-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Günther Greindl on 01/07/2008 12:57 PM: > thanks for taking the time to write such a long response, here some > comments:
And thank you for pursuing it. Since I'm only slightly versed in RR, I enjoy the opportunity to talk about it. It helps me think clearly. >>> complexity (=uncomputability in the Rosen sense) >>> For my problems with his "uncomputability" see below. >> Living systems are just the particular example set of the (possibly very >> large) category of complex_rr* systems. It doesn't _start_ with life. >> Life just happens to be what RR (Robert Rosen) was interested in. > > Ok - but are all Rosenites sure about this? > complex_rr is a thesis which I find scientifically ok because it does > not introduce an arbitrary distinction between matter in different > organizational forms (animate vs inanimate), although I disagree (with > complex_rr) ;-) Hmmm. I guess that depends on what you mean by "Rosenite". [grin] But off the top of my head, I'd say "no". Most Rosenites I've talked to seem to hang the distinction clearly between living and non-living systems. In many cases, I just didn't have the chance to dig deep enough to find out whether they, too, believe that living systems are just a sub-set of complex_rr systems. In the end, I don't know the distribution of Rosenites who think life and complexity_rr are tightly correlated. > (see the excellent book by Torkel Franzen: Gödel's Theorem: An > Incomplete Guide to Its Use and Abuse > http://www.amazon.com/Godels-Theorem-Incomplete-Guide-Abuse/dp/1568812388/ref=pd_bbs_sr_1?ie=UTF8&s=books&qid=1199723625&sr=8-1) I second that recommendation. It's a very cool (and small!) book. >> A better exposition comes in Penrose's work, in which tries to argue >> that math (as done by humans) regularly involves hopping outside of any >> given formal system in order to catch a glimpse of a solution, then >> hopping back inside the formal system in order to develop a formal >> proof. And in this regard, RR's rhetoric is not inconsistent. > > The Penrose/Lucas argument has been debunked many times. In the Torkel > Franzen book above, in Rudy Rucker's "Ininity and the mind" (another > excellent book, recommended reading, good fun and educational) and in a > number of philosophy papers. But it still sticks around :-)) The argument sticks around because the "debunking" (at least what I've seen) _merely_ targets the validity of the argument, not the truth or falsity of the conclusion. So, it's true that the argument is INVALID; but that doesn't mean the conclusion is false. In any case, my point was that RR's argument is _like_ Penrose's argument. And, to the best of my knowledge, RR's argument hasn't been formalized to the point where we could show it to be invalid. (Kudos should go to anyone who makes their statements clear enough so that their rhetoric can be shown invalid!) > I like the holarchy idea, I think this is important, but I don't see why > this should not be capturable via computation. We can model formal > systems in other formals systems (indeed is being done in foundations of > math, as ZFC is currently seen as basis for math together with classical > logic, but that is another discussion entirely). I agree completely. In fact, to the best of my knowledge, no Rosenite put that idea into my head. I think it came to me after listening to a presentation by Terry Bristol (entitled "Carnot's Epiphany") wherein Bristol tried to make the case that the universe is an engine composed of sub-engines. His talk of symmetry and of energy being an asymmetry had already reminded me of RR because of the central role symmetry plays in RR's work. Then when I asked Terry what happens at the very top of the engine of sub-engines, he said something about it folding back down to the tiniest sub-engines (and vice versa). At that point, I began thinking of RR's efficient causation band-aides as a holarchy. The important point being that if you try to talk to a Rosenite about this "holarchy of formal systems", they may not know what you're talking about since it might merely be my extrapolation. [grin] Sorry for my lack of scholarship. > Either one is strictly materialist like Dawkins, then natural selection > is indeed enough of an explanation. > (that what can stay will stay, because if it couldn't it wouldn't) :-)) > > Or you assume a purpose to the universe, maybe something like Teilhard > de Chardin's Omega point (which "draws" evolution toward it). > > Maybe Rosen is somewhere in between? Hmmm. I reject the false dichotomy of "materialist or not". There is no "either-or", here. I can't defend my opinion; but, when people start extrapolating from the very tiny amount we know for sure out into the huge universe of which we're mostly ignorant, my warning bells go off too loud for me to think. So, if anyone (including Dawkins) claims he knows how the universe works well enough to be strictly materialist (or strictly _anything_), then that person loses most credibility in my book. I'm more ignorant than many; but I'm pretty sure that nobody here on earth can rigorously defend any statement that starts with "The Universe is ...". Given that, I'm open to all sorts of wacky ideas about how things may or may not work. And from that perspective, RR is simply trying to toss out a few ideas for why living systems seem so different from non-living ones. >> It's this "why" that leads him to consider "final cause". He takes the >> most prevalent answer to the why question seriously: living systems do >> what they do in order to benefit _themselves_. But how can an organism >> at time t_0 know what actions will benefit that organism at time t_100? > > It does not know. It it chooses wrongly, it will not be here to complain. Well, actually, it might "know" enough to estimate. And their estimate will be caused by the events in their history and their current interaction with their environment. RR calls this "anticipation". In order to talk about things like this estimate or a unit's ability to guide themselves toward a (imaginary?) goal, we need a name. "Final cause" is that name. >> The question he asks specifically is: "How can we have organization >> without finality?" I.e. How can we say that an activity of an organism >> is purposeful without some external _agent_ declaring the purpose of the >> organism? In the end, he comes to the idea that effects cause their >> causes, which is obviously cyclic. > > So he not also challenges the "mechanist/computationalist" thesis but > also standard neo-darwinism? I don't think RR provides an explicit challenge to neo-darwinism. However, I do think he would say that neo-Darwinism is either incomplete or just a (small?) part of the theory (of life) we will eventually develop, because it doesn't fully explain the organization of living systems. I think this "challenge" (were he to make it) would not invalidate neo-Darwinism; but it would posit that neo-Darwinism is only part of the explanation for biology as we know it. > Sounds a bit like converging toward an attractor - that is a nice idea > (and would also fit nicely with the Omega point) - but one does not need > any final causation for that - rather it is normal causality which > inevitably produces a result. Like a stone which is dropped on the Earth > will fall toward the Earth and not, say, to the moon. Not quite. This positive feedback would modify not only the state of the system, but also the ontology in which the system sits. So, it's not like two gravitational bodies falling toward one another because both bodies (rock and earth) are slavishly imprisoned within the ontology (gravitational physics). Any two biological units, on the other hand, can fundamentally change the rules by which they interact. The "physics" is mutable and is (purposefully) changed by the components of the system.... at least that's the position I think RR would take. I suppose they would do this by changing the assembly of formal systems and their respective semantic groundings. At heart, RR seems (to me) to assume a type of "fluidity" (or "logical abstraction") for the biological units that allows them to change their mechanisms but achieve the same (or similar) phenomena/outcome. This "fluidity" hinges on the "modeling relation" between causal and inferential entailment. (See * below.) > But the new iterations are not new axioms; Well, I'm claiming that (some subset of) the new iterations WOULD be new axioms. The whole idea of involving a the Goedel-related ideas (I think) into RR's conception is to talk explicitly about "level jumping", when a formal system becomes inadequate, you jump out of it and plug any holes with new axioms, then jump back in and continue on with your "inference". Of course, any new axioms might just be assumed temporarily in order to escape some temporary ambiguity. Or they might be permanent. > If you adopt an ultrafinitist stance, it only makes sense within a > strong coupling to reality: the claim that reality is in the end > discrete (QM, loop quantum gravity, holographic principle etc are all > theories which give hints in this direction) > Ultrafinitism is an extrapolation of the physical world (at least in my > interpretation, I am sure one can also hold it as a pure philosophy of > math, although it loses much of it's appeal then I would say). Ahhh. I get it now. If the world _is_ finite, then every state of the world can be achieved by an effective method (a dumb machine that just chugs mechanically along). When I talk of "computable", I tend to think in a limited way and only refer to the truth value (and decidability) of any given statement (or hypothesis). Something is incomputable if it's truth value cannot be determined within the formal system. And in that usage, reality is totally unrelated.... it's about the validity and NOT the soundness of any given statement. "Computer", as a term, belongs to the realm of thought not reality. A computer is an abstract machine. An HP ze4510 is a concrete machine. Hence, "computable", as a term, belongs in the lexicon of thought and inference, whereas "HP ze4510" belongs in the lexicon of reality and cause. Ultimately, what you're talking about is the _accuracy_ of any given formal system in describing the real world. If the world _is_ finite, then we could (in principle) discover/invent a language that describes the world _perfectly_. Of course, this might be true even if the world isn't finite ... * basically (back to RR) if causal entailment is equivalent to inferential entailment in the right ways, then we can accurately describe cause (reality) with inference (thought). RR's main premise is that inferential entailment is missing some fundamental attributes that causal entailment has. Hence, our languages are incapable of representing the world in important ways (namely, the ability to handle self-referencing loops). Of course, I'm not personally opposed to the idea that thought and matter are identical. But, RR (and most people actually) are steeped in dualism. So, for the purposes of this conversation, we have to preserve the dual. >> I do believe that there are certain processes in reality >> that are noncomputable in terms of what we now call "computation". > > Ok for the computability issues - I can't build a computer which solves > the halting problem; but what I was speaking about above was the > assumption that the universe could _be_ a computation (Seth LLoyd, Max > Tegmark, Jürgen Schmidhuber come to mind). Or do you think there are > physical processes which rule this conclusion out? (if yes, I would be > very interested to hear about this, because I am currently researching > this issue) Good question. I believe (emphasize _believe_ ;-) reality is NOT equivalent to computation as we currently know "computation". But, I'm no authority and please take anything I say as random chatter. But, before I go on to explain which parts of reality I think cannot be captured by computation, I want to be as clear as possible about what "computation" means. Computation, as the term is commonly used, is very close to the concept of "effective calculability", a set of well-defined steps (requiring NO intuition, special skills, or intelligence) that's guaranteed to stop and guaranteed to behave correctly, including adhering to the valence of any given operator. This is basically the way our computers work. You tell it what to do and it stupidly obeys you, eventually resulting in an infinite loop, a final answer, or a crash. And if you tell it that there are only, say, 10 possible answers, it will _merely_ produce one of those prescribed 10 possible answers. (I live for the day when I ask a computer: "Is this true or false?" And it answers: "Neither, it's _blue_!" ;-) - -- glen e. p. ropella, 971-219-3846, http://tempusdictum.com A government which robs Peter to pay Paul, can always count on the support of Paul -- George Bernard Shaw -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.6 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org iD8DBQFHg4tZZeB+vOTnLkoRAnTGAJ4vvyZONzNjlZJLUHtDxZ950X8u3gCglPPl +N0E5dwHhTCluOPdd0PRcFI= =yhEI -----END PGP SIGNATURE----- ============================================================ FRIAM Applied Complexity Group listserv Meets Fridays 9a-11:30 at cafe at St. John's College lectures, archives, unsubscribe, maps at http://www.friam.org
