Serendip is an independent site partnering with faculty at multiple colleges and universities around the world. Happy exploring!

Irreducibility without dualism: chaos or indeterminacy?

Paul Grobstein's picture

Interesting discussion in the emergence group this past week, based on a presentation/paper by Mikio Agaki to be continued next week.  Here's my my read of what Mikio is about, why it matters to all of us, where I am currently worried he may get into trouble, and what the implications are for the future.

Mikio's primary concern is how to deal with the relation between mind and brain.  This though is a specific case of the general problem of how to account for "emergent" phenomena "synchronically" in the terms Alan Baker defined last week, ie to account for a property (in this case "mind") that does not exist in any of the parts of the system (neurons or sets of neurons less extensive than the entire brain).  As such, Mikio's effort is directly relevant not only to those of us who happen to care specifically about the relation between mind and brain but also to those who instead (or in addition) want to defend the proposition that analyses at some particular level of organization (eg economics, social science, etc) are not "reducible" to analyses at some lower level of organization, ie that someone interested in social organization can make useful contributions to understanding that could not, in principle, be deduced logically from studies at a lower level of organization (because neither economics nor sociology nor mind can be "reduced" to physics nor, for that matter, to neuroscience narrowly understood as the study of brains in isolation from other brains).

An important constraint that Mikio has set himself is to avoid "dualism" and this is expressed in his commitment to what he calls "physicalism" and his effort to find a "narrow crack between reductionism and the rejection of physicalism."  In my own terms (Grobstein, 1988; perhaps too simple?), what Mikio is trying to come up with is a way to think about things such that phenomena at a particular level of organization ("mind" or "social organization") are permitted but not required by phenomena at a lower level of organization (neurons, brains", individuals, ultimately in all cases the chemical and physical properties of matter).  And hence they not only exist in their own right as legitimate subjects of inquiry but also have causal significance (can effect not only each other but also have top-down influences on lower levels of organization). 

I think there is indeed a way to think about things that has the characteristics Mikio is looking for (cf From Complexity to Emergence and Beyond and Inquiry as Emergence, both in the specific case of the relations between mind and brain and more generally.  It depends though on accepting that some form of genuine indeterminacy (stochasticity) is an admissible property within the "physicalism" constraint.  Given indeterminacy/stochasticity, phenomena at one level of organization are necessarily not fully reducible to phenomena at lower levels of organization and hence may have independent causal significance and warrant study in their own right. 

For reasons worth exploring further, many people  are reluctant to  accept indeterminacy/stochasticity as falling within the rubric of "physicalism" and, perhaps for this reason, Mikio prefers to try and achieve the same logical independence between levels of organization by appealing to non-linear dynamical systems, and particularly those with strange attractors.  My concern is not only that this introduces an unnecessary complexity but, more importantly, that it actually vitiates the program I would like to think Mikio has set himself.  Yes, systems involving strange attractors display sensitive dependence on initial conditions and so may produce phenomena at the next level of organization that are ill-mannered, not in practice deducible from lower order phenomena.  They are, however, in principle so deducible (see Variability in Brain Function and Behavior for the distinction between "ill-mannered" and "probabilistic"). This may be enough for many people.  It is not enough for me, for several different reasons.

The first point is the general one that dissociating phenomena at different levels of organization by using non-linear dynamical systems is really no more satisfying than the simpler argument that one works at higher levels of organization because  the complexity of interaction at lower levels necessary to account for the upper level phenomena are too complex to do the necessary calculations on.  A legitimate reply is that one needs then to acquire the capability to do the necessary calculations (or, in the linear dynamical systems case, to characterize the equations that yield the appropriate strange attractor).  In short, the dissociation doesn't eliminate the reductionist option; it simply makes it harder to pursue.  To put it differently, the dissociation doesn't really give us a new way to think about things but allows us to go on thinking about them as we have, content that "strange attractors" accounts for our difficulties with the levels of organization problem and we can hence regard that as crossed off our list of problems.  I'd prefer to have it persist as a reminder of our ignorance and hence as a goad to finding new ways to think about things.

The second point is more specific to the mind/brain relation.  Here too I'd prefer to have the problem sit as a goad to finding new ways to think about things, but there is something even more important at stake (for me at least).   The mind/brain relation is not only an instance of the levels of organization problem but also bears importantly on the  different but equally interesting problem (for me at least) of whether our behavior can be fully accounted for given adequate information about our history (our genes and our experiences).  If the distinction between brain and mind is regarded as solved by the strange attractor approach, that leaves us with no strong argument that we are ourselves semi-independent shapers of our own lives.  Like the narrator of Dostoyevsky's Notes from the Underground, I'm not happy as a "piano key", insist on some measure of "free will", and believe it can be found through a more subtle understanding of the mind/brain relation, one that involves a "physicalism" that embraces indeterminacy.  Given indeterminacy and a bi-partite brain organization, I believe we can understand how an assembly of matter such as ourselves can not only surprise others but ourselves as well, bringing into the world things that would not otherwise exist (are not simply puzzling due to strange attractors but genuinely new, ie emergent in the strongest sense). 

Needless to say, all this in turn relates to the issue of what inquiry is about, whether it is narrowing to the best available hypothesis or expanding to novel ones.  Mikio has, it seems to me, well-defined a problem that encapsulates much of our discussion over several years: "to find the narrow crack between reductionism and physicalism" or, in my terms, to understand how organized matter can generate the genuinely unpredictable.  If I've misunderstood the direction he intends to take to define that crack, much of the preceding is irrelevant.  Its also irrelevant to the extent that Mikio (and others) are content with "in practice" justifications for their preferences to work at higher levels of organization and "in practice" distinctions between mind and brain.  To the extent the ambition is broader, I still think one is going to have to admit indeterminacy into physicalism. 

That said, I freely admit that there does not currently exist a set of observations that will "disprove" either "deterministic physicalism" (as per Mikio and others (cf Wolfram)) or physicalism with indeterminacy (as per me). That doesn't bother me, and I hope it doesn't bother others either.  Popper notwithstanding, the falsification of alternative understanding/stories is only one part of science/inquiry.   The simultaneous existence of several incompatible ("incommensurate") unfalsified stories is a common occurrence in science/inquiry (Kuhn), and can be viewed not as a problem but as the grist that motivates both new observations and new stories. 

I also freely admit that in such circumstances, one might turn to Occam's razor as an adjudictor and that, for some people, application of Occam's razor might seem to favor "deterministic physicalism" if one regards indeterminacy as an add on to physicalism.  A different reading though might treat indeterminacy as a basic feature of physicalism and determinism as the odd on.  And, in any case, Occam's razor is demonstrably not the best approach in a number of circumstances.  The upshot?  One is free to place one's bets on whether deterministic physicalism or physicalism with indeterminacy will prove over time to be the more generative perspective.  We'll see


Alfonso Albano's picture

Emergence, reductionism and strange attractors

I realize that I am jumping into the middle of a discussion, not having
been to Mikio's presentation. I am just reacting to Paul's blog with
comments as far away as you can get from the brain-mind problem -- but
then, again, thermodynamics is far simpler than the mind. Anyhow, here are
my reactions:

Is it reductionist to claim that the laws of thermodynamics can be
“explained” starting from Newtonian mechanics? One emergent aspect of
thermodynamics is that things tend to equilibrium (there is an arrow of
time) while Newtonian mechanics is time-reversible (ain’t no arrow of
time). Boltzmann achieved the transition from the Newtonian to the
thermodynamic by introducing stochasticity. His description went from a
precise specification of positions and momenta to the probabilities that
positions and momenta lie within certain ranges.

Strange Attractors:
I don’t think that strange attractors can be considered as “emerging” from
nonlinear dynamics in the same way that thermodynamics “emerges” from
Newtonian mechanics. Strange attractors are families of solutions of the
nonlinear equations. Their properties are contained in the equations; they
are no more surprising than the elliptical orbits that arise from the
nonlinear equations describing the motion of a planet around the sun.