Then you have no idea what the theory of abiogenesis is.
Crick was directly quoted from one of his books.
Prove adding noise to a signal adds information to a signal.
Prove adding noise to a signal can improve the signal .
Random Mutation + Natural Selection + Time = Design
That is the math of evolution. Whether were considering the evolution of falcons or any other organism, the changes over time have to come from changes in the DNA. The entire plan for a falcon is contained in its DNA, a molecule with a 4-letter alphabet. A strand of DNA can have anywhere from 500 thousand letters (in the case of the smallest known parasite) to 3 billion letters (man and large animals).
Darwinism says that Random Mutation (copying errors in the DNA alphabet) produce modified falcons, and that natural selection (survival of the fittest as inferior falcons die out and superior falcons dominate) weeds out the losers. What remains is new innovations in falcon design.
Thats what Darwinism says. Its elegantly simple. Almost intuitively obvious.
Can it be verified?
No, and yes.
One of the difficulties with Darwinism is that it evidently takes millions of years and many billions of falcons to produce significant change over time. This makes it very hard to empirically prove Darwinism in the short lifetime of a human being. Practically speaking, Darwinism doesnt even provide us with very many testable hypotheses. Thus the vast majority of the evidence for evolution is anecdotal.
Anecdotal evidence is unreliable evidence based on personal experience that has not been empirically tested, and which is often used in an argument as if it had been scientifically or statistically proven. The person using anecdotal evidence may or may not be aware of the fact that, by doing so, they are generalizing. (From Wikipedia) An example of anecdotal evidence would be My grandma smoked and drank whiskey every day and she lived to be 95, so cigarettes and whiskey are good for you. Not all anecdotal evidence is misleading, of course, but its not proof.
But with Darwin, the principle itself should be easy enough to demonstrate. I know as an engineer with a strong math background that in principle it should be easy enough to statistically answer the question:
Is the formula
Random Mutation + Natural Selection + Time = Design
Mathematically true, or false?
Hey, if were going to get answers to these questions, we have to ask questions that actually have a possibility of being answered in the first place. Most arguments for and against evolution are argued at an intuitive level, with only anecdotal evidence. The way most questions are asked in the origins debate, they are unanswerable. The question as I posed it is much, much simpler and can certainly be verified.
Is Natural Selection Valid?
I shouldnt have to spend much time defending the idea that Natural Selection is a perfectly valid concept that we see proven all the time. We all know and observe every day that winners win and losers lose. We know that babies with severe birth defects often do not survive, much less thrive. We know that Natural Selection weeds out losers. Natural Selection works.
The real question is, does Random Mutation produce winners? Does it create plans for new muscles in the eye of the falcon? Does it add information to the code of DNA? Thats the question I set out to answer.
Like I said, DNA is a blueprint for life. Its a code and a language. It has an alphabet (A = Adenine, G = Guanine, T = Thymine, C = Cytosine) and that alphabet spells out the instructions for everything. So the question is:
Can Random Mutation add information to DNA?
I decided the best way to answer this question is to make it more general: Can random mutation add information to any code or language? Can it make it more meaningful, such that Natural Selection will weed out the garbage and leave us with better and better information?
My reasoning was this: Even if Darwinism couldnt be empirically proven in the lab (since we dont have millions of years available to conduct an experiment), we should still be able to investigate some other part of the world where languages and codes are used.
We should be able to experimentally determine if Random Mutations add information. This could come from any number of fields linguistics, digital signal processing, computer networking, computer aided design, language translation.
Information Theory Sheds Light on this: 1948
The questions Im asking here are answered in a field known as Information theory which really began with a paper written by Bell Labs scientist Claude Shannon in 1948. His paper was called The Mathematical Theory of Communication and its one of the most important papers ever published in the history of Electrical Engineering and electronic communication. Shannons paper literally gave birth to the digital age we now live in not because Shannon invented digital communication, but because he defined what it was capable of.
Claude Shannons paper tells you how much hi-fidelity music you can put on a CD (74 minutes), or how much data you can squeeze through a 56K computer modem (not much). It tells you how much information you can successfully transmit, given a certain amount of noise and a certain speed of sending your data. It discusses error correction schemes and even defines something called Information entropy which is the degradation that happens when you add noise to a signal.
Shannons book is not for the layman, but certain parts of it are understandable by anyone and it deals with things that most people are at least somewhat familiar with. He defines various layers of information which in laymans terms would be things like alphabet, spelling, grammar and meaning and how the upper layers are built on top of the lower layers, and how we can use this knowledge to detect and correct errors.
And in my research into the origins question, I quickly discovered that everything Claude Shannon discusses in his paper applies to DNA. (I have not encountered any geneticist or bio-informatics researcher who disagrees with this.)
DNA is a molecule, a data storage medium and digital communication protocol all rolled into one. It has a certain amount of memory. Your own DNA, in every cell of your body, contains about the same amount of information as a compact disc.
But heres where things get interesting: In Claude Shannons world, Random Mutation in DNA is exactly the same as Noise in an electrical communication system.
This really struck a chord with me because I already knew a great deal about digital communication systems. I spent six years of my life selling exotic networking equipment to factory engineers, Ive published dozens and dozens of magazine articles about communication networks, and In 2002 I published the book Industrial Ethernet. Now in its 2nd edition, this book explains the operation of Ethernet networks, TCP/IP protocol (the language that runs the Internet), and various networking languages that are used in modern equipment installations.
The world of electronic communication is a world of languages and codes. Every different kind of file on your computer a Microsoft WORD document, an Excel spreadsheet, a web page in HTML, a JPG or GIF image the difference between all these things is the language theyre written in. Microsoft WORD isnt just a brand, its a language thats been defined by Microsoft, for writing and storing documents. Same with everything else on the list its a language thats been defined by someone for a specific purpose.
So another way of asking the question is this:
Is there any instance in the field of computers, electrical engineering, radio, TV or any other aspect of communication where noise is added to a signal to increase its quality? Can adding noise increase the information in a signal?
I searched and I searched and I searched.
And the answer to that question, oddly, is a qualified no.
There is no instance in the field of computers, electrical engineering, radio, TV or any other aspect of modern communication where noise is added to a signal to increase its quality. There is no example where noise increases the information in a signal.
None.
But remember, I said the answer is a qualified no. The answer is still yes, sometimes.
Let me explain.
Lets take the sentence
The quick brown fox jumped over the lazy dog
I could randomly mutate this sentence, and if I got really lucky, the sentence could become
The quick brown fox jumped over the lazy dog.
(I added a period.)
The period did add information to the sentence; it corrected a grammatical mistake.
Now this sentence is 45 characters long. With the possibility of 26 lower case letters, 26 upper case letters, spaces and periods (excluding numbers and any other characters), there are 54 possible characters.
So how many combinations of letters are possible in this sentence? A probability textbook tells us that there are 4554 possible combinations. If you punch that into your scientific calculator, you get 1.9 x 1089 possibilities. (In other words, a very big number with ninety (!) zeros. Bigger than any number anyone ever uses in everyday life.)
Which is to say that there is one chance in 1.9 x 1089 one chance in a trillion trillion trillion trillion trillion trillion trillion or so that a single random mutation could correct the grammar of this short sentence.
So yes, random mutation (noise) can add some information to a signal. Parts per trillion trillion trillion
That is the qualification.
Obviously it is a very small qualification so small as to be thoroughly trivial.
I thought this was way too simple.
I said to myself, Geez, is the answer to the whole evolution question that cut-and-dried? Falcons cant evolve because DNA mutations can only destroy information, not increase it? That seems too easy, too obvious.
I spent several months hunting for answers all over the place, scouring more books and the Internet, looking for discussion on this topic. Surely there must be more to this question.
Claude Shannon does address it in his book. Not DNA specifically, but noise in general. Shannon says noise causes entropy. In physics, entropy is the process of useful energy becoming useless. Entropy is what happens when a candle burns down to nothing and you cant light it anymore.
One of Shannons contributions to science was showing that the math equations for information entropy are exactly the same as for heat entropy. And in both cases, entropy is an irreversible process. In other words, once you add noise to a signal, it is permanently corrupted and cannot be recovered, much less improved.
Noise always degrades a signal. Always.
No exceptions.
Anybody whos spent much time recording music knows what I mean. Cassette tapes are going out of vogue, but back in the day we used to record our CDs onto cassettes so we could listen on our Walkman or in our car. Cassettes always add noise (tape hiss) to the music.
Companies like Dolby and DBX devised ingenious methods of noise reduction Dolby B, Dolby C and so on to combat this problem. They were fairly effective, but never perfect. Actually the way noise reduction works is this: The signal is boosted and equalized before its put on tape, then its equalized the opposite way and cut back to its original volume when you play it back. All it does is lessen the effects of the tape hiss; it doesnt actually take it away.
And again, once the noise is there, it is absolutely impossible to get it back out.
And Ive never met any engineer who ever said the signal could be better after you added noise to it. The only exception to this is something called dither which does add noise to the signal before its recorded, but that is done to neutralize distortions in the recording equipment. Its dither in digital recording, and bias in analog recording. But it does not increase the information; it degrades the signal, albeit in a useful way.
So Im hunting for a flaw in this theory. Can anyone show that noise increases the useful information in a signal?
Now I am far from the first person to discover or discuss this, and I did find people debating this topic. I found some interesting misconceptions.
For example, Claude Shannon discusses how the addition of noise increases the information in a signal. But you have to be very careful to understand what he means when he says this.
Lets say you take your favorite CD and record it onto a cassette tape. Now you have added some noise to your favorite music. You can hear the tape hiss when you play it.
Well lets say you get a CD burner and you play the tape back and copy the taped version back to a new CD. Now you have a CD of a tape of a CD. A copy of a copy, with tape hiss thrown in.
Well the new CD does actually have more information than the old one. It has not only the music, but the tape hiss too. Instead of silence between the songs, youve added tape hiss. Of course the CD player doesnt care what it is, it just plays it. From the CD players point of view (CD players being totally dumb objects), yes, there is more information to send to the speakers.
But from a human point of view, there is obviously less useful information. The useful information has been compromised. Fine details you used to be able to hear are covered up by the tape hiss, never to be recovered.
All arguments you may find that cite Claude Shannon in saying that noise increases information are really saying that the tape hiss is an increase in information. Well obviously its an increase in useless information at the expense of the useful information.
So What is Darwinism Really Saying?
If we go back to the falcon question, what Darwinian theory is saying is this:
Noise gets added to the signal in the DNA of billions of falcons.
Most of the time, it produces harmful mutations.
Some of the time, it adds useful information.
Natural selection weeds out the harmful mutations and only the useful ones are left.
The useful mutations make the new falcons more fit to survive.
They proliferate and then more mutations make their progeny better adapted, more competitive, with enhanced features and ability to survive.
And the evolutionary dance continues.
Sounds pretty plausible, right? I certainly thought it did.
But lets use an analogy of something more familiar: Those cassette tapes.
You have a CD, and you make billions of cassette tape copies of it. Each copy is slightly different, because each one has different microscopic bits of tape hiss.
Most of the time the tape hiss is bad, but sometimes it is good.
People buy the good copies and return the bad ones to the music store, so only the good ones survive.
Billions more copies are made of the good ones, and the process repeats. Every few thousand generations of tapes and tape hiss, a new musical feature is added, so primitive tribal music evolves into modern jazz fusion.
Do you buy that?
I dont.
We could make a very similar analogy with photocopies. Weve all seen documents that were a copy of a copy of a copy of a copy. They look horrible. After enough generations, they become unreadable. A copy of a copy of a copy of instructions on how to get to my house never evolve into a superior plan for getting to my house.
Information is only destroyed by noise, never enhanced by it. (Devolution, not evolution.)
So if noise is bad for music and photocopies and FM radio, how can it be good for falcon DNA?
DNA, Computers, Human Language Have Error Correction Built-In
Youd be interested to know that virtually all communication systems, including DNA, have error correction mechanisms built in. The English language is about 50% redundant if half the words and letters are missing, you can still read it and figure out what is being said. Ethernet and TCP/IP have sophisticated error detection / correction mechanisms. If youre downloading a file and some of the bits get corrupted, your computer detects that and tells the other computer on the Internet to re-send those bits. And DNA has sophisticated error correction mechanisms, too. Errors are always bad, never good. DNA is designed to detect mutations and correct them.
I Start Presenting This Challenge to Darwinists
There are still a whole bunch of questions we havent really discussed yet, but this in itself gets down to the bottom of what Darwinian evolution claims to be true. So has any advocate of Darwinian evolution ever proven that random mutation can increase information? I wanted to find out.
So in addition to extensive searching and reading, I started having email exchanges with proponents of Darwinism. I would say Show me an example where random mutation actually increases information and they would try. And Boy, would they ever try!
The conversation would go something like this:
They would say, Richard Dawkins shows how evolution works in his methinks it is like a weasel evolutionary computer program. (Dawkins, in his book The Blind Watchmaker, shows how a random letter generator can evolve a complete sentence, if the computer program rejects erroneous letters and accepts correct ones, and it only takes 30-40 steps to do so.)
Id say But the desired result is pre-programmed into the experiment in the first place. All he has demonstrated is that his computer works properly, thats all.
So the conversation would turn to all sorts of other Genetic Programs which is an entire category of software where random mutations generate variations which are then selected according to certain criteria and improvements are observed. Fascinating stuff.
But thats not the Darwinian math formula. Remember, the Darwinian formula is
Random mutation + natural selection + time = Design
But genetic programs work on this formula:
Random mutation + deliberate selection + time = Design
Deliberate selection and natural selection are not the same thing. Look, if the falcons could say Lets randomly mutate only the small part of our DNA that designs our eyeball muscles and nothing else for a few million years, and lets just keep the ones that pass a quality control test then it would work. But thats not how evolution is claimed to work. Mutations (noise) are not selective about where they occur.
Darwinism: The Math Just Doesnt Work
In these debates we would go back and forth about this, and sometimes the conversations would get quite emotional as the person I was talking too kept hitting dead ends in their argument.
Sometimes theyd finally sigh and say Just because we havent discovered the answer to this question doesnt mean we wont someday.
My reply would be Yes, thats absolutely right, nobody can predict what science may discover tomorrow. But for right now I cant see how the fundamental premise of Darwinian evolution can even be said to be scientific, because the math simply does not work.
Most of the time Id eventually get silence. Much as we tried to make these conversations factual and friendly (they were rarely hostile), eventually the person would just stop replying to my emails. They were stumped.
Dawkins Cant Answer It, Either
Its interesting to note that the fanatical atheist Richard Dawkins, perhaps the most famous living proponent of Evolutionary theory, has never answered this question either in fact he has studiously avoided it. Theres an article
- The Problem of Information - where this question was posed to Dawkins, and even though six years have gone by, he never answers it. The answer he does give is a smoke-and-mirrors example at best.
There is a lot of material on this particular web page and there is a great deal of discussion about it. I did a bunch of backwards searching on Google to find every last reference to this article on the entire Internet, and nobody has successfully answered the question that it raises: Can you produce an example of a mutation or evolutionary process that led to an increase in information?
Many do claim to have answered it with genetic programs and the like, but if you examine their experiments carefully, you will see that none of these programs are actually examples of true Darwinian evolution. All of them without exception (including the much ballyhooed Avida softare program) use un-random mutation or un-natural selection in some way. Genetic programs are extremely useful, instructive, well worth studying. But all are examples of intelligent design, not evolution
Darwin: Brilliantly Half-Right; Tragically Half-Wrong