Proof of god via DNA and evolution (merged)

[wdog]

I have been emailing back and forth with Perry Marshall, the author of this site

http://www.cosmicfingerprints.com/ifyoucanreadthis.htm

and since it quickly expanded in scope I invited him to come here and present his 'evidence' and proof. You might find the site amusing anyway. Feel free to critique his statements as i am sure he may at least read this since i will make him aware of this thread.

please be polite. thanks




Moderator's Note:

Welcome to this very long-running thread. If you are new here, and would like to see some highlights (rather than trudging through all 50+ pages (as of this writing)), you might want to begin with these posts, suggested by the participants:

Post 78 by theyeti, a good reply to Perry back when he was actually participating here.
Post 1273 by Zenobias, a good summary of many posts.
Post 1278 by boba123, Bob Allen, on his challenge to Perry Marshall.
Posts 1296 and 1297 by boba123, summaries of Marshall's shifting syllogisms.
Post 1301 by Chemist99a with more challenge to Perry.
Post 1314 by marcdraco on his spoof site, cosmicfingerpuppets.com.

One might also find good information in the Wiki post on Syllogism

If anyone finds more posts they'd like to nominate for this little index, please PM them to an EC Mod.
Barefoot Bree
EC Mod

[RBH]

The site claims

Natural Selection is perfectly valid and has been proven time and time again. But most people will be very surprised to discover that no one has ever actually demonstrated that random mutation can create new information. Information theory shows us why this is so: In communication systems, Random Mutation is exactly the same as noise, and noise always destroys the signal, never enhances it. (Bolding original)

Mr. Marshall suffers from the common creationist misapprehension that evolution is fundamentally a random process. It isn't. It is undirected, in the sense of not having long-term goals, but it is not random. The process of evolution proceeds through the interaction of random (with respect to fitness) mutations and natural selection. The latter is quite definitely a non-random process. Once again, evolution arises from the interaction of stochastic and non-stochastic processes, and is therefore not purely random.

Further, Marshall's claim that "noise always destroys the signal" is factually false, and he should know it given his field of expertise. There are circumstances in which adding stochastic noise to a system enhances the system's performance. See here, here, and here for examples. The latter is in a communication framework, by the way. More generally, I suggest Marshall look up "stochastic resonance" for more examples.

I suggest that Mr. Marshall acquaint himself with Evolutionary Computing. In any number of applications, random mutations coupled with selection generate distinctly non-random outcomes. My company uses evolutionary algorithms employing random mutations (a Mersenne twister generates them) and selection to produce very non-random adaptations of artificial agents that must operate successfully in a complex adaptive system out in the real world. If they didn't operate successfully I'd have some very unhappy clients.

RBH

[anthrosciguy]

I used his "random mutation generator" (which simply switches some symbols around in text you write, and inside of a couple hundred (ie. hardly any at all) "mutations" got 4 different words... which of course I could've "selected". So even according to his "random mutation generator" there's just loads of mutations that could be selected and be useful.

I'm certain there's plenty other objections out there. (One I have is sound on web page; yuk, why do people foist that on you? At least it should be opt-in.) That and "Big Bang equals Jesus Christ" (actual quote from one of the presentations there) is just a bit specific, godwise.

[RBH]

Looking further into a search I performed for my previous post, I found this:

Stochastic resonance is a phenomenon whereby the transmission of a signal by certain nonlinear systems can be improved by addition of noise. We propose a brief overview of this effect, together with an extension based on information theoretic concepts. We analyze various conditions of nonlinear transmission where the input–output Shannon mutual information, the input–output Kullback divergence, or the input–output Fisher information can receive improvement from noise addition, demonstrating different forms of noise enhanced transmission.

Directly contradicts the "noise" claim quoted from Marshall's site above.

RBH

[theyeti]

The random mutation generator might be more useful it if had a population greater than n=1. If there was never more than one single organism on the Earth, I'd probably agree that mutation wasn't going to generate biological information. But dude makes the sophomoric error of ignoring the fact that evolution is a population level phenomenon.

That's me being as polite as I can.

theyeti

[theyeti]

This is only tangenitally related, but I got a kick out of it nonetheless:

The Atheist's Riddle - "So simple, any child can understand; So complex, no atheist can solve." An airtight inductive proof for the existence of a Superintelligence.

Someone better call in the philosophers. We've got the first person in history to construct a logical proof based on induction.

theyeti

[espritch]

In communication systems, Random Mutation is exactly the same as noise, and noise always destroys the signal, never enhances it.

However, self replicating organisms adapting to changing environmental demands is a very different thing from information being transmitted from one point to another. Attempting to apply a theory about electronic data transmission to biological systems is exactly like trying to apply quantum mechanics to planetary motion. It just isn't applicable.

[RBH]

Originally Posted by espritch

However, self replicating organisms adapting to changing environmental demands is a very different thing from information being transmitted from one point to another. Attempting to apply a theory about electronic data transmission to biological systems is exactly like trying to apply quantum mechanics to planetary motion. It just isn't applicable.

And as I showed via several references, the claim isn't even true in its original information transmission context.

RBH

[RBH]

Gack! Here's another one

“Did the Antelope evolve into the Giraffe?� According to Darwinian evolution, the necessities of the environment, random mutation and natural selection working together caused the antelope to grow a longer neck and become a giraffe. OK, then what does communication theory say about that hypothesis?

The answer to the question is "No". See the phylogeny here. Looks to me like antelope and giraffes had a common ancestor but are cousins (in different families), not members of the same lineage.

RBH

[wdog]

I hope the author registers, but in any event I think it is useful to get a critique broader than my own (not mention saving me some time). Obviously the author has only been preaching to the choir on this topic and if he really believes he has real proof than I can't understand why he would not share it on an atheist web site. After all, I have been looking for such proof ever since I decided to question my own religion and reject Christianity and this should therefore cause a christian to jump at the prospect of saving souls and proving their case.

I also also hope that the author understands that since he is the one making the lecture circuit and creating public web sites, then I feel it is ok to take his exchange with me to a public forum.

In communication systems, Random Mutation is exactly the same as noise, and noise always destroys the signal, never enhances it.

But there is no culling algorithm in communication systems (no natural selection), secondly DNA is not a communication system so the analogy is false. This gets to the heart of the authors argument (at least with me). The parallel drawn here is that DNA is a code like the code used between two intelligent and conscious entities to exchange information. Clearly on first examination this analogy is incorrect as there is no intelligent and conscious 'reciever' to take the 'message', whatever he imagines that message to be. The author assumes god is the sender, or creator, but again even granting that- clearly the purpose is not to encode and send a message, it is to create self replicating biological machines.

The fact that random mutation can in fact increase the information content (In DIRECT contradiction to the author's claim he made during his lecture) of DNA does seem to destroy the whole argument as well since then we not only have a changing code but one that clearly has evolved over time to become more complex. Self-evolving and non-conscious DNA also brings up the issue of whatever 'message' there originally was is now gone anyway. So how could there be any intent on the part of an intelligent and conscious sender? For the sake of argument, assume random mutations are destroying information. Then that also means that part of the original message is gone, so what was the point of sending it if after a while it is gone? what was the message? can't god create a non-mutating message if this is to be the proof of its existence?


In the lecture Perry Marshall also claims that the very use of the word evolution is always used in context with an intelligently guided process, except in darwinian evolution. Cleary that is not the case as i am sure many examples could be found, the evolution of a hurricane for instance.

[Sven]

Originally Posted by RBH

Mr. Marshall suffers from the common creationist misapprehension that evolution is fundamentally a random process. It isn't. It is undirected, in the sense of not having long-term goals, but it is not random. The process of evolution proceeds through the interaction of random (with respect to fitness) mutations and natural selection. The latter is quite definitely a non-random process. Once again, evolution arises from the interaction of stochastic and non-stochastic processes, and is therefore not purely random.
[snip]

I've just read a nice analogy in a physics textbook (college level, I think): A laser.
Light emitted by excited particles normally covers a relatively broad range of wavelengths (due to Heisenberg's uncertainty principle). We can call the possible wavelengths "random" (over a certain range). But as soon as we introduce mirrors on both sides of the cavity (an "environment"), only certain wavelengths lead to standing waves - and these stimulate more radiation of the same wavelength to be emitted (they "reproduce"). Other wavelengths don't reproduce they disappear quickly ("natural selection").

What do you think of this analogy?

[Sparrow]

I never fail to be amazed that individuals, with no apparent training in biology, think they can come up with a simplistic argument that would cause tens of thousands of biologists to say "I never thought of that! I never looked at it that way! Wow you're right; evolution is obviously flawed."

[Graculus]

In Shannon IT the quantity of "information" is determined by comparing the source and the recieved. Shannon information is completely self-referential within the system. Anything that is not in the system to begin with is "loss" or "noise". It's a circular argument, based on the hope that people will mistake Shannon "information" with "meaning" or "functionality". If a system generated a more highly functional/meaningful received, it would still count as a loss in Shannon IT.

This comes back to interpretation. There is nothing in Shannon IT to determine "quality" of the original, and compare it with the "qualtiy" of the recieved. Of course, in regards to organisms, there is such an interpreter... selection.

[pmarshall]

Originally Posted by wdog

I have been emailing back and forth with Perry Marshall, the author of this site

http://www.cosmicfingerprints.com/ifyoucanreadthis.htm

and since it quickly expanded in scope I invited him to come here and present his 'evidence' and proof. You might find the site amusing anyway. Feel free to critique his statements as i am sure he may at least read this since i will make him aware of this thread.

please be polite. thanks

Gentlemen:

The starting point of this discussion is my central thesis, which is:

1) DNA is not merely a molecule with a pattern; it is a code, a language, and an information storage mechanism.
2) All codes are created by a conscious mind; there is no natural process known to science that creates coded information.
3) Therefore DNA was designed by a mind.

If you can provide an empirical example of a code or language that occurs naturally, you've toppled my proof. All you need is one.

Perry Marshall

[NottyImp]

OK, then what does communication theory say about that hypothesis?

Er, who cares? I'd rather find out what evolutionary biology has to say about it.

[The Palace]

Welcome to IIDB, Mr. Marshall.

Though I am a layman in both evolutionary biology and information theory, I would like to hear your thoughts on the following:

Originally Posted by pmarshall

1) DNA is not merely a molecule with a pattern; it is a code, a language, and an information storage mechanism.

When categorization is asserted, my usual means of verification is to ask for a usable test of the categorization; therefore, please tell me how one would be able to tell a molecule with a pattern that is merely a molecule with a pattern from a molecule with a pattern that is also a code, a language, and an information storage mechanism.

Originally Posted by pmarshall

2) All codes are created by a conscious mind; there is no natural process known to science that creates coded information.

Also, I would appreciate it if you could provide the exact meaning of the word "code", as you use it, since I believe many have slightly different usages of that word. What is a "code"? How would we be able to tell if something is a code from if it's not one?

[Monad]

Interesting. This is virtually identical to the so called proof offered by dionysus932 here:

http://www.iidb.org/vbb/showthread.php?t=136014&page=1

Please scope out the following deductive argument:

1. DNA is imbued with information and is a component of information processing
2. Information and information processing are only engendered by intelligent agents
3. Therefore, DNA is the product of an intelligent agent

Is it sound? I'm pretty sure it's valid. I suppose an evolutionist would take exception to proposition #1, although it would probably be a difficult position to defend. Proposition #2 seems above reproach. Any comments?

coincidence???

[Monad]

Originally Posted by pmarshall

If you can provide an empirical example of a code or language that occurs naturally, you've toppled my proof. All you need is one.

Perry Marshall

Bee waggle dances. Bird song. Whale song. Ant communication by pheromone (very complex - yet to be unravelled). Vervet monkey proto language etc etc.

[Monad]

Not to mention human language (though a creationist would never admit it)

[Graculus]

Originally Posted by pmarshall

Gentlemen:

Hi

Welcome to the shark tank, we'll try to mind our table manners

1) DNA is not merely a molecule with a pattern; it is a code, a language, and an information storage mechanism.

Code and language are abstract, symbolic systems that are completely arbitrary. DNA cannot be a code because it is not abstract or arbitrary.

2) All codes are created by a conscious mind; there is no natural process known to science that creates coded information.

But DNA isn't a code, so...

3) Therefore DNA was designed by a mind.

.. does not folow

If you can provide an empirical example of a code or language that occurs naturally, you've toppled my proof. All you need is one.

We don't need any, because your first premise is incorrect.

[HRG]

Originally Posted by pmarshall

Gentlemen:

The starting point of this discussion is my central thesis, which is:

1) DNA is not merely a molecule with a pattern; it is a code, a language, and an information storage mechanism.

A. DNA is a code only in the metaphorical sense that a 1/r^2-force is a code for conic sections (the orbits under it are ellipses, parabolas or hyperbolas), and hydrogen is a code for oxygen to make water.

B. DNA is not a language. A language needs a grammar, which distinguishes statements from mere sequences of symbols. But any sequence of bases results in some protein.

2) All codes are created by a conscious mind; there is no natural process known to science that creates coded information.

False. A speck of radium creates coded information (the location of the emitted alpha particles "code" for the times of decay of the nuclei).

Of course, this is a code only in the metaphorical sense. But so is DNA.

[anthrosciguy]

Originally Posted by pmarshall

If you can provide an empirical example of a code or language that occurs naturally, you've toppled my proof. All you need is one.

Perry Marshall

Consider it toppled then, since you've no doubt read the dionysus932 thread that monad pointed out.

[RBH]

As is obvious, the question turns on what is meant by the term"code", and whether its use (which is frequent) with reference to DNA is simply metaphorical or is technically identical with its use in, say, human language. Since Mr. Marshall hasn't addressed that question, his assertion that

Codes, however, do not occur without a designer. Examples of symbolic codes include music, blueprints, languages like English and Chinese, computer programs, and yes, DNA. The essential distinction is the difference between a pattern and a code.

at this point is merely that -- an unsupported assertion. It is begging the question until Marshall provides a definition of "code" that circumscribes his examples but rules out stuff like the other examples given by Monad above.

RBH

[Senlatheil]

Will he return? *holds breath*

[Sparrow]

In what sense can human language be said to be designed? Would it not be more accurate to say that language emerged?