OUTLINE: From: GEMS::HMARAGH 29-FEB-1996 19:38:15.82 Subj: summary From: John Wilson - Radiation Biology 1-MAR-1996 17:07:26.11 Subj: chest pain Date: Sat, 02 Mar 1996 22:27:19 -0400 (EDT) From: DON MIKULECKY Subject: thermodynamics and complexity From: Dianne Lewis 3-MAR-1996 09:20:44.50 Subj: Recent Complexity Class Date: Sun, 03 Mar 1996 14:40:20 -0400 (EDT) From: DON MIKULECKY Subject: some thoughts on Dianne's comments Date: Sun, 3 Mar 1996 17:59:47 -0500 (EST) From: Jerry Chandler Subject: Re: thermodynamics and complexity Date: Mon, 04 Mar 1996 11:20:51 -0400 (EDT) From: DON MIKULECKY Subject: some further exchange on thermodynamics From: "Jeff Prideaux" 5-MAR-1996 13:31:11.63 Subj: thoughts generated by Wednesday's discussion group Date: Wed, 06 Mar 1996 10:43:06 -0400 (EDT) From: DON MIKULECKY Subject: math is a complex system for sure! From: Jeff Prideaux 6-MAR-1996 14:27:19.34 Subj: Was Hermann Hess a constructivist? Date: Wed, 06 Mar 1996 19:48:55 -0400 (EDT) From: DON MIKULECKY Subject: some further dialectic! Note on Layers: > > > > > 05Mar96 18:25 Mikulecky's question to John Wilson > > > > 05Mar96 19:58 John Wilson's reply > > > 06Mar96 11:48 Chandler's comment > > 06Mar96 13:23 Mikulecky's response > 06Mar96 19:12 Chandler's reply Mikulecky's response Date: Wed, 6 Mar 1996 21:27:07 -0500 (EST) From: Jerry Chandler Subject: Last Call! Re: some further dialectic! > > > > 06Mar96 11:48 Chandler's comment > > > 06Mar96 13:23 Mikulecky's response > > 06Mar96 19:12 Chandler's reply > 06Mar96 19:48 Mikuleck's rebuttal Chandler's final comments From: "Jeff Prideaux" 7-MAR-1996 10:50:39.43 Subj: Posting on time > > > 06Mar96 19:12 Chandler's reply > > 06Mar96 19:48 Mikulecky's rebuttal > 06Mar96 21:27 Chandler's final comments Prideaux's comments Date: Fri, 08 Mar 1996 10:22:56 -0400 (EDT) From: DON MIKULECKY Subject: time marches on???? Date: Fri, 08 Mar 1996 13:19:00 -0400 (EDT) From: DON MIKULECKY Subject: the complexity of complexity > From: "Jeff Prideaux" 8-MAR-1996 11:27:46.15 > Subj: different usages of the word "complexity" ___________________________________________________________________________________ From: GEMS::HMARAGH 29-FEB-1996 19:38:15.82 Subj: summary The nursing profession faces a changing environment. In order to survive it must change the "dominant logic" which is at equilibrium and is rigid and isolated. It must move from equilibrium to "edge of chaos" where the system changes and adopt to the new environment. The elements of the system must change educational and job pursuits to be flexible in the" new order." Problem solving must involve all elements of the system and strategic planning must be limited to shorter time periods to allow a greater flexibility. However, "are we throwing out the baby with the bath water?" The was no suggestion of directionality, except self-organization, and is this wise? Will we progress or fail? This may depend on the stimulus for change, cost-effectiveness or profiteering? So jobs are cut, care reduced, strategies of care not tested etc. Consequence of NASA strategies-explosion. Is the "Paradigm Lost"? ___________________________________________________________________________________ From: GEMS::WILSONJD "John Wilson - Radiation Biology" 1-MAR-1996 17:07:26.11 Subj: chest pain Don...I don't know if I made the connection between those two folders that I shoved at you earlier this afternoon. The director of the Williamson Institute is Ken White's collaborator/thesis director, James Begun. One of the institute's faculty members, Louis Rossiter, is a speaker at the chest pain management conference. The conference itself is about trying to address the organizational problems we were just talking about Wednesday, but from the perspective of delivering emergency care in the area of chest pain. Look at the topic for the dinner speech (Friday, 7:30). I just talked to Jim Tatum, Chairman of Nuclear Medicine and one of the prime movers in setting this conference up, to try to wheedle a tuition waiver. I was unsuccessful in this endeavor ("Radiology has plenty of money, get them to pay for you!") but we got off onto a wonderful discussion much like the one we had at Wednesday's meeting. Tatum is a guy who's bored with the practice of medicine in its current state, dabbling with law school in his free time, and heavily into how to restructure the practice of medicine to meet future needs. His views were right in line with those expressed in Ken's paper, however, he brought up an interesting aspect which also had occurred to me after hearing Ken speak. And that is that the perspective taken is entirely reactionary; an unstable, or chaotic environment is simply accepted as inevitable and all modification must involve the restructuring of the organization to favor more rapid adaptation to external pressures. Has any consideration been given to finding ways to drive the environment toward a more stable state, or is the inertia just too great for this to be a practical solution? Last semester we talked a little about controlling chaotic states as I remember; I gave you a paper from Science in which it was shown that it was possible to drive a chaotically arrhythmic heart into a stable state. Tatum claims that this would result in reestablishing control, which is one of the first things that Ken said had to be given up. At the end of our conversation he commented on how much more he liked to talk about these kinds of things than the everyday clinical fare (duh!). He might be an interesting guy to get to talk to us. He's not shy about expressing his views; I would imagine he could stir the pot in much the same manner as Ken did. ___________________________________________________________________________________ Date: Sat, 02 Mar 1996 22:27:19 -0400 (EDT) From: DON MIKULECKY Subject: thermodynamics and complexity Some further thoughts on the Thermodynamics of Complexity: Upon a rereading of Rosen's "Cooperation and Chimera" in "Cooperation & Conflict in General Evolutionary Processes", (J. L. Casti and Anders Karlqvist, eds.), Wiley, 1995, I was struck by the following realizations as a result of his arguments and demonstrations. In all our presentations of thermodynamics, we focus our attention on the simple and throw out the complex. This starts at the most basic level, when we define what we mean by a "function of state" such as the entropy. It thus is no accident that as we progress from classical thermodynamics to non-equilibrium thermodynamics to far-from-equilibrium thermodynamics, we loose touch with the real world in some crucial ways. One problem I have known and talked about for years is "Meixner's Paradox". This is the fact that unless one has a COMPLETE state space representation of a nonlinear, non-equilibrium system, there is no way to uniquely define entropy. In other words, what we are ALWAYS faced with in the real world is the necessity to see some aspects of any system as being within a "black box" for which we only have input-output information. Once we recognize this, Meixner is able to demonstrate that there are relatively uncomplicated electrical circuits which are made only of resistors, capacitors, and switches (the nonlinear elements) for which the distinction between dissipation in the resistors IN the box and non-dissipative storage on the capacitors IN the box is impossible to discern from without if the controls for the switching elements IN the box are unknown to us. Hence all the beauty and cleanness of Caratheodry's proof is lost to us here as well as much more! Should this have been a surprise? In fact, few people know about or understand Meixner's paradox. Why is this so? One reaction to the paradox is a relieved sigh and the statement: "That's all right because IN PRINCIPLE we can ALWAYS have a state space description of any system." Yes, maybe in simple systems. However, in complex systems, this is demonstrably impossible, since the notion of unambiguous, well defined states also dissapears as Rosen has so elegently shown. One way in which he does this is through the analysis of the effect of nested levels of inhibition and activation such as we see in enzyme networks, neural networks and ecological networks. In the present reference he extends this argument to evolution. What strikes me about this is the way we mislead students of biology when we teach them thermodynamics! We quickly demonstrate the fact that heat and work are not functions of state because they depend on the path by which a change in state is carried out. ( They are therefore CLEARLY CONTEXT DEPENDENT!) We reassure the students (and ourselves) by demonstrating the existence of a thermal state variable called the ENTROPY which is path independent. Of course the fact that the states being considered are equilibrium states and the path we are worried about is the path from one equilibrium state to another is of no consequence. Or is it? In biology, we seldom, if ever, are interested in equilibria. And there is the rub! We are PRIMARILY interested in path dependent systems as the argument about activation/inhibition clearly demonstrates. Thus, it would seem, that the thermodynamics we need for complex systems is the opposite in character from what has historically been presented! We should see path independent state transitions between equilibrium states as the exception and the path dependence of heat and work as the norm! Once again the heavy hand of the reductionist has stifled us and led us down a path leading further and further from the reality of complexity and swindled us into seeing a simple world which cannot exist! I find this notion liberating, and wonder how others see it? ___________________________________________________________________________________ From: Dianne Lewis 3-MAR-1996 09:20:44.50 Subj: Recent Complexity Class In thinking over this past Wednesday's class, it occurs to me that the schematic diagram for nursing paradigms that the speaker gave does not take into account the environment in which that dynamic exists. My pondering and musings have led to the following questions/ideas: Neither ordering of systems nor disordering of them occurs in linearity. Physiology of even the simplest organism demonstrates that order and disorder occur in a larger cyclic fashion. Please picture an oval model with an eccentric circle inside where the larger space represents the more chaotic phase and the smaller space represents the more ordered phase. This, then, leads to the question: Is there a scale to the potential order within a system? I.e., at its most ordered point in the cycle, an element of chaos remains and at the most chaotic phase some order is detectable. If this is true, is there a measurable equilibrium? Also, what is the mechanism for transition between the phases. One idea is based (loosely) on my limited understanding of electricity. If there are two "well-ordered" systems in proximity, does all the matter/energy between them become disordered as a result of the difference in "order potential." Please forgive my lack of vocabulary, I am a new student of this subject and am probably not using proper terminology. Also, regarding a class Carlisle told be about on the topic of spirituality. It seems to me that spiritual matters are elusive and non-testable because they occur in the more chaotic phase of the system (where the human organism/experience is the system). We cannot accurately measure spirituality because our current measurement methods apply only to the more ordered phase. I believe that spritual awakenings and experiences occur as information is "reorganized" in the more chaotic phase. Learning, with its inherent accummulation/assimilation dynamics would also occur in the more chaotic phase, possibly explaining why our testing methods in this area are so inadequate. Other thoughts: perhaps the gap between Eastern medicine and Western medicine could be bridged if we applied this model in an effort to reconcile them -- Eastern medicine focusing on the chaotic Chi and Western medicine focusing on the ordered physiology. Both have world-wide applications, hence neither can be dispensed with. They cannot be considered simultaneously, however, because they have so few elements in common (except, of course, they work on the same organism). I'm pressed by mid-term stress and cannot hold this train of thought very long but am excited and would enjoy an opportunity to discuss it with you further. If my fledgling revelations are a bore, perhaps you could recommend someone who would be tolerant enough to talk through these ideas with me. ___________________________________________________________________________________ Date: Sun, 03 Mar 1996 14:40:20 -0400 (EDT) From: DON MIKULECKY Subject: some thoughts on Dianne's comments First, thanks! Nothing wrong with using your own words and certainly nothing boring there! If anyone has trouble with terminology...we can discuss that further! Cycles: Cyclic processes seem to be at the heart of organization and certainly self-organization. If nothing else it is a way of breaking free from linear time. Many good studies on this. They include Eigen's hypercycles, autocatalytic processes in evolution, steady states, chaos, ecosystem organization and others. Mechanism for transition between phasae: Generally, this is not an algorithmic (mechanistic) process. Hence it introduces complexity. Disorder as a result of potential differences: This is the progression of transitions which occur as a system is moved away from equilibrium. (As the difference in potential is increases). In many complex systems the progression is: equilibrium--->ordered steady state---->self organization----->chaos each one is "ordered" in its own way. The transitions are not mechanistic. Often each stage can be characterized by dynamic systems theory, but most of the transitions can not. They are outside that paradigm. Many features of complex systems defy "measurement" in the classical sense. Among them are spirituality, quality, love, beauty, etc. This is why we are exploring constructivism and other approaches. Eastern thought is replete with clues for this sort of thing. Certainly Eastern approaches to medicine and other ways to rise above the body as machine model are valuable. I use two books in my honors modules which are good as an introduction: Capra: "The Turning Point" and Briggs and Peat: "Turbulent mirror". The influence of environment on chaotic systems is a key question. Rosen devoted a whole paper to this in:"Some Random Thoughts on Chaos and Some Chaotic Thoughts on Randomness" J. Biol. Systems. 1:19-26(1993). ___________________________________________________________________________________ Date: Sun, 3 Mar 1996 17:59:47 -0500 (EST) From: Jerry Chandler Subject: Re: thermodynamics and complexity On Sat, 2 Mar 1996, DON MIKULECKY wrote: > Some further thoughts on the Thermodynamics of Complexity: > Don Mikulecky > > Upon a rereading of Rosen's "Cooperation and Chimera" in "Cooperation & > Conflict in General Evolutionary Processes", (J. L. Casti and Anders > Karlqvist, eds.), Wiley, 1995, I was struck by the following realizations > as a result of his arguments and demonstrations. > > In all our presentations of thermodynamics, we focus our attention on > the simple and throw out the complex. You are exactly right! But, you limit your major premise to thermodynamics. Isn't this the philosophy of physics in general? "The art of finding the simple within the complex" is almost the definition of physics, isn't it? Consider the atomic table, which is a representation of a pattern among all elements and hence of all matter. Isn't the atomic table a focus on the simple elements of complexity? > This starts at the most basic level, > when we define what we mean by a "function of state" such as the entropy. Again, you are exactly right. But you limit your major premise to "function of state". But isn't this a deeper problem which is based on how we chose to represent observations in mathematical form? In other words, how we chose to use the concept of heat in explaining natural observations? In the late 1980's I concluded, via a related set of arguments developed from biochemical considerations, that the concept of "heat" was vastly over - rated when it comes to living functions. Certainly, thermal energy plays a substantial role in living systems. But, "heat" constraints are set of constraints which living systems have learned to cope with by superior organization of structures. > It thus is no accident that as we progress from classical thermodynamics to > non-equilibrium thermodynamics to far-from-equilibrium thermodynamics, we > loose touch with the real world in some crucial ways. One problem I have > known and talked aboutb for years is "Meixner's Paradox". Could you provide a reference for this paradox? > This is the fact > that unless one has a COMPLETE state space representation of a nonlinear, > non-equilibrium system, there is no way to uniquely define entropy. You are exactly right! The emphasis must be placed on the COMPLETENESS of the representation. This follows rather directly from the third law of thermodynamics and chemical representations. It is also closely related to understanding of spectra and molecular "Mechanics". > In other > words, what we are ALWAYS faced with in the real world is the necessity to > see some aspects of any system as being within a "black box" for which we > only have input-output information. Once we recognize this, Meixner is able > to demonstrate that there are relatively uncomplicated electrical circuits > which are made only of resistors, capacitors, and switches (the nonlinear > elements) for which the distinction between dissipation in the resistors IN > the box and nondissipative storage on the capacitors IN the box is impossible > to discern from without if the controls for the switching elements IN the box > are unknown to us. Well, things are not so bleak in biochemistry and larger scale systems. The isotopic tracers (such as C14 -glucose or N15-aminoacids) have provided a detailed map of the biochemical dynamics of living systems. Such metabolic tracer experiments were conducted in wholistic experiments in the '40s and '50s. These experiments are one source of the "CASA" hypothesis. The cellular consituents are "Classified" by structure, the "Analysis" of the isotopic flow patterns provided a map of the relationships and also the pathways for "Synthesis" of macromolecules and other cellular components. By varing the external milieu, the cells "Actions" could also be monitored. These wholistic, isotopic cellular studies in the 50's led directly to the first demonstration of cellular feedback mechanisms at the chemical level (about 1958) and subsequently to the identification of the roles of t-RNA and m-RNA in protein synthesis. > Hence all the beauty and cleanness of Caratheodry's proof > is lost to us here as well as much more! > > Should this have been a surprise? In fact, few people know about or > understand Meixner's paradox. Why is this so? One reaction to the paradox is > a relieved sigh and the statement: "That's all right because IN PRINCIPLE > we can ALWAYS have a state space description of any system." In my experience, the claim of "IN PRINCIPLE" is often invoked as a defense mechanism when the representation of the system is incomplete and the theory does not predict the observations. This scientific tradition can be traced all the back to Archimedes and the claim, "Give me a long enough lever and I will move the earth." For the past few years, the folks who give top priority to "CHAOS" seem to have the most "IN PRINCIPLEs" going for them. ( I would note that a colleague and I teach a class listed " Chaotic, Cyclic and Oscillator Processes in Biology and Medicine.") > Yes, maybe in > simple systems. However, in complex systems, this is demonstrably > impossible, since the notion of unambiguous, well defined states also > dissapears as Rosen has so elegently shown. > > One way in which he does this is through the analysis of the effect of nested > levels of inhibition and activation such as we see in enzyme networks, neural > networks and ecological networks. In the present reference he extends this > argument to evolution. > > What strikes me about this is the way we mislead students of biology when > we teach them thermodynamics! We quickly demonstrate the fact that heat and > work are not functions of state because they depend on the path by which > a change in state is carried out. ( They are therefore CLEARLY CONTEXT > DEPENDENT!) This is an extremely important point because it focuses on the basic relationship between language usage and scientific principles. It gets at the heart of the problem of going from the local observations to the global level of conceptualization. In other words, changing the system changes the context and, in the new context, old rules may or may not apply. > We reassure the students (and ourselves) by demonstrating > the existence of a thermal state variable called the ENTROPY which is > path independent. Of course the fact that the states being considered > are equilibrium states and the path we are worried about is the path from one > equilibrium state to another is of no consequence. Or is it? In biology, > we seldom, if ever, are interested in equilibria. And there is the rub! > We are PRIMARILY interested in path dependent systems as the argument > about activation/inhibition clearly demonstrates. > > Thus, it would seem, that the thermodynamics we need for complex systems is > the opposite in character from what has historically been presented! We > should see path independent state transitions between equilibrium states > as the exception and the path dependence of heat and work as the norm! A remarkable statement! Yes, indeed, a remarkable statement. I am not certain that I can accept this conclusion. Perhaps you can help me understand what you are after? What would you do if you had logically rigorous thermodynamics based on the path dependence of heat and work? > Once again the heavy hand of the reductionist has stifled us and led us > down a path leading further and further from the reality of complexity > and swindled us into seeing a simple world which cannot exist! This statement is less remarkable and also less open to understanding. Are you approaching 'reductionism' from the perspective of newtonian mechanics and this mechanical world view which exaggerated it powers? Or, are you approaching 'reductionism' from the perspective of the disassembly of complex systems into simpler components and gaining some knowledge of how complex systems work? The question of 'simple vs complex' contains a huge semantic component. For example, in thermodynamics, we may wish to study P-V-T relationships with a simple inert gas (neon) or we may wish to study P-V-T relationships with a mixture of reactive gases - say with several of the nitrogens oxides. Is the latter study a "complex study"? Compare the preceding example of simplicity and complexity with the following: A heart surgeon may do a simple bypass replacement or a quadruple replacement. Should we consider the later to be complex? (It is these sorts of questions which are located within the context of the scientific languages used to represent the system, that forced me to look for and define a systematic classification of complexity. In other words, what concepts are important in enumerating the attributes of complexity verus what words are used to describe the attributes of complex behaviors? This is part of the question Jeff was addressing with the distinction between syntax and semantics.) > I find this notion liberating, and wonder how others see it? > Liberating is a good word to describe the mental and emotional impact of discovering the limitations of some of the classical notions of science. However, my experience is that after one has 'liberated' oneself from a particular classical view, one is immediately faced with a difficult problem. The problem is to re-construct a plausible pattern of scientific logic and language which is equally satisfying as the classical view. This is a non-trivial exercise. :-) !! In short, Don, I guess that we operate on several of the same wavelenghts. ___________________________________________________________________________________ Date: Mon, 04 Mar 1996 11:20:51 -0400 (EDT) From: DON MIKULECKY Subject: some further exchange on thermodynamics Layer Notation: > > 02Mar96 22:27 Don Mikulecky's original thoughts > 03Mar96 17:59 Jerry Chandler's reply Mikulecky's reply to Chandler's reply > > Some further thoughts on the Thermodynamics of Complexity: > > > > Upon a rereading of Rosen's "Cooperation and Chimera" in "Cooperation & > > Conflict in General Evolutionary Processes", (J. L. Casti and Anders > > Karlqvist, eds.), Wiley, 1995, I was struck by the following realizations > > as a result of his arguments and demonstrations. > > > > In all our presentations of thermodynamics, we focus our attention on > > the simple and throw out the complex. > > You are exactly right! But, you limit your major premise to thermodynamics. > Isn't this the philosophy of physics in general? "The art of finding the > simple within the complex" is almost the definition of physics, isn't it? > Consider the atomic table, which is a > representation of a pattern among all elements and hence of all matter. > Isn't the atomic table a focus on the simple elements of complexity? Yes I WAS limiting my COMMENTS to thermodynamics. That's because we have already said so much about the rest of physics. > > This starts at the most basic level, > > when we define what we mean by a "function of state" such as the entropy. > > Again, you are exactly right. But you limit your major premise to > "function of state". But isn't this a deeper problem which is based on > how we chose to represent observations in mathematical form? In other > words, how we chose to use the concept of heat in explaining natural > observations? In the late 1980's I concluded, via a related set of > arguments developed from biochemical considerations, that the concept of > "heat" was vastly over - rated when it comes to living functions. > Certainly, thermal energy plays a substantial role in living systems. > But, "heat" constraints are set of constraints which living systems have > learned to cope with by superior organization of structures. Yes, again, this was a VERY limited comment! No disagreement here. > > It thus is no accident that as we progress from classical thermodynamics to > > non-equilibrium thermodynamics to far-from-equilibrium thermodynamics, we > > loose touch with the real world in some crucial ways. One problem I have > > known and talked aboutb for years is "Meixner's Paradox". > > Could you provide a reference for this paradox? Rheologica Acta 7:8(1968). [In German] Summarized in English in the editorial "Networks in Nature" Nature 234:380-381. > > This is the fact > > that unless one has a COMPLETE state space representation of a nonlinear, > > non-equilibrium system, there is no way to uniquely define entropy. > > You are exactly right! The emphasis must be placed on the COMPLETENESS > of the representation. This follows rather directly from the third law > of thermodynamics and chemical representations. It is also closely > related to understanding of spectra and molecular "Mechanics". please say more on this > >In other > > words, what we are ALWAYS faced with in the real world is the necessity to > > see some aspects of any system as being within a "black box" for which we > > only have input-output information. Once we recognize this, Meixner is able > > to demonstrate that there are relatively uncomplicated electrical circuits > > which are made only of resistors, capacitors, and switches (the nonlinear > > elements) for which the distinction between dissipation in the resistors IN > > the box and nondissipative storage on the capacitors IN the box is impossible > > to discern from without if the controls for the switching elements IN the box > > are unknown to us. > > Well, things are not so bleak in biochemistry and larger scale systems. > The isotopic tracers (such as C14 -glucose or N15-aminoacids) have > provided a detailed map of the biochemical dynamics of living systems. > Such metabolic tracer experiments were conducted in wholistic experiments > in the '40s and '50s. These experiments are one source of the "CASA" > hypothesis. The cellular consituents are "Classified" by structure, the > "Analysis" of the isotopic flow patterns provided a map of the > relationships and also the pathways for "Synthesis" of macromolecules and > other cellular components. By varing the external milieu, the cells > "Actions" could also be monitored. These wholistic, isotopic cellular > studies in the 50's led directly to the first demonstration of cellular > feedback mechanisms at the chemical level (about 1958) and subsequently > to the identification of the roles of t-RNA and m-RNA in protein synthesis. interesting! > > Hence all the beauty and cleanness of Caratheodry's proof > > is lost to us here as well as much more! > > > > Should this have been a surprise? In fact, few people know about or > > understand Meixner's paradox. Why is this so? One reaction to the paradox is > > a relieved sigh and the statement: "That's all right because IN PRINCIPLE > > we can ALWAYS have a state space description of any system." > > In my experience, the claim of "IN PRINCIPLE" is often invoked as a > defense mechanism when the representation of the system is incomplete > and the theory does not predict the observations. > This scientific tradition can be traced all the back to Archimedes and > the claim, "Give me a long enough lever and I will move the earth." > For the past few years, the folks who give top priority to "CHAOS" seem > to have the most "IN PRINCIPLEs" going for them. ( I would note that a > colleague and I teach a class listed " Chaotic, Cyclic and Oscillator > Processes in Biology and Medicine.") can you tell us more? > > Yes, maybe in > > simple systems. However, in complex systems, this is demonstrably > > impossible, since the notion of unambiguous, well defined states also > > dissapears as Rosen has so elegently shown. > > > > One way in which he does this is through the analysis of the effect of nested > > levels of inhibition and activation such as we see in enzyme networks, neural > > networks and ecological networks. In the present reference he extends this > > argument to evolution. > > > > What strikes me about this is the way we mislead students of biology when > > we teach them thermodynamics! We quickly demonstrate the fact that heat and > > work are not functions of state because they depend on the path by which > > a change in state is carried out. ( They are therefore CLEARLY CONTEXT > > DEPENDENT!) > > This is an extremely important point because it focuses on the basic > relationship between language usage and scientific principles. It gets > at the heart of the problem of going from the local observations to the > global level of conceptualization. In other words, changing the system > changes the context and, in the new context, old rules may or may not apply. > > We reassure the students (and ourselves) by demonstrating > > the existence of a thermal state variable called the ENTROPY which is > > path independent. Of course the fact that the states being considered > > are equilibrium states and the path we are worried about is the path from one > > equilibrium state to another is of no consequence. Or is it? In biology, > > we seldom, if ever, are interested in equilibria. And there is the rub! > > We are PRIMARILY interested in path dependent systems as the argument > > about activation/inhibition clearly demonstrates. > > > > Thus, it would seem, that the thermodynamics we need for complex systems is > > the opposite in character from what has historically been presented! We > > should see path independent state transitions between equilibrium states > > as the exception and the path dependence of heat and work as the norm! > > A remarkable statement! Yes, indeed, a remarkable statement. > I am not certain that I can accept this conclusion. > Perhaps you can help me understand what you are after? > What would you do if you had logically rigorous thermodynamics based on > the path dependence of heat and work? Rosen has already gone a long way in showing how the same character and path dependence we see with heat and work is demonstrable in biochemical networks, neaural nets and other activation/inhibition systems. > > Once again the heavy hand of the reductionist has stifled us and led us > > down a path leading further and further from the reality of complexity > > and swindled us into seeing a simple world which cannot exist! > > This statement is less remarkable and also less open to understanding. > Are you approaching 'reductionism' from the perspective of newtonian > mechanics and this mechanical world view which exaggerated it powers? > Or, are you approaching 'reductionism' from the perspective of the > disassembly of complex systems into simpler components and gaining some > knowledge of how complex systems work? what I am saying is that the influence of reductionism has limited ALL our thought (here Thermo) not just in mechanics. > The question of 'simple vs complex' contains a huge semantic component. > For example, in thermodynamics, we may wish to study P-V-T relationships > with a > simple inert gas (neon) or we may wish to study P-V-T relationships with a > mixture of reactive gases - say with several of the nitrogens oxides. Is > the latter study a "complex study"? This is an example Rosen says a lot about. I need to get the references > Compare the preceding example of simplicity and complexity with the > following: > A heart surgeon may do a simple bypass replacement or a quadruple > replacement. Should we consider the later to be complex? The SYSTEM sure is > (It is these sorts of questions which are located within the context > of the scientific languages used to represent the system, that forced > me to look for and define a systematic classification of complexity. > In other words, what concepts are important in enumerating the attributes > of complexity verus what words are used to describe the attributes of > complex behaviors? This is part of the question Jeff was addressing with > the distinction between syntax and semantics.) Remember that we have focused on one clear definition of complexity. Rejecting all others. We use Rosen's > > I find this notion liberating, and wonder how others see it? Liberating is a good word to describe the mental and emotional impact of discovering the limitations of some of the classical notions of science. > However, my experience is that after one has 'liberated' oneself from a > particular classical view, one is immediately faced with a difficult > problem. The problem is to re-construct a plausible pattern of > scientific logic and language which is equally satisfying as the > classical view. This is a non-trivial exercise. :-) !! > > In short, Don, I guess that we operate on several of the same wavelenghts. So far, the non-trivial job of a new construction looks like it is going to be a lot of fun!!!! ___________________________________________________________________________________ From: "Jeff Prideaux" 5-MAR-1996 13:31:11.63 Subj: thoughts generated by Wednesday's discussion group About the talk on complexity and the challenges to the Nursing profession: To me, the whole theme centered around the idea that in times of instability there is uncertainty (or unpredictability), and in such situations, one should concentrate on short-term planning and increase options (for example by further training). Only in times of "stability", can one formulate a reasonable dynamic model of the future and plan accordingly. In "unstable" times, it is impossible to anticipate exactly what the "equations of motion" will be in the future...that is, the future "equations of motions" are not derivable from the current "equations of motion" and the values of the current state variables. I then ponder "Why is this so?" I see at least three explanations (or view points) to the question "Why do we have unpredictability?" I am predominately interested in the third explanation listed below. The first two explanations (in my opinion) require no paradigm shift. I also think the first two may be "red herrings" Explanation #1: The future is, in principle, predictable but for technical reasons we can not gain access (or measure to enough precision) the necessary observables. Therefore, any models we form will (for purely technical reasons) be inadequate to predict the future (but not that there is anything wrong, in principle, with those models). The easiest example for this case is a description of something consisting of a system of coupled differential equations. Consider the case where the system of equations manifests a chaotical attractor which, by definition, is VERY sensitive to initial conditions (or any particular state values). This means that if you are off (in your model) in the value of a state variable by just a tinny amount (like .0000000000000000000001), your model (after enough time) may very well come up with very different values on those state variables than what the real system would have. Your model would only be able to accurately predict the future for a while, then its prediction would deviate from what is really happening. Chaotic systems are considered to be (for all practical purposes) random (or indeterminate) because of the practical limits to data storage on computers. For example, if the real-world state variable required an irrational number, then it would be impossible to store that number on a computer (it would require infinite storage space). The computer's dynamic system would deviate from the real-world's dynamic system (assuming the system manifested a chaotical attractor). For point attractors or limit cycles, the precision of data storage is not as important. Explanation #2: One could also form an argument for non-predictability by using environmental randomness. This explanation also assumes that dynamic descriptions are, in principle, applicable although unpredictability comes out of the use of the environmental randomness. This explanation (although I won't go into it) is a subtle variation of #1 above. I view both of the above explanations as an attempt to explain unpredictability while retaining the basic tenants of the mechanistic world view. In short, the above explanations provide a way of accounting for unpredictability without adopting a new (and different) paradigm. I don't consider the difference between a chaotic attractor and a point attractor to be a paradigm shift. I view it as simply different behavior modes a dynamic system. Both behavior modes (chaotic and non-chaotic) are easily simulatable by computers. The difference is that the computer cannot represent the necessary precision to accuratey represent a chaotical system over time. Although, a different issue entirely would be a context change which resulted in a change in description from one to the other. See #3 below. Explanation #3: This explanation rejects the premise that structure exclusively entails function. While structure (once it exists) is context independent, function is context DEPENDENT. For example, what is the function of a hammer. You may say "To pound things". Well, a hammer could also be used as a paper weight. How you classify something depends on how you interact with it (the function depends on the context in which it is embedded). The function is not an intrinsic property of the object. Bio-chemists say "structure, then function.....structure, then function". What they are really saying is that once a particular structure is embedded in a particular context, then it is possible to determine, experimentally what the function is...and once we know how that particular structure acts in that particular context, we can know how that structure (or one just like it) will act when it gets in an identical context in the future. An interesting observation is that genetics has basically given us a theory of structure (but not of function). Genetics basically says "Structure A plus function entails structure B". Here, structure A pertains to the DNA sequence; function pertains to the known action (or behavior) of the various proteins (and RNA's) involved in transcription and translation (the function of the genetic machinery); and structure B is the resulting protein. With the knowledge of the function of the genetic machinery and the structure of the DNA, one can predict the primary sequence of the resultant protein. But that is all. If I were to arbitrarily write down a previously uncharacterized sequence of 300 amino acids, no biochemist would be able to tell me its function or (or know for sure that it had no function) without performing an experiment to determine its function in a certain context. Consider the possibility that this protein has the function (in the current context) of participating in the genetic machinery. Also consider that this protein wasn't previously being used. Now, it is possible that the presence of this new protein can alter the way that the genetic machinery looks at the DNA (maybe by shifting the reading frame over by one, by skipping a nucleotide, by repeating a nucleotide, or by some splicing mechanism. Because of this, additional new proteins can be produced (of which we don't know (can't know) their function before hand. With these new proteins, the whole context of the system could change. Additionally, old existing proteins (or protein types) may start to behave in different ways (because the system context has changed). From our viewpoint, the system would behave completely unpredictable because we would have no way of knowing how these new proteins would behave in advance. In summary: - Context is defined by the organization (network) of functions - Function is an emergent system property (as opposed to a property of individual structures) that occurs when structures are embedded in the system. Function can not be defined independent of context. Function emerges at the time the structure becomes embedded in the system. There is a simultaneous change in context and establishment of new function. - Function + structure -> new structure. This circularity completely defies mechanistic thought for such a function producing system. A mechanism requires, at some level, a static or unchanging context. The concept of state variables requires a static context. The above explanation (#3) describes unpredictability without even mentioning randomness or chaos. Of course, some may view the cause of unpredictability as irrelevant for forming policy. They may say that the important fact is THAT the future is unpredictable. But it may also be true that better policy decisions can be reached if it is understood exactly WHY the future is unpredictable. This is why I'm interested in distinguishing between different causes of unpredictability (such as #1, #2, and #3). ___________________________________________________________________________________ Date: Wed, 06 Mar 1996 10:43:06 -0400 (EDT) From: DON MIKULECKY Subject: math is a complex system for sure! > Here are news bulletins from The Chronicle of Higher Education > for Wednesday, March 6. > > MAGAZINES & JOURNALS > > A glance at the March/April issue of "The Sciences": > > For more than 300 years, mathematicians puzzled over how to > prove Fermat's Last Theorem, the elusive equation that the > French mathematician Pierre de Fermat wrote he had proved in > 1637. But Fermat never explained how he'd done it. Last June, > two scholars presented what they said was a completed proof of > the theorem, which states that there is no way to add two > numbers raised to a power greater than two and produce a third > number raised to the same power. Now Dorian Goldfeld, a > professor of mathematics at Columbia University, questions > whether his colleagues actually proved the theorem. Mr. > Goldfeld suggests that the proof in question advances a broader > theory of Diophantine equations of which Fermat's is only an > "incidental consequence." (The magazine may be found at your > library or newsstand.) > > Copyright (c) 1996 The Chronicle of Higher Education, Inc. Doesn't this sound like what Rosen was talking about with number theory? As you try to algorithmize a piece of the theory, you end up having to extend the theory beyond its old bounds. Goedel is at work here I suspect! ______________________________________________________________________ From: Jeff Prideaux 6-MAR-1996 14:27:19.34 Subj: Was Hermann Hess a constructivist? Kampis talks about the importance of history in component systems and that history is really technically absent from algorythmic systems. This is basically another statement of the "shuttle principle". A thought popped into my mind as I was reading this. I remember from Hermann Hess's "The Glass Bead Game" that the main character (Magister Ludi) started taking seriously the importance of history (which was basically an unmentionable in Castalia). For this he became a secret radical and eventually left Castilia to "find" something on the "outside". ___________________________________________________________________________________ Date: Wed, 06 Mar 1996 19:48:55 -0400 (EDT) From: DON MIKULECKY Subject: some further dialectic! Note on Layers: > > > > > 05Mar96 18:25 Mikulecky's question to John Wilson > > > > 05Mar96 19:58 John Wilson's reply > > > 06Mar96 11:48 Chandler's comment > > 06Mar96 13:23 Mikulecky's response > 06Mar96 19:12 Chandler's reply Mikulecky's response > Date: Wed, 06 Mar 1996 19:12:52 -0500 (EST) > From: Jerry Chandler > Subject: Efficient Causality and "Fire" > > > > Date: Wed, 06 Mar 1996 11:47:06 -0500 (EST) > > > From: Jerry Chandler > > > Subject: Re: closed to efficient cause > > > > > > On Wed, 6 Mar 1996, DON MIKULECKY wrote: > > > > > > > From: John Wilson - Radiation Biology" 5-MAR-1996 19:58:07.43 > > > > To: MIKULECKY > > > > Subj: once more, with feeling > > > > > > > > > From: DON MIKULECKY 5-MAR-1996 18:25:39.19 > > > > > To: WILSON > > > > > Subj: RE: help > > > > > > > > > > This morning as I was carrying out my customary ablutionary > > > > > rituals, a nagging problem pushed it's way into my thoughts: I > > > > > still have no idea what the phrase "closed under efficient > > > > > causation" means, except that it is VERY important because you use > > > > > it over and over to distinguish organisms from machines. I can't > > > > > tell you how many times I've pondered these words. True, like the > > > > > dinosaurs, I have a brain the size of a walnut, but that's never > > > > > been a disadvantage in the radiology department. Suddenly the room > > > > > was bathed in a strange light and over what I could swear was choir > > > > > music a distant voice intoned, "Don't flatter yourself, your brain > > > > > is the size of a peanut, but I'll give you a hint...think of how a > > > > > computer works." I don't know how long I was "out", long enough to > > > > > squeeze the entire contents of a tube of toothpaste into the sink, > > > > > but I was left with what seems to be such a simple answer that, if > > > > > correct, suggests that even "peanut" was being generous. Does > > > > > "closed under efficient causation" simply refer to the fact that > > > > > certain systems are able to function without the intervention of > > > > > outside forces, that all the information necessary to sustain the > > > > > functional characteristics of the system is contained within the > > > > > system itself (self-sustaining is the descriptive term that comes > > > > > to mind)? In contrast, a system as complex as this computer could > > > > > not send even as simple a message as this unless some old, peanut- > > > > > brained fart told it to. Why? Because it's a machine! I know > > > > > you're busy so here is a selection of answers to my query which you > > > > > can simply refer to by number: > > > > > > > > > > 1. close enough, have a cigar > > > > > 2. close, but no cigar > > > > > 3. hopeless, consider career change > > > > > 4. cigarette, blindfold, "Tombstone pizza, Pierre!" > > > > > > > > #1...that earns a good stogie: > > > > Brings to mind ol' Kipling: "a woman is just a woman, but a > > > > good cigar is a SMOKE!" > > > > Efficient Cause is that cause among the four Aristotelian > > > > Causes which has to do with making the critter. Thus "artificial life" > > > > prorams which duplicate themselves, etc. do not qualify. > > > > The causalities there do not match up as you so cleverly deduced. > > > > If you keep thinking, we'll have to promote you to guru! > > > > > > Huh? > > > Aristotle classified all causes into one of four categories: > > > > > > formal, material, efficient and final. > > > > Right, Rosen bases almost EVERYTHING on these causes. He insists > > that the positivist "how" has steered us of course and that we need to > > go back to the Aristotealean "why". In LIFE ITSELF most of the argument > > is involved in the question of "entailment" and whether or not things > > are "entailed". Entailment is a way of speking about whether a cause has > > been identified or not. Things which are unentailed need external causes. > > Machines have an outstanding need for outside causes, because so much is > > left unentailed. Organisms do not suffer this need. > > But, but, but... > > Rosen's semantics serves him very well! He has a special knack for > clarifying many issues related to biology and I have learned much from > him. But, a basic problem of biology / medicine is the structure of > living systems. In my view, the term 'entailment' or 'to entail' leaves > me with a warm, fuzzy feeling without knowing either how or why. Rosen does not deal in vacuous semantics...shame on you! Everything he says is expressed ALSO in rigorous mathematics. > With regard to an organism's needs, it clearly includes a niche - a niche > implies that it (the species) contributed to the development of the > niche. Isn't that the origins of eco-systems? In life itself he talks briefly about this and evoluition. But it took him three books to get to the point where he could rigorously define organism. I think at this point you had better start giving specific examples of where you think his arguments fall down. Otherwise I am at a loss to understand you! > > > Teleology emerged from the notion of "final" cause. > > > > Not in Rosen's hands! He makes a clear distinction between final > > cause and teleology. In Rosen's way of formulating things, all complex > > systems are "anticipatory". This means that there are "future" events > > incorporated in models WITHIN the system that determine its PRESENT > > behavior. (The cephalic phase of Gasric secretion or the step > > in glycolysis where a substrate for the first enzyme in a chain of reactions > > activates an enzyme downstream). This not simply "teleology, but much more. > > Final cause is a much more sophisticated notion than teleology. Rosen > > wrote an entire book om this idea (Anticipatory Systems) for this reason. > > Anticipatory systems is another remarkable effort. The word > 'anticipatory' is very aptly chosen. But it is a semantical > representation of the phenomena, not a syntactical representation. And, > it does not, to me, get to the issues surrounding emergence which are > associated with genetic systems. You missrepresent again! The book goes through a step by step development of both. I an getting the feeling you are not remembering the content of these books? > With respect to the terminology that "there are 'future' events > incorporated in models within the system that determine the present > behavior", I am rather at a loss for words. The occurrance of phenomenon > is in 'time present'. The potential of 'time present' to form an event > in 'time future' is contextual, is it not? sorry, try this again. I don't follow. look at his biochemical example to see both clear syntax and a real life example of what is meant. > I would prefer to use the term 'organization' (which is related to formal > cause) to express the potential of a system to repetitively generate > comparable states within the flow of time. Perhaps we are saying the > same thing in different vocabularies. no you simply then dissagree with Rosen. On what grounds I do not know > > > Rosen's conclusion with respect to mechanism is basically a philosophical > > > conclusion, is it not? > > > > Hmmmm. Not sure what you mean here? Surely it is philosophical, but if you > > mean ONLY philosophical, then I think somethink has been lost. > > The epistemological and ontological implications of all this are at the level > > of foundations. > > I mean that reasoning is fundamentally philosophical in nature such that > I can not make practical predictions for real-world systems and test the > predictions in laboratory experiments or in clinical trials. I certainly > do not understand category theory, but the level of abstraction of category > theory does not lend itself to dealing with structures of interest to me. > How can we make a bridge between categorical relationships and practical > applications? My suspicion is that the mathematicians will need a great > deal of help to create a link between the mental world and the world of > the laboratory. so in other words, you really are not interested? > > > Secondly, since Aristotle had exactly four densely intermeshed / > > > intertwined categories of causality, I presume that one has to place > > > Rosen's > > > conclusion into the context of the other three causes in order to examine > > > the semantic and syntactical content of his assertion. > > > > Rosen goes through GREAT LENGTHS to convince us that in machines, there are > > only three of the causes at work (final cause is absent) and that they are > > entirely seperable. In complex systems all causes are there and, as you say, > > deeply intermeshed and intertwined. > > Clearly, in Rosen's mind, he has achieved this separation. But, could he > persuade Aristotle? Since Aristotle is not on this net, we should be > cautious in our interpretation of the interrelationships among the four > causes. From other discussions, I have learned that philosophers prefer > "motion" / ""fire" over "structure" (material cause). Structure lacks > pizazz as a philosophical topic. Rosen's usage of 'efficient cause', may > be closely related to the notion of 'change' or 'source of motion' or > 'fire'. The closest scientific term that I know is the term "energy" - > which can be defined as a contributing source of a dynamic. huh? please stop putting words in his mouth. especially words he would never say > Two questions: > Would it be wrong to state that an equilibrium constant is a final cause > for representation of a chemical reaction? yes...no relationship whatsoever > Is Rosen's conclusion exactly equivalent to stating that organisms need > energy from the ecosystem in order to live? (This latter question is > only half in jest! ;-) ) no! > > > If my memory > > > serves me correctly, Rosen sought to bring all of Aristotelian causality > > > under the notion of integration, did he not? > > > > It is the entire basis for all he does > > > But, the time evolution of a system is remarkedly clouded in Rosen's > writings. Fuzzy. Difficult to make connections between real world stuff > and abstract symbolism. another misrepresentation. he talks about the role of dynamics as clearly as any physicist. he also explains carefully why time does not enter relational models > > > Question: Must one also believe in teleology to accept Rosen's conclusion > > > with respect to mechanism? > > > > No, one must use final cause and the notion of "anticipation though. > > Why? Because you can not go from a machine to an organism without it. > > It is precisely what allows an organism to be closed under efficient cause. > > Here the intertwining is crucial! > For alternative views of causality, one should read Mario Bunge's works > on causality. His work has helped me attach to this very difficult term > at a different valence level. I need to look again, but is not Bunge an apologist for positivism? > Cheers -Jerry I hope if we are going to discuss Rosen's ideas further that some care is given to talking about what he actually said! Respectfully, Don Mikulecky ___________________________________________________________________________________ Date: Wed, 6 Mar 1996 21:27:07 -0500 (EST) From: Jerry Chandler Subject: Last Call! Re: some further dialectic! > > > > 06Mar96 11:48 Chandler's comment > > > 06Mar96 13:23 Mikulecky's response > > 06Mar96 19:12 Chandler's reply > 06Mar96 19:48 Mikuleck's rebuttal Chandler's final comments Don: One more response then I really must get on to other things! You are right in that I have not read Rosen's work's recently and should go back and study them again. But please do not mis-understand me. Rosen has done a great job of structuring a detailed argument which has deeply influenced me for several years. But I need help on specific scientific questions with regard to dose - response relationships and drug design. Can you help me over this gap? What am I missing? Other category theorists ( Baas,McLain, Kainen) are equally if not more difficult to read. > > But, but, but... > > > > Rosen's semantics serves him very well! He has a special knack for > > clarifying many issues related to biology and I have learned much from > > him. But, a basic problem of biology / medicine is the structure of > > living systems. In my view, the term 'entailment' or 'to entail' leaves > > me with a warm, fuzzy feeling without knowing either how or why. > > rosen does not deal in vacuous semantics...shame on you! Everything he says > is expressed ALSO in rigorous mathematics. Sorry, Don, but those are feelings I experience. I expect that I would need to develop a deeper understanding of category theory before I could grasp Rosen's meaning. Perhaps you can help the rest of us over this hurdle. I agree that Rosen does not deal in vacuous semantics. Nevertheless, there exists a clear distinction between semantics and syntax (as Jeff and many others point out.) Perhaps we differ in our views of mathematics. I attempt to place all levels of knowledge on equal ground. By this I mean that no superiority is given to knowledge gained from spiritual, mathematical, physical, chemical, biological, medical and other domains. Secondly, high priority is given to prediction as the basis of science and experience. In the medical arts, it is critical that we do not harm others. This, in my opinion forces one to higher standard than exists in the physical sciences. > > > > Teleology emerged from the notion of "final" cause. > > > > > > Not in Rosen's hands! He makes a clear distinction between final > > > cause and teleology. In Rosen's way of formulating things, all complex > > > systems are "anticipatory". This means that there are "future" events > > > incorporated in models WITHIN the system that determine its PRESENT > > > behavior. (The cephalic phase of Gasric secretion or the step > > > in glycolysis where a substrate for the first enzyme in a chain of reactions > > > activates an enzyme downstream). This not simply "teleology, but much more. > > > Final cause is a much more sophisticated notion than teleology. Rosen > > > wrote an entire book om this idea (Anticipatory Systems) for this reason. > > > > Anticipatory systems is another remarkable effort. The word > > 'anticipatory' is very aptly chosen. But it is a semantical > > representation of the phenomena, not a syntactical representation. And, > > it does not, to me, get to the issues surrounding emergence which are > > associated with genetic systems. > > You missrepresent again! The book goes through a step by step development > of both. I an getting the feeling you are not remembering the content of > these books? I certainly do not intend to mis-represent Rosen. Rather, my view of the biochemistry starts with the notion that each chemical molecule is an independent entity. The behavior of glycolysis is very deeply studied at the molecular level. Each of the 14 (?) enzymes has been purified (crystalized?) and studied in detail from both a thermodynamic and kinetic view. Detailed balance equations and kinetic constants have been developed. Consequently, detailed calculations of the free energies and entropies for the reaction chain have been developed. It is probable that substantial knowledge of both the gene DNA sequence and protein sequences exists (but I do not know if this is so or not). Others have formulated differential equations for the biochemistry of glycolysis and simulated the dynamics, which can be, depending on the parameters, highly chaotic. I conclude from these studies: 1. The biochemistry of the 'parts' / 'components' of life tell us very little about the whole organism. 2. As we progress to finer and finer detail, our focus requires us to change languages and adopt the language of the smaller domain. Thus, the scale of language goes from 'glycolysis' to 'enzymes' to sequences of enzymes to individual bond changes within substates to specific vibrational frequencies during bond-breaking processes. 3. The syntactical representation of the the dynamics can be defined in terms directly connected to observables (in the Rosen sense.) 4. In order to place glycolysis into a semantical context which makes sense to the biomedical community, I must 'glue together' a large number of syntactical rules and describe the chemistry in terms of the attributes that have emerged. This to me, is a problem in hierarchy. Is hierarchy fundamentally different from nested 'activation and inhibition'? I am uncertain at this time whether this is an important distinction. What do you think? > > With respect to the terminology that "there are 'future' events > > incorporated in models within the system that determine the present > > behavior", I am rather at a loss for words. The occurrance of phenomenon > > is in 'time present'. The potential of 'time present' to form an event > > in 'time future' is contextual, is it not? > > sorry, try this again. I don't follow. > look at his biochemical example to see both clear syntax and a real life > example of what is meant.] ( I just finished a long discussion of the nature of 'time' and do not have the energy needed to begin another. My current feelings about the nature of time is one of near 'hopelessness'.) > > I mean that reasoning is fundamentally philosophical in nature such that > > I can not make practical predictions for real-world systems and test the > > predictions in laboratory experiments or in clinical trials. I certainly > > do not understand category theory, but the level of abstraction of category > > theory does not lend itself to dealing with structures of interest to me. > > How can we make a bridge between categorical relationships and practical > > applications? My suspicion is that the mathematicians will need a great > > deal of help to create a link between the mental world and the world of > > the laboratory. > > so in other words, you really are not interested? exactly the opposite! The joys of thinking emerge only after tough problems are addressed! > another misrepresentation. he talks about the role of dynamics as clearly as > any physicist. he also explains carefully why time does not enter relational > models But, other category theorists have put time into relational models. An time is critical to the frequencies of cyclic processes. Not being trained in either mathematics or physics, I have a very difficult time defining where the observational world ends and the underlying philosophy of the mathematician enters into the symbols. Experience indicates, however, that philosophy enters into the work of mathematics in ways that are are completely foreign to me as an experimentalist. Very clear examples of this are found in the area of health risk assessments for carcinogensis and environmental toxicity where different mathematicians calculate widely different estimates of risk. Perhaps some day I will have a clearer understanding of the underlying philosophical roots of the Aristotelians, the Platonists and the empiricists, and the implications this has on how they do the mathematics. At present, given the range of claims, I chose to remain a skeptic. I hope that this is not an incurable disease! _________________________________________________________________________________ From: "Jeff Prideaux" 7-MAR-1996 10:50:39.43 Subj: Posting on time > > > 06Mar96 19:12 Chandler's reply > > 06Mar96 19:48 Mikulecky's rebuttal > 06Mar96 21:27 Chandler's final comments Prideaux's comments > > > But, the time evolution of a system is remarkedly clouded in Rosen's > > > writings. Fuzzy. Difficult to make connections between real world stuff > > > and abstract symbolism. > > > > another misrepresentation. he talks about the role of dynamics as clearly > > as any physicist. he also explains carefully why time does not enter > > relational models. > > But, other category theorists have put time into relational models. > An time is critical to the frequencies of cyclic processes. > ... > ( I just finished a long discussion of the nature of 'time' and do not > have the energy needed to begin another. My current feelings about the > nature of time is one of near 'hopelessness'.) I'm curious who these category theorists are (that put time in relational models)!!!! Perhaps one would be Kampis. Kampis, in his book SELF- MODIFYING SYSTEMS IN BIOLOGY AND COGNITIVE SCIENCE, says that time is abstracted away in set theoretical reasoning. That is, in set theoretical reasoning, there is no time distinction between elements in a set. All elements are assumed to already exist. For example, on page 50, Kampis says: " In a set-theoretic definition time cannot be incorporated. A set- theoretic definition makes past and future coexistent by treating then as data'." Kampis admits that all of our modeling tools are based on set-theoretic principles. Then, (given that) the trick is to properly map the time- structured observables over to the time-unstructured set theoretical world. In his Constructivist approach, the encoding serves precisely to do this. The key point to this whole enterprise lies precisely in having the RIGHT encodings. Kampis claims that most "modelers" don't handle time properly and, in essence, throw time away because they don't pay proper attention to the encoding step. On page 74-75 Kampis says: "Take any mathematical description. It operates on classical sets of data...These sets, by being sets and nothing else, do not have any internal structure...Observations, however, are bound to a strong temporal structure. An element of a classical set is available any time; this is not so with empirical observations and natural qualities.... It follows that to form a model is to map structured sets to structure- less sets...In traditional, essentially set-theoretic methods of modeling, this is done in such a way that the respective structure, embodied in the empirical information, and in Nature, IS SIMPLY THROWN AWAY. Thereby WE LOOSE SOMETHING that may be crucial. In the suggested constructive method, the mapping is done in such a way that the structures are eliminated by the encodings between information sets and data sets. The encodings therefore take over the role of the temporal structures. It is these transformations we need to study". Ironically, Kampis uses the term "anticipatory" in a different sense than Rosen. That is, they are not talking about the same thing when they use the term "anticipatory". Rosen, as Don points out, refers to "anticipatory" as an internal predictive model within a system. For example, I (being a legitimate system) can anticipate going on vacation and pack a suitcase. My anticipation of the vacation doesn't involve actual knowledge of the future (for something could come up and I would have to cancel my vacation)...It just involves an internal anticipatory model of a possible future event (the vacation) based on present information. Kampis, on the other hand, uses the term "anticipatory" to refer to equivocations on the part of the modeler in establishing state variables. The mere establishment of a state variable (that can be used to predict the future) requires knowledge that the state is actually defined in the future. Kampis claims that for actual self-modifying systems in biology (and cognitive science) we can NOT be guaranteed that the state variables that currently exist will exist in the future (or that new state variables won't come into existence in the future. Kampis claims that all of our dynamic system descriptions (that involve the concept of state) are only applicable (or usable) for those special systems where the existence of the states actually are time-invariant. Rosen would call such special systems MACHINES or MECHANISMS. As I see it, the whole point of CONSTRUCTIVISM, as Kampis develops it, is to handle time properly in our models. Both Kampis and Rosen say that we probably need to give up the traditional concept of STATE for truly self- modifying systems. Rosen, as I understand him, says that a "closed to efficient cause" system is something other than a mechanism...and the techniques applicable to mechanisms fail for these non-mechanistic systems. Kampis, as I understand him, says that "truly self-modifying systems" defy our conventional set-theoretical reasoning which leads to a dynamical, or state-space description. In short, time is VERY VERY important. And we traditionally have not handled time properly. Both Kampis and Rosen put the emphasis on the encoding and decoding steps in model construction. It is these steps, they say, that have been traditionally over-looked. Of course, the conventional (classical) view is that "life" is just epiphenomena of an underlying mechanistic process...that we are all just elaborate (very complicated) machines...completely understandable by the scientific strategy of reductionism...and modelable by dynamic systems using the concept of state variables. The classical view is that biological systems are NOT "truly self-modifying" in Kampis's sense...and that biological systems are NOT "closed to efficient cause" in Rosen's sense. For the skeptical reader (we all must be skeptical if we are to obtain true understanding...otherwise we are just memorizing and regurgitating) it is a prerequisite to understand clearly what Kampis means by a "truly self-modifying system"...and what Rosen means by "closed to efficient cause". These are closely related (if not the same). I hope to be of some service in the future by sharing my own emerging understanding. ___________________________________________________________________________________ Date: Fri, 08 Mar 1996 10:22:56 -0400 (EDT) From: DON MIKULECKY Subject: time marches on???? Some comments on time, Rosen, etc. This is a follow-up on Jerry's astute comments about the frustrating problem of dealing with time in our models. Also to augment some of Jeff's comments. In "Anticipatory Systems" in the chapter on "The Concept of a Natural System" Rosen reiterates some of what he develops in greater depth in "Fundamentals of Measurement". He begins by talking about "observables" [ I refer you back to an earlier posting where I summarized some of these writings] and their relation to our sensory inpressions, or "percepts". As you may recall, he is able to give us a rigorous definition of what he means by observables. Then he comes to time. As you may also recall, in physics, we are always taught that mass, length, charge and time are the fundamental properties from which all other things such as force, energy, etc. can be derived. Now, mass, length, and charge are observables in the sense that Rosen is speaking. But TIME is not! from p 49: "There now remains one more basic notion to be introduced before we can proceed with a more comprehensive discussion of the linkage relations which can obtain between the observables of a system. This is the notion of TIME, which is of course central to any discussion of change, and hence is a fundamental ingredient in any discussion of change, and hence is a fundamental ingredient in any discussion of natural systems of the kind sketched above. [here he inserts a note and a reference to; Withrow, G. (1980) "the Natural Philosophy of Time", Nelson, London.] "...... "The notion of time is a difficult and complex one. We may recall the plaintive words of Saint Augustine, who in his "Confessions" remarks: "What then is time? If no ond asks me, I know: If I wish to explain to one that asketh, I knoe not." He then points out that our "sense'of time is a percept. It has two aspects: 1) the perception of cotemporality or "simultaneity" 2) the perception of "precidence", involving the distinction between past and future. "It should be noted at the outset that both these aspects of time perception involve RELATIONS bewtwen percepts. For this reason, it does not seem appropriate to treat time directly as a "quality" or observable belonging directly to the external world. Indeed we have interpreted all such quatities as potential capacities for interaction, as manifested especially in the moving of meters. It does not seem that either aspect of time perception is of this character." He then challanges our intuition that time can be "measured". He argues that what we see as a time meter or clock is a refection of our mind's handling of simultaneity and precidence in such a way that we use observables and the relations between them to come up with our concept of time. The clock merely is a generator of labels not a measureing device in the usual sense. this is worth pondering for a while before I go on. More to come! ___________________________________________________________________________________ Date: Fri, 08 Mar 1996 13:19:00 -0400 (EDT) From: DON MIKULECKY Subject: the complexity of complexity > From: "Jeff Prideaux" 8-MAR-1996 11:27:46.15 > Subj: different usages of the word "complexity" > > Some comments on Complexity. > > Although they are (apparently) dealing with the same subject > matter, Rosen and Kampis use the word "complexity" in different > ways. Rosen refers to a dichotomous distinction between "simple" > and "complex" systems... were "simple" implies a single largest > model (or one single way of interacting with the system. Rosen's > "complex" implies the necessity of multiple ways of interacting > with the system...where no single largest model is adequate. > > Kampis, on the other hand, uses the word "complex" in a more > conventional way. "Complex" has to do with the length of the > minimal description of somthing (as related to the observer). > Kampis's main point is that the "complexity" (defined in this way) > of an algorithmic process (or state desription) does not change > over time. Although, the "complexity" of a component-system (or > truly-self-modifying system) can change over time. Kampis then > makes a dichotomous distinction between systems that have a set > fixed complexity and those systems that have a changing complexity. > > In sumary, both Rosen and Kampis make a dichotomous distinction > concering systems. I am currently trying to ascertain whether > Rosen's "Complex" is the same as Kampis's "Changing (or increasing) > complexity".... And whether Rosen's "Simple" is the same as > Kampis's "Unchanging complexity". Jeff seems to be right on the mark here. I like to think that I dealt with this in my review article (however sketchy) in the section where I try to distinguish between the complexity of systems "out there" (of which we have percepts) and the complexity of the formal systems which we tie our precepts to in some sort of modeling relation. My sense is that Kampis, along with Bruce Edmonds (of Pricipia Cybernetica Project [PCP], and others in PCP are focusing on the complexity of formal systems. Note that the Scientific American Article "Complexity or Perplexity" uses the fact that they were able to find over 30 definitions of complexity to poke fun at the whole idea. This,in my way of looking at it, is because the issue is a mess when it comes to formal systems. With numers, for example, they say that a number like 121212121212121212121212 is less complex than a random number because it can be represented in a simpler way: [(12) 12 times] for example. Chaitin is THE expert on what is called "algorithmic" or "computational" complexity. Here it gets cleaned up considerably, but then Chaitin is a genius so what does that prove? Rosen does noy want to wade through that swamp so he goes after what he thinks is the essence:What have we done to our heads? (The punch line from Jerry Farber's "The Student as Nigger" from the 60's: It is not reall y even what Mr. Charlie has done to you. It is what he has done to your head!") By the Newtonian/reductionist/mechanist/positivist (enough?) blinders, we have convinced ourselves that simple machines exist and that they are everywhere we look! Remove those blinders and lo and behold all the world is complex. Remember that in my letter to him, I asked Rosen if he thought EVERYTHING in the natural world is complex. Here's his amswer: "Yes, I've come to believe that all natural systems are complex." Any sign of hidden meaning there?" My heart was gladdened by that answer because it made me believe that I was understanding him. One can get involved in the classification of formal systems if one likes. I prefer to get involved in making a new science which recognizes that simple machines and mechaisms are a construct which has limited use, if any, and which takes off with the idea that we must learn a new way of looking. Any questions? ___________________________________________________________________________________