|
Author
|
Topic: Karl D. Stephan: What Does Evolutionary Computing Say About Intelligent Design?
|
Moderator
Administrator
Member # 1
|
posted 09. November 2002 12:22
What Does Evolutionary Computing Say About Intelligent Design?
by Karl D. Stephan
ABSTRACT—When Charles Darwin proposed his theory of variation and natural selection in The Origin of Species, he unwittingly provided the intellectual foundation for what has recently become a thriving subfield of computer science. Evolutionary computing arose from artificial intelligence research in the 1960s by computer scientists such as Lawrence Fogel and John Holland. Holland was one of the first to show that the genetic mechanisms of variation, reproduction, and selection are peculiarly suited to deal with certain problems whose characteristics make them difficult to solve with conventional approaches (Holland, 1975; Holland, 1992). In the decades since, dozens of books and hundreds of papers have been published on the subject, ranging from theoretical advances in the genetic algorithms used in evolutionary computing to practical applications in molecular modeling, mechanical and electrical engineering, and even investment strategies.1 Evolutionary computing today is much more than a twinkle in the eyes of a few computer scientists. In its wide range of successful applications, it is one of the relatively few concepts in artificial intelligence which has actually lived up to the hopes of its early proponents.
Does the practical success of evolutionary computing say anything about the theory of natural evolution? Beyond that, we can ask a more specific question: does evolutionary computing give support to the contention that variation and natural selection alone are inadequate to explain the origin and variety of life? This contention is often referred to as the “intelligent design” hypothesis, since it denies that nature, unaided by intelligence, is capable of producing the plethora of present life forms on Earth. Before we begin to answer either of these questions, we shall first describe evolutionary computing in enough detail to elucidate some of its basic approaches. Then we shall point out the parallels and differences between the trajectory of information in evolutionary computation, and the way information is stored, modified, and used in in natural biological systems such as bacteria. At that point, we will be prepared to address the two questions above.
To read the entire paper, please click here
[ 09. November 2002, 12:26: Message edited by: Moderator ]
IP: Logged
|
|
Frances
Member
Member # 169
|
posted 10. November 2002 17:22
An interesting paper about the use of analogies to argue that the mind is a requirement for both evolutionary algorithms and natural selection. I would like to raise another possibility based on the same data. The author argues that it is the mind which is present in evolutionary algorithms suggests that it needs to be present in natural selection. But lets explore for a moment what is more fundamental: the mind or natural laws? Natural laws came into existence at the formation of the universe, but did the mind? One may argue that the mind is merely an extension of natural law. Thus rather than argue that the common thread needs to be a mind, it may be argued that the common thread is 'natural law'.
The mind in instances of evolutionary algorithms is just like the natural environment to biological evolution, it 'directs' through interactions where the natural laws take us. So are the mind and the environment interchangable? And which one is a metaphor for the other one? I suggest rather than assuming that the mind takes precendence, it may be the environment/natural laws.
Analogies can be helpful in detecting similarities but they may not be helpful in detecting causal direction. In fact, I would argue that we may not be able to detect the differences between what (evolutionary) algorithms can do and what ID can do. Take the example of the traveling salesman, can we distinguish between a solution that is reached by intelligent design versus a solution that is reached by what is essentially random mutations and selection. Thus if CSI can be introduced by either evolutionary algorithms/natural laws/intelligent design then how do we venture to distinguish between them? Is the mind a requirement for CSI or is it merely a substitution for the environment as it interacts with natural law?
So in other words are ID and natural law/environment merely different ways to infuse complex specified information?
IP: Logged
|
|
RBH
Member
Member # 380
|
posted 10. November 2002 21:25
Stephan has written an interesting version of the argument from analogy for design in nature. Several aspects of his paper deserve some extended comment.
Stephan's major effort in the paper is to attempt to strengthen the case for an argument from analogy in general so as to use evolutionary algorithms as evidence for a non-natural designer of physical and biological phenomena. That is, he tries to show that most of science is in effect nothing more than elaborate arguments from analogy. He does this by claiming that the use of mathematics in science is nothing but the use of analogy. Mathematics, in Stephan's argument, is nothing more than an analogy for the entities and their properties and the processes and relationships of the real world, and so the utility of mathematics in science lends credence to the general use of arguments from analogy.
The Use of Math in Science
While an excursion into the philosophy of mathematics would take us down many a garden path, it is worth mentioning that there are two quite different views of the role of math in science. One is what might be called "mathematical essentialism." Mathematical essentialism is the view that the physical universe is 'built' on the basis of principles of mathematics; that it has an inherent mathematical structure that science attempts to discover. The other view might be called "instrumental mathematics." In this view, math is a formal language invented by human beings to describe and model entities, processes, and relationships in the physical universe. Newton, for example, is said to have invented the calculus of infinitesimals to represent the dynamical behavior of bodies moving in gravitational relationships.
In either case, regardless of whether one is an essentialist or an instrumentalist, in science the terms and operators of math are taken to represent objects, properties, processes, and relationships of the physical universe. Stephan, however, views math as being merely analogous to the physical world. For Stephan, when we manipulate equations containing terms and operators, we are merely constructing analogies for the physical world, and our confidence in our conceptions of the physical world is somehow stronger if the analogy is expressed in formal mathematics.
In fact, however, as I noted above, science uses math to represent (aspects of) the physical world. When one performs the syntactical (content-independent) operations of math to derive new mathematical expressions, one is generating hypotheses in the form of the new relationships implied by the derivation. But those new hypotheses are not automatically accepted because the analogy with math lends them veracity; they are accepted if one then goes out into the world and tests those hypotheses. That in fact is how much theory testing proceeds in the mathematicized sciences: Represent a theory of phenomena in math symbols and operators, manipulate the symbol strings syntactically to derive new expressions, interpret those new expressions in terms of the theory, and go look at the world to see if that interpretation is in fact the case.
In order to do that, the physical phenomena must be mapped into math. The objects, properties, processes, and relationships of the physical domain of interest must be mapped into the terms and operators of math. That mapping provides the means of interpreting the outcome of purely syntactic manipulations of symbols according to the rules of math as predictions of observations of phenomena in the world, the predictions of the theory that are tested against the physical world. Theory guides the mapping of phenomena into math. It is theory that says given force, acceleration, and mass, we are to map force onto F, acceleration onto A, and mass onto M, and relate the three by the formal equation F = M*A. Given that mapping, we can plug various values into M and A and calculate F. We can then do experiments, varying M and A and measuring F and thereby assess whether F = M*A is a veridical representation of (or formal model of) the relationship among the three physical concepts.
It is theory that gives meaning to the outcome of math manipulations. So math in science is not merely analogy, it is model. Stephan claims that he has quote:
established that all mathematical sciences (other than pure mathematics itself) are based on the principle of analogy ... (p. 8)
but that's not the case. He subtly misrepresents the role of mathematics in science, conflating "model" and "analogy."
What do analogies really tell us?
Stephan has high regard for the evidential and inferential value of analogies. He asks quote:
...does evolutionary computing give support to the contention that variation and natural selection alone are inadequate to explain the origin and variety of life?" (p. 1, emphasis added)
And quote:
An analogy is the setting up of a comparison between a first thing and a second thing, in the hopes that information known about the first thing will also provide information about the second thing. (p. 7, emphasis added)
Then after a description of the use of genetic algorithms to generate designs for antennae, he says quote:
In particular, because the computational evolutionary process used a small amount of randomness within a highly structured framework designed by minds, I claim that this fact provides evidence that mind is necessarily involved in the natural biological process generally termed evolution. (p. 9, emphasis added)
In all these cases, analogy is to provide information or evidence for the "thing" being analogized to. But that's not the way analogies function in science. What analogies do in science is generate ideas and hypotheses for test. They do not provide information or evidence, they generate questions.
Analogies can lead us astray. The "information" or "evidence" they are alleged to provide can be erroneous or even actively deceptive. For example, consider the analogy Stephan uses as an example: quote:
For example, a common analogy used in physics is the similarity between water waves in a still pond and other kinds of waves such as sound waves. Both waves share a number of common characteristics: periodicity in time and space, a specific relationship between wave amplitude and distance traveled from the source, and so on. But there are also differences: water waves are basically an interfacial phenomenon, sound waves generally propagate through a volume of space, and so on. (p. 7)
That very same analogy generated the hypothesized "ether," a substance held to fill space to provide a medium in which electromagnetic waves could propagate. Because wave phenomena in water and air require a medium in which the oscillations occur, so it was believed that electromagnetic waves required a medium. Michelson and Morley tested that hypothesis and found it wanting: the ether as a physical medium for electromagnetic 'waves' does not exist. The analogy did not provide information about the ether or evidence for its existence; it provided the hypothesis of the ether, and experimental test showed the hypothesis to be false.
Referring to the heavily mathematicized disciplines, Stephan writes quote:
Those [mathematical] analogies which are successful in representing the physical world with mathematics command respect simply because they work, often extraordinarily well. And in this pragmatic age, anything which works well commands respect.
But mathematically expressed scientific theories do not command respect merely because they "work," unless "to work" is taken to mean "to organize phenomena in a domain of interest, to provide causal explanations for those phenomena, and to provide corroborated predictions of new observations of phenomena." A mathematical representation that merely curve-fits data is not highly respected. A coherent theory is necessary.
A brief remark on evolutionary algorithms
Stephan's account of evolutionary algorithms focuses solely on their use in directed search in engineering and optimization. While I will not develop this argument here, I will develop it later on ARN in the "What Genetic Algorithms Can Do" thread. However, I will say here that the analogy from directed search applications by humans seeking to solve a problem is different from the theoretical principle of undirected evolution in nature. Natural biological evolution is not an optimization process in the same sense that "optimization" is used engineering applications by humans, so the analogy between biological evolution and Stephan's presentation of evolutionary algorithms is therefore suspect from the very start. Human-designed applications of evolutionary algorithms to solve human-defined problems embody human goals, but that by no means makes the case that biological evolution embodies the goals of a Designer. It may raise that speculation or conjecture, but it does not provide evidence for it. (I eschew the term "hypothesis" here because a hypothesis is much more specific than "A designer is responsible for the biological phenomena we observe.")
One can use evolutionary algorithms to model aspects of biological evolution, but not in the same way that EAs are used in engineering optimization applications. EAs can both generate hypotheses and when appropriately configured can serve as models and even test beds with which to test those hypotheses about biological evolution, but they are not in themselves "evidence" for or against the design conjecture.
Stephan's argument is a standard one in ID: Since human-designed objects require a human designer, so do natural objects. His attempt to strengthen the force of the argument from analogy fails. Analogies may generate hypotheses, but they cannot provide evidence. Only the systematic testing of hypotheses generated by analogies that have been transformed into theoretical models can do that. Stephan actually sort of recognizes that when he writes quote:
The purpose of making analogies is to use knowledge developed in one domain to analyze and predict phenomena in a second analogous domain. (p.17)
But there is no mention there of testing those predictions. Instead, after an extended discussion, he writes quote:
If we have thousands of cases of exhaustively known computational evolution going on around us on computers, and we know for a fact that minds were involved in the origin of every single one, the conclusion that mind was involved in the origin of naturally occuring species is an obvious one. (p. 22, emphasis added)
But once again, an analogy cannot provide a conclusion; it can only provide speculations, conjectures, or perhaps hypotheses. This is the same inductive argument Dembski offers for the "no false positives" claim made for the Explanatory Filter: In every case of human design studied, he says, when the EF says "design," the object was in fact (human) designed. Once again, this is analogy from humans rather than systematic evidence for non-human design.
Stephan attempts to answer some anticipated criticisms. For example, he write quote:
Just as there are some who will refuse to accept the implications of an analogy simply because it is an analogy, there are others who insist on definitions which allow them to maintain their point of view.
It is not necessary to be a hidebound reactionary to observe that the "implications" of an analogy are not logical, mathematical, or formal implications; they are at most suggestions. I'm not aware of a formally developed "suggestive logic."
Finally, Stephan identifies the core scientific problem for ID: quote:
If instead, we adopt the hypothesis that mind directed and directs biological development by some as-yet-undiscovered means, we have not solved the problem of the incredible order and complexity of biology. We have simply moved it back one step to the question of the nature of the mind which brought it about. Unless we possess independent data about the nature of such a mind, we are basically in the same position as those who believe biological evolution has no purpose. So the bare statement that mind is behind evolution is no threat to the postulate of randomness behind evolution, until we begin to attribute characteristics to the mind we postulate. And unless such characteristics can be determined within the empirical framework of science, the hypothesis of mind will turn out to be as scientifically barren as William James regarded the hypothesis of the soul in psychology.(p. 25, emphases added)
Just so.
RBH
[ 10. November 2002, 21:39: Message edited by: RBH ]
IP: Logged
|
|
Mark Szlazak
Member
Member # 391
|
posted 11. November 2002 01:21
Frances you wrote that:
quote:
But lets explore for a moment what is more fundamental: the mind or natural laws? Natural laws came into existence at the formation of the universe, but did the mind? One may argue that the mind is merely an extension of natural law. Thus rather than argue that the common thread needs to be a mind, it may be argued that the common thread is 'natural law'.
What do you mean by "natural laws" and "mind"? I hope you don't mean classical physics since what you said would be just plain fantasy or nonsense ... no matter what you define "mind" to be.
[ 11. November 2002, 01:23: Message edited by: Mark Szlazak ]
IP: Logged
|
|
Frances
Member
Member # 169
|
posted 11. November 2002 12:20
Dear Mark
When you suggest that if I mean classical physics that this would be plain fantasy or nonsense, why would you thus assume that I meant classical physics?
I do not see natural law being constrained to classical physics. Why should I?
IP: Logged
|
|
Mark Szlazak
Member
Member # 391
|
posted 11. November 2002 13:14
Hi Frances.
OK, then I'm more with you on this...but maybe not entirely. So, what do you mean by "natural laws" and which ones are you talking about?
Just as a heads up, I hope they're not information theory or biological information ideas, but if they are lets hear them (one more time) so some "unpacking" or possibly even some "deconstruction" can be done on them. ![[Wink]](wink.gif)
IP: Logged
|
|
Noel Rude
Member
Member # 516
|
posted 11. November 2002 13:58
RBH brings up a very interesting subject -- one that divides mathematics and pretty much defines physics: The Use of Math in Science.
The two camps, more commonly designated "formalist" and "Platonist", divide on what math IS. For the formalist mathematics is merely a human invention (design?) -- it need not even "represent" anything in the world, it is simply the product of mathematicians' minds, a game with rules they set. Hard core formalists are not perturbed by "the extraordinary effectiveness of mathematics".
Physicists, on the other hand, tend to be Platonists -- for them mathematics is real, out there to be discovered (as opposed to invented), and it sits higher on the hierarchy of reality than the physical laws which, though likely contingent, are written in the language of mathematics. Even God (if there is a God) cannot defy mathematics/logic -- that realm is not even of his invention. And thus the many worlders can study other possible worlds, but never for a moment assuming that logic and math would differ in them -- only the laws would differ.
Physics popularizers like Paul Davies and Roger Penrose talk about this, even speaking of the Platonic ream as setting the ultimate standards for logic, esthetics, and ethics. But bring it up among biologists or, say, in my field of linguistics, and you generally get a blank stare. I never know whether folks are simply ignorant of this grand philosophical issue, or simply don't want to talk about it.
Linguistics can be defined as the search for all the ways that languages differ and for all the ways that languages are the same. For the past forty years the emphasis has been on the search for universals, with the main argument being over whether such universals are "hardwired" neurologically or simply the result of "general cognitive functions". But if the physicists are correct and if mathematics/logic is out there to be discovered -- why not Language?
Physics sits atop the pyramid of scientific prestige and physicists discuss topics that are taboo in the lower eschelons of professional academe -- also how can physicists not believe in the reality of their principle tool?
Nevertheless this larger philosophical issue that divides mathematics and defines physics -- is it not pertinent to all the information sciences? How so to ID? Well, for the hard core formalist there is nothing that William Dembski might say that will impress -- he's just playing games like all mathematicians. The mathematical Platonist, however, will likely be convinced that design is detectable in all possible worlds.
IP: Logged
|
|
brauer
Member
Member # 398
|
posted 11. November 2002 14:19
Noel,
Well put. Do you think that the "formalist/platonist" distinction mirrors the "empiricist/realist" distinction in some meaningful way?
I bring this up not because I pretend to know anything about these issues (I am, after all, merely a biologist with a blank stare!). However, Bas Van Fraassen seems to specifically address the concept of design (or at least religion which, while not the same, shares some features with design) in his new book The Empirical Stance.
Following is a review reposted from a recent Science. (I have the book itself on order, but am not sure when I'll get to it.)
quote:
Empiricism, Realism, and Religion
A review by Paul Thagard*
-------------------------------------------------
The Empirical Stance
Bas C. van Fraassen
Yale University Press, New Haven, CT, 2002. 302 pp. $30, £22.50, €36. ISBN 0-300-08874-4.
-------------------------------------------------
One of the most active debates in current philosophy of science is between empiricism and scientific realism. Realism is the view that science aims to produce theories that are at least approximately true, along with the claim that it often succeeds in doing so. Theories go beyond sensory observation by hypothesizing the existence of nonobservable entities such as quarks, chemical bonds, genes, and mental representations. According to scientific realists, we are justified in believing that atoms and other theoretical entities exist because the theories that hypothesize their existence are the best available explanations of experimental results and other observations.
In contrast, empiricists such as Princeton University's Bas van Fraassen argue for a more modest view of the aims and accomplishments of science. They claim that it is too risky to believe in the existence of nonobservable entities and that we should not believe that theories are true, only that they are, at best, adequate for predicting what is observed. Atoms, quarks, and other such entities are not to be taken as parts of the world but merely as convenient means for predicting or redescribing observable phenomena such as those that result from experiments in physics.
Empiricism might seem inherently more antagonistic to religion than scientific realism. Various religions assume the existence of a host of entities that are not observed, such as gods, souls, angels, and heaven. If it is not legitimate to believe in the existence of atoms or electrons, for which there is an enormous amount of empirical evidence, surely one should not believe in the existence of gods, souls, and angels. On the other hand, the methodology of scientific realism seems to open the door for a reconciliation of science and religion. If science can justify the existence of atoms because the theories that postulate them provide the best explanation of observations, so religion might be able to justify the existence of God because this hypothesis provides the best explanation of phenomena such as the origin and design of the universe, the prevalence of religious belief, and the contents of scripture.
Surprisingly, however, empiricism and religiosity can coexist. Pierre Duhem, the most distinguished empiricist of the early 20th century, was a devout Catholic. And in The Empirical Stance, van Fraassen, the most influential empiricist of recent decades, combines a penetrating discussion of empiricism in science and philosophy with a sympathetic discussion of religion. According to van Fraassen, empiricism is not a doctrine but a stance, which is a cluster of attitudes, commitments, and approaches. This stance urges scientists to perform experiments and observations in order to evaluate the empirical adequacy of hypotheses, while avoiding issues concerning their truth. Empiricism does not claim that atoms exist, but it also does not assert that they do not exist. Similarly, empiricism says nothing about the existence of God.
In contrast, scientific realism can give rise to intense skepticism about religious claims. Past science has proposed theoretical entities, such as phlogiston, vital force, and the ether, that are now considered non-existent because the theories advocating them have been superseded by alternatives that better explain observable phenomena. For example, chemists no longer believe in phlogiston because Lavoisier's oxygen theory provides a much better explanation of combustion, respiration, and other phenomena. Similarly, religious hypotheses such as divine design are no longer the best explanation of the complexity of the world, for we now have well-supported scientific theories of cosmology, geology, evolution, genetics, and so on.
Hence the allegiance between empiricism and religion is surprisingly natural; it enables one, in Kant's phrase, to deny knowledge in order to make room for faith. Just as science lacks knowledge about the existence of oxygen or phlogiston, it has nothing to say about the existence of God. According to van Fraassen, science is not the only approach to understanding ourselves and the world we live in, but should be supplemented by what he calls "an abiding astonishment not allayed by the fruits of scientific inquiry." He explores this approach by discussing what existentialist theologians such as Martin Buber and Rudolf Bultmann have said about the distinction between secular and religious approaches to life. Van Fraassen concludes that the crucial distinction between the secular and the religious lies in a certain attitude to how we approach the world and relate to our own experience. He does not go so far as to designate this attitude the "religious stance," but he clearly sees it as a valuable supplement to the empirical stance that he thinks best fits the secular, scientific side of life.
From a realist perspective, however, religion is the wrong place to look for a supplement to what science has to tell us about the nature of our lives. Inference to the best explanation provides us adequate grounds to judge that atoms and many other kinds of theoretical entities exist and serious reasons to doubt whether theological entities are real. Science does not by itself tell us how to live meaningfully and ethically, and many people have thought that religion could fill in the philosophical gaps. But we cannot reasonably complete our lives with wishful thinking about mythical beings. Reflection on meaning and values can proceed philosophically by means of discussion of secular theories of ethics, enriched by insights from literature, history, and sciences such as psychology and anthropology. Hence van Fraassen's eloquent examination of empiricism and religion does not undermine scientific realism.
IP: Logged
|
|
RBH
Member
Member # 380
|
posted 11. November 2002 15:54
Noel wrote quote:
RBH brings up a very interesting subject -- one that divides mathematics and pretty much defines physics: The Use of Math in Science.
The two camps, more commonly designated "formalist" and "Platonist", divide on what math IS. For the formalist mathematics is merely a human invention (design?) -- it need not even "represent" anything in the world, it is simply the product of mathematicians' minds, a game with rules they set. Hard core formalists are not perturbed by "the extraordinary effectiveness of mathematics".
I knew there had to be names for them! ![[Smile]](smile.gif)
My main point in the long post above, however, is indifferent to that difference, if I may put it that way. Whether physics embodies math or math is merely a way of talking about physics, neither view does anything to strengthen the general approach of arguing from analogy. Whether math is merely an analogy used by physics (clearly a formalist position) or is the inherent organizing deep structure of physics, neither increases or even glancingly affects the (lack of) force of Stephan's analogy between computational evolutionary algorithms running on Sparc stations and biological evolution running in populations of living organisms.
RBH
IP: Logged
|
|
warren_bergerson
Member
Member # 262
|
posted 11. November 2002 17:05
The conceptualization of the relationship between abstract mathematics and the uses of mathematics in scientific analysis is an important element in analyzing biological design. It is, quite obviously an oversimplification to reduce the possible conceptualization to a choice between to formalist Platonist or realist and empiricist. It is also a mistake to assume that the conceptualization that fits best with the analysis of design will be found in the existing literature.
In looking at the relationship between math and scientific analysis of design the following three items should be kept in mind:
1. Design, particularly biological design involves complex causal relationships with ‘unusual’ forms such as dynamic causal relationships, and teleological or intelligent causal relationships.
2. Systems and processes which simulate/model design must create information. Analysis suggests that such systems or subsystems must involve interactions with other systems (closed systems can not create information).
3. Concepts and techniques taken from physics have not been effective or useful in the scientific analysis of biological systems.
These issues can not be addressed in terms of simplistic binary choices.
In his article, Karl Stehan, raises some good questions- Quote: "Does the practical success of evolutionary computing say anything about the theory of natural evolution? Beyond that, we can ask a more specific question: does evolutionary computing give support to the contention that variation and natural selection alone are inadequate to explain the origin and variety of life?"
As he points out, there are good reasons to recognize that evolutionary computing is very different from biological processes. A closer more technical analysis shows that the differences between evolutionary computation and biological processing are very substantial. Such finding, however, do not provide evidence of design by an external source, but rather the inadequacies of evolutionary algorithms.
IP: Logged
|
|
Noel Rude
Member
Member # 516
|
posted 11. November 2002 17:44
Matt Brauer asks, "Do you think that the 'formalist/platonist' distinction mirrors the 'empiricist/realist' distinction in some meaningful way?" Interesting question and, though I'm only a linguist trying to camoflage my blank stare, I'd imagine that -- as this dichotomy is defined in the review you quote -- most actual scientists would be realists: Who wants to be engaged in constructing useful fictions. And though surely all mathematical Platonists are realists, probably not all realists are mathematical Platonists.
Which brings up my fear after posting my last post (and which W. Bergerson attempts to remedy): What does all this have to do with the original thread of Evolutionary Computing. Let me try to recover by noting that evolutionary computing obviously requires a computer (which is designed!) and a designer who can implement for his specified purpose whatever insights might come from the algorithms and random functions provided by his computer. If we believe that the universe has an algorithmic component (ultimately answerable to the Platonic realm) and a random component (operative at the quantum level?) AND design -- then would not evolutionary computing (if at all fruitful in technology) support this belief?
Any practical success of evolutionary computing would suggest that chance and necessity do play a role in design (how many inventions have been inspired by chance events and unforseen reactions to new exigencies?). But unless someone can point to evolutionary computing having created a whole new complex entity (with specified complexity) -- all unguided -- then there is no support to the notion that chance and necessity alone are adequate to explain the origin and variety of life.
IP: Logged
|
|
warren_bergerson
Member
Member # 262
|
posted 12. November 2002 11:43
Noel,
Quote: But unless someone can point to evolutionary computing having created a whole new complex entity (with specified complexity) -- all unguided -- then there is no support to the notion that chance and necessity alone are adequate to explain the origin and variety of life.
I don’t entirely agree with the terminology you use, but I think you capture the basic requirement that must be satisfied if one is to offer a ‘scientific explanation of the origin and variety of life’.
As a starting point, you need to replace the term evolutionary computing with ‘a definable set of deterministic/materialistic process’. The term evolutionary computing is ambiguous and ill-defined. At times, the term evolutionary computing refers to a process involving natural selection and random mutation/variation. At other times the term refers to any mathematical/logical process which produces change. In the absence of a precise mathematical/logical definition, the term ‘evolutionary computing’ is meaningless/useless in scientific analysis.
The phrase ‘chance and necessity’ is redundant. In scientific terminology both chance -stochastic fluctuation and necessity refer to ‘produced by deterministic processes’. The term chance as in ‘unexplained miracle of chance’ is not a scientific/deterministic concept.
Finally, if you are going to offer a scientific explanation, it must be a scientific explanation of a precisely defined phenomena. If someone is to offer a scientific explanation of the ‘origin and variety’ of life, then they must first offer a precise definition of life.
Correcting for terminology, your statement becomes, - "Unless someone can point to a precisely defined set of deterministic/materialistic processes which can create new forms of precisely defined life from either existing life forms or from non-life forms, then one can not claim that the origin and variety of life can be explained by deterministic/materialistic processes."
It will be noted that this restated requirement for ‘claiming a deterministic/materialistic explanation’ is not a requirement of a peer review science such as evolutionary biology. Peer review biology can and does claim that at least the variety of life can be explained by deterministic processes even though it can not identify a process which could create a new life form. Peer review biology goes a step further and claims it has a scientific model or theory of evolution which can explain the variety of life even though this model or theory can not explain, model or simulate the modification of a single life form to create a new life form.
It is probably not possible/practical at the present to create new or artificial life forms, but it does appear possible to 1)offer a precise mathematical logical definition of life, and 2)identify and precisely define a set of mathematical processes representing deterministic materialistic processes, which are capable of modifying one life form to create another life form and are capable of modifying certain non-living systems into a living system. It is thus possible to demonstrate mathematically that a certain type or class of mathematical process have the capacity to create and modify the mathematical logical equivalent of a living system. [This is based on defining a life form as a set of evolving/maintaining teleological causal relationships and ‘information generating machines as the type of mathematical logical processes which capable of modifying and generating life.]
As your comments point out, it is not reasonable to claim the EA’s can explain the origin and variety of life if it is not possible to demonstrate the validity of the claim. It is, however, equally unreasonable to suggest that the inadequacies of EA’s can be used to justify inferences other than the inadequacy of EA’s and the scientific standards supporting EA’s.
IP: Logged
|
|
Noel Rude
Member
Member # 516
|
posted 12. November 2002 18:10
Warren Bergerson is right: quote:
The phrase ‘chance and necessity’ is redundant. In scientific terminology both chance -stochastic fluctuation and necessity refer to ‘produced by deterministic processes’. The term chance as in ‘unexplained miracle of chance’ is not a scientific/deterministic concept.
Except that this was Einstein's argument when he asked, "Does God play dice ...?" The Darwinists were seemingly unaware that physics had banished chance, but then Bohr brought it back -- which is good -- because we can imagine three types of explanation (with combinations thereof): chance, necessity, and design. Now ID is bringing back design -- which is even better -- and ID does not deny chance and necessity.
Also I'd say that defining life may be both easy and difficult -- it just depends on how many essentials we wish to ignore.
[ 12. November 2002, 18:14: Message edited by: Noel Rude ]
IP: Logged
|
|
Noel Rude
Member
Member # 516
|
posted 12. November 2002 20:36
Seeing my sin (posting without having read the article which initiated this thread) -- I've now read it. Very interesting. Have always told my friends that evolution is evidence -- evidence for design -- because the only observable evolution is of man's making. Therefore, rather than letting the materialists cite evolutionary computing as favoring a materialist story of origins, better that ID capitalize on this example of human creativity.
ID is often chastized for inferencing the unobservable when it's really the other way around. Materialist accounts of "the appearence of design" can only be inferred -- we cannot observe the process. But design is constantly occurring everywhere that we are -- and therefore we can study it. I say design is much better informed by the study of human agency than by the reasoning of the theologians. At least we can observe the former in action.
The article brings up the question of an increase in information that might arise in evolutionary computing.
Information increases via feedback from the environment. Both men and machines (which are made by men) can learn -- so the question is: Can chemicals learn? And if so, what are the limits re evolution? Also perhaps it behooves IDists to have another look at some of the old Lamarckian claims, and then of course there's Rupert Sheldrake. Lots to think about ...
The article also suggests the question: Is randomness necessary for design?
Some have speculated that the randomness of the quantum realm is the root cause of human consciousness -- but this makes sense only if the mind is completely random. Yet, on the other hand, if quantum randomness cannot account for human agency, might it be where the soul effects its will? If we really cannot bend spoons (a la Uri Geller), then maybe this is because we do not have authority over the laws of physics. Perhaps we can only effect a material outcome at the quantum level -- hence its indeterminacy. Thus my soul does not even have direct control over my arm -- cut the nerve and I cannot bend it -- but I can effect the firing of neurons in the brain that will result in a chain reaction which culminates in my arm bending.
Maybe design is impossible without chance and necessity -- the necessity ultimately of the Platonic realm (to which the mind has access according to the mathematical Platonists) and the contingency provided by the quantum realm.
And who knows -- perhaps even the Designer ordinarily inputs information only at the quantum level. Hasn't William Dembski suggested another benefit of inputting information at this level? -- this avoids inputting energy to the cosmos.
IP: Logged
|
|
RBH
Member
Member # 380
|
posted 13. November 2002 00:45
Noel wrote quote:
Have always told my friends that evolution is evidence -- evidence for design -- because the only observable evolution is of man's making. Therefore, rather than letting the materialists cite evolutionary computing as favoring a materialist story of origins, better that ID capitalize on this example of human creativity.
and quote:
The article brings up the question of an increase in information that might arise in evolutionary computing.
Information increases via feedback from the environment. ...
I commend to his attention my remarks on the generation of information by evolutionary algorithms in the absence of "feedback from the environment" on ARN at this URL. I refrain from repeating the post here since it runs over 2,000 words.
RBH
IP: Logged
|
|
|
Printer-friendly view of this topic