[FRIAM] Any non-biological complex systems?

[glen ☣]

I don't want to distract from the new thread, which I think is more important than this part of this thread.  So, feel free to ignore this one.  But I'll answer now because I have the time now.  FWIW, I'm not annoyed.  But I'd like to see us (everyone) make some progress, even if it's not toward a paper or other artifact.  (To answer Owen, I am willing to help organize the content.)  Anyway, the rest below:

On 06/07/2017 09:31 AM, Nick Thompson wrote:
> [gepr] Below is what I wrote in response to Stephen's suggestion that any physical system might qualify as complex.
> 
> On 05/26/2017 12:40 PM, glen ☣ wrote:
> 
>> [gepr] Yeah, but you're relying on the ambiguity of the concept.  A system that is only complex for very short spans of time, or under very special conditions wouldn't fit with _most_ people's concept of "complex".  To boot, unadulterated oscillation wouldn't satisfy it either.  And, as has been said earlier in the thread, allowing any an all physical systems to be called "complex" when they're placed under special circumstances defeats the purpose of the concept.
> 
> [NST==>I am keen to know what is meant by “the purpose of the concept”.  Peirce says the meaning of a concept is the difference in the observations (experiments) that it leads us to make.  What difference does calling something complex make for you.  Forgive me please if this requires you to direct my attention to earlier posts.  <==nst] 



It means nothing to _me_.  Whenever someone says it, I a) abandon the conversation, b) ask them what they think it means, or c) impute what I think they mean.  But I can tell you what I think it means to most.  "Complex" is like "obscenity".  And this is why the discussion of the "order parameter" is important (and why so many ABMs are qualitiative and not quantitative).  When other people use the term, I think they mean "I can't tease out the causes of the phenomenon I'm looking at."  Sometimes they might (also) mean "It would take me a lot of effort to discuss what we're looking at."  Hence, the purpose of the concept of complexity (better termed "plectics") is to identify that "interesting region" of the phenotype ... or perhaps to distinguish one phenotype from another.

So, Stephen's suggestion that complexity is common defeats that purpose.  It's like when someone asks me if I believe in extraterrestrial life.  I say yes.  But if someone asks if I believe in extraterrestrial intelligence, I say no.  Further, to say that a system is complex, when only tiny regions of its behavior space exhibit complexity is equally useless.  It's like saying that all the air molecules in a room _could_ condense up into the corner, "in principle".  Sure, to people who think hard about theories and ideas, that helps explain what the theory means.  But to most people, it's either misleading or useless.



> [gepr] I'm used to thinking in terms of things being closed to one operation and open to others.  I'm also used to using it as a verb, e.g. "closing a set" by, say, adding the limit point to it.  Or more appropriate to this context, closing a region by wrapping a boundary around to itself like you would by tying a string to itself to get a loop.  Further, a region of some "space" can be "partially closed" by being very convex ... like some champagne flutes, bulging in the middle with a small-ish opening at one end ... or maybe think of a pinhole camera or a black body radiation device.
> 
> [NST==>But doesn’t this bring us back to the question that I got into a tangle with Steve about: Whether a system can close it self?  Or whether closure has to be be provided “from the outside” (=petri dish).  <==nst]


Yes, but we don't _have_ to go there.  Closure can be done by an outside actor (lab grad student setting up a BZ reaction in a dish).  Russ states it well: "around which one can draw a boundary".  That is a type of closure.  A hollow sphere or a shallow dish or whatever.  We can even attribute/draw a boundary around a hurricane.  But where and how we "close" the collection of thermals/molecules/trees/trucks/etc matters.  It's a study of how things are closed or open.

A self-referencing closure, a boundary defined by the bounded, is a fantastic sophisma, well beaten but not dead by Rosen et al.  But you don't have to leap all the way from ordinary closures to "teleological" or anticipatory closures.




>> [NST==>Huh?  You mean I don’t get to be in charge of whether I am 
> 
>> naïve, or not?  Are you some kind of behaviorist?  <==nst]
> 
>  
> 
> 8^)  Right.  You can't sa[NST==><==nst] by you're being naive and then get all hyper-technical about whatever naive statement you made.
> 
> [NST==>Naïve in the sense of standing apart from the technical details.  Would “phenomological” be better?  I am a great believer in the importance of outsiders to the health of science – the Emperor’s New Clothes metaphor. Outsiders can be really annoying and slow things up, but they also have an essential role to play.  Hard to know when one is playing that role, or merely being annoying. <==nst] 


Aha!  Yes, phenomenological would be better.  I agree about outsiders and (as Marcus pointed out) using language with wiggle room so that even if you can object to any given argument, perhaps you can still grok the gist.  But we still have to "listen with empathy".  And to do that, we have to model the other discussants and try to talk about what they're talking about, regardless of the language used.  So, a sophist[†] feigning naivete is annoying, whereas _actual_ naivete is essential.


[†] You may have also missed that I don't use "sophist" in a derogatory sense.  It is the identifier for those of _us_ who like to use paradoxical or koan-like issues to spark discussion and disambiguate perspectives.  Sophistry is that practice.  It's fun and useful (e.g. the claim that all the air molecules could condense into the corner of a room).


-- 
☣ glen