From toms Mon Sep 6 23:13:13 1999 Subject: Re: TrueOrigin Feedback Response To: twallace@trueorigin.org (Timothy Wallace) Date: Mon, 6 Sep 1999 23:13:13 -0400 (EDT) In-Reply-To: <37C80128.1C70F7E0@trueorigin.org> from "Timothy Wallace" at Aug 28, 99 11:32:56 am X-Mailer: ELM [version 2.4 PL24] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 30750 Tim: | So then do you retract your assertion that I "object" to dS>= dq/T | when written as -ds <= -dq/T? Do you now retract your assertion | (repeated multiple times) that I do not understand that the entropy of | an unisolated system can decrease, at the expense of its | surroundings? Or are you blithely skipping along to other topics, as | if bearing false witness is a perfectly acceptable practice, requiring | no retractions or apologies on the part of the perpetrator? Please be polite. No need to get all upset. Just make clear statements of what you are thinking and we can discuss them. If you want to do verbal attacks (which the above feels like to me) then we can just terminate this discussion. Otherwise we may both have somethings to learn from each other. This was one of the clearer, although still indirect, statements that you have made. Previous statements were not made when you said you would, then you made the wrong statement. It has been like extracting teeth to get a clear statement from you. If you insist on making indirect statements, then you will inevitably be misunderstood. Though you now indirectly imply that you agree that -ds <= -dq/T is valid, my understanding is that you think it rarely happens. You have avoided answering my previous questions along those lines. In particular, you have not answered my question about a simple jar of water (http://www.lecb.ncifcrf.gov/~toms/twallace/#9 [corrected from #5]). Are you willing to make a clear positive statement about the Second Law? | I have no choice but to ask questions if I want you to substantiate | your statements. I suppose I could respond in the form of an | accusation, but I suspect that would be counterproductive. The above was rather full of accusations. "blithely skipping along", "bearing false witness", "the perpetrator". I surmise that either this is an intentional tactic to throw me off, which you have used with other people, or you are very scared and angry. If so, what are you scared of? What would happen if you were wrong? Your nasty words seem inconsistent to me since I thought you were coming from a philosophy that teaches that god is love, and that one should love ones fellows. Was I wrong? | > Ok, we've beaten "G2L" into the ground. Looks to me like it's | > just another form or various specializations of the Second Law. | | Thank you for finally acknowledging the simple fact. Ok, then let's just call it the Second Law, and drop "G2L", which sounds fancy but no special meaning. May I suggest you simplify and clarify issues by saying "The Second Law of Thermodynamics" on your web site? As for the various forms, Jaynes' lovely paper is really a neat to read! http://www.lecb.ncifcrf.gov/~toms/glossary.html#Second_Law_of_Thermodynamics | > ...there is no precise distinction between micro and macro | > evolution. | | That's not true. There is a very significant distinction. | "Micro-evolution", by definition, is the same thing as genetic | variation (the shuffling of existing genetic information). No, variation is not evolution. Biological evolution also requires selection. | It is both | observable and observed, measurable and measured, repeatable and | repeated -- in short, it has been scientifically verified as a natural | phenomenon. However, in every single case, the organism that has | undergone the variation is the same kind of organism. What do you mean by "the same kind"? Here are two sequences: Query: 5 tggattttggcgtaggtttggtctagggtgtagcctgagaataggggaaatcagtgaatg 64 ||||||||||| ||||||||||||||||| ||||||||||||||||| |||||||||||| Sbjct: 6558 tggattttggcataggtttggtctagggtatagcctgagaatagggggaatcagtgaatg 6499 Query: 65 aagcctcctatgatggcaaatacagctcctattgataggacatagtggaagtgggctaca 124 ||||||||||||||||| |||||||||||||||||||| || ||||||||||||||||| Sbjct: 6498 aagcctcctatgatggcgaatacagctcctattgatagaacgtagtggaagtgggctacg 6439 Query: 125 acgtagtacgtgtcgtgtagtacgatgtctagtgatgagtttgctaatacaatgccagtc 184 |||||||| ||||||||||| |||||||||| |||||||||||||| ||||||||| || Sbjct: 6438 acgtagtatgtgtcgtgtagcacgatgtctaatgatgagtttgctagtacaatgccggtt 6379 Query: 185 aggccacctacggtgaaaagaaagatgaatcctagggctcagagcactgcagcagatcat 244 ||||||||||||||||| |||||||| || |||||||||||||| ||||| ||||||||| Sbjct: 6378 aggccacctacggtgaagagaaagataaaccctagggctcagagtactgcggcagatcat 6319 Query: 245 ttcatattgcttccgtggagtgtggcgagtcagctaaatactttgacgccggtggggata 304 ||||||||||||||||| ||||| ||||||||||| ||||||||||||||||| || ||| Sbjct: 6318 ttcatattgcttccgtgaagtgtagcgagtcagctgaatactttgacgccggtaggaata 6259 Query: 305 gcgatgattatggtagcggaggtgaaatatgctcgtgtgtctacgtctattcctactgta 364 || |||||||||||||||||||||||||| ||||| |||||||||||||| ||||||||| Sbjct: 6258 gcaatgattatggtagcggaggtgaaataggctcgggtgtctacgtctatccctactgta 6199 Query: 365 aatatatggtgtgctcacacgataaaccctaggaagccaattgatatcatagctcagacc 424 ||||||||||||||||||||||||||||||||||||||||||||||| |||||||| ||| Sbjct: 6198 aatatatggtgtgctcacacgataaaccctaggaagccaattgatattatagctcaaacc 6139 Query: 425 atacctatgtatccaaatggttcnnnnnnnccggagtagtaagttacaatatgggagatt 484 || ||||| |||||||||||||| |||||||| ||||||||||| ||||| ||| Sbjct: 6138 atgcctatatatccaaatggttctttttttccggagtaataagttacaatgtgggaaatt 6079 Query: 485 attccgaagcctggtaggataagaatataaactttcaggtgaccgaaaaatcagaaatag 544 |||||||||||||||||||||||||||||||||| |||| || ||||||||| |||| Sbjct: 6078 attccgaagcctggtaggataagaatataaacttcggggtggccaaaaaatcag-aataa 6020 Query: 545 gtgtgggtatagaat-gggtctccttctncggcggggtcgaagaa 588 |||| | |||||||| ||||||||| | |||| ||||||||||| Sbjct: 6019 gtgttgatatagaatagggtctcctcccccggctgggtcgaagaa 5975 They are the same at 530 out of 585 bases (90%) with two gaps. Are they the same kind? How can I tell? By "how", I mean by what precise procedure or algorithm that I could program into a computer please. If you want to say something like "90% means they are of the same kind, then on what basis do you pick 90%? That is, I'm looking for an algorithm that does not have arbitrary constants. | The distinction is both precise and significant. To blur the | distinction is to show contempt for empirical science and mix fact | with fancy. Please stop being nasty. If it is so precise, please apply it to the above two sequences to answer whether they are the same kind or not. | Frankly, you are engaging in semantic subterfuge. Again, please keep the nasty implications out of this discussion. | "In the meantime, the educated public continues to believe | that Darwin has provided all the relevant answers by the | magic formula of random mutation plus natural selection -- | quite unaware of the fact that random mutations turned out | to be irrelevant and natural selection a tautology." | [Koestler, Arthur, Janus: A Summing Up (New York: Vintage | Books, 1978) p. 185] Random mutations irrelevant? How is it that there is such a diversity of people if we came from only two people? (Is that what you believe?) | "It has been estimated that those chance errors occur at a | rate of about one per several hundred million cells in each | generation. This frequency does not seem to be sufficient | to explain the evolution of the great diversity of life | forms, given the well-known fact that most mutations are | harmful and only very few result in useful variations." | [Capra, Fritjof, The Web of Life (New York: Anchor Books, | 1996) p. 228] Oops. He must have made mistakes. Aside from the fact that he may have forgotten that somatic cells are not germ cells (it depends on the context) there are lots of progeny and there are populations of animals. There is lots of variation as a result. (The mutation rate in HIV is one per genome per generation! Typical numbers are 1 in 10^6 per gene locus, which is quite sufficient to do all kinds of fancy selections on bacteria in the lab since one can easily have 10^8 bacteria per mil. Of course it can go WAY up when a transposon is active, or maybe even under stress conditions.) Rock layerings (the grand canyon!!) and radioactive decay rates and other evidence indicate that there was also a lot of time available. | "It should be clear that the claim for an inherent | evolutionary increase in entropy and organization is based Oops! This person didn't understand that entropy decreases correspond to "organization". They got it backwards. | on an arbitrary model which shows signs of having been | constructed simply to yield the desired result. There is | nothing in evolutionary or developmental biology that | justifies their assumptions that a successful mutation | (which seems merely to mean a selectively neutral one in | their model) Well that seems silly, but it is clearly taken out of context Tim! | is always associated with an increase in some | global measure of phenotype. Ok, good, it will be worth publishing on this. | Nor is there anything to | support the assumption that new species arise as the result | of single gene mutations and are initially genetically | uniform. If these assumptions are removed, the whole edifice | collapses." Generally it is not thought that a single mutation could cause speciation; it is probably a cluster of them. | [Charlesworth, Brian, "Entropy: The Great Illusion," review | of Evolution as Entropy by Daniel R. Brooks and E. O. Wiley | (Chicago: University of Chicago Press, 1986, 335 pp.), | Evolution, vol. 40, no. 4 (1986) p. 880] Oh yea. Brooks and Wiley. They confused digital noise with evolution! In other words, they didn't know what they were doing. Are they creationists? How surprising! (Or did you taking them out of context Tim?) | "The fruit fly has long been the favorite object of mutation | experiments because of its fast gestation period (twelve days). | X rays have been used to increase the mutation rate in the | fruit fly by 15,000 percent. All in all, scientists have been | able to 'catalyze the fruit fly evolutionary process such that | what has been seen to occur in (fruit fly) is the equivalent | of many millions of years of normal mutations and evolution.' | Even with this tremendous speedup of mutations, scientists | have never been able to come up with anything other than | another fruit fly." | [Rifkin, Jeremy, Algeny (New York: Viking Press, 1983) p. 134] With legs coming out of its head? With FOUR wings? Four wings would be considered another species, like a dragon fly! How about eyes all over its body? :-) Rifkin should do some more reading, he chose a bad example! Why would it be so easy to get 4 wings? Because the back two were repressed into the halter during evolution. The ancestor was multi segmented; later genes repressed things in different segments. It's rather lovely genetic work, and a beautiful system. Worth reading about, Tim. You have segmentation and homeo box genes too. | "The proof of the occurrence of mutations is by no means a | proof of a current evolution. The most important the | inescapable question, is whether the mutations are fully vital, | so that they are able to survive in natural stands. A review of | known facts about their ability to survive has led to no other | conclusion than that they are always constitutionally weaker | than their parent form or species, and in a population with | free competition they are eliminated." | [Nillson, Heribert, (Lund, Sweden: Verlag CWK Gleerup, 1953), | (English summary) p. 1186] So how does one account for the new strains of influenza that pop up every fall? What about the changes in HIV during infection? Golly that's an old reference. | > There is no such thing as "genetic potential", just as there is | > no limit to the number of sentences. | | Indeed just as there are distinct limits to what constitutes a | meaningful sentence, there are also distinct limits to viable genetic | arrangements. Meaningful sentences aren't random mixtures of words, | letters, and spaces, and nor has genetic code ever been shown to be a | random mixture of unordered genetic material. Genetic variation is | not a mix-n-match free-for-all. Oh of course not. Even with the demand for a functional system, the sequence space is still vast! Life on earth could not have possibly explored even a tiny fraction of the possibilities. The size of the sequence space does not prevent organisms from evolving. | By "genetic potential" I simply mean the range of genetic variations | inherent in the subject population. The population's gene pool | contains a finite set of genetically determined features. Well I have to agree it is finite. But what does that mean in this context? E. coli is 4639221 base pairs long. So the (finite) number of (DNA based!) organisms this size is 4^4639221 = 10^2793089 = 1 followed by 2.7 million zeros, which is horrifically big. Most probably won't work. Given the variants of E. coli that are known, perhaps 70% of these would be functional. (How did I compute that? You'll have to read my papers.) But if we streak purify a bacterium (do you know any basic biology?), there is only one to start with. There is initially no variation. Yet I can get sequence variants. So 'genetic potential' must be bigger than just the existing variants. What would limit where the changes occur? Certainly the DNA polymerase can't do that, it doesn't have enough brains. So there is no "genetic potential", no limit on an organism on where it can get to from where it currently is. If you still don't understand that, go and streak purify some bacteria several times and then *select* some mutants. | The | "genetic potential" of that population (while surely unknown, due | largely to limits in the scope of man's knowledge) consists of the | entire set. It's a subset, as discussed above. | That limited set does indeed exist, just as the number of | potential sentences also has a limit, since the number of words, | meanings, and combinations, while so vast as to be unknown (if not | incomprehensible) to man, is nevertheless finite. It is big but not incomprehensible. People play with much bigger finite numbers. Though finite, it is so large that it does not restrict. Your concept of limited "genetic potential" is full of holes. That's why modern molecular biologists don't use it. | > It is also wrong to think that selection can't produce | > unequivocally new functions, as that has been demonstrated | > many times by things like SELELX. | | It isn't clear what you mean by introducing the term "functions", but | I invite you to cite a basis for thinking that selection, in and of | itself, can yield an increase in quantity and quality of genetic | information and -- therefore -- any unequivocally new genetic traits. Read the SELEX literature and read the literature in the next few years. But for now, a new function can be for a molecule to stick to another specifically or for a molecule to do some enzymatic reaction. These have been done starting with random sequences by SELEX and other techniques. | I also invite you to cite a basis for believing that SELEX produces an | increase in quantity and quality of genetic information and -- | therefore -- unequivocally new genetic traits. Oh that's easy, read any SELEX paper. Compute the information content of the binding sites. | > The pool doesn't contain the new or "more complex" information | > at first. It appears in variants. | | This seems like more semantic subterfuge. Please keep the nasty implications out of the discussion, it implies to me that you are unable to respond rationally and so chose to attack me instead as a distraction. This is not an attack, it is an observation about what you are doing. | Let's keep our terminology | clear: Variation is limited to the manifestation of variables | inherent in the genetic code. Mutation is limited to (degenerative) | changes to the genetic information itself. There is no reason to suppose that mutation (or more precisely, sequence changes) are always degenerative. The SELEX experiments show clearly that there is no such limitation. There is a problem with the terminology. When we say 'mutation' we usually imply a bad thing for the organism. But that is our term. I work with lots of people who are sequencing human genes. They find a change in the gene and would like to know if it is a polymorphism or a mutation. We do a splice junction analysis and can often make clear statements about what's going on. Other times we can't. In either case more experimental would could be done to determine what the effect is. So we have to be very careful with our language. That whole project started because someone mistakenly stated that a certain T->C change caused a colon cancer. Our analysis said it was unlikely. Later someone else sequenced 20 normal people and 2 of them had this change. So it was a polymorphism. Language can fool you. Early this year on Burt Vogelstein, who did the 2/20 experiment told me that some company had sunk over a million dollars into an assay for this polymorphism! What a waste of money! (The reference is http://www.lecb.ncifcrf.gov/~toms/paper/colonsplice/index.html) | While it is a popular | practice to treat these two as synonymous, they are not: one is an | empirically established natural phenomenon in healthy populations, the | other is an empirically established source of disease and defect, the | effects of which are largely weeded out by selection (a process for | which we should be thankful!). Well, though the mutation was interesting, I rather am sorry that the little 9 year old boy whose DNA sequence we analyzed has Xeroderma Pigmentosum. It's sad to see him suffer. It was sobering, last January, to finally see the faces of the people whose DNA I had helped analyze. I am not thankful for this, aside from my observation that there is evidently no one to "thank" for it other than an errant DNA polymerase molecule and a bit of thermal noise or a photon. (The reference is S. G. Kahn, H. L. Levy, R. Legerski, E. Quackenbush, J. T. Reardon, S. Emmert, A. Sancar, L. Li, T. D. Schneider, J. E. Cleaver, and K. H. Kraemer. Xeroderma Pigmentosum Group C splice mutation associated with mutism and hypoglycinemia - A new syndrome? Journal of Investigative Dermatology, 111:791-796, 1998.) But how are we to think of mutations in the hemoglobin gene that cause sickle cell anemia but *also* protect the heterozygote against malaria? Is it a "bad" mutation or beneficial one? | > The DNA polymerase makes mistakes when it copies and | > there are lots of rearrangements by transposons... | | This has a corruptive, degenerative effect on the genetic code. Only most of the time! But let's be really clear that the function of transposons is *to mess up the genome*! On occasion a really good mess up is advantageous. That's why we still have transposons, because they give us the ability to evolve. (No engineer would intentionally put transposons into a computer, now would they?) So there you go. Along with the problem of pain, you now have the problem of transposons to deal with. | Errors do not create new information; they damage existing | information. There is no empirical basis for postulating new or more | complex information from a process that degenerates existing | information. See above about SELEX. Oh it's easy. Suppose I put contact glue on the tail side of 1000 coins. Then I flip them by jostling the box and let them glue down. Then I repeat the process. Roughly how long will it take before all of them are heads up? (Assume that the coins are spread out and don't glue to each other.) How about with 4.7 million coins? | > > These, again, are examples of genetic variation (animal | > > and plant breeding is selection for specific, already | > > existing traits or combinations of already existing | > > traits, and laboratory selection of biological systems | > > likewise selects that which already exists). | > | > No, when people do chemostat experiments on bacteria (to pick | > an example) the first thing that one should do is streak out | > the bacteria twice to get a pure genetic strain. Then this | > is frozen down as a record of the start point. All further | > variation comes from that one strain. So in any decent | > experiment (ie, publishable) there is NO variation initially. | | Any subsequent changes in such a strain of bacteria will still result | only from either inherent genetic variables or mutations. If a | specific trait has been effectively bred out of the strain by | artificial selection, it will not re-appear unless re-introduced | through exogenic contamination of the strain or through mutation. In | either case, no increase in the quantity or quality of available | genetic information has been effected. Right. Read the literature in the next few years. | > > > So "evolution" does not have a problem with the Second Law. | > > It has not been empirically and unequivocally demonstrated | > > that known natural processes can account for the generation | > > of new and more complex genetic data, new and more complex | > > organs and traits, and new and more complex energy conversion, | > > transport, and storage mechanisms. Specific natural processes | > > to which can be attributed the entropy decreases necessarily | > > associated with the generation of these things have similarly | > > not been empirically and unequivocally demonstrated. | > | > That's very different from having a problem with the Second Law!! | | On the contrary, that IS the problem evolution has with the Second | Law... Hunh? The Second Law allows for decreases in entropy (see start of this email and your previous one). So there is no problem. If you still think this, show me how precisely in the equations. Just because the Second Law doesn't (alone) explain the information gain does not mean that there is no information gain. | "One problem biologists have faced is the apparent | contradiction by evolution of the second law of | thermodynamics. Systems should decay through time, | giving less, not more, order..." | [Lewin, Roger, "A Downward Slope to Greater Diversity," | Science, vol. 217 (September 24, 1982) p. 1239] Why did you quote this? By doing so are your now saying that you agree with this statement? Do you have a specific position on this or are you going to change it all the time? So to respond specifically to this quote: Nope. In the previous email you (only!) apparently agreed that -ds <= -dq/T. This is the same as saying that an increase of information does not violate the Second Law since -ds <= -dq/T is equivalent to KTln(2) <= -q/R. So if heat leaves a system (ie -q) then the information change of the system (R) can be positive. See my second molecular machine paper: http://www-lecb.ncifcrf.gov/~toms/paper/edmm/ | "The greatest puzzle is where all the order in the | universe came from originally. How did the cosmos | get wound up, if the second law of thermodynamics | predicts asymmetric unwinding towards disorder?" | [Davies, Paul C., "Universe in Reverse: Can Time | Run Backwards?" Second Look (London: King’s College, | September 1979) p. 27] Spreading out cools things down. Total disorder is "hidden" in heat. There is no puzzle. Again, you have indirectly (only!) accepted that such decreases DO NOT VIOLATE THE SECOND LAW. So why do you quote this here? Was it a mistaken oversight? | "We are faced with the idea that genesis was a | statistically unlikely event. Based on one case? We can't speak very clearly about statistics in such a circumstance. | We are also faced with | the certainty that it occurred. Was there a temporary | repeal of the second law that permitted a 'fortuitous | concourse of atoms'? It is unnecessary to postulate this. The Second Law just means the spreading out of heat; it can't be 'repealed'. | If so, study of the Repealer and | genesis is a subject properly left to theologians. Or | we may hold with the more traditional scientific attitude | that the origin of life is beclouded merely because we | don't know enough about the composition of the atmosphere | and other conditions on the earth many eons ago." | [Angrist, Stanley W., and Loren G. Hepler, Order and Chaos | (New York: Basic Books, Inc., 1967) p. 205] That's a simple enough hypothesis. Let the facts come forth. | > You vaguely imply that SELEX doesn't work! | | You vaguely imply that SELEX generates previously non-existent genetic | information. I suppose you'll have to go read the papers! Maybe you should do some work in a laboratory and try it yourself instead of this arm chair stuff. | > Many tiny changes over millions of years could mold the | > organism enormously. This is often, granted, an | > extrapolation. | | "Could"? On what empirical basis? It is ONLY an extrapolation, and | it has NO empirical support. Oh only the observation of the huge variety of organisms. Only the observation of DNA sequence variations. Only the fossil record layered in a way that no flood could have produced. Only computer simulations, mathematical models and theory ... lots of geology too. The nice thing about all these observations is you don't have to believe anybody about them, you can go check for yourself. When you yell that there is no empirical support, you make it appear that you haven't ever looked closely at a road cut. There is one near you at Sidling Hill, MD. http://www.berkeleysprings.com/ecotourism.htm Sidling Hill Mountain and Visitor Center -- The U.S. interstate highway system blasted a cut through the mountain in order to make way for Interstate 68. This cut reveals a cross-section of a geosyncline-concave layers of sedimentary rock at various layers of the earth's crust. Interpretive exhibits are available free at the visitors center. Approximately 15 miles west of Berkeley Springs on I-68. A picture: http://www.geocities.com/Yosemite/Geyser/2368/dayone2.html | > ...There isn't a better explanation around. | | That's strictly a matter of opinion. And even in the absence of a | "better explanation", a scientifically unsubstantiated explanation | should be neither touted as fact, nor assumed to be true. You know a better, simpler explanation that fits the facts and does not lead to infinite regress? "God" is not an answer because: 1. it is much more complex, 2. it doesn't fit the observable facts; 3. it gives infinite regress. Also, historically it has failed as a predictive hypothesis. One I've been thinking about recently is the question of why people still don't insist on saying that god throws lightning bolts down from the sky. This was a clear hypothesis - Zeus did the throwing. Why isn't that idea still held by most people? | > I see what you are asking for. The question is: what kind | > of data would satisfy you, besides time travel? | | The same kind of data that should be required by any self-respecting | scientist: Please watch your language. You imply, without knowing me, that I have no self respect. Whether or not it is true, please strip such implications from your messages. While it is true I am a scientist, you denigrate this on your web site by putting quotes around it. A loving person would not do this. Please remove them. You have yet to answer my question (July 30, 1999, http://www.lecb.ncifcrf.gov/~toms/twallace/#11) as to what you are. | nothing less than that which would satisfy the requirements | of empirical science. That's not said anything so far, and science is broader than just measurements. You keep pounding on "empirical science". What about theoretical science? | That is, data that demonstrate unequivocally | that natural processes alone can increase the quantity and quality of | available genetic information. You take a rather limited view of how science works. Scientists do more than just make measurements. How should we measure genetic information? Will Shannon's information theory do? It's lasted 50 years and has allowed us to build a world wide communications system. Is that good enough? What do you mean by "quality"? Will quantity alone satisfy you? | > If I measure the two short legs of a right triangle, can I | > figure out the length of the long side without measuring it? | | Yes. Why? Not because of mere extrapolation, but because you know | the unchanging laws of mathematics (not unlike the laws of | thermodynamics) yield consistent results. The unknown data is | calculated from the known data and is determined precisely and | absolutely through a precise and absolute process. No, the answer - to my surprise!! this was not a trick question! - is that one *can't* predict the long side! One has to make a postulate about the number of lines through a point not on a line that are parallel with the line. In other words, one has to define the space, and it may be noneuclidian. This discussion point will have to wait for my publication. | This is a far cry | from assuming the existence of an unknown natural process for which | there is no empirical evidence -- based solely on an empirically known | natural process, the similarity of which begins and ends with the fact | that their names share the same root word! There a ton of evidence, and more coming in all the time you just somehow ignore it. Why do you ignore the evidence (eg, those diagonal layers on Sidling Hill)? The question is how can we explain the tons of empirical evidence that we already have at hand? Tom Dr. Thomas D. Schneider National Cancer Institute Laboratory of Experimental and Computational Biology Frederick, Maryland 21702-1201 toms@ncifcrf.gov permanent email: toms@alum.mit.edu http://www.lecb.ncifcrf.gov/~toms/