather than the first circle being the outside to the second?

    I had a similar problem explaining this to another guy. For some people it’s immediate and for others it’s not.

    Anyway, it’s just a way to point out the relativity of a concept in respect to the point of view of an observer. Another example is that if you are on a train moving, you can see yourself (and the train) as standing still, while it’s someone standing outside that is moving away. But if you’re outside it’s the train and the guy on it that are moving.

    Or you simply draw a line, a demarcation, and you consider the “inside” the place you are, and the outside the other, but if you move over, they swap. They are all concepts relative to the position of the observer.

    It’s like certain Escher drawings, where something goes up, something else goes down, but you turn the drawing around and they swap. It’s all relative.

    Seems to deny our estimating/heuristic nature. We live in estimates, not absolutes.

    Yes, the “we” defines an opposite level to a bottom-up reductionist fundamental level. Your human idea of “reductionism” is just a rough approximation of how the reality truly operates, so it’s still an high level concept that tries to get closer to fundamental reality. But it’s still human language, and so it’s still far away from that fundamental reality and its accurate description.

    Basically “reductionism” is a placeholder idea of a description of reality we simply cannot achieve. It’s an hypothesis that is very vague, because we cannot truly define it.

    I already stated two basic point. One is that you won’t even be able to reach a fundamental description of reality, not even in thee future. But on the other side language CAN become more accurate and progressively approach a fundamental level. But without reaching it.

    I’m not sure why you read SAP that way? You were saying he was talking in terms of fundamental truth?

    No, the opposite. The inner circle is the human one. And as I said human beings cannot reach any fundamental truth.

    The fundamental truth, for an human being, is solely an hypothesis. It defines “something”, like a box, but you cannot describe what the box contains, because I said we cannot reach that level. We imagine there’s a fundamental truth, some basic laws of physics that are complete. But we also know it’s not something we can grasp (we can go closer, though) It’s an hypothesis today and it will be an hypothesis forever because it cannot be verified. We cannot go there.

    Hence subjective reality becomes ontology. Because the fundamental reality can never step in to invalidate subjective reality. It remains a remote utopia.

  70. Tom (#56),

    “I’m wondering if it’s been proven that in principle it’s impossible to predict my behavior in the next 24 hours without using laws above the micro-physical level.” My contention is “merely” that it’s impossible in practice – and is extremely likely to remain so. For one thing, to predict 24 hours ahead we would need to know all the conditions in a 24-light-hour sphere around you. Ouch, that’s most of the solar system. Then there’s the computations.

    The rule “it’s real iff it’s necessary for successful prediction” is already too narrow even if it only means “necessary in practice”. Taking that rule with “necessary in principle” would be folly. Theoretical dualities pose a dilemma for it.

    But I guess that’s not you’re concern. You just wanna know, has downward causation been ruled in, out, or neither? On that, I’ll just echo SelfAware (#59): it hasn’t been ruled out, but I’m skeptical. It would be a major complication – and simpler theories are more credible, when both fit the available data.

    Abelieno (#55),
    I’m not trying to make anything a “fundamental truth”. Truths are truths, no need to be fundamentalist about them. I actually agree with much of what you say in #63.

  71. Paul: “You just wanna know, has downward causation been ruled in, out, or neither? On that, I’ll just echo SelfAware (#59): it hasn’t been ruled out, but I’m skeptical. It would be a major complication – and simpler theories are more credible, when both fit the available data.”

    Intuitively, it seems to me that higher level laws of say biology and behavior would constrain the location of the sub-atomic particles of an organism in a way that makes their location unpredictable just using microphysical laws. From whence the confidence that they *would* be predictable just using those laws, I wonder?

  72. Tom,
    Why “in a way that makes their location unpredictable using just microphysical laws”? I assume you mean “in principle” since seem less interested in unpredictable-in-practice. Bear in mind that laws of behavior seem only to be statistical, so their constraints are loose.

  73. Thanks for the retrieval, Peter! I think there’s possibly philosophy to be had in the lost post – the feeling of loss and how a rewrite just looses the vigor of the first and certainly in no way can we duplicate it at a near word for word level. Why do words ‘dry up’ the second time we try to write them? That said I’m taking a copy of this post… 🙂

    Abalieno,

    I guess I just don’t immediately get why the first, inner circle would treat the second circle as the outside, but the second wouldn’t do the same and treat a third outer circle as its outside?

    And it would seem to fit your idea much better that the first circle looks to something inside it as it’s outside as much as the second circle looks to the first as its outside.

    I mean, the second circle was looking to the first as the ‘outside’ because it was ignorant of a third circle around it, right?

    On reductionism you originally said ‘it’s all or nothing’.

    But what if people start using impure reductionism – and it gets results in the world? Advertising campaigns center around an impure reductionism of human beings and human behaviour – and become markedly more successful? Does ‘It’s all or nothing’ matter when the impure reductionism is changing the world?

    It defines “something”, like a box, but you cannot describe what the box contains, because I said we cannot reach that level.

    But you already have in part reached that level, surely? By definitely knowing you can’t? That “can’t” is on the same level as that fundamental truth.

  74. Callan:
    I guess I just don’t immediately get why the first, inner circle would treat the second circle as the outside, but the second wouldn’t do the same and treat a third outer circle as its outside?

    And it would seem to fit your idea much better that the first circle looks to something inside it as it’s outside as much as the second circle looks to the first as its outside.

    Because it’s an example in the context of bottom-up model of upward causation versus downward causation of conscious minds. It’s binary. These are two directions and my example analyzes a single system of reality from within. There can be a theoretical “outside” to this system representing a third circle, but it’s out of context of the example as we don’t question here what could exist outside our reality.

    I’ve simply drawn a demarcation within a system to make a differentiation within the system itself. It’s like drawing a circle in the sand on a beach. So now you have the larger system of the beach made of sand, and a small circle still made of sand representing a smaller system. Since we talk epistemology within this system (reality/beach) there’s no need to take into consideration what could be outside the system of reality, like the sea.

    So we have a subjective point of view, from within the smaller circle, building an idea of “self” thanks to the new boundary I just drew. And then there’s a description from the outside, the larger system. The third person impersonal description. From within the system of reality there cannot be any other point of view, because you cannot exit reality.

    On reductionism you originally said ‘it’s all or nothing’.

    But what if people start using impure reductionism – and it gets results in the world?

    I stated both in different contexts. I said a complete reductionist description (foundationalism) is impossible to reach for human beings stuck this side of reality.

    But I also said human knowledge can get closer and closer to that foundational level. So of course “impure reductionism” can work. Same as all science can have a dramatic impact. But it’s always not enough to reach the foundational level. It’s a lot but still not everything.

    Paul Torek:
    Truths are truths, no need to be fundamentalist about them.

    Nope, because truth is relative to point of view, and if that point of view can’t be surpassed, being ontologically closed, then subjectivity CAN be considered a truth.

    There are many different levels of truth, exactly because human knowledge is scattered across many levels. Every semantic level has its contextual truths.

Leave a Reply

Your email address will not be published. Required fields are marked *