Tracking Rabbits Through Cambrian Fields
by BradfordTracking rabbits leads us to John Timmer's The Cambrian explodes and fish get limbs. Genetic tools in place and ready to explode. Sounds like the type of prose one would expect from one who has read The Design Matrix. But this refers to a pivotal event in natural history and some preparations that preceeded it. Let's have a look at a quote:
Erwin also described animal life that was poised to explode. The prior era was filled with the Ediacaran Fauna, which he described as, "no eyes, no appendages, lots of fronds, and maybe some guts." But that era also generated fossil embryos that suggest that bilateral animals predated the Cambrian. More telling, however, has been the findings of modern genomics and evo-devo. Genomic studies reveal that many of the genes involved in producing complex animals predate animals themselves, and some of the key regulators of bilateral animal development exist in Cnidarians, which don't share that body plan. Other work has revealed that genetic networks of regulatory genes that are used in appendage and body plan specification probably predate the origins of either limbs or a body plan.
Genetic networks predating the existence of that which they regulate. This watchmaker looks like he not only sees but has good long distance vision as well. And this:
In short, the genetic tools were in place were in place for millions of years before the Cambrian, but it took the Cambrian's unique combination of environmental challenges and opportunities to force organisms to deploy them in new adaptive combinations.
Indeed.
March 2nd, 2008 at 11:14 am
Hi Bradford,
Thanks for that link!
Comment by MikeGene — March 2, 2008 @ 11:14 am
March 2nd, 2008 at 11:43 am
After all, humans are 'just' elaborated Deuterostomes. A tube with appendages to stuff food into one end. Microevolution.
More particularly, bilaterians have segmentation and a set of homeobox genes. These genes are often specialized for specific axial domains. The distal-less gene pertains to protusions or extensions. They are common to bilateria. We would expect that certain fundamental genes, such as those that control basic body organization, might remain fundamentally similar, with other genes coevolving later that regulate the expression of these traits. There is still some uncertainty about the exact history. But the concept of protusions on segmented organisms evolving into more elaborate structures doesn't seem too far a stretch.
Like I said, microevolution.
Comment by Zachriel — March 2, 2008 @ 11:43 am
March 2nd, 2008 at 1:55 pm
Zachriel:
Easy of enough to say after the data flows in. Would we expect that: "Other work has revealed that genetic networks of regulatory genes that are used in appendage and body plan specification probably predate the origins of either limbs or a body plan?" Why would we expect that of a neo-Darwinian perspective?
Comment by Bradford — March 2, 2008 @ 1:55 pm
March 2nd, 2008 at 3:59 pm
Because cooption and multi-purposing are the norm under the neo-Darwinian perspective? There is nothing that is not cooptable. You are assuming these genes can only regulate appendage and body plan specification and have no other purpose. The neo-Darwinian perspective would assume they first played some other role and were coopted into their current role.
Comment by Todd Berkebile — March 2, 2008 @ 3:59 pm
March 2nd, 2008 at 4:36 pm
Echoing Todd Berkebile, under an evolutionary perspective, all new structures are adaptations or cooptions of existing structures. The particulars can be very interesting and surprising, though.
Comment by Zachriel — March 2, 2008 @ 4:36 pm
March 2nd, 2008 at 6:47 pm
Perhaps, but the theory itself does not lead to this. Novelty could arise without cooption too. We are left with little from NDE that would guide us in a choice aside from ad hoc adjustments based on incoming data. Why would genetic networks of regulatory genes preceed the existence of structures they regulate? Is that a pattern with regulatory genes or merely applicable to this instance? NDE has us reacting, not anticipating a goal or mechanisms geared to steer toward it. In any case novelty must preempt cooption at some point in the process. Cooption assumes the existence of that which will be modified.
There is little empirical support for that broad sweeping statement.
Comment by Bradford — March 2, 2008 @ 6:47 pm
March 2nd, 2008 at 6:59 pm
Bradford:
They didn't. Network A regulated structure X and later a modified network A' regulated structure Y. Your formulation is a rather transparent distortion. Let's be generous and call it wishful thinking.
Comment by Raevmo — March 2, 2008 @ 6:59 pm
March 2nd, 2008 at 7:24 pm
Rather than structure x preexisting the Cambrian with network A and becoming later modified so as to be compatible with regulatory network B. Why wasn't that the scenario? These are all vague, ad hoc descriptions without reasons to presuppose the actual order of development.
Comment by Bradford — March 2, 2008 @ 7:24 pm
March 3rd, 2008 at 1:19 am
Hi Todd,
After-the-fact, sure. But why didn't the "neo-Darwinian perspective" predict this? "I was surprised to learn that so much of animal biology was in place before the origin of animals," King says. "And I think that's what motivates most scientists - not learning that you were right, but learning that you were wrong."
Yet neo-Darwinian evolution can occur without cooption and multi-purposing.
How do you know this? Would you also argue that all gene products are equally cooptable?
Comment by MikeGene — March 3, 2008 @ 1:19 am
March 3rd, 2008 at 1:45 am
Are you asking why the "neo-Darwinian perspective" didn't predict every possible piece of evidence before that evidence was discovered? How about because they are only human. In this case what surprised King was that apparently more coption than he expected had occurred. We don't really know exactly which evolutionary mechanisms are responsible for what percentage of any particular life form. As such you should expect a lot of surprises as people investigate this further. If we already had such a perfect model that we could predict everything then further research in biology wouldn't be very fruitful, but no one is claiming that existing theories are that perfect. The very fact that so much research is going on in this area is proof that there's a lot we don't know.
To claim something is cooptable for another purpose all you need is the imagination to dream up some other possible purpose. Still, I'm willing to retract my statement and instead state, "We know of nothing that is not cooptable." And I would not even begin to argue that all gene products are equally cooptable.
Comment by Todd Berkebile — March 3, 2008 @ 1:45 am
March 3rd, 2008 at 2:05 am
Hi Todd,
No, I said nothing about predicting every possible piece of evidence. But when an evolutionary biologist is trying to understand the evolutionary origin of animals, why would she be "surprised to learn that so much of animal biology was in place before the origin of animals" when you claim "the neo-Darwinian perspective would assume they first played some other role and were coopted into their current role." You make it sound as if the neo-Darwinian perspective was superfluous to her research. Shouldn't she have said, "And we found exactly what the neo-Darwinian perspective led us to expect?"
I think you are trying to rationalize here.
Then why did you claim, "cooption and multi-purposing are the norm>?" Wouldn't you use "the norm" to make predictions?
It's not a question of any theories being perfect. You are trying to engage in some revisionism to make these findings sound like "business as usual." The Duck absorbs all data.
So you need someone to prove a negative for you?
Indeed. So it would be reasonable to assume that some gene products are rarely coopted. And from there, perhaps some gene products have never been coopted. The nice thing about front-loading is that it encourages a new perspective that in turn encourages new thinking and questions. While you seem satisfied with appealing to cooption as some brute given, declaring it a "norm" that can't be used to predict, and insisting it's just the duck quacking, I wonder if there are factors that are likely to make one gene more cooptable than others. The Rabbit beckons.
Anyway, I have to go to bed and dream about cooption. Since you are new here, I should also point out that my time is limited and I am unlikely to engage in argument during the week. So I'll likely talk to ya next weekend.
Comment by MikeGene — March 3, 2008 @ 2:05 am
March 3rd, 2008 at 4:54 am
I would like to submit the idea that co-option is actually evidence of front loading. Give me two seconds and hear me out.
A functional set of proteins is determined by their specific shapes. Furthermore, those shapes are determined by the laws of physics and chemistry, thus causing a potential template based on these laws. In fact, recent work has shown that this template of functional proteins among the full sequence space is extremely constrained.
But we are still missing a step here. We can't arrive at life just because these potential functional protein forms are encoded within the laws of physics. What we need is something to coax out these forms. That is done with the processing of an instructional tape. When these laws of physics act in accordance with a tape of nucleotides being operated upon by a processing system, a protein form can be created.
But there's a catch. The specific sequence on that tape of nucleotides, which codes for each and every protein, is not caused by any physical laws of attraction between the nucleotides on the tape. Furthermore, the arrangement of the system which processes the tape is not defined or predicted (that is, necessitated) by any laws of physics or chemistry. ie: put all the ingredients in a test tube and they won't self assemble based on the laws of physics. So, we either have an accidental pre-determined template of forms and their potential matches (co-opted or not) based on fine tuned physical laws awaiting a fortuitous informational accident to discover them, or we look at options other than the fortuitous accident.
Consequently, an evolutionary pathway and its subsequent molecular machine is the result of specific laws of physics and chemistry being organized by a sequence not determined by laws of physics on an instructional tape. Thus, at the most basic level, natural selection has nothing to do with the cause of molecular machines and their protein forms. It only selects (based on environmental factors) from an already available template that has been handed to it by extremely fine tuned laws of physics and chemistry being organized by non physical/chemical tape of instructions.
What does this have to do with co-option and front loading? Just as the evolution of a molecular machine must have a pathway of pre-existing templates — based on finely tuned natural laws being organized by non lawful tape of instructions — to follow, so must co-option follow a pre-existing template being operated upon by non lawful (in the sense of being not physically determined) instructions.
To just waive one's hands and say, "well look, we have co-option therefore evolution must be a process that takes a form that's already available and use it again," is to miss the point and not ask the scientific question of cause that counts. "What must be in place for co-option to even take place?" There must be a template already in place to even allow for the option of co-option, since this ultimately involves the re-matching of one form in an instance to other forms in another instance. As already explained, the matching of form require certain parameters which are caused by the laws of the system in question.
What causes this available co-option template and the necessary processing system and non-lawful instructions to discover it? Is co-option inherently a process which is void of foresight? Is feasible given an accidentally generated set of instructions and replication? As I have stated above "we either have an accidental pre-determined template of forms and their potential matches (co-opted or not) based on fine tuned physical laws awaiting a fortuitous informational accident to discover them, or we look at options other than the fortuitous accident."
It seems obvious, based on our understanding of the foresight necessary to create information processing systems, instructional tapes not defined by physical law, and evolutionary algorithms, that an instance of co-option within this program is best explained by intelligent front-loading of a re-usable template and the necessary non lawful instructions that will aid in unfolding the pre-determined potential co-opted forms.
This is quite similar to how all the information necessary to create a human containing approx 3 lbs. of the most complex structure in the universe with the ability to learn to write essays on philosophy and science, learn to interact socially and raise a family is contained within the fertilized egg and its programmed response to its environment. Indeed, the development of the brain, especially with parts of the brain being co-opted for different functions especially in the case of injury (read "The Brain that Changes Itself"), is an excellent example of evolution and co-option being guided by a pre-existing informational template in response to the environment.
Is there any evidence that a random set of laws and variables (absent intelligent programming) being operated upon by chance can even come up with a use-able template for pathways of co-option, must less generate the non physically/chemically defined instructions necessary to discover that template?
P.S. Resorting to any variant of a "natural selection" argument is to pre-suppose what you are attempting to explain, since these same laws which cause the template and merely *allow* (not define or cause) the processing of the non-lawful instructions also determine the environment which causes the natural selection. As already explained above, I am dealing with a level deeper than natural selection, as this level of templates of form and function and non-lawful instruction tape exists pre-natural selection. "It [natural selection] only selects (based on environmental factors) from an already available template that has been handed to it by fine tuned laws of physics and chemistry being organized by non physical/chemical tape of instructions."
That is indeed the key. The pre-determined template and kit of instructions was just waiting for natural selection to catch up and allow (not cause) that template and instructions to unfold, similar to how the template of an intelligent human being unfolds from a fertilized egg because the environment at the time allowed for it to happen.
Comment by CJYman — March 3, 2008 @ 4:54 am
March 3rd, 2008 at 9:13 am
Most genes evolve stepwise by structure and by function, adaptation. But there are known examples of entirely new functions evolving in microbiology. A famous example of cooption is the evolution of nylonase. An existing plasmid was duplicated and then subjected to a frameshift creating an entirely new function and new functional information (Ohno, 1984). The strongly supported hypothesis that this is due to various random mutations. The alternative suggestion that this was due to some sort of preplanning billions of years before the invention of nylon is not supported by the evidence.
If we look at macroscopic evolution, we see many examples of similar processes, e.g. air bladders evolving into lungs. The intermediate organisms were well-adapted to their own environments without regard to the future course of evolution.
As I said, all new biological structures are adaptations or cooptions of existing structures. The particulars can be very interesting and surprising, though.
Comment by Zachriel — March 3, 2008 @ 9:13 am
March 3rd, 2008 at 10:11 am
CJYman:
Very true. All one has to do to demonstrate that function is not dictated by chemical properties is scramble the order of the nucleotides. DNA remains DNA. The nucleotides are all there. Only functions are disabled because information is lost.
To look for other options is to look for reasons- rationality- the essence of science.
Very good CJYman. Attempts to use natural selection to explain the origin of what is needed to set off a process of adaptive genetic changes is a sign of desperation. The round peg in the square hole phenomenon.
These functional templates are cognitive end products. There's the design- right at the outset.
Comment by Bradford — March 3, 2008 @ 10:11 am
March 3rd, 2008 at 10:52 am
Bradford:
Chemical properties of DNA molecules are certainly not independent of their nucleotide sequences. Randomizing sequences therefore affects both function and chemical properties.
Indeed, there is some evidence (discussed on this very blog some time ago IIRC) for chemical affinity between codons and their corresponding amino acids.
Comment by Raevmo — March 3, 2008 @ 10:52 am
March 3rd, 2008 at 11:12 am
Raevmo:
The "affinity" consisted of a slight increase of corresponding AA codons in tRNAs which does not explain why proteins and nucleic acids became sequentially ordered so as to enable biological functions.
Comment by Bradford — March 3, 2008 @ 11:12 am
March 3rd, 2008 at 4:07 pm
Zachriel:
"As I said, all new biological structures are adaptations or cooptions of existing structures. The particulars can be very interesting and surprising, though."
Exactly, and as I've shown, the particulars of co-option — necessary templates and forms pre-determined by physical laws working in conjunction with non-lawful instructions — provide evidence that co-option is indeed a result of front loading.
Someone once said that a little bit of philosophy leads to atheism, but a deeper comprehension of philosophy leads one away from atheism. It is also true that a cursory examination of an investigation which scratches the surface of biology may lead one to the conclusion that foresight, front loading, and telos is un-necessary and merely an invocation of agents beyond necessity. However, upon a deeper investigation into cause, law, and the foundation of biology one discovers himself being drawn away from mere chance and necessity (randomness, law, and environmental selection) into a world rich in necessary intelligence and front loading.
Again, Bradford has already provided the key:
That is indeed the key. The pre-determined template and kit of instructions was just waiting for natural selection to catch up and allow (not cause) that template and instructions to unfold, similar to how the template of an intelligent human being unfolds from a fertilized egg because the environment at the time allowed for it to happen.
Comment by CJYman — March 3, 2008 @ 4:07 pm
March 3rd, 2008 at 4:12 pm
Hey Bradford, as an aside have you ever thought of writing a book on the consequences of templates for form and function caused by physical laws and the necessary non-lawful instructions to coax out this template?
Ever thought about hiring on a co-author?
Comment by CJYman — March 3, 2008 @ 4:12 pm
March 3rd, 2008 at 4:15 pm
… and convergent evolution is extremely relevant to this topic, providing evidence against Gould's assertion that the tape of life doesn't replay itself. Again … front loaded forms and functions being discovered time and again by evolution.
Comment by CJYman — March 3, 2008 @ 4:15 pm
March 3rd, 2008 at 5:56 pm
In vats of artificial chemicals, we find bacteria that can digest nylon, a material never before seen on Earth.
We have a hypothesis that suggests that new traits are due to variations in genes that are then selected for by the environment. From that hypothesis, we expect that the new trait should be due to naturally occurring variations. Investigation determines that a genetic changed occured (more than once), on the order expected from random processes.
We have another conjecture that this was "front-loaded". The specific and distinguishing prediction being … Can you provide a valid test of your conjecture, or is it just puddle thinking?
"”
This is rather as if you imagine a puddle waking up one morning and thinking, 'This is an interesting world I find myself in - an interesting hole I find myself in - fits me rather neatly, doesn't it? In fact it fits me staggeringly well, must have been made to have me in it!' "” Douglas Adams
The puddle can measure the hole to a vast degree of accuracy and verify that the hole fits staggeringly well. But this doesn't distinguish the conjecture that the hole was designed from a fit due to some sort of unknown symmetry.
Quite so. Convergence indicates that at least some aspects of evolution are adaptive and not contingent.
Comment by Zachriel — March 3, 2008 @ 5:56 pm
March 3rd, 2008 at 7:26 pm
Hey Zachriel,
I understand what you are saying. Some aspects of evolution may be the effects of random processes once evolution is already set in motion.
These examples have been discussed in Mike Gene's book "The Design Matrix". However, even these examples of lack of foresight can not exist apart from an evolutionary process and evolutionary algorithms are inherently teleological as I've already explained. In fact, there must be some actual targets in order for evolution to proceed. However, once these targets are achieved then there can be some degradation of the target or limited or cyclical random variation from that point.
Furthermore, the evolutionary process involved with the immune system and even antibiotic resistance is not truly random, since the toolkits and pathways necessary for self defense against future antigens is already front loaded into the operation of the organism. So, even though the immune system or bacteria is combating a never before seen enemy and its solution may not be a result of foresight of the agent that caused the system, the very defense mechanism that causes that solution is front loaded into the organism long before it came against the novel threat and the necessary solution. Will the defense mechanism come about without being a pre-programmed target?
There are other examples of forms and functions that may have been necessarily pre-programmed into life. This is where convergent evolution comes into play.
So, we need to be able to separate targets from non-targets while recognizing that even the non-targets could not have come about absent a teleologically caused system. This is excellent further ID research.
Zachriel:
"We have another conjecture that this was "front-loaded". The specific and distinguishing prediction being "¦ Can you provide a valid test of your conjecture, or is it just puddle thinking?
"”
This is rather as if you imagine a puddle waking up one morning and thinking, 'This is an interesting world I find myself in - an interesting hole I find myself in - fits me rather neatly, doesn't it? In fact it fits me staggeringly well, must have been made to have me in it!' "” Douglas Adams
The puddle can measure the hole to a vast degree of accuracy and verify that the hole fits staggeringly well. But this doesn't distinguish the conjecture that the hole was designed from a fit due to some sort of unknown symmetry."
For starters, I propose that convergent evolution, defensive systems which non-randomly guide a system to defend against a never before seen enemy (akin to the operation of the immune system) based on a pre-existing toolkit, and IC systems (remove an element and it doesn't work) are not puddles fitting into a random hole.
1. Convergent evolution shows that separate examples of evolution are guided toward the same end functions and forms (providing a more law like order based on pre-existing templates).
2. The evolutionary operation of defensive systems is inherently non-random no matter the foreign "enemy" or "environment."
3. IC systems require a base of complex instructions to develop them in a specific order. No matter the evolutionary pathway, will the non-random pathway necessary to construct an IC system evolve without any previous information?
The distinguishing prediction being that an information processing system that isn't guided to the aforementioned systems will not produce those effects. This and other relationships between EAs and targets and developed non-targets can be examined while experimenting with evolutionary algorithms.
Of course, the Design Matrix is extremely useful here as well.
Now, all that *I* am asking is for a distinguishing prediction that would set a non-teleological evolutionary process (from a holistic view) apart from the simulated and observed teleological evolutionary process. Furthermore, do you even have evidence that an evolutionary process can occur absent intelligent programming?
Comment by CJYman — March 3, 2008 @ 7:26 pm
March 3rd, 2008 at 9:25 pm
Which is adequately explained by natural selection sorting variations of existing traits. Something we can observe.
So? No one thinks evolution is necessarily random.
Another term that has had several incarnations, Irreducible Complexity.
There are a variety of evolutionary mechanisms that can result in an Irreducibly Complex system, including cooption and optimization.
One of these days you'll figure out that that is not a valid scientific prediction. I need to know where to observe, when to observe, how to observe, and for how long. Your 'prediction' requires that I observe every possible information processing system (including those never yet dreamt of), under every possible configuration, from every possible angle, with every possible analytical tool, for all time.
Comment by Zachriel — March 3, 2008 @ 9:25 pm
March 3rd, 2008 at 11:15 pm
I was with you completely right up until the "it seems obvious that some telic explanation must be the answer" part. This is not at all obvious. The points you lay out before that seem correct and well spoken, but you then jump to a conclusion that simply doesn't follow. Telic explanations have always been the default explanations of the unknown since the dawn of civilization, and yet to date not a single one of those telic explanations has been proven correct. I'm not sure why it would be obvious that the origin of life is the one exception to this trend. I don't think science has much of a clue as to how life originated, but that ignorance doesn't make a telic explanation the obvious alternative.
Keep in mind evolution is not abiogenesis. Co-option is an evolutionary process; it doesn't necessarily play any role in abiogenesis. Maybe the concept applies in some form, maybe it doesn't. Science knows a great deal about evolution even though there is still a great deal more yet to learn. Yet your various posts here seem focused on the problem of the "boot-strap" process that would generate the first life which can then be subject to evolutionary mechanisms such as co-option. You seem to be effectively saying, "evolution can't explain abiogenesis therefore life must have first been designed." I don't think anyone is claiming that Darwinian evolution is responsible for abiogenesis. I don't think science has much to offer beyond some speculative theories to explain abiogenesis. But ignorance does not justify a telic conclusion.
Comment by Todd Berkebile — March 3, 2008 @ 11:15 pm
March 3rd, 2008 at 11:29 pm
Todd:
Could you give some examples of what would justify a telic conclusion with respect to evolution? Are you allowing for the possibility or is it ruled out a priori due its philosophical nature i.e., it's scientifically superfluous. I get the "its not fundamentally scientific" but not the "we've already showed it [evolution] is not telic."
So, assuming such a question is scientific - what would you look for or expect to see were evolution an inherently telic process?
Comment by Rob R. — March 3, 2008 @ 11:29 pm
March 4th, 2008 at 12:38 am
I have previously stated that I accept that teleology is a possibility but so far I have seen no observation where it would be a requirement. I suspect it might be beyond the ability of science to recognize teleology, you could always claim that a sufficiently intelligent designer intentionally covered his tracks to avoid detection when he, say, front-loaded life or whatever. You could claim that an all-powerful deity crafts life by controlling the quantum, in effect such a being could "be" the random in random mutation, science can never disprove that. An all-powerful deity might have just created the universe two nanoseconds ago and we couldn't tell the difference. Obviously if a designer wanted to be discovered then he would have left epistemological evidence where we'd find it. So how might you prove a laissez-faire designer operating with human-like intelligence? I don't think you can. I haven't seen an approach to this question yet that's better than the one Mike offers in the DM. Even though I find some basic parts of Mike's method to be flawed its the best method I've seen out there, that's why I read his blog.
Well, first I think any answer to this question is little more than pointless speculation, but I'll play along. Evolution by definition is not a telic process, so I'll substitute Front-Loaded Evolution so that this makes some amount of sense. Now on with the wild speculation! First, I would expect to see a larger toolkit. If I was designing life with the knowledge that it would evolve to fit future environments I cannot predict then I would start out with multiple vastly different templates in the hopes that some of them might "stick." Heck, I might even try a few prototypes that use Silicon instead of Carbon just to see what happens. Second, I would prototype many varieties of genetic information storage instead of just designing one, DNA, and hoping that one works out bug free for four billion years. Third, I would design thousands of amino-acid proteins because you never know which ones might work out the best four million years later. Fourth, if I wanted, say, multi-celled organisms I'd go ahead and make them from the start rather than making just their various parts but scattering those parts among single-celled organisms in the hope that they just happen to show up millions of years later in a multi-celled life forms. I could go on, but its bed time.
Comment by Todd Berkebile — March 4, 2008 @ 12:38 am
March 4th, 2008 at 1:18 am
That's a strange reaction given the fact that nature seems to have hidden an alternative non-telic pathway as well. What one should do is recognize that scientific evidence flows from experiments which in the case of life's origins are inherently difficult. That is not to say though that evidence appears hidden. Teleologists can point to information rich molecules and parallels to computers as their opponents are able to point to organic chemicals and theorized pathways. The inability to make a decisive case either way in no way implies hidden supportive data.
Comment by Bradford — March 4, 2008 @ 1:18 am
March 4th, 2008 at 1:49 am
Zachriel:
"One of these days you'll figure out that that is not a valid scientific prediction. I need to know where to observe, when to observe, how to observe, and for how long. Your 'prediction' requires that I observe every possible information processing system (including those never yet dreamt of), under every possible configuration, from every possible angle, with every possible analytical tool, for all time."
Incorrect, as it is based off of understanding of information theory (NLF, COI, and CSI especially). You don't need to look at every single case of what could happen to *understand* that perpetual motion free energy machines do not occur.
Furthermore, ID Theory is based off of observation and experimentation of the type of system in question: information processing systems and EAs.
To state that certain informational systems are the necessary result of intelligence is to merely state a hypothesis of cause and effect. It just so happens that this one is consistent with observation and understanding of the conservation of information.
You know Zachriel, you are seriously excellent at blowing smoke. Obfuscation after obfuscation. I can see why some people get so frustrated with you.
But, oh well …
The rest of your comment has no relevance to what I am saying as I, in no way, deny that evolution has occurred.
For the "umpteenth" time:
"Now, all that *I* am asking is for a distinguishing prediction that would set a non-teleological evolutionary process (from a holistic view) apart from the simulated and observed teleological evolutionary process. Furthermore, do you even have evidence that an evolutionary process can occur absent intelligent programming?"
Comment by CJYman — March 4, 2008 @ 1:49 am
March 4th, 2008 at 2:05 am
Todd:
"Telic explanations have always been the default explanations of the unknown since the dawn of civilization, and yet to date not a single one of those telic explanations has been proven correct. I'm not sure why it would be obvious that the origin of life is the one exception to this trend."
This is the only issue since the dawn of civilization where man has had to face the fact of where information and intelligence comes from. This is also the first time in the history of civilization where man has discovered that the foundation of life is based on the processing of an instruction tape that is not defined by physical or chemical laws. This is also the first time that man has discovered the necessary correlation between intelligence and information.
This is the first time in history where we are dealing with a system that is not fundamentally caused by laws of physics and chemistry. In fact, this information that we are dealing with is not caused by laws at all. Yet, as I have already explained elsewhere it is the foundation of natural intelligence and intelligence routinely produces it.
So we have our choice:
"As I have stated above "we either have an accidental pre-determined template of forms and their potential matches (co-opted or not) based on fine tuned physical laws awaiting a fortuitous informational accident to discover them, or we look at options other than the fortuitous accident."
How about we at least start with observation and what we have gleaned from mathematical information theorems and with what we have discovered while working with information processing systems and evolutionary algorithms.
Todd:
"I don't think science has much of a clue as to how life originated, but that ignorance doesn't make a telic explanation the obvious alternative."
But they do have a clue as to what is necessary in order to cause an information processing system (not defined by physical or chemical laws) to generate. Furthermore, they also already know that evolutionary algorithms are caused by previous information guiding the program to an end target and that this process never creates its own information. If this isn't teleological I'm not sure what is.
If you could point me to some evidence that a system not programmed by intelligence (ie: random set of laws and variables) will produce an information processing system and an evolutionary algorithm that would be great.
What other option is there?
Comment by CJYman — March 4, 2008 @ 2:05 am
March 4th, 2008 at 8:06 am
I asked for a specific and distinguishing prediction. This is what you provided.
Your prediction requires that we look at *all* conceivable processing systems, forever.
I didn't ask about your *understanding*, but for valid scientific predictions. The valid scientific predictions concerning the Laws of Thermodynamics are that we can examine *any* system. 1) In an isolated system not in equilibrium, the amount of energy to do work will tend to decrease over time. 2) It is impossible to extract heat from a hot reservoir and use it all to do work. 3) It requires work for heat to flow from a colder body to a warmer body. I can verify these predictions. Yours I can't.
We can directly observe evolution, so it is no longer an open question. The evolution of nylonase is a case in point.
We can directly observe evolution, and we can point to very specific mechanisms that can explain the data. All you have left is the assumption of your conclusion, petitio principii.
Comment by Zachriel — March 4, 2008 @ 8:06 am
March 4th, 2008 at 4:56 pm
I think evolution suggests that: Noise + selection -> information. Are you arguing that any selection process is inherently intelligent? If yes then that's a very different definition of intelligence than I'm used to but one that interests me. I've been pondering a lot recently about whether a simple iterative process devoid of consciousness or foresight can be considered "intelligent." Evolution has amazing problem solving skills and such skills are typically thought of as a sign of intelligence in humans. Evolution does all its problem solving though trial-and-error, but if the memory-prediction model of human intelligence is correct than all human problem solving basically amounts to trial and error too. We're just able to perform our trials abstractly in our minds.
This is completely false. We have witnessed "new information" in the form of unique DNA sequences performing new functions forming without intelligent intervention. That is shown in the nylonase example that Zachriel sited. You seem to be claiming that because evolution cannot explain abiogenesis (again, which no one is claiming it can) that means evolution cannot create new information. But we have directly observed the creation of new information by an evolutionary process. You seem to be claiming that the origin of life inherently requires intelligence and thus anything that happens afterwards can also be considered intelligent. If so, that is simply wrong. To use a telic example, it takes intelligence for a human to write a computer program but the computer doesn't have to be intelligent to run that program.
Comment by Todd Berkebile — March 4, 2008 @ 4:56 pm
March 4th, 2008 at 5:12 pm
Todd:
Indeed. It could nevertheless be argued that in a sense there is a principle of conservation of information operating. There is information in the environment of a population of organisms, and there is information in the population's gene pool. One might argue that natural selection transfers some of the information in the environment (the part that is relevant for fitness) into the gene pool. Without changing the total amount of information present in environment and population. OK, this is just an approximation, but maybe it's helpful to think along these lines when talking about information.
Comment by Raevmo — March 4, 2008 @ 5:12 pm
March 4th, 2008 at 6:27 pm
Ah, I see, that makes more sense. Still, that model of COI seems to divorce information completely from intelligence. What is the connection between information and intelligence? Why would that kind of "information" require an intelligent origin?
Comment by Todd Berkebile — March 4, 2008 @ 6:27 pm
March 4th, 2008 at 6:45 pm
Todd:
It doesn't. The environment transfers info into genomes, and in turn genomes become part of the environment of other genomes and they - in a sense - transfer information between them. My point is, the argument against evolution being capable of building up info into genomes without "intelligent" input is demonstrably BS.
Comment by Raevmo — March 4, 2008 @ 6:45 pm
March 4th, 2008 at 7:16 pm
Yes, I've taken a click over to your blog. I've been meaning to engage that conversation.
Yes, those are common signs of intelligence. Rather slow learners, though. Literally takes generations.
Trial-and-error can be very powerful when running numerous diverse experiments and when combined with a type of averaging over long experience.
That's a very significant difference. Minds can build from the foundation up, so to speak. Biological evolution is tightly constrained by its history.
Comment by Zachriel — March 4, 2008 @ 7:16 pm
March 4th, 2008 at 7:43 pm
I vote: foresight necessary, consciousness not necessary.
Comment by kornbelt888 — March 4, 2008 @ 7:43 pm
March 4th, 2008 at 7:50 pm
Of course not. The program nevertheless is a product of intelligence.
Comment by Bradford — March 4, 2008 @ 7:50 pm
March 4th, 2008 at 7:58 pm
True.
Depends on what you mean by "explain" and what you assume in addition to the available empirical evidence.
If you extrapolate from empirical microevolutionary processes the ability to achieve all the known macro level features, regardless of an actual empirical demonstration, well then, fine. Some of us aren't so impressed.
Time and again I've asked you for a demonstration of any empirical process shown to generate novel cell types, tissue types, organs or body plans. Your answer always amounts to something like the rates of variations are sufficient enough to explain all the macro level changes. However, implicit in that response is that the empirical microevolutionary variations are the right kind of changes required by novel cell types, tissue types, organs and body plans. You've never demonstrated that and nobody else has. Whenever I ask for a demonstration that they are the right kind of changes, no answer is forthcoming.
Assuming what you're trying to prove is called petitio principii. "Circular reasoning" or "begging the question" to us common folk.
Comment by kornbelt888 — March 4, 2008 @ 7:58 pm
March 4th, 2008 at 8:04 pm
One could say the program was endowed with intelligence from the creator, with insight and goal-seeking to further the creator's purposes. To call such systems and processes "intelligent" is common in engineering circles.
Comment by kornbelt888 — March 4, 2008 @ 8:04 pm
March 4th, 2008 at 8:42 pm
So you need to directly witness differentiation of two species but you do not need to directly witness the designer in action? That's quite the double standard.
Time and again you've asked for one specific example that you know doesn't exist yet. Time and again people have given examples of significant changes that are well documented but you want to see a massive change with your own eyes and do not believe extrapolation of the evidence is valid. And yet time and again when people ask for epistemological evidence of a creator you claim that sort of evidence is not required?
Yes, exactly the distinction I'm interested in. I want to know if CJYman thinks evolution is intelligent for the same reason some people would call that running computer program intelligent. This is a different definition of intelligent than I had been assuming so its important to know his definition in order to understand his claims. It seems to me that he attributes all the "real" intelligence to the origin of life and then carries over the term to apply it to evolution in the same way as calling the running program intelligent.
Comment by Todd Berkebile — March 4, 2008 @ 8:42 pm
March 4th, 2008 at 10:00 pm
Predictive.
Most cell types are deeply rooted in the phylogenetic tree. I would suggest starting with more recently diverged taxa where the evidence is strongest. But if you want to delve right in, you could try looking for papers on metazoan axial development or the evolution of vascularization. Another interesting area is the biphasic life cycle. Better yet, you might take a look at embryonic development of e.g. nematodes and the nested hierarchy of cell differentiation. I'm not sure the level of detail you are interested in, but Schmidt-Rhaesa provides an overview.
These cites aren't meant to be complete, but just to provide you some general direction for a more complete search. However, understanding most of them require an understanding of evolutionary fundamentals, such as cladistics.
Comment by Zachriel — March 4, 2008 @ 10:00 pm
March 4th, 2008 at 10:22 pm
You're the one making positive claims about evolution here. It's encumbent upon you do the research and post it here for us to see. Boil it down, write a report, and present it here (or somewhere.) It's your big chance to shut the mouth of a few skeptics.
Otherwise all you have is a vacuous claim.
Comment by kornbelt888 — March 4, 2008 @ 10:22 pm
March 4th, 2008 at 10:55 pm
Hahaha, because kornbelt888 is not willing to look at the evidence that makes Zachriel's claim vacuous! Wow, that's rich. So it's his responsibility not only to lead the horse to water but to then forcibly pump a few gallons into his stomach. And to be fair Zachriel has mentioned the Reader's Digest version of a lot of this information but you dismiss that information too. If all you want is to believe in something based on faith or to confirm what you already believe then there is no reason even participate here. Personally I come here specifically to read evidence that counters my current position. I can only accept the things I believe by trying to disprove them and failing.
Comment by Todd Berkebile — March 4, 2008 @ 10:55 pm
March 4th, 2008 at 11:04 pm
If we stick to the rules of methodological materialistic science, I am agnostic on both questions.
You can either demonstrate it or you cannot. Thanks for the admission. But Zechriel apparently feels it can be demonstrated. I challenge him to try.
What do you have in mind?
I want empirical evidence that the kind of evolution that we empirically see is the right kind that can lead to novel cell types, tissue types, organs and body plans, yes. Asserting that the empirical evidence we have scales to that level is not merely an extrapolation, it is an extrapolation based partly on an added assumption.
A valid extrapolation goes something like this: if I interview 1000 people at random using standard polling methods and 1/3 of them have blue eyes, I should expect the number of blue-eyed people to be around 3333 if I interview 10000 people. If I were to take that original poll and extrapolate that 1/3 of the people also like vanilla over chocolate, I've made an invalid extrapolation.
We have empirical knowledge what certain microevolutionary changes can accomplish. It is invalid to extrapolate these microevolutionary changes to novel cell types, tissue types, organs and body plans- we do not know what kind of variation is required to get the job done. There's a logical disconnect. I would never ride in an airplane designed by someone who didn't understand that.
Zechriel previously tried to use a bogus analogy that if a car can take you across the street it can likewise take you across the country, even though you may not know in advance the exact route or terrain. Of course, this is ridiculous. His analogy is rigged to be successful, because he assumes there are no oceans or other such impassible barriers. For example, unless you have a reasonable map of the terrain, you may encounter an ocean before you reach your destination milage. Cars can't travel across oceans. They're not the right kind of mechanism to get you across.
I posted an analogy myself using the Rubics Cube, which fell on deaf ears, but here goes again. If you take a standard Rubics Cube and randomly shuffle it, you can be certain there is a path to the normalized state of uniform colors on all sides. But if I secretly pull off the colored stickers and put them back on in a way that makes it impossible to get to the normalized state, then, by golly, it's impossible. A casual player might not realize why he can never get to the normalized state, tee hee. But the reason is that there is no logical path.
The question is: is it possible for the kinds of mutations we have empirical evidence of the right kind of mutations that can produce novel cell types, tissue types, organs or body plans? Extrapolation won't get you there unless you're willing to make a logical leap.
I claim no such thing scientifically. (And we are talking science here, I assume.)
I can only speak for myself. I consider the cellular machinery to be strong intuitive evidence for a system designed to evolve. However, scientifically, I am agnostic.
Comment by kornbelt888 — March 4, 2008 @ 11:04 pm
March 4th, 2008 at 11:07 pm
I'm not sure which specific claim you are referring to as the quote you included in your comment doesn't include any claims. But I assume you are questioning the Theory of Evolution, especially as regards to the origins of "novel cell types, tissue types, organs or body plans".
Oddly, you had asked for empirical evidence, but now you don't seem interested. Instead, you want a "report", by which I guess you mean a "boiled down" explanation.
That's fine. We start with the Theory of Common Descent. The vast majority of taxa form a singular nested hierarchy of morphology, individual traits and genomes. From this we can make a variety of predictions, including the placement of fossils, the plausible configurations of traits and genomes. We know with reasonable certainty that life has a common origin, and that for the vast majority of taxa, any two organisms share a common ancestor.
(There is some ambiguity at the root of the tree of life. The evidence is scant, but a consensus is growing that the origins of cellular life primarily involved horizontal processes.)
Do you feel comfortable with this aspect of the evidence, or do we need to layout this important foundation?
Comment by Zachriel — March 4, 2008 @ 11:07 pm
March 4th, 2008 at 11:13 pm
Citing the names of works of proffered authorities without citing relevant text from those authorities is bad form, and is not an argument. If Zechriel has a case to make, let him make it.
I have a different analogy in mind, but nevermind.
For example?
Comment by kornbelt888 — March 4, 2008 @ 11:13 pm
March 4th, 2008 at 11:18 pm
Yes, the a-telic Modern Synthesis to be more precise.
To clarify my position, there is much about the Modern Synthesis that is compatable with a telic position. My position is that the MS is not yet sufficient to explain the origin of novel cell types, tissue types, organs and body plans.
You're the one making a positive claim here. Why would you expect anyone to accept it without you actually laying it out? If the books you cite actually make a solid case, you should be able to summarize the logical steps. If that looks promising, I, and others no doubt, would probably be interested in obtaining copies of those works and digging in to the excruciating details outselves. But so far, I've seen nothing here that even remotely indicates that.
So far, so good.
Could be. Much less convinced about this. We'll have to see.
Yes.
Comment by kornbelt888 — March 4, 2008 @ 11:18 pm
March 5th, 2008 at 2:08 am
Hello Todd, it is great talking to someone who is actually willing to engage in discussion without blowing smoke for a change.
Todd:
"I think evolution suggests that: Noise + selection -> information.
First, what is necessary in order to have this selection:
1. Information processing system.
2. Instructional tape which is not defined by laws so that is has freedom to carry information.
3. Template of forms and functions pre-set by laws of physics for selection to choose from.
4. Pre-set Linked pathways in the search space to get from one form and function to another.
5. According to NFL and COI, previous information is needed in order for these forms and functions to be discovered at better than chance performance.
In fact, evolutionary algorithms are pretty much an example of artificial intelligence and vice versa.
Todd:
"Are you arguing that any selection process is inherently intelligent? If yes then that's a very different definition of intelligence than I'm used to but one that interests me. I've been pondering a lot recently about whether a simple iterative process devoid of consciousness or foresight can be considered "intelligent.""
Great minds think alike!! Not "every selection process." But, the type of selection process that results in better than chance performance and the unfolding of information. These types of selection processes are basically artificial intelligence. In fact, I have come to the conclusion over the past year that life is indeed an example of an artificial intelligence network.
A "learner… that achieves at least mildly than better-than-chance performance, on average, … is like a perpetual motion machine - conservation of generalization performance precludes it."
–Cullen Schaffer on the Law of Conservation of Generalization Performance. Cullen Schaffer, "A conservation law for generalization performance," in Proc. Eleventh International Conference
on Machine Learning, H. Willian and W. Cohen. San Francisco: Morgan Kaufmann, 1994, pp.295-265.
Basically, you get out from EAs and AI what you put in. Its all about the programming for future targets, optimization, and solutions..
CJYman: Furthermore, they also already know that evolutionary algorithms are caused by previous information guiding the program to an end target and that this process never creates its own information.
Todd:
"This is completely false. We have witnessed "new information" in the form of unique DNA sequences performing new functions forming without intelligent intervention. That is shown in the nylonase example that Zachriel sited. "
Who said anything about "intelligent intervention." I am a methodological naturalist (although not a philosophical naturalist or a materialist).
And, actually, your example of "new" information is incorrect. And this is where the majority of the confusion lies. That is not an example of the creation of new information. "New" information is an impossibility. You only get out as much as you put in. You must understand that information is a measure of probability and more specifically a decrease in uncertainty. You will not get a learner to consistently learn (discover solutions) at a rate of probability higher than the probability of the information that you program into the learner. Evolution and artificial intelligence must follow this rule (in fact, it is connected to the foundational rule of all of science "“ thermodynamics).
Your comment would be like saying that your artificially intelligent servant robot creates new information whenever he discovers the solution to something. But, that is only an example of the discovery of information that is only possible because of previously inputted information.
I have discussed this on my blog under "NFL Theorems …" and "Philosophical Foundations of ID Theory" linked to at the top of the left hand side bar of my blog under "My Case for a Naturalistic Hypothesis of Intelligent Design." My blog is linked to my handle.
"… no operation performed by a computer can create new information."
– Douglas G. Robertson, "Algorithmic Information Theory, Free Will and the Turing Test," Complexity, Vol.3, #3 Jan/Feb 1999, pp. 25-34.
"The [computing] machine does not create any new information, but it performs a very valuable transformation of known information."
–Leon Brillouin, Science and Information Theory (Academic Press, New York, 1956).
Evolutionary algorithms and artificial intelligence will only solve problems that they are programmed to solve. Information in, information out. It is the conservation of information.
Todd:
"You seem to be claiming that the origin of life inherently requires intelligence and thus anything that happens afterwards can also be considered intelligent. If so, that is simply wrong. To use a telic example, it takes intelligence for a human to write a computer program but the computer doesn't have to be intelligent to run that program."
Incorrect. I am claiming that certain effects, which are necessarily targets, are also the result of previous intelligent front-loading. We need to discover which solutions are indeed targets, since evolution only proceeds if characteristics of specific problems are programmed into the nature of the search procedure.
In an evolutionary scenario, according to the No Free Lunch Theorem, for a search algorithm to perform efficiently (find targets at better than chance) which is exactly what an evolutionary algorithm does, problem specific information must be front loaded into the program. Problem specific information = characteristics of the search space and targets programmed into the algorithm, thus guiding the search to pre-determined targets. This is where my discussion re: pre-determined templates of form and function and the instruction tape which coaxes them out from the laws of physics comes into play.
Once it hits these targets, it can then randomly search the surrounding space for anything that it can find. These may result in minor effects and cyclical variations.
Evolution is intelligent (as in artificially intelligent) however it is the result of a program guided by intelligently (conscious intelligence) inputted information towards pre-determined targets.
To sum up: Artificial intelligence (the simulation of foresight) must come from intelligence (true foresight). Better than chance performance out means problem specific information in.
Comment by CJYman — March 5, 2008 @ 2:08 am
March 5th, 2008 at 4:11 am
Nice story in this morning's NYT:
http://judson.blogs.nytimes.co...
The money quote:
Clearly the evolution of flying birds was anticipated, but first some excess DNA had to be removed. I would like to suggest a new technical term for this kind of removal: front-unloading.
Comment by Raevmo — March 5, 2008 @ 4:11 am
March 5th, 2008 at 9:33 am
Raevmo, the smaller genomes are linked to traits birds have like high metabolism. In addition the author stresses that genome size is not to be confused with complexity.
Comment by Bradford — March 5, 2008 @ 9:33 am
March 5th, 2008 at 11:39 am
Interesting. I've never heard of the "A-telic Modern Synthesis Theory".
Your concerns seem primarily within the urmetazoan lineage, which is monophyletic. So that shouldn't be a problem.
Good. We need to establish a few other basics.
FIRST-ORDER APPROXIMATION
Do you understand the concept of a first-order approximation? A first-order approximation of the Earth's shape is a sphere. A closer look reveals that it is an oblate spheroid (as predicted by Newton). And of course, there is a very minor roughness on its surface called "mountains".
The relevance of a first-order approximation is Darwin's claim that evolution proceeded as a slow trend over time (though he did know that the rate may have varied). A closer look reveals that the process has long periods of stasis, many small changes often about a mean, a few large changes, and very rare revolutions. This comports with what we know today about the evolution of complex systems, scale-free networks, and how organisms radiate into new niches.
And if we look very, very closely, we might notice some discontinuities at the level of species which may require additional explanation. Just as a smooth table is actually granular from a microscopic vantage.
The relevance of understand the concept of a first-order approximation is to avoid the trap of insisting upon inappropriate universals.
VARIATION & NATURAL SELECTION
All populations have natural variation that typically exhibit a normal curve. If we select for a trait, the normal curve tightens and moves towards the desired trait. But novel variations occur over time. Horses evolved from much smaller organisms, for instance. Do you have a problem with the evolution of the horse by natural selection? If so, we need to establish this before proceeding.
Today, we know that many such variations are due to genetic mutations. Because of the modular nature of genomes, we know that some genetic mutations can cause very significant changes (especially those that regulate development). But most mutations have little if any effect. A few have significant impact, and successful revolutionary changes are very rare. Again, consistent with modern models of network evolution.
We can measure the rate of evolution, such as in darwins. The rate of morphological change can be observed to be much, much faster than that required to evolve a mouse to an elephant in just a few million years. Do you have a problem with measure rates of evolutionary change? If so, we need to establish this before proceeding.
"”
I understand you might think these may be tangents, but it is very difficult to argue concerning the most ancient transitions where the evidence can be very scant if you reject the evidence for evolution where the evidence is strongest.
Comment by Zachriel — March 5, 2008 @ 11:39 am
March 5th, 2008 at 12:24 pm
CJYman,
OK, I think I'm starting to understand your viewpoint but I'm still fuzzy on some parts. For example, humans are intelligent but within that framework it seems that humans can't create new information either. We can only take the baked in functional pathways of our brain and discover information? So total universal information is constant and intelligence plays no role in creating information? Information would simply be a fundamental property of the universe just like energy. I don't see where that would imply that initial intelligence is required, in fact it almost seems like under this view there can only be "artificial intelligence" and no real intelligence.
I'm also not sure that this basic concept of COI makes any sense. If two people learn the same fact then that single piece of information has been duplicated. That both people can now recall this information means that the information has been stored into their brains, but the original information also still exists. What information is lost in this process in order to maintain COI? It seems to me that "memory" is specifically a duplication of information.
There is also the axiom that evolution has done a "better than chance" job. Can this be demonstrated? If it is shown that evolution only performs equally well with chance or perhaps even worse than chance that would disprove the theory. Naturalists already claim that the rate of evolution is consistent with mutation rates, this suggests that enough coin tosses have been performed to explain the results.
Comment by Todd Berkebile — March 5, 2008 @ 12:24 pm
March 5th, 2008 at 12:24 pm
Yes.
Horses apparently evolved by variation and natural selection.
No.
Comment by kornbelt888 — March 5, 2008 @ 12:24 pm
March 5th, 2008 at 12:55 pm
Todd Berkebile: it seems that humans can't create new information either. We can only take the baked in functional pathways of our brain and discover information?
That seems to me to be an odd way of describing what engineers do when they design, say, an aircraft. Sure, preexisting laws of nature are discovered, but do you consider the specification for an airplane to be the result of known natural laws?
I guess is depends how you view human thought, free will vs determinism. Either way there are interesting questions to be answered.
If this is true, and the universe had a beginning, what was the source of this highly specific implicit information content?
Like I said to Zechriel, it depends on what you're willing to accept as an explanation. According to Allen MacNeill, there are close to 50 known sources of variation that can be sel