Posted By Caulimovirus on October 22, 2008
“Darwin’s Unpaid Debt”
.: If you pick up today’s Lariat and turn to the back page, you’ll find this paragraph:
Dembski was introduced to the audience by an officer of the aspiring chapter of the American Scientific Affiliation, junior Sam Chen, who said Dembski’s research as “absolutely phenomenal.” Although, after Dembski’s lecture, some audience members seemed in disagreement with his intelligent design theory.
.: About Sam Chen we all remember, but who were those audience members that seemed in disagreement with Dembski’s intelligent design theory? One was sitting right next to me, and the other was me. Had the talk been advertised in the Baylor Science Building like all the previous lectures in this series, I imagine I would have had more people sitting next to me.
.: The talk was titled “Darwin’s Unpaid Debt”, and I figured it would more or less cover the same material in his 2001 talk of the same name as reported by Dave Thomas. It consisted of two prongs:
.: Dembski’s stated goal for the night was not to argue for alternatives to evolution, but to simply offer a critique. Curiously, he went out of his way to note that often in science a theory isn’t simply displaced but rather replaced (cue Kuhn and “paradigm shifts”). Despite this, he repeated his intent to only critique evolutionary theory.
.: He gave a rather accurate characterization of evolutionary thought at the start: evolution is a blind, instant gratification process that has as its primary fuel random variation and natural selection and does not rely on teleological innovation. He then quoted some famous figures in biology, such as Francisco Ayala (“Darwin’s greatest accomplishment was to show that you can get the creative organization of living things without the need for a creator.”) All of this sounded just swell to me.
.: Now, despite all these new facts and ideas scientists have discovered, Dembski maintained that we have not “transcended Darwin.” We biologists, according to Dembski, still cling to the idea of natural selection acting upon random variation as the sole source of biological innovation. This of course brings us to what Dembski calls Darwin’s “promissory note”: natural selection and random variation can explain the emergence of biological information.
“I’m using ‘biological information’ loosely right now.”
.: Evolutionists say natural selection is adequate to bring about the great diversity of life, but must this be the case? Francis Collins thinks so, especially when he writes “this must be the case.” Well, announced Dembski, if it must be the case, then that sounds like mathematics.
.: Cue Darwin’s Black Box, self-replicating robot factories, transportation motors, vesicles and monorails, bacterial flagellum, bidirectional motor propeller, all this biological complexity that Darwin had no idea about (remember, we haven’t transcended him, so if he didn’t know it we don’t know it either).
[Side note: Dembski said the flagellum came before the Type III secretary system and not vice versa. Multicellular organisms, he says, have only been around for 500-600 million years, but the need to swim around in water has existed for billion years. Has Dembski made his views on the age of the Earth clear before, or is this just news to me?]
.: So the challenge to the Darwinian community, then, is for them to present plausible evolutionary pathways of natural selection acting upon random variation to bring about novel biological structures and processes. So far, sayeth Dembski, all that has been presented is wishful thinking and speculation. No tests have been performed that show step-by-step mutations of simple systems into more complex systems, and if they have, then that more complex system isn’t irreducibly complex, and the challenge still stands.
.: More names were named and quotes quoted. James Shapiro said no detailed accounts exist, only wishful speculations. Franklin Harold said the same thing. Even a certain process theologian with no dog in this race has claimed whenever he speaks with biologists, they are never able to provide him with detailed accounts.
.: Detailed accounts. Think about what that entails. It’s not enough for Dembski to show him two similar proteins with different functions. You have to show him the exact mutation and substitution and the environment in which all of these changes arose, and the different proteins need to be part of a complex system for the change to be deemed significant, and you have to show that each step was beneficial, because — remember? — evolution is really only about natural selection acting upon random variation. So what if you do research and discover things that you didn’t previously know but which align perfectly with the theory you’re operating under. So what if scientists uncover a few more details — they’ll never uncover enough, because they’ll never uncover them all; what Franklin Harold said in a book from 2001 and what James Shapiro said five years earlier will forever be true. Forever.
“I’m not sure I could pull out more recent quotes [of scientists stating no details account exist]. Not that the field has gone any further than what I have described [inaudible] but I think the biological community, in its effort to squash intelligent design, is now just ignoring the point.”
.: But all of this is small beans, because really the ultimate irreducibly complex system is the molecule of life itself: DNA. You need not only DNA but also RNA and proteins for the system to function properly. “What could be a simpler system?” asked Dembski. “When you take that system, and erase, as it were, protein synthesis mechanisms, you don’t have life at all; you don’t have viability.” Life itself could not logically emerge, it seems, because before life came about there would have to be non-living precursors, and evolution obviously can’t work on non-living things. You can’t evolve if you’re not alive. Or something like that.
.: That all was just prong number one of Dembski’s critique of evolution; the rest of the talk focused on a more proscriptive approach. Cue John Stuart Mill and the method of differences: if you want to explain different effects, you must identify the differences in causes. Common causes, then, cannot explain the differences in effect. So, take the causes of evolution: replicability, heritability, random variation, and natural selection. Now look at some of the effects we observe in the world: organisms gain complex information over time rather than stay stagnant. Assume that natural selection and random variation are doing most of the work in these cellular contexts. What kind of tests have been performed with these causes in mind?
.: Dembski mentioned Sol Spiegelman’s polynucleotide synthesis experiments. The idea here is that whole viral RNA genomes were dumped into a mixture with RNA polymerase (something that copies RNA molecules), and the replication process was carried out over several generations. The original molecule was 3,300 nucleotides long, but shorter versions of the molecules replicated more efficiently and soon took over the population, resulting in a 470 nucleotide strand known as “the little monster”. Clearly, Dembski stated, this shows not a gain of complex information but a loss of information — and that’s with all the same causes (replicability, heritability, random variation, and natural selection) in place! So if we have the same causes in the lab as we do in the wild, and we see different results (gain of information in the wild, loss of information in the lab), then we don’t actually have the same causes in the wild that we do in the lab; something is missing. In other words, evolution isn’t the whole picture. Confused yet? Good, because Dembski’s retelling of Spiegelman’s little monster story is wrong.
.: For starters, what was Spiegelman actually testing? He was looking for for ways to replicate RNA directly, since the bacteriophage he was working with only had RNA, not DNA. He wondered how one little strand of RNA could manage to copy itself in a cell that was already replete with DNA and thousands of other RNA molecules. Spiegelman surmised that the polymerase which bound to the viral RNA did not bind to the other RNA already in the cell. He found such a polymerase, and demonstrated conclusively that it bound specifically his viral RNA. But he did something more.
Because he could produce unlimited generations of RNA, Spiegelman was able to watch Darwinian selection in action at the molecular level. By varying the selective pressure on the system (such as temperature, nutrient mix, or the reproduction time allowed) he could “create” mutant RNA molecules with a wide range of properties. One of these, for instance, gradually trimmed itself down from the original 3,300 nucleotides to 470–just big enough to grab the replicase enzyme and copy itself. It had apparently jettisoned sections that coded for protein coats and other components it didn’t need in the artificial environment. (source)
.: The key point to realize here is that the function of the RNA in Spiegelman’s experiment was not survive in various environments, invade a bacterium, hijack its machinery, and replicate yourself. Rather, it was be replicated. It didn’t have to do most of the things required of viruses in the wild, like code for a coat protein or even code for its own replicase. Indeed, later research was able to trim down the length of the polynucleotide from 470 to double digits, and it was shown that this remaining length of the original molecule was simply the recognition sequence that bound to the polymerase. So, the causes are not the same in both examples; Dembski’s conclusion is invalid.
.: From there, Dembski expounded on what he considers the central idea of biology, information. Information is the key to successful reproduction. If you’re looking for buried treasure on an island, it does no good to start digging at random. You’ll only succeed if you have a treasure map. So, the question that arises is, whence the treasure map?
.: Richard Dawkins’s weasel program, Dembski argued, could only generate the target phrase in 40 generations because the “information” was smuggled into the programming, i.e., it was placed there by an intelligent agent (Dawkins himself). For more robust genetic algorithms, Dembski went on to argue, it still holds that solutions are only as good as the fitness functions used to evolve them. The role of evolution, then, is to simply shuffle and combine the genetic information already made available to the organism or the environment, not to create new information.
.: For his conclusion, he reformulated Theodius Dhobzhansky’s famous statement into “Very little of microbiology makes sense in the light of evolution”; he stated the latest work in information theory says information can’t be self-generated by genetic algorithm; and he asked the ultimate question, “Where did the information come from?”
Q & A
.: I hate being the first to ask a question at a lecture, but since nobody else put their hand up, I had a go at it:
“On one of your last slides, you mentioned that information could be embedded in the environment. I’m wondering if, say you have an environment like the tundra where everything is white, doesn’t that give an organism the information or operating principle it needs – ‘Be more white’ – so it can evolve towards camoflague?”
.: I must admit I cleaned up my question here, not because of any foul language, mind you, but because it sounds much more disorganized when written down. At any rate, Dembski understood my question well enough to respond at length before shifting goal posts on me. To wit: yes, that’s certainly plausible, but that’s not a rich information source (by what criteria did he make that evaluation, one wants to ask), and that scenario still leaves us with the question of where the organisms that could evolve towards camoflague came from in the first place. The scenario just pushes the information back one more step. “The information has to come from somewhere.”
.: I persisted:
“It doesn’t seem to me to specify any kind of design. There’s a billion different way you can get a white creature — you can remove pigments or grow hairs that are slightly different. But simply having this information imparted from the environment into the organism (“be more white”), nobody has to tell the organism that, the organism’s not listening–”
.: He cut me off to repeat that it (“be more white”) was still low level information. Moreover, the changes that take place in those scenarios don’t involve new genetic information; the genetic potential was already there.
.: Dembski was then asked by the guy sitting next to me if he had read the Lenski paper, meaning the recent one that’s thrown conservapedia into apoplexy. Dembski thought he was referring to Lenski et al’s 2003 paper in Nature on genetic algorithms and talked for a few minutes about it instead. He found it sneaky that Robert Pennock, one of the co-authors, cited the paper in his testimony in Dover as refuting the work of Michael Behe, but at the same time the paper itself does not actually cite any work by Michael Behe. Doing so would automatically include Behe in the Science Citation Index, and we Darwinists are careful to avoid accidentally lending credence to the views of apostates.
.: Fifteen or so minutes of friendly questions ensued before I got a chance to speak again:
“As recently as 2003, there was a lot written about the flagellum and a bunch of its proteins. Out of forty of its proteins, it was shown only about 10 or so had homologies in other systems. And then, three years later it was shown that the homologies are actually greater in number, and they discovered that, of the 42 proteins, there’s only two of them that are absolutely indispensable to function that don’t have homologies elsewhere. The thing about that that strikes me, is it doesn’t have to be that way. It could really be this disparate system where nothing matches up and nothing makes sense. But when the scientists actually did get to looking for these pathways and steps, they found something. Not everything, but they found something to lead them on and reinforce the idea that this whole evolution thing is a pretty good framework to work in.”
.: Ah, that’s evidence but what is that evidence for? Homology at best suggest evolution by common descent. Cue talk of convergence and the unreliability of attributing similarities to homologies. But, even supposing they do find detailed pathways and homologies, contended Dembski, that only tackles the first prong; it leaves untouched the second prong (“Where did the information come from?”). “It’s not evidence for the efficacy of natural selection.”
“I was told, when I first took biology here, that the best way to learn biology is to learn everything first and then go from there. It seems to me that you want to look at these incremental pathways and plausible scenarios, and you want detailed step-by-step formulations of how something came to be, well, you gotta start somewhere. But it happens to be that when they do start looking, though they don’t have a complete picture of every single system everywhere, they do find that things are starting to fit together. And it doesn’t really have to be that way, but it is. It doesn’t have to be that the flagellum has these homologies, but it does.”
.: He responded that if every organism was completely different from one another, that would be good evidence for design and against evolution. Explaining every similar feature as homology is, however, a kneejerk response from evolutionists which doesn’t address the problem of mechanism.
“Suppose that they do find a plausible pathway for something like the flagellum or something even simpler like the type III secretory system or the F1F0 subunits of ATPase. Suppose they do find this detailed pathway; can’t you just move on to another cellular process and say, ‘Okay, figure that one out too, and until you have, you haven’t proved anything.’?”
.: Dembski made another appeal to Franklin Harold’s proclamation that no detail pathways exist. But, even if they did, that still wouldn’t address prong 2 of his talk. Moreover, proving these pathways is actually a way of showing intelligent design is testable!
.: The most incredible thing he said in this exchange, though, was the way he framed the problem of moving from a type III secretory system to flagellum. “We’re probably talking about thousands of structural changes, additions, subtractions, and functional changes to get from the type III secretory system to the flagellum.” In the same breath, he then anounced that scientists have been looking for twelve years to find an answer, and the type III secretory system is the best they have.
.: Twelve whole years! To figure out probably “thousands” of changes!
.: At this point Sam Chen stopped the Q & A session and most everybody left. My new friend and I stuck around to talk to Dembski afterwards. I wish I kept my recorder running, because I’m having difficulty recreating what followed. Normally, when I recreate a chain of events, I can do so because everything follows a logical progression. But when people use different meanings of the same word, things get hazy. Consequently, reconstruction of a past event is no longer a logical matter; one must rely on brute memory.
.: I tried to explain how I don’t think it makes sense to refer to genetic information in the same sense as regular information. If I take the sentence “The dog is red” and write it down on a piece of paper, it has the same meaning as when I type it onto a computer. Moreover, if I tell you the character “#” stands for “d” now, the sentence “The #og is red” still retains the original meaning, despite its different appearance. For all intents and purposes, this bit of information does the same thing in every scenario (e.g., it tells you what color the dog is). At the same time, it’s an entirely arbitrary representation of that information.
.: But, if you take a sequence of the genetic code, say ATG, the meaning of that sequence is not embedded in our representation of it. What the sequence “ATG” actually does is dependent entirely on the structure of the adenine, thymine, and guanine bases; it’s not arbitrary in any sense. I can’t take an adenine base and change it around like I did with the “d” character from “#og”. If I alter the adenine in any fashion (say, by adding a methyl group to it), I change its meaning — I change its function.
.: Dembski, however, still thought the comparison between human generated information and genetic information was perfectly apt because — get ready for this — you can patent genes. This inspired an undesired digression into the topic of what can and cannot be patented which utterly sidetracked me from the point I was trying to make, and I never did recover the conversation.
.: Now, I could be wrong in my retelling of this last bit. I didn’t have the recorder on, and if something doesn’t make sense initially it’s even more difficult to recall accurately at a later time. It could be I simply misheard him, or he misheard me and responded to the point he thought I made. I don’t know for sure, which is why I plan to ask him more questions at the complementary speaker lunch next Thursday.