|
Author
|
Topic: A physics student's view of evolution (response to Barham)
|
Erik
Member
Member # 160
|
posted 20. June 2002 09:06
I made a challenge to James Barham in the thread Darwinism as an Idol of the Theatre, and among other things got a counter-challenge in response. I will address this counter-challenge here. It reads as follows:
quote:
James Barham: Could I challenge you as well---to come up with a discussion of natural selection that has some empirical content, and is not just invoking the mantra? And remember, I am not just talking about speculation about evolutionary scenarios, I am talking about explanations of putative evolutionary scenarios in terms of the RM/NS mechanism, that give a convincing mechanistic explanation without surreptitiously relying upon teleological factors.
Because of the deep misunderstandings of terms like "natural selection" that persist in anti-evolutionary circles, I must begin with some comments on terminology. Consider a population of individuals that (for simplicity) have one of two genotypes, called A and B.
(i) If A-individuals have a greater chance than B-individuals of reaching the reproductive age (and all other things are equal), then A-individuals will have a greater chance of reproducing. Unless B-individuals are produced by genetic necessity in populations of A-individuals, the A-genotype will take over the population. Biologists use the term natural selection to describe the process.
(ii) If A-individuals are generally considered more attractive than B-individuals (and all other things are equal), then A-individuals will have a greater chance of reproducing. Unless B-individuals are produced by genetic necessity in populations of A-individuals, the A-genotype will take over the population. Biologists use the term sexual selection to describe the process.
(iii) If A-individuals tend to produce a larger number of offspring than B-individuals (and all other things are equal), then A-genotypes will take over the population, unless B-individuals are produced by genetic necessity in populations of A-individuals. I must admit that I'm not sure which term biologists use to describe the process, but my guess is "natural selection".
(iv) If A-individuals tend to produce offspring faster than B-individuals (and all other things are equal), then A-genotypes will take over the population, unless B-individuals are produced by genetic necessity in populations of A-individuals. I must admit that I'm not sure which term biologists use to describe the process, but my guess is "natural selection".
(v) If a different species in the ecology acts in some sophisticated way (not just predating or competing for resources) with the net effect that A-individuals have a greater chance of reproducing that B-individuals, then A-genotypes will take over the population, unless B-individuals are produced by genetic necessity in populations of A-individuals. I must admit that I'm not sure which term biologists use to describe the process, but my guess is "natural selection". Biologists use the term artificial selection to describe the process. Some examples may be in order. Example 1: Some ant species breed some species in the superfamily Aphidoidea to collect food for them (alternatively, some species in the superfamily Aphidoidea breed some ant species to protect them). Example 2: There is a species called Homo sapiens that has some peculiar psychological traits. Most individuals of this species find neotenic features enjoyable and cute. This psychological trait is exploited by a number of different species that are kept as pets by H. sapiens. The most extreme case is some populations of Canis familiaris, which are fed, bred and protected by H. sapiens in exhange for walking around a looking cute. This has led to an artificial selection for cuteness and a relief of the natural selection for hunting abilities.
The process of growth of the genotype frequencies through any of the processes (i-v) I call differential reproductive success. I call the speed of this growth (excluding mutation and death rates) fitness. In general, the fitness of an individual depends on its genotype, all other individuals in ecology, and the environment (but this is of course inconvenient to model, so biologists tend make simplifying assumptions). As the populations in the ecology evolve, the fitnesses of different genotypes will, in the general case, change as well. It is important to realize that while fitness is an important factor that influences the trajectory taken through genotype frequency-space, it is not the only one. There is also genetic drift, biomechanical and developmental constraints, externally caused environmental changes, and higher-order fitness effects of which the most spectacular seems to be evolutionary suicide, the non-Darwinian mechanism horizontal gene transfer, etc.
It is also important to understand what is considered established and what is not in the mainstream view of natural history. This view is summed up in Theodosius Dobzhansky's essay Nothing in Biology Makes Sense Except in the Light of Evolution:
quote:
"Let me try to make crystal clear what is established beyond reasonable doubt, and what needs further study, about evolution. Evolution as a process that has always gone on in the history of the earth can be doubted only by those who are ignorant of the evidence or are resistant to evidence, owing to emotional blocks or to plain bigotry. By contrast, the mechanisms that bring evolution about certainly need study and clarification. There are no alternatives to evolution as history that can withstand critical examination. Yet we are constantly learning new and important facts about evolutionary mechanisms."
Thus, what is generally called "the fact of evolution" is that living organisms have been transformed over time and that they share a common ancestry. The currently known mechanisms are constitute our best explanation, but I think biologists would be open to adding new mechanisms to the current evolutionary biology if a good reason came up.
After this long introduction, it is time to figure out exactly what I'm supposed to do and do it. I must remember to not just speculate about evolutionary scenarios. I must write about explanations of putative evolutionary scenarios in terms of the RM/NS mechanism, that give a convincing mechanistic explanation without surreptitiously relying upon teleological factors. Creationists are reliable in the sense that they can be counted on make assertions about a barrier between "microevolution" and "macroevolution". Genealogy, transformations, and differential reproductive success are accepted in the form of "microevolution" and denied in the form "macroevolution", and the difference between the two is impossible to define for creationists. Anti-evolutionary philosophers like James Barham are a different matter. Barham seems to accept the fact of evolution, but claims that the current evolutionary models are incoherent. Is it therefore enough to show that differential reproductive success really is an important mechanism determining the trajectory through genotype-frequency space? It seems to me that it is, so that's were I'll begin.
It is trivial to prove that differential reproductive success is a coherent idea, because it has been simulated on a computer and that allows the us to precisely specify what kind of interactions that may happen. This has been done by people working with artificial life. One of the many examples is the Tierra platform, which is a platform that allows self-replicating computer programs to compete for the finite system resources. Everything that happens in Tierra simulations is driven by truly (pseudo-)random mutations, differential reproductive success, and some limited ecological interactions. Ecologies of hosts, parasites, and hyperparasites have been shown to emerge in Tierra simulations. Mutations (by that I mean any change that is directly passed on to offspring) and differential reproductive success happen in biological populations too and is understandable in terms of known mechanisms. (Whether or not biological mutations are biased in some direction is a fairly uninteresting question. Our current models of evolutionary dynamics does not require such biases, but there is room to add such biases to the models if a good reason should come up. Until then, mutations will be considered biased in the sense that they may depend on external factors and that they are not drawn from uniform probability distributions, but unbiased in other senses.)
I thereby consider myself to have shown that mutations and differential reproductive success occur in biological populations. When it comes to large-scale evolution (i.e. evolution above the species level), quantitative models based on conventional mechanisms are emerging. The startling thing about these models is how much they explain and how few mechanisms they invoke (the beauty of mathematical modelling, no doubt). It seems to have begun in 1993 with the Bak-Sneppen model, which is a toy model for explaining the power-law feature of paleontological extinction curves in terms of the internal evolutionary dynamics. Since then more sophisticated and realistic models have been formulated. In some of them, the evolutionary dynamics results in self-organized criticality, but not in all (a subdebate is whether. The fact that we have evolutionary models that generate the general features of extinction curves and food webs is prime evidence that the known mechanisms are major factors in large-scale evolution. No existing non-evolutionary models are able to do that.
I will end here for now, and let Barham set the direction for the future discussions in this thread by disagreeing with parts of this posts.
Erik
PS. Oh, I almost forgot. All thread-initiating posts must make a positive contribution to the ID movement. Based on my above post, I conclude that Barham's conception of evolutionary change is misguided and that ID advocates better look for models that give the standard evolutionary mechanisms more attention than Barham's ideas, since they are demonstrably important factors in evolution. DS.
[ 20 June 2002, 09:15: Message edited by: Erik ]
IP: Logged
|
|
Erik
Member
Member # 160
|
posted 20. June 2002 09:45
Since I brought up the models of evolving ecologies that are mainly the creation of the physics community, I should also point out that the physicists working with SOC and non-SOC models of evolving ecologies are pretty confused about the meaning of "fitness" and "punctuated equilibrium". For instance, the "fitness" of the Bak-Sneppen model is badly misnamed because it is not the fitness of evolutionary biology (a better name is "ecological adaptedness"). This confusion is restricted to inappropriate naming of parameters and some attacks on straw-men. The models themselves are fine.
Erik
IP: Logged
|
|
James A. Barham
Member
Member # 50
|
posted 20. June 2002 11:24
Erik:
I am interested to discover that we have more areas of agreement than I would have guessed. I, too, find the work on self-organized criticality extremely suggestive, although as pointed out elsewhere on another thread, it cannot be applied directly to living systems, since it is just a result of energy minimization under particular constraints.
As a physicist, would you at least agree that living cells are doing something other than merely minimizing energy? If so, how would you explain the overall coordination of the many thousands of individual chemical reactions in the cell? Did all of this really happen "by chance", even given a stepwise Darwinian ratchet? (I don't deny the ratchet, by the way, I say it is fundamentally teleological in character.) Surely, as a physicist, you grasp the problem here.
Now, obviously, we differ violently on our interpretations of the significance of the new work in physics for evolutionary theory. I don't especially want to repeat everything I've said on other threads all over again, so let my question to you in the previous paragraph suffice for now to shape the future discussion between us (if that is all right with you).
Finally, I must respond to two of your remarks, in particular. First, I am not anti-evolution, I am anti-Darwinian. There is a distinction.
Second, I love it how Darwinians always say that their critics are just too stupid to grasp the subtleties of natural selection. Would you say that Rene Thom, winner of the Fields Medal and by common consent a brilliant and creative mathematician (whatever one thinks of the applications of his ideas), is too stupid to understand Darwinian thinking? Here is what he has to say:
"If it is true that Darwinism is the only theoretical area in biology, it's because it is the only one to introduce a virtual, the ensemble of possible evolutions of a species in a given time and place. But this virtual is uncontrolled, one can say nothing about it . . .
"Darwinism thus offers the example of a theory that everyone can understand, which is the obvious reason for its success. But as soon as one wishes to make precise, to however small an extent, the concepts it uses . . . all of these concepts apparently so simple dissolve under analysis into a reality of a complexity defying all description. There is no doubt that precision and rigor are necessary to clarify the basic postulates of the Darwinian schema. One may legitimately wonder, however, what will be left of them on the day when this clarification will be accomplished." ("Darwin apres cent ans," Apologie du Logos, Paris, Hachette, 1990, pp. 599--605; the passages cited are on p. 599 and on p. 605; my translation)
[ 20 June 2002, 11:34: Message edited by: James A. Barham ]
IP: Logged
|
|
Jesse
Member
Member # 112
|
posted 20. June 2002 22:53
quote:
If so, how would you explain the overall coordination of the many thousands of individual chemical reactions in the cell? Did all of this really happen "by chance", even given a stepwise Darwinian ratchet? (I don't deny the ratchet, by the way, I say it is fundamentally teleological in character.) Surely, as a physicist, you grasp the problem here.
Funny you should bring this up--I've just been reading about the topology of protein interaction networks, and what it may tell us about their evolution, in Linked: The New Science of Networks by Albert-Laszlo Barabasi. Apparently when you map out all the different reactions, you find a "scale-free" network similar to what we see when we map the topology of other self-organized networks, like social networks or the internet. In scale-free networks, the number of links per node also follows a power law, so a few nodes--the "hubs"--have a lot of links (corresponding to proteins that interact with a lot of other proteins), but they're rarer than those with fewer links. The author explains the current general theory of how these scale-free networks arise: it involves starting from a small network and adding new links gradually, which naturally results in older nodes having more links than younger ones. To this one must also add the idea of "preferential attachment"--new nodes are more likely to link to nodes that already have a lot of connections, just like new webpages are more likely to know about, and link to, webpages that already have a lot of links to them. Together, the ideas of growth and preferential attachment are enough to reproduce the features of scale-free networks, although one can also include other features in the model like different nodes having different "fitness" values (some webpages are more likely to be linked to because they're better/more useful than others).
So, in explaining the topology of the protein interaction networks, the author suggests an evolutionary model which includes growth and preferential attachment:
quote:
Taken together, the similar large-scale topology of the metabolic and protein interaction networks indicated the existence of a high degree of harmony in the cell's architecture: Whichever organizational level we examine, a scale-free topology greets us. These journeys within the cell indicate that Hollywood and the Web have only rediscovered the topology that life had already developed 3 billion years earlier. Cells are really small worlds that share the topology of many other nonbiological networks, as if the architect of life could design only these.
How did life arrive at this architecture? Almost as soon as we asked the question, we had an answer. Approximately a half year after the publication of our findings on the topology of the protein interaction network, I received three e-mails within about a month. Each of them contained a manuscript by a different research group. Amazingly, each of the three research groups independently offered the same simple and elegant explanation, claiming that the cell's scale-free topology is a result of a common mistake cells make while reproducing.
Cells reproduce by duplicating their content and dividing into two. The details of these processes may vary for simple bacteria and more complex human cells. Certain steps are universal, however. First, in order to produce a genetically identical offspring cell, the DNA must be faithfully replicated. This process is not free of errors, however. Although the cell's intricate copying mechanism insures that DNA sequences are inherited with extraordinary fidelity, about one letter in a thousand is randomly changed every 200,000 years. Another common error is gene duplication. Through a rare accident in the copying process, gene duplication can occur when the ends of broken DNA molecules join together. As a result, segments of varying length of the parent DNA will appear twice in the offspring's genome. Such copying mistakes sometimes kill the cell. In other cases, multiple copies of the same gene have evolutionary advantages and are passed on to future generations. Hemoglobin is a well-known example.
Originally cells had only one hemoglobin gene. About 500 million years ago, during the evolution of the higher fish species, a series of gene duplications occurred, resulting in four copies of the hemoglobin gene scattered along the genome. Today each of these genes encodes one of the four components of the hemoglobin protein complex.
Gene duplication has a significant impact on the cellular network. It results in two identical genes, which produce identical proteins, that in turn interact with the same protiens. A new node thus has been created, the protein generated by the duplicated gene. Its neighbors, the protein with which the duplicated protein interacts, will each now interact with both the parent and the identical offspring protein. Therefore, each protein in contact with the duplicated protein gains an extra link. In this game highly connected proteins have a natural advantage: They are more likely to have a link to the duplicating protein than their weakly connected cousins. It's not that hubs duplicate more often. Rather, since the hubs are in contact with more proteins, they are more likely to have a link to the duplicating node, which offers them an extra link, a subtle version of preferential attachment.
The most important feature of this explanation is that it traces the origin of the scale-free topology back to a well-known biological mechanism, gene duplication. It does so by showing that gene duplication can simultaneously lead to both the growth of the protein network by adding an extra protein and to preferential attachment by adding new links at a higher rate to the more connected proteins. It is too early to determine if this is the only explanation, since it is conceivable that different mechanisms, yet unexplored, could generate the same topology. It is unclear if it explains the scale-free structure seen in the metabolism, as well. Nevertheless, it demonstrates that mechanisms present in the cell can generate the scale-free topology.
All this suggests that protein interaction networks and metabolic networks grew in a haphazard way like the internet, through the constant addition of new nodes/links to an existing network, rather than being designed all at once in a top-down manner.
Some papers on the subject of scale-free networks, including one on protein networks (#14 on the list) can be found at:
http://www.nd.edu/~networks/papers.htm
And a paper on how this network may have evolved through gene duplication can be found here:
http://www.santafe.edu/sfi/publications/wpabstract/200108041
And here's a more popular article on the study of scale-free networks:
http://www.ukpoliticsmisc.org.uk/usenet_evidence/scale_free.htm
[ 20 June 2002, 23:03: Message edited by: Jesse ]
IP: Logged
|
|
James A. Barham
Member
Member # 50
|
posted 21. June 2002 08:13
Jesse:
Thanks for your very interesting post. I have the book, but haven't had a chance to look into it much, yet.
I, too, am extremely interested in physical approaches to the cell. But, in the context of this thread and this forum, the larger question is, What do such approaches mean for standard Darwinian theory?
There are two points I would like to make. First, the most important question is whether such ideas allow us to reduce the cell to pure mechanism, or whether a non-mechanistic (note, I am NOT saying non-natural) principle needs to be postulated, as well.
I personally believe that all analogies to computers and manmade networks, while they may throw light on the problem, at the end of the day will not do the job. Why do I believe this? Simply because all manmade machines are inherently brittle because there is no connection between the matter itself and the functional state. In a manmade machine, WE have imposed the functional configuration on the matter, so there is nothing internal to the matter that is opposing the second law in degrading the functional configuration. I have expressed this by saying that "machines just don't give a damn" (the Rhett Butler Problem).
Well, what would it mean for matter to "give a damn"? It would mean that there is something about living matter itself that is causing it to spontaneously strive to preserve its functional state (AKA as "being alive"). I think it ought to be obvious that something like this is required, just on statistical mechanical grounds, even if we don't yet understand it.
Here is an example. What we are looking for is a principle to explain the coherence of life. In a laser, we get coherence of light by manipulating the boundary conditions, pumping in energy from the outside, and fine tuning everything. So the laser would be a good analogy of the coherence of life from an ID point of view.
Now, the late Italian physicist Giuliano Preparata developed a theory of "superradiance" in which certain states of matter can "lase" internally, without population inversion, and also in a self-contained way via reflection at the boundary (this is crudely put---I am no physicist---see "QED Coherence in Matter", Singapore: World Scientific, 1995; see, also, Charles P. Enz, "On Preparata's Theory of a Superradiant Phase Transition," Helvetica Physica Acta, 1997, 70: 141--153).
Now, I am not saying that Preparata's theory is right or wrong (that would be absurd of me to try to do). But I am saying that IF it (or something similar) is right, then that gives us a handle on a form of coherence that is internal to certain states of matter and spontaneous. If life were "superradiant", rather than laser-like, then we could begin to understand how autonomy and purpose are possible from a naturalistic point of view.
The second point I mentioned above is the following. The Darwinian may well ask me, OK, say we accept that a naturalistic explanation of purpose based on QFT (as opposed to the machine metaphor) is possible. So what? This is still something we can accommodate within selection theory.
My reply is, Sure ANYTHING can be explained by selection theory. You could "explain" nucleosynthesis by saying that the more complex atoms are "selected" inside stars. But my point is that chemistry constitutes a deeper explanation of the properties of the periodic table than such a selectionist pseudo-explanation would. In the same way, a physical explanation of the intelligent agency of life would constitute a deeper explanation than the pseudo-explanation of natural selection.
[ 21 June 2002, 08:16: Message edited by: James A. Barham ]
IP: Logged
|
|
Moderator
Administrator
Member # 1
|
posted 21. June 2002 13:44
Erik said:
quote:
PS. Oh, I almost forgot. All thread-initiating posts must make a positive contribution to the ID movement.
I just want to clarify. All threads-initiating posts must make a positive argument regarding complex systems. The argument need not be in defense of intelligent design. What we do not want is a thread where the only interest is in critiquing other arguments.
IP: Logged
|
|
Mike Gene
Member
Member # 149
|
posted 22. June 2002 11:43
Jesse: All this suggests that protein interaction networks and metabolic networks grew in a haphazard way like the internet, through the constant addition of new nodes/links to an existing network, rather than being designed all at once in a top-down manner.
Of course there is much middle ground between these two extremes. That is, original cells could have been designed all at once in a top-down manner front-loaded to evolve in such a manner. For example, consider the bottom line here: "The most important feature of this explanation is that it traces the origin of the scale-free topology back to a well-known biological mechanism, gene duplication. " Yet I have explained in the past why gene duplication play such a pivotal role in front-loaded evolution. In this case, the designer chooses a set of genes that will serve as the stem population from which the duplicates will arise. Evolution is thus necessarily channeled. Furthermore, don't be confused in thinking that gene duplication as an "accident" means it is a simple affair. Gene duplication depends on the coordinated activity of some rather sophisticated processes. An explanation that begins by assuming gene duplication from the start leaves out the most interesting piece - the origin of the gene duplication process and its stem population.
BTW, while the concept of scale-free networks applies nicely to metabolic pathways, I'm not sure how much help it is with molecular machines (like the flagellum). But that's another topic.
[ 22 June 2002, 11:44: Message edited by: Mike Gene ]
IP: Logged
|
|
Erik
Member
Member # 160
|
posted 24. June 2002 14:50
quote:
James Barham: I am interested to discover that we have more areas of agreement than I would have guessed. I, too, find the work on self-organized criticality extremely suggestive, although as pointed out elsewhere on another thread, it cannot be applied directly to living systems, since it is just a result of energy minimization under particular constraints.
If you look at the Bak-Sneppen SOC model, it's more sophisticated successors, and competing non-SOC models, you will find that none of them is based on energy minimization. They are based on very simple postulated statements about evolution and ecological interactions.
quote:
James Barham: As a physicist, would you at least agree that living cells are doing something other than merely minimizing energy? If so, how would you explain the overall coordination of the many thousands of individual chemical reactions in the cell? Did all of this really happen "by chance", even given a stepwise Darwinian ratchet? (I don't deny the ratchet, by the way, I say it is fundamentally teleological in character.) Surely, as a physicist, you grasp the problem here.
As a mathematics and physics student, I think it depends on the level of detail. I am one those people who is not ashamed to be open to the possibility that reductionism holds in our universe. If someone wanted make an in principle argument that organisms can be described by a state vector that changes in time according to the Schrödinger equation (or whatever equation of motion that may replace it), then I would have no conceptual objection. I agree, however, that organisms are doing something more than energy minimization if viewed macroscopically.
As for the second question, I don't know whether to interpret it as a question about the origin of life or as a question about the subsequent biological evolution. If the latter, then I think Jesse gave a good answer. By now there have been lots of articles published on growing networks and their statistical properties. Duplications of genes, chromosomes and genomes provide well-established mechanisms for growth of biochemical networks and therefore the growing network models are applicable under the assumption that duplications really were a major mechanism in the evolution of biochemical networks. The fact that the predications fit our data is a provisional validation of the assumption. If, on the other hand, you are asking about the origin of life, then I don't know. Partly because I have not studied the OOL literature in detail and partly because there are no complete and well-established answers available.
quote:
James Barham: Now, obviously, we differ violently on our interpretations of the significance of the new work in physics for evolutionary theory. I don't especially want to repeat everything I've said on other threads all over again, so let my question to you in the previous paragraph suffice for now to shape the future discussion between us (if that is all right with you).
Finally, I must respond to two of your remarks, in particular. First, I am not anti-evolution, I am anti-Darwinian. There is a distinction.
Exactly what does it mean to be "anti-Darwinian"?
quote:
James Barham: Second, I love it how Darwinians always say that their critics are just too stupid to grasp the subtleties of natural selection. Would you say that Rene Thom, winner of the Fields Medal and by common consent a brilliant and creative mathematician (whatever one thinks of the applications of his ideas), is too stupid to understand Darwinian thinking? Here is what he has to say:
This is the ISCID, not the Society for Fields Medal Winners. Considering the context (e.g. you had just posted comments showing a misunderstanding of natural selection), I do not think that my introduction was an ad hominem argument. I was trying to minimize the number of exchanges necessary to explain terminology, which, given your own suggestions that natural selection as an idea is incoherent, seems like an eminently reasonable thing to do. I have made no statement about your intelligence nor about the intelligence of other members of the ID movement.
As for Rene Thom, I am not familiar with his writings. I am, however, quite capable of drawing my own conclusions about the basic concepts of evolutionary biology and therefore I am not even slightly intimidated by the fact that Thom has won the Fields Medal. The Fields Medal is awarded to people who are brilliant mathematicians without regard for their understanding, or lack thereof, of evolutionary biology. Evolutionary biology is not an area that is untouched by mathematicians. A number of accomplished mathematicians have made contributions. Sewall Wright and Ronald Fisher were quite mathematically minded persons, for instance. Even John von Neumann and Gregory Chaitin did some work on evolution (although Chaitin's work is a bit simplistic). There are also a number of journals devoted to mathematical modelling in biology, and at least some of them are authored by mathematicians writing about biological evolution. So why should I care if you can find one excellent mathematicians who perhaps disagree with all, or parts of, our current evolutionary biology?
quote:
"If it is true that Darwinism is the only theoretical area in biology, it's because it is the only one to introduce a virtual, the ensemble of possible evolutions of a species in a given time and place. But this virtual is uncontrolled, one can say nothing about it . . .
"Darwinism thus offers the example of a theory that everyone can understand, which is the obvious reason for its success. But as soon as one wishes to make precise, to however small an extent, the concepts it uses . . . all of these concepts apparently so simple dissolve under analysis into a reality of a complexity defying all description. There is no doubt that precision and rigor are necessary to clarify the basic postulates of the Darwinian schema. One may legitimately wonder, however, what will be left of them on the day when this clarification will be accomplished." ("Darwin apres cent ans," Apologie du Logos, Paris, Hachette, 1990, pp. 599--605; the passages cited are on p. 599 and on p. 605; my translation)
If you want me to judge whether or not Thom understands the basic concepts of evolutionary biology, then you have provided insufficient information. It seems from the translated quotes that he is view not favourable to mainstream evolutionary biology, but no argument is provided in the quotes. If all Thom wants is to have mathematical definitions of key concepts, then that can be provided. von Neumann's work provides another formal framework, as do the more experimental computer simulations of self-replicating computer programs.
quote:
James Barham (in reply to Jesse): The second point I mentioned above is the following. The Darwinian may well ask me, OK, say we accept that a naturalistic explanation of purpose based on QFT (as opposed to the machine metaphor) is possible. So what? This is still something we can accommodate within selection theory.
Do you realize that quantum field theory is a mechanistic theory?
quote:
James Barham (in reply to Jesse): My reply is, Sure ANYTHING can be explained by selection theory. You could "explain" nucleosynthesis by saying that the more complex atoms are "selected" inside stars. But my point is that chemistry constitutes a deeper explanation of the properties of the periodic table than such a selectionist pseudo-explanation would. In the same way, a physical explanation of the intelligent agency of life would constitute a deeper explanation than the pseudo-explanation of natural selection.
You are confusing different levels of explanations and models. I think it is perfectly legitimate to hold the view that biology, ecology, and evolution can be reduced to particle physics (although I am surprised to see this mechanistic and extremely reductionistic view coming from someone who is concerned about the "deleterious impact of Darwinian thinking, and mechanistic and reductionist thinking generally"). I also agree that if this is the case, then it would be a deeper explanation than the current evolutionary biology. That implication is uncontroversial. But evolutionary biologists wouldn't become redundant any more than chemists became redundant with the arrival of quantum physics and statistical physics.
Most of your posts that I've read leaves me with the impression that the main cause of your disagreement with evolutionary biology is that you have misunderstood the terminology used and the scope of the models.
Erik
IP: Logged
|
|
James A. Barham
Member
Member # 50
|
posted 25. June 2002 19:18
Erik:
I guess that what counts as a satisfactory explanation is relative to one's interests. My main interest is in understanding how purpose and intelligence fit into the greater scheme of things.
Now, most Darwinians seem to believe that the theory of natural selection has made it possible to explain the structure, physiology, and behavior of living things without resorting to purpose (or teleology).
My claim is very simple. It may be summarized as follows:
(1) A trait must already exist before it can be "selected."
(2) Therefore, the functional or purposive or teleological (or whatever you want to call it) aspect of the trait must already exist before it can be selected.
(3) Since no theory can explain what it presupposes, the theory of natural selection cannot explain the teleological character of biological phenomena.
You claim that I "misundersand" natural selection, but nowhere do you actually address my main claim. Is it that you dispute the simple point that I am making, or is it that you simply do not feel that it is relevant?
One way the Darwinian might want to reply is to say that, of course, natural selection does not explain the functional organization of the cell---that is what molecular biology etc. are for. The only problem with this reply is that molecular biology does not explain the functional organization of the cell, either. It only explains individual mechanistic pieces, not how they can possibly be coordinated together in the fantastically complicated way that they are.
As for QFT and "mechanism," I guess it depends on what we mean by "mechanistic." I simply mean a linear, or one-to-one mapping of initial states ("causes") onto final states ("effects"). This means that mechanism is by definition time symmetric (reversible).
Any theory that encompasses nonlinearities will be by this definition non-mechanistic. Both nonlinear dynamics and QFT (not quantum mechanics, mind you) incorporate nonlinearities, thus providing for many-to-one or "equifinal" phenomena which are time asymmetric.
Maybe you don't like my calling this "nonmechanistic." But the real issue is the one regarding determinism and reductionism. Most high-energy physicists are reductionists, but most condensed-matter physicists are not. The latter (e.g., Philip Anderson, Robert Laughlin, Max Dresden, Michael Fisher, Kenneth Wilson, and others) tend to believe that reality consists of emergent levels, each with some sui generis order parameters that make the operative laws non-deducible from laws at lower levels.
What I believe is necessary to understand the teleology that is immanent in life is some form of cooperativity deriving from living matter (presumably, having something to do with the immense size of proteins) in much the same way that rigidity is a cooperative property of the crystal lattice. As Philip Anderson likes to say, "More is different." Therefore, reductionism is false.
IP: Logged
|
|
Erik
Member
Member # 160
|
posted 27. June 2002 04:42
quote:
James Barham: I guess that what counts as a satisfactory explanation is relative to one's interests. My main interest is in understanding how purpose and intelligence fit into the greater scheme of things.
It seems that metaphysics rather than science should be your playground then.
quote:
James Barham: Now, most Darwinians seem to believe that the theory of natural selection has made it possible to explain the structure, physiology, and behavior of living things without resorting to purpose (or teleology).
Yes, most evolutionary biologists think that evolutionary biology makes it possible to explain the structure, physiology, and behaviour of living organisms without assuming an artificially imposed purpose. They probably don't care about metaphysical kinds of intrinsic purpose, except to note that they do not require it.
quote:
James Barham: My claim is very simple. It may be summarized as follows:
(1) A trait must already exist before it can be "selected."
No, that's too idealized to be correct. For most traits, the line between existence and non-existence is fuzzy, because they are the result of interactions between many genes and there is a whole multi-dimensional spectrum of similar traits. Where's the line between the sprawling gait of early synapsids and the more upright gait of modern mammals (Paleobiology 27 : 14-38), for instance?
quote:
James Barham: (2) Therefore, the functional or purposive or teleological (or whatever you want to call it) aspect of the trait must already exist before it can be selected.
I want to call it functional, because that is the correct word. ("Functional" is not a synonym for "purposive" or "teleological"!! Anyone is of course free to argue that all or some functional structures are purposive or teleological, but to take the words themselves as synonymous is to beg the question.) Second, I take it that you mean that a trait must be both functional and present in a population before it can be selected. That depends a bit on what you mean by "selected". Think of selection as a factor in deciding the direction for evolution and mutations as taking steps in this direction (steps are of course in general taken in all directions, but those that are in the wrong directions are discarded). The direction set in this way can lead to traits that are functional, but not yet presented in the population.
quote:
James Barham: (3) Since no theory can explain what it presupposes, the theory of natural selection cannot explain the teleological character of biological phenomena.
The theory of differential reproductive success mainly sets out to explain the preferred directions leading from and to traits. The actual steps in those directions are assumed to be unbiased in most ways because biases don't seem to be required.
quote:
James Barham: You claim that I "misundersand" natural selection, but nowhere do you actually address my main claim. Is it that you dispute the simple point that I am making, or is it that you simply do not feel that it is relevant?
Well, I did provide an introduction where I tried to explain the basics of differential reproductive success. I do dispute the simple point you are making, because it relies on unrealistic idealizations.
quote:
James Barham: One way the Darwinian might want to reply is to say that, of course, natural selection does not explain the functional organization of the cell---that is what molecular biology etc. are for. The only problem with this reply is that molecular biology does not explain the functional organization of the cell, either. It only explains individual mechanistic pieces, not how they can possibly be coordinated together in the fantastically complicated way that they are.
I have not studied cell biology very much, so I don't know to which extent it is explained and unexplained. I note, however, that the models of network evolution is a very large step toward explaining the evolution of the cell's organization.
quote:
James Barham: As for QFT and "mechanism," I guess it depends on what we mean by "mechanistic." I simply mean a linear, or one-to-one mapping of initial states ("causes") onto final states ("effects"). This means that mechanism is by definition time symmetric (reversible).
Curiouser and curiouser! How extraordinarily absurd to write "mechanistic" when you actually mean a linear or one-to-one mapping of states to states. You might as well write "red" when you mean "green", or "tall" when you mean "fat". Where to begin?
First, a linear mapping is not the same as a one-to-one (=injective) mapping. The mapping
f(x,y) = x - y (x,y real numbers)
is linear, but not one-to-one. The mapping
g(x) = x * x * x (x a real number)
is one-to-one, but not linear. I take it, therefore, that you label any mapping that is linear or one-to-one by "mechanistic".
quote:
James Barham: Any theory that encompasses nonlinearities will be by this definition non-mechanistic. Both nonlinear dynamics and QFT (not quantum mechanics, mind you) incorporate nonlinearities, thus providing for many-to-one or "equifinal" phenomena which are time asymmetric.
Pretty much all physical theories encompass nonlinearities. The equation of motion may be a linear differential equation (as in QM and pre-relativistic classical mechanics), but there can still be plenty of nonlinearities. One theory that is definitely not mechanistic, by your unconventional redefinition, is general relativity. Another is population genetics! Classical thermodynamics is yet another example.
quote:
James Barham: Maybe you don't like my calling this "nonmechanistic."
Indeed, I find such useless and misleading word games unbefitting of someone who claims to do science.
quote:
James Barham: But the real issue is the one regarding determinism and reductionism. Most high-energy physicists are reductionists, but most condensed-matter physicists are not. The latter (e.g., Philip Anderson, Robert Laughlin, Max Dresden, Michael Fisher, Kenneth Wilson, and others) tend to believe that reality consists of emergent levels, each with some sui generis order parameters that make the operative laws non-deducible from laws at lower levels.
So it's not really quantum field theory, but rather condensed-matter physics, that you hope will provide a basis for your ideas?
Erik
IP: Logged
|
|
James A. Barham
Member
Member # 50
|
posted 27. June 2002 10:37
Erik:
Your wrote:
"It seems that metaphysics rather than science should be your playground then."
I have never claimed to be "doing science." What I am trying to do is absorb as much science as I can to come up with a better understanding of the place of purpose in nature.
As for metaphysics being my playground, I will make a deal with you. When the Darwinians stop doing metaphysics, so will I.
On the meaning of the word "function." In biology, at least (not in mathematics, of course), the clear implication of "function" is something that happens for a purpose. There is no getting around this, and you are the one indulging in word games, not I.
I appreciate the help with the math. I will compare what you have written with the writings of Ian Stewart, Ivar Ekelund, Bruce J. West, Robert Rosen, Alwyn C. Scott, and others, from whom I have gleaned my admittedly rudimentary understanding of these matters. I don't doubt that I may have gone wrong, but I want to see better exactly where. Your corrections were a little too peremptory to put my mind completely at ease.
In one respect, I can already see where I went wrong. I was initially reacting, I believe, to your statement that you were willing to countenance the in-principle reduction of everything to the Schrodinger equation. That is one issue---reductionism (epistemic) or Laplacian determinism (ontological). The status of life is a separate issue, which my response probably blurred. Of course, the falsity of Laplacian determinism is proven by the existence of time asymmetry, friction, symmetry breaking, stability, and a host of phenomena that have nothing to do with life. How to account for life is an additional problem. But first we have to clear the decks by getting Laplacian determinism (which is often unconsciously assumed) out of the way.
Then, once we admit a variety of emergent levels, each with is own sui generis forms of dynamics and stability, we can begin to ask the question, What is required to make sense of the intelligent agency of life? Here is where some people (Mae-Wan Ho, Giuseppe Vitiello) have speculated that quantum coherence may be required in a new way that we do not yet really understand. It is not that life is "reducible" to QFT in its present form, it is that QFT may need to be extended in such a way as to encompass the sort of long-range coherence and coordination we find in living matter.
As for your distinction betweeen QFT and condensed-matter physics, unless I am very much mistaken (always a possibility!) a glance into any condensed-matter physics textbook will show that QFT is widely used in that discipline. Maybe I am missing something here, and not invoking QFT in quite the right way, or with the right qualifications. If so, a genuine attempt to explain where I am going wrong would be much appreciated.
The context I am working within here is provided by Robert Batterman, Mae-Wan Ho, Giuseppe Vitiello, Tian Yu Cao, Silvan Schweber, Philip Anderson, Robert Laughlin, and others.
[ 27 June 2002, 11:05: Message edited by: James A. Barham ]
IP: Logged
|
|
James A. Barham
Member
Member # 50
|
posted 27. June 2002 18:14
Erik:
I have spent the day going through various books, trying to think through the issues you raise, and the objections you make to my statement regarding "mechanism." If we are to move the debate beyond a shouting match, I think we need to try to disentangle various issues. This is the best I have been able to come up with, as a layman. See what you think of the following as a "typology" of problems all lurking within the problem of "mechanism," or in its vicinity:
(1) Determinism. Is it sensible to believe that everything that exists will ultimately be reducible to an equation that can be written on a T-shirt ("Theory of Everything")? What would the evidence for such a belief look like? The main evidence for determinism traditionally has been the fantastic success of Newtonian mechanics, yet Poincaré showed over 100 years ago that already the 3-body problem is nonintegrable, meaning, I gather, that it cannot be solved using Newtonian methods (because nonlinear equations are neither analytically solvable themselves, nor resolvable into linear components). If even the 3-body problem defeats the theory that is our best reason for believing in determinism in the first place, what reason do we have left? I say determinism is wrong.
(2) Math and reality. This raises the issue of the relationship between our mathematical equations and reality itself. Should we say that apparent failures of our equations to capture reality are merely due to our own epistemic limitations (maybe because of practical limitations on how much information we can gather, or on how precisely we can make measurements), or do they point to the fact that the behavior of reality is richer than our equations are allowing for? A closely related issue is whether failures in prediction (as in chaos theory) are only practical limitations (epistemic) or are in-principle failures (ontological) due to the combination of sensitivity to initial conditions together with quantum uncertainty. I say that our empirical findings should drive our metaphysical interpretation of our equations, not the other way around.
(3) Quantitative vs. qualitative techniques. Poincaré introduced the use of qualitative or topological methods to characterize dynamical behavior globally when Newtonian methods failed. These methods have exploded over the past three decades with the introduction of computer modelling. Is this a stopgap, or is it a new and legitimate way of doing science? I say that qualitative techniques are legitimate, and will no doubt continue to be developed into increasingly powerful scientific tools.
(4) The arrow of time. The fundamental equations of physics (both Newtonian mechanics and QM) are time invariant. And yet the universe seems to clearly have a definite arrow of time, which manifests itself in the Second Law for closed systems, in the cosmic expansion, in evolutionary complexification, in the anticipatory behavior of living things, and in countless other ways. How to reconcile these facts with fundamental physics? I don't have any idea, of course, but I say the arrow of time is real. It is no illusion.
(5) Friction. What is the role of energy dissipation in structuration or complexification, in both cosmic and organic evolution? Is a rigorous nonequilibrium thermodynamics possible? (I gather that Prigogine's claims to have achieved this are hotly disputed.) I am not quite sure what to say about the role of "dissipative structures" in life. I have seen very strong statements on their indispensability and also on their irrelevance. Intuitively, it seems plausible to me that they are playing an important role, but I am basically neutral on this one.
(6) Nonlinearity. If I was too quick to identify "mechanism" with "linearity," still it seems to be the case that linear processes spontaneously disperse, whereas nonlinear ones spontaneously cohere. This connects back to (3) above. Is the distinction between linear and nonlinear equations merely a reflection of our epistemic limitations, or does it say something deep about the way reality is---that linear approximations are idealizations of a spontaneously self-cohering or "striving" reality? I say the latter.
(7) Active vs. passive matter. This is closely related to (6). Should we think of matter in the Democritean fashion as inert atoms blindly bumping into each other in the void, or rather as having its own intrinsic causal powers? Robert Rosen expresses this by speaking of "inertial" vs. "gravitational" views of matter. In other words, is it enough to think of life in terms of impressed forces, as in a machine, or do we have to take the intrinsic physical properties of the actual matter constituting life into account? I say the latter.
(8) Structure and function. In a machine, the function arises from the boundary conditions imposed on the structure from the outside by an external intelligent agent. Should we think of the cell as a machine---with the boundary conditions "tinkered" together by natural selection---or does function in the cell arise somehow directly out of the structure? I say the latter. Of course, this only makes sense on the assumption that we opt for living matter being active in (7) above. Stuart Kauffman writes: "Physicists bury the consideration of the organization of process in their initial and boundary conditions. But where did the initial and boundary conditions come from? This is typically not answered by the physicists..." ("What Is Life?," in J. Brockman (ed), The Next Fifty Years, Vintage, 2002, p. 132--133). That's okay for physicists, but biologists cannot avoid this question---it is the heart of their problem.
(9) Energy minimization vs. directed work. One of the key differences between a living system and an inorganic dissipative structure is that the organism is not just minimizing energy, but is actively using energy to achieve specific goals. Can this fact be explained by simply appealing to "chance" and the all-powerful Darwinian ratchet, or does it require a deeper physical explanation? I say the latter. Kauffman writes: "the cell does thermodynamic work to construct constraints on the release of (here chemical) energy down particular pathways. And once released, that energy can do work to construct further constraints." (ibid, p. 134). This is the real chicken-and-egg problem at the heart of biology, even more than the DNA-protein problem. Somehow, we have to explain how work cycles derive directly from the structure of the cell, conceived of as active matter. The only speculations I know of that point in this direction are those of Hans Frauenfelder, Gerald Pollack, John Watterson, and a few others which relate function directly to (1) the immense size of proteins, which gives rise to special properties like "frustration," and (2) the specific nature of the protein-ordered water-phosphate gel. I believe that further elucidation of these phenomena will probably require recourse to QFT (à la Herbert Frohlich, A.S. Davydov, Mae-Wan Ho, Emilio Del Giudice, Giuseppe Vitiello, and others), but who knows?
(10) Emergence. Is the success of asymptotic methods (the renormalization group, and the "effective field theory program" generally) in condensed-matter physics a mere gimmick, a way around our epistemic limitations, or is it rather evidence that reality itself is layered, and that genuinely novel structures come into existence together with the laws governing them at each new level? This relates back to (2) above. It is very important, because if we allow that the effective field theory program has a hold on reality, then we get a picture of the universe that is totally at variance with (1) above. I say we have every reason to believe the condensed-matter physicists and to disbelieve the high-energy physicists on this point. Reality is layered, and novelty is real, not merely apparent---ontological, not merely epistemological. Against this background, "vitalism" begins to make sense, and to lose its mystical connotations. Novelty continues to arise, of course, throughout the evolutionary process. Kauffman expresses this by saying that life alters its own phase space as it goes along. But the loss of determinism and precise prediction does not mean nothing can be said. This is where qualitative methods come in again.
(11) Formal vs. material emergence. Is life fundamentally a matter of an abstract organization? Is there a certain level of complexity which causes a system to "take off" and live, regardless of its material instantiation? Are A-life simulations really alive? Could a robot be alive? (Call this "formal emergence.") Or is there a real distinction between the simulation of a behavior and its occurrence due to the causal powers of the matter in which the behavior is "instantiated"? (Call this "material emergence.") I say the latter. Robert Rosen says that life involves "impredicativities" (basically self-referentiality, I gather) that transcend any possible syntactic or algorithmic formalism. He identifies this property with Aristotle's "final cause." I agree that what I am calling "teleology"---that is, the functional striving and intelligent agency of the cell---has this property, and should be understood as arising as a result of material emergence, not formal emergence. This also relates back to points (7) and (8) above.
Taking the above points together, I say that this view of life and its metaphysical background transcends "mechanism", however you define it.
So, Erik, if you could tell me which (if any) of these points you agree with, and which you don't, then we could narrow down the discussion and make it more productive. If I have formulated any of them badly, or if you feel there are other issues I have omitted, I look forward to your corrections.
--- James
IP: Logged
|
|
Erik
Member
Member # 160
|
posted 02. July 2002 15:04
quote:
James Barham: I have never claimed to be "doing science." What I am trying to do is absorb as much science as I can to come up with a better understanding of the place of purpose in nature.
What is your role in the ID movement then? I was under the impression that the ID movement claimed to be a scientific movement whose lack of scientific recognition is the result of a world-wide conspiracy.
quote:
James Barham: As for metaphysics being my playground, I will make a deal with you. When the Darwinians stop doing metaphysics, so will I.
First, you've recently failed to support an assertion about "just so"-stories. Don't get into a similar failure by making assertions about the use metaphysics that you cannot support. The metaphysical implications of evolutionary biology have been discussed as a philosophical question by philosophers, but they are very rarely discussed in the scientific literature by evolutionary biologists. (The philosophers and the evolutionary biologists may of course be the same individuals--I'm just pointing out they usually separate their roles as philosophers and scientists.)
Second, I don't necessarily want you to stop doing metaphysics. I think it is perfectly alright for (professional and amateur) philosophers to speculate on the existence of intrinsic teleology, consciousness, and intelligence in nature as long as they realize that they currently must play the role of a philosopher (rather than, say, a scientist or a plumber) when they address such issues. (The same applies to authors of popular science like Dawkins and Gould.)
quote:
James Barham: On the meaning of the word "function." In biology, at least (not in mathematics, of course), the clear implication of "function" is something that happens for a purpose. There is no getting around this, and you are the one indulging in word games, not I.
Function has to do with what something does right now. Teleology has to do with why something originated. For instance, your nose and ears can function as holders of a pair of glasses, but that doesn't necessarily mean that your nose and ears originated because they can perform this function. Another example is the early attempts human flight. People would attach mechanical wings to their bodies and try to fly. These devices were teleological, but they did not function. Clear up your confusion by reading what Dobzhansky, Ayala, Stebbins and Valentine had to say about teology and functions.
You assert that I am indulging in word games, yet you offer no argument at all, and of the two of us I am certainly the one being careful with definitions.
quote:
James Barham: I appreciate the help with the math. I will compare what you have written with the writings of Ian Stewart, Ivar Ekelund, Bruce J. West, Robert Rosen, Alwyn C. Scott, and others, from whom I have gleaned my admittedly rudimentary understanding of these matters. I don't doubt that I may have gone wrong, but I want to see better exactly where. Your corrections were a little too peremptory to put my mind completely at ease.
Just out of curiousity, do you have any kind of mathematical understanding of QFT, QM, self-organization, population genetics, etc.? Or are your arguments (i.e. the ones you actually understand) just what physicists would derogatorily classify as "word arguments"?
quote:
James Barham: As for your distinction betweeen QFT and condensed-matter physics, unless I am very much mistaken (always a possibility!) a glance into any condensed-matter physics textbook will show that QFT is widely used in that discipline. Maybe I am missing something here, and not invoking QFT in quite the right way, or with the right qualifications. If so, a genuine attempt to explain where I am going wrong would be much appreciated.
It seems to be like this: Quantum field theory is a result of the attempt to make the highly successful non-relativistic quantum mechanics into an even more successful relativistic theory. First, this was accomplished for electromagnetic interactions (QED) and later the weak and strong forces were incorporated into the description (QFT). QFT is essentially the tool of nuclear and sub-nuclear physics.
The objective in condensed-matter physics is to study the collective proporties of large collections of atoms and molecules in the condensed (i.e. liquid or solid) state. If condensed-matter physicists could solve the Schrödinger equation for, say, 10^22 molecules in a crystal, they would. But this is absurdly difficult, even with the help of modern supercomputers. Therefore they must rely on approximations. They ignore some interactions and make idealized models of the rest to obtain some (hopefully) accurate approximations to the description provided by quantum mechanics. There are a number of techniques for doing this. One is to formulate a so-called mean-field theory, which means that local properties are replaced by local averages of the same properties. When the mean-field theory is unsatisfactory, one can resist the simplification provided by the averaging and instead study fields describing some (probably approximate) properties at each point in space. It turns out that such field theories have some formal similarities to QFT (e.g. path integrals), which makes particle physicists feel quite at home. Note, however, that the field theory of condensed-matter physics is a result of simplifications of non-relativistic quantum mechanics. The motivation for these simplifications is simply that it would be too difficult without them.
------------------------------------------------
And now onwards to the second post in response to my latest post. I will use the same numbering as you.
(1) Although you titled this item "determinism" it does not deal with determinism. Determinism means that the same initial state always results in the same final state. You may regard determinism as either a directly counter-factual claim ("initial state S0 evolved into the final state S and no other final state could have resulted") or, as I prefer, as an ensemble concept ("all systems in a large ensemble of systems prepared in the initial state S0 will evolve into the state S"). Determinism can be true or false depending on how much detail we want incorporate. For instance, in Las Vegas we may regard a die in someone's hand to be in the same initial state depending on the orientation and position of the die, and the final state to be the face of the die that faces upwards. In this case, dice-throwing is definitely not deterministic, since the initial state does not uniquely fix a final state. Or one could insist that the dice-thrower, the surrounding air, etc. also be included in the initial state, in which case dice-throwing could be argued to be deterministic (Laplace would've argued that). Determinism is very difficult to verify or falsify empirically. It has definitely become unfashionable with the emergence of quantum mechanics, but it is not dead (after all, QM is deterministic except for the still poorly understood "collapse of the wave function", and philosophical theories like Bohm's interpretation and no-collapse interpretations preserve determinism at all times). On the other hand, general relativity suggests that time is on the same footing as space, and we live in a static 4D world, rather than a dynamic 3D world. In this static world moments in time exist in the same way as points in space. Sir Isaac Newton still exists (just not right here in the vicinity of my space-time location) in the same sense that the Adromeda galaxy still exists. In such a static world, determinism would seem necessary if we just insist that a sufficient level of detail is included in the initial state.
(1') Your item (1) was actually about predictions, rather than determinism. First, the general 3-body problem is not solvable analytically, but that doesn't mean that it cannot be studied nor that final states are impossible to predict. It is true that the precision with which we need to know quantities characterizing the initial state is too high for practical predictions over long time scales in systems of many particles. But that is not necessarily the same thing as indeterminism.
(2) Mathematics is such a general language that it is extremely hard to conceive of anything that is not describable in terms of mathematics. It is certainly the most accurate language we have to describe nature. Quantum mechanics applied to Josephson junctions has lead to predictions accurate to one part in 10^-19, an accuracy that is so superior to any non-mathematical description that one can only stand in awe, admiring the effectiveness of mathematics in general and QM in particular. At the very least, mathematics is superior to all other kinds of descriptions.
(3) Of course it is legitimate to study other mathematical properties of the differential equations describing the dynamics of physical systems, than just their exact solutions.
(4) The equations of motion in Newtonian and quantum mechanics are indeed time symmetric (not "time invariant", which is true but not what you meant). I also agree that this is a problem for thermodynamics. Physicists are often very eager to employ symmetry arguments of all kinds, yet one tends to be very quiet about the time symmetry of microscopic physics in thermodynamics and statistical mechanics. When I took the course on statistical mechanics, I used to terrorize the lecturer with questions of this kind. I argued that because of the time symmetry of QM, any argument that entropy increases forwards in time must be equally (in)valid in the backwards direction of time. Much to my disappointment he did not understand what I meant. If there was ever a good argument against reductionism, I think this is it: Microscopic physics is time symmetric. The macroscopic second law of thermodynamics is not. Therefore reductionism can not always hold.
On the other hand, statistical mechanics is based directly on QM. The solutions to the Schrödinger equation constitute the states, and the entropy is in fact a measure of how many states a system coupled to a thermal bath wanders between. Thus, we derive statistical mechanics from properties of the Schrödinger equation.
(5) I don't know.
(6) That makes no sense at all. The following gem could have appeared in Sokal's brilliant fake paper: "Is the distinction between linear and nonlinear equations merely a reflection of our epistemic limitations, or does it say something deep about the way reality is---that linear approximations are idealizations of a spontaneously self-cohering or "striving" reality? I say the latter." The distinction between linear and nonlinear equation is actually quite rigorous, and it is possible to program a computer to classify equations into "linear" and "non-linear". Furthermore, the "failure" of Newtonian models to allow (in practice) the computation of their solutions is nothing compared to the difficulties one meets in non-linear models. Indeed, it is common to make linear approximations of non-linear models in order to simplify the computations.
(7) I think the distinction is physically uninteresting, in the sense that it nothing is changed in our physical models either way.
(8) I don't understand the question.
(9) I think it is the result of quantum mechanics (yes, plain QM without reference to QFT) in much the same way that autocatalytic chemical reactions are ultimately the result of QM. That doesn't necessarily exclude differential reproductive success from being the main mechanism.
(10) I would say that the successful approximations to QM made in condensed-matter physics is quite definitely not evidence that reality "layered" in any deeper sense than that the appearance of reality is "layered" when not seen through physics (the beauty of physics is that it provides simple models that unifies large classes of seemingly different phenomena).
(11) I do not believe there is a sharp line between life and non-life. I also don't think one should confuse the concepts of life and consciousness. It is fine to argue that all life is conscious and that everything that is conscious is also alive, but they are not equivalent by definition. Personally, I could agree to call anything that is both a process and thing alive.
quote:
James Barham: So, Erik, if you could tell me which (if any) of these points you agree with, and which you don't, then we could narrow down the discussion and make it more productive. If I have formulated any of them badly, or if you feel there are other issues I have omitted, I look forward to your corrections.
I think that above all you are deeply confused about the terminology. Conceptual clarity is a virtue, and I think you would to well to try to write down precisely what you mean by different terms.
Erik
IP: Logged
|
|
Micah Sparacio
Member
Member # 6
|
posted 02. July 2002 15:20
Hey Erik,
Not that this is relevant to the technical issues, but I thought I'd help clear up a misunderstanding. You said:
quote:
What is your role in the ID movement then? I was under the impression that the ID movement claimed to be a scientific movement whose lack of scientific recognition is the result of a world-wide conspiracy.
I've only just recently met James Barham via the net. From what I know, he has no role in the ID movement and would never claim to be an ID person. From the conversations I've had with him, I would venture to say that he is a generalist with a vast store of knowledge. If labels are necessary, I would classify him as a philosopher with interests in biology and self-organizational theory. In particular, he is interested in seeing the development of a robust theory of biology based on physics. In order to see this play out, he feels that Darwinism needs to humble itself a bit, and is associated with ID only so far as it offers critiques of the Darwinian mechanism. However, if you ask him, I'm sure he would instantly tell you that he is by no means an ID theorist.
IP: Logged
|
|
James A. Barham
Member
Member # 50
|
posted 02. July 2002 19:48
Erik:
You are quite right that I am engaged in "word arguments." That is what all of us are doing, except for the mathematicians among us.
I never meant to imply any disrespect for mathematics. I agree with you, of course, that it is the most precise form of reasoning that exists, where it is applicable. I merely meant to raise the issue of how we should go about deciding what is the truth when our equations fail to conform to our experience (as they often do, I am sure you will admit).
I guess I would be happy simply accepting the pronouncements of scientists from on high, were it not for three things: (1) Not everything can be explained mathematically; (2) Science advances over time, and what is widely believed today is often no longer believed tomorrow; and (3) Even today, scientists often disagree with each other. Under these circumstances, alas, I am forced to think for myself.
Now, as I said, my personal interest in all of this is to understand the place of purpose (or teleology) in nature. On this subject, it is certainly not the case that there is some one scientific authority I can safely turn to to find the answers I seek. So, what would you have me do except read as widely as possible, and try to understand things as best I can? If my articulation of what I have understood then necessarily takes the form of "word arguments," what alternative do I have? Of course, I still want to make the word arguments as precise as I can, and to that end, I am grateful to you for your efforts to help me get the words right. But I don't see the use of "word arguments" as in itself something blameworthy.
Here are some miscellaneous replies to some of the points you raise, drawing on the scientific sources that I have found the most persuasive. If I am still misunderstanding what I have read (which I admit is possible), then I would be grateful for your further indulgence in setting me straight.
On Determinism:
"Consider an analogy. A rigid steel rod of, say, 1 cm diameter is almost unbendable if it is only 10 cm long. But suppose it is 10n miles long? Very minute impacts would cause it to wave about like a blade of grass. The actual constitution of matter is not such that there is complete rigidity over indefinite lengths; rigidity is an idealization. Similarly, I suggest, the 'determination' of one state of affairs by another is also an idealization and is only to be taken as a good approximation to the extent that, in any actual instance, prediction over finite time intervals is found in fact to be a good approximation." (Kenneth G. Denbigh, Three Concepts of Time, Springer, 1981, p. 85)
Unless I very much mistaken, Denbigh disagrees with you (and, of course, the majority opinion that you articulate) about determinism. So what am I, as a layman, to do? I listen to what both of you have to say, and I have to decide which of you makes the better case. In my view, Denbigh's argument makes more sense than yours. Why do I feel this way? Because I believe that in the end determinism is a metaphysical notion inferred incorrectly from the great success of math in certain limited situations.
I agree with you very strongly that it is imperative to be careful not to confuse ontological with epistemic issues here. But things are not quite as clear-cut as you suggest. After all, our cognitive access to the world is our only evidence for what the world must be like in itself. Therefore, at the end of the day, we have to ask ourselves, Is our inference to what the world is like in itself better grounded in mathematical idealizations, or in experience (success or failure of predictions)? I believe, the latter.
Here is another quote that perhaps says what I am getting at better than I can:
"One of our strong reasons for believing in science, and in physics in particular, is the ability that its laws give us to make precise predictions. For example, the laws of mechanics and of gravity accounted for the locations of the planets and of comets over long periods of time. There are, however, other physical phenomena, such as the weather, for which we obviously have only short-term predictive power (say, two or three days at best), rather than the long-term predictive power we seem to have for the solar system. The traditional view was that systems consisting of only a few parts (for example, the planets moving about the sun) are amenable to the precise calculations necessary for meaningful prediction, whereas complex systems, such as a collection of gas molecules (or the atmosphere) are simply beyond our calculational abilities. However, no in-principle difference was seen between these two types of systems as regards determinism. One was just too complicated to handle computationally---a mere practical limitation.
"This traditional view, or intuition, was based on an examination of the (relatively few) physical problems that we could solve analytically (as in the case of the two-body central-force problem above). Such integrable systems are, essentially by definition, the ones that can be treated by the methods of (classical) mathematical analysis. The exactly soluble problems of classical mechanics usually turn out to be separable. This means, roughly, that the equations describing their behavior separate into sets of individual one-body problems. That might already have given us some hint about how special, and not surprisingly, atypical, these cases are. Nevertheless, we took our 'old' insights, formed on the basis of those systems we were able to handle analytically, and assumed them to be typical of all physical systems. (Of course, what else could one reasonably do but form a picture of the physical universe based on a theory that appeared to be enormously successful?) But, as we have just indicated, such integrable systems turn out to be very special. Our intuition, or general picture of the world, was based on a poor induction from too narrow a range of systems. For nearly 300 years we thought we understood classical mechanics, but we didn't." (James T. Cushing, Philosophical Concepts in Physics, Cambridge UP, 1998, pp. 171--172)
In short, we now have good reason to believe that the world is a fundamentally different place from the one described by Newton's equations. Therefore, we have good reason to reject Laplace's inference from Newton's equations to universal determinism. I am not saying that I believe this is so because I have special knowledge of the relevant mathematics; I am saying that I believe the scientists who say this is so. I find their view more persuasive than the majority view.
On the layered model of reality:
"In this case we can see how the whole becomes not only more than but very different from the sum of its parts." (Philip W. Anderson, "More Is Different," Science, 1972, 177: 393--396, p. 395)
"In some sense, structure---functional structure in a teleological sense, as opposed to mere crystalline shape---must also be considered a stage, possibly intermediate between crystallinity and information strings, in the hierarchy of broken symmetries." (Ibid., p. 396)
"This, then, is the fundamental philosophical insight of twentieth century science: everything we observe emerges from a more primitive substrate, in the precise meaning of the term 'emergent,' which is to say obedient to the laws of the more primitive level, but not conceptually consequent from that level." (Philip Anderson,"Historical Overview of the Twentieth Century in Physics," in Laurie M. Brown et al., eds., Twentieth Century Physics, Volume III, New York: American Institute of Physics Press, 1995, pp. 2017--2032, p. 2020)
"The ideas and concepts of symmetry breaking, renormalization group and decoupling that stemmed from a deeper study of relativistic field theoretical models adumbrate and justify a picture of the physical world that is hierarchically layered into quasi-autonomous domains, with the ontology and dynamics of each layer essentially quasi-stable and virtually immune to whatever happens in other layers." (Silvan S. Schweber, "The Metaphysics of Science at the End of a Heroic Age," in Robert S. Cohen et al., eds., Experimental Metaphysics, Kluwer Academic, 1997, pp. 171--198, p. 172)
"Renormalization-group methods in condensed matter physics gave new insights into why the details of the physics of matter at microscopic length scales and high energy are inconsequential for critical phenomena. What is important is the symmetry involved, the conservation laws that hold, the dimensionality of space and the range of interactions." (Silvan S. Schweber, "Physics, Community and the Crisis in Physical Theory," Physics Today, November 1993, Vol. 46, No. 11, 34--40, p. 37)
"This example shows further that this is not just a matter of convenience and practice; but that the formal structure of a general framework must be supplemented by numerical information, the accuracy required and estimates and bounds on the neglected terms. This is necessary before such formalisms can be used to describe realistic physical situations. The supplementary information is precisely what defines and specifies the level." (Max Dresden, "Reflections on 'Fundamentality and Complexity'", in Charles P. Enz and Jagdish Mehra, eds., Physical Reality and Mathematical Description, Reidel, 1974, pp. 133--166, p. 153)
"The conclusion to be drawn is that the behavior of large aggregates of elementary atoms should not just be understood in terms of a simple extrapolation of the properties of the system with just a few atoms. There seem to be many levels of complexity, and at each such level entirely new properties begin to appear." (Ibid., p. 161)
"This unpredicability has disturbing implications for the traditional ways of relating theory and experiment. Competition among two or more principles of organization is common in nature. When it occurs, the deductive paths from the microscopic model assumptions to the observed behaviour become unstable at low energy scales and essentially impossible to solve accurately. The theory may describe the behaviour or it may not, but whether it does is unknown because the theory cannot be solved with sufficient accuracy to tell. In other words, the predictive power afforded by knowledge of the underlying equations of motion is blocked for reasons having nothing to do with complexity. This then implies that the theory is under-constrained by experiments at low and intermediate energy scales and cannot be falsified by them even in principle . . . Thus there is reason to suspect that the laws of self-organization of matter may render the most fundamental equations of the universe both unknowable and useless." (Robert B. Laughlin et al., "The Quantum Criticality Conundrum," Advances in Physics, 2001, 50: 361--365, pp. 363--364)
"Details are required to account for why a given instance of a pattern can arise, but such details obscure and even block understanding of why the pattern itself exists. Physicists have a technical term for these patterns of behavior. They call them 'universal' . . . Asymptotic methods such as the renormalization group provide explanations for this remarkable fact. They do so by providing principled reasons grounded in the fundamental physics of the systems for why many of the details that genuinely distinguish such systems from one another are irrelevant when it comes to the universal behavior of interest." (Robert W. Batterman, The Devil in the Details: Asymptotic Reasoning in Explanation, Reduction, and Emergence, Oxford UP, 2002, p. 4)
"Emergent properties are universal. It is legitimate to search for, and expect, explanations of their universality. Contrary to received opinion, such properties are not brute and inexplicable features of the world. As we will see in several places throughout the book, reduction and explanation, when properly understood, do not march in lock-step. Asymptotic explanations are possible even for phenomena that are in an important sense irreducible and emergent." (Ibid., p. 6)
------------------------
I could continue, but I am sure I have already tried everyone's patience to the limit, as it is. I have resorted to these extensive quotes because I have apparently done such a poor job of articulating their contents myself.
At any rate, Erik, I hope you appreciate my dilemma. I cannot just accept what you say on these deep issues on your say-so, in spite of your much greater technical command, because, as you see, there are others with technical command equal to yours who disagree with your metaphysical interpretation of the math.
---James
[ 02 July 2002, 19:56: Message edited by: James A. Barham ]
IP: Logged
|
|
|
Printer-friendly view of this topic