Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Top-down vs. bottom-up causation

  1. Mar 14, 2009 #1
    What is your preferred view of causation? There seems to be more and more talk about top-down causation in terms of the consciousness. What do you think, is top down-causation a viable interpretation in regards to current physics?
  2. jcsd
  3. Mar 14, 2009 #2
    I'd say the jury is still out, mainly because our notion of causation is as vague as our notion of consciousness. I don't think 'physics' will give us a definitive answer. Although there are wonderful things happening in neuroscience of late.
  4. Mar 15, 2009 #3
    I wonder if a top-down causation interpretation of QM would support Bohmian mechanics? Does any body have an opinion on this? JoeDawg?
  5. Mar 15, 2009 #4
    Not sure what you mean by 'top-down' in relation to QM.

    On the level of consciousness, its a very touchy subject. If it does occur, my understanding would be that it has to do with information structure and/or complexity affecting the behaviour of individual neurons. Like how dancing with a partner impacts both individuals or playing in a symphony affects individual musicians.

    As to the physics, I suppose this might be in line with the nonlocality of entanglement, but the math is way beyond me, and collapsing wavefunctions appeal to the anarchist in me.
  6. Mar 15, 2009 #5


    User Avatar
    Science Advisor
    Homework Helper
    Gold Member

    I'd have to disagree that our notion of causality is vague. Certainly at a classical level, our notion of causality is quite specific. We understand fluid dynamics (Navier Stokes), we understand classical thermodynamics, electromagnetic interactions, etc... In each case, at a classical level, there's a fundamental premise - a philosophy of how these things interact. The interactions are causal, but they are also local. These local/causal interactions don't depend on other interactions some distance away, they are independent of nonlocal interactions. Any causal, local interaction must propogate through space and other mediums at a velocity dependant on the physical interactions.

    Finally, the local interactions are 'indistiguishable'. By that I mean that given two identical volumes of space with identical physical states, then a given causal action operating on either of those two identical volumes of space will produce the same change in physical state.

    What this says is that each interaction within any classical mechanical system is dependent only on the local interaction and the causal action will produce a reaction dependent only on that particular causal, local interaction. This is "bottom-up" causation.

    There have been attempts to dethrone this philosophy, but they really aren't mainstream as near as I can tell. They are attempts such as Alwyn Scott for example, who has proposed that any non-linear physical system is in some way 'emergent' and therefore, subject to 'top-down' causation. I think this non-linear attack on bottom-up causation fails however as all non-linear systems can be broken down into finite regions of space and the non-linear phenomenon can be largely duplicated. An extreme example of this non-linear attack is by Robert Bishop (See "Downward Causation in Fluid Convection") in which he argues that Rayleigh-Benard convection is an example of a nonlinear system that supports downward causation (ie: top-down causation). Yet here we are as physicists and engineers using conventional FEA approches (ie: bottom-up approaches) to calculating the phenomenon of Rayleigh-Benard convection, which we can do quite successfully.

    The only reputable 'top-down' approach to causation I'm aware of is at a quantum level. Quantum mechanics is fundamentally holistic in a way that classical mechanics is not.
  7. Mar 15, 2009 #6
    We certainly use the idea of causality. But the 'problem of induction' still means we have no rational justification for it. What really occurs is that we experience 'correlation' and infer causation.

    Its really the same way with locality, IMO. On the everyday level we experience locality.
    Well thats one model of causality, but it doesn't seem to be the case when you have very small 'volumes' of space. What I mean is, if we don't have an understanding of causality that applies to QM and our more classical world... then we really don't have a good understanding of it, we just have a bunch of consistent observations.
    Yes and scientist types tend towards a reductionist mentality, because the devil is in the details.
    And this is the problem in my view, it works when it works. And it doesn't when it doesn't.
    And conciousness seems to work both ways.
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook