The information theory argument

[scientia]

tigert,

I don't mind discussing Information Theory but it seems pointless to go around in circles on it. I can tell you what the theory says and then we can see if there is any point that can be made from it.

A long time ago people understood that a free energy machine or perpetual motion machine was not possible. However, someone then proposed question. What if you made a detector that could tell the difference between an air molecule moving with higher than average velocity and one with lower than average velocity. That sounds possible. Then what if you had some kind of door that you could open and close very quickly. Then what if you used these together so that each time a high speed molecule was heading towards the door, you opened it; but if a low speed molecule was heading that way, you closed it. The container would then get warmer than room temperature and theoretically you would have a free energy device. At the time, no one could think of a counter argument even though this seems to violate the laws of thermodynamics. Eventually, someone came up with the Information corollary which says that any energy advantage you gain by information will cost at least as much in terms of energy to obtain it. In other words, information is not free; it requires energy to obtain. Therefore, you can't build a free energy machine by using information.

The second part (your reference to Shannon) is a different topic; it has to do with the nature of communication and information in this context is not exactly the same thing. When the word encode is used, it refers to abstraction. This is important because communication is impossible without abstraction. Any communicated information must be abstracted. This can actually be multiple levels of abstraction. For example, the letters in this sentence are an abstract representation of vocal sounds. The letters form words which are abstract representations of actions, objects, descriptions, and concepts. However, you can't make these symbols on a computer so letters are abstracted as bits represented by energy states of gates. If these are transmitted as they are right now over my wireless LAN connection, they have to be modulated as radio frequencies. The particular type of abstraction is not really important as long as you understand what the abstraction is. This is where you refer to an encoder and a decoder. Keep in mind that these can include your speech and someone's hearing or hand gestures and someone's eyes. These can also include things like emotional expressions. Of course, this area is a little different because it seems that emotions are innate to our species. All humans seem to understand what sadness, happiness, and anger look like in terms of facial expression whereas an alien would not. This does not remove the abstraction; it just makes the abstraction innate.

Still on the topic of communication, there are limits to compression. This basically means that you cannot summarize the unabridged Oxford English dictionary with one vocal sound, while others think of this in terms of data compression using various algorithms. But, they are basically the same. Continuing, there are limits based on signal to noise ratio. You can approach this topic from many different angles. For example, what would be the smallest planet that could be detected in a nearby solar system? How far away from Earth could a space probe get and still be within communication range if it had a ten watt transmitter and ten meter diameter parabolic dish antenna? What would be the minimum volume difference between someone standing next to you speaking and surrounding conversation so that it could be understood? If you are familiar with computers then this gets into things like parity bits, checksums, gray code, GCR, MFM, RLL, EFM. If you prefer music, we could talk about Dolby noise reduction. Essentially, there are limits as to how much information can be maintained and recovered based on these things. There are limits to the bandwidth of USB and ethernet or the bandwidth of a hard drive controller.

So, how does all this relate to DNA? Well, DNA is a process code. I recall a number of science fiction stories that were based on the idea that RNA could be used to directly store abstract information such as memories. None of those are actually true, however, neurons do have special provisions to move mRNA the longer distances. Is DNA information? Yes, it is indeed. Is it true that only intelligent things can create information? No, that would not be true. A simple example of this would be the machines that create random power ball numbers. There is no intellect used in this process; these machines are as random as possible. Yet, the exact number that is generated is a piece of information.

I assume the point you are trying to bring up is whether or not DNA is a functional abstraction and whether or not abstraction requires intelligence. DNA is a functional abstraction. That is fairly easy to see since DNA gets converted to transfer RNA which is then used to create messenger RNA which is used to create proteins (process simplified). Also, there are control sequences in the DNA which determine how the sequences are acted upon. That is, to the best of my knowledge, there is no direct, necessary correlation between the sequences of three base pairs that make up one piece of tRNA (codon/anticodon) and the specific amino acid that it relates to. As far as I know, the codon sequences are arbitrary, although I could certainly be wrong and this could be corrected by someone who knows more about organic chemistry. So, again, to the best of my knowledge, this is functional abstraction. To explain the difference, there is a toy that consists of a tube with disks inside of it that divides it into sections. And these disks have holes in them that are in order by size from one end to the other. So, if you hold it with the smallest holes on top, the holes get larger and larger toward the bottom. And, it contains balls of varying sizes. If you put the small holes up and shake it, all the balls will go to the bottom. Then if you turn it over and shake it, the balls will fall through each section until they won't fit through the holes and you'll end up with the balls sorted by size. Is this information? Yes, it's not really any different than if you used a sort algorithm (such as bubble sort, shell sort, heap sort, or quick sort) to put data into order. Is it abstraction? No, because each ball is a necessary correlation to the size of the hole. It's just a physical property. It's the same as a key fitting in a lock is not an abstraction. Abstractions are arbitrary.

So, this is where we come to the crux of the discussion. BTW, when others have said that crystal patterns or other inanimate patterns are information, they were correct. You can create information just by rolling dice. The specific pattern is information. However, a snowflake pattern is not abstracted information -- and perhaps this is where the confusion comes in. This is a natural result of trying to combine two different theories where the term "information" is used in different ways.

So, are you saying that abstraction requires intelligence? Can you create an argument for this?

 

[he_who_is_nobody]

False. I don't begin defining DNA to be code that can only be created by a mind so that I can automatically arrive at that conclusion.

" Francis Crick received the Nobel prize for discovering DNA. The following is from the first paragraph of Francis Crick's Nobel lecture on October 11, 1962. Note his use of the word "code" and "information," emphasis mine:

"Part of the work covered by the Nobel citation, that on the structure and replication of DNA, has been described by Wilkins in his Nobel Lecture this year"¦ I shall discuss here the present state of a related problem in information transfer in living material, that of the genetic CODE, which has long interested me, and on which my colleagues and I, among many others, have recently been doing some experimental work"¦"
The following quotes are from atheist Richard Dawkins' book The Blind Watchmaker:

"Every single one of more than a trillion cells in the body contains about a thousand times as much precisely-coded digital information as my entire computer.

"Each nucleus, as we shall see in Chapter 5, contains a digitally coded database larger, in information content, than all 30 volumes of the Encyclopaedia Britannica put together. And this figure is for each cell, not all the cells of a body put together."
Having quoted Dawkins here, it's interesting to note that neither he, nor any materialist has ever provided any scientific (i.e. empirical, testable, falsifiable) explanation for the origin of information. For a very interesting and extensive read on this subject, read "The Problem of Information For The Theory of Evolution" by Royal Truman. If you carefully trace every reference and rebuttal to this article on the internet, you'll discover that not one person has ever supplied a scientific response to the questions raised here, nor provided any examples of materialistic processes that produce coded information."

"Information, transcription, translation, code, redundancy, synonymous, messenger, editing, and proofreading are all appropriate terms in biology. They take their meaning from information theory (Shannon, 1948) and are not synonyms, metaphors, or analogies." (Hubert P. Yockey, Information Theory, Evolution, and the Origin of Life, Cambridge University Press, 2005)"

"If there are pebbles below a rapids, there are pebbles below a rapids. There is no coded information associated with them, unless you measure their size, in which case you have created information to describe the pebbles, based on your chosen symbols and units of measurement. Same with orientation of sand dunes, layers of hailstone. Those objects represent only themselves; there is no encoding and decoding mechanism within these material objects, such as there is in DNA. If someone says the layers of a hailstone are an encoding mechanism, I reply that there is no convention of symbols, nor is there a decoding mechanism."

You are doing exactly what the video I posted said is wrong with this argument. You are equivocating the word information with a different definition. The quotes you provided by researchers and scientists are not using information as you defined it in this post.

In order to prove this wrong, please demonstrate how DNA is a code by the definition of information you first provided. You will be unable to do this because DNA is a pattern based on the definition of information you have provided. DNA is only thought to be information when the definition of information is expanded to include patterns, but by doing this, it makes all the qualities of information you are seeking to promote no longer apply.

 

[tigert]

No I'm not trying to expand the definition of info to contain patterns. Information & designs will always contain patterns but patterns never contain design - design always invokes symbolism and language which represents something other than itself, which is what DNA does, or Chinese or a pile of sticks re arranged to spell your name. Patterns don't. A snowflake is a snowflake is a snowflake. Or a sand dune. Or a rock etc etc, these will come about when the requisite natural forces occur.

 

[Visaki]

No I'm not trying to expand the definition of info to contain patterns. Information & designs will always contain patterns but patterns never contain design - design always invokes symbolism and language which represents something other than itself, which is what DNA does, or Chinese or a pile of sticks re arranged to spell your name. Patterns don't. A snowflake is a snowflake is a snowflake. Or a sand dune. Or a rock etc etc, these will come about when the requisite natural forces occur.

I admit that I haven't dwelved into information theory since my programming lesson day over ten years ago, and even then just skimmed it, but am I the only one that finds the underlined part a little... funny?

 

[tigert]

Visaki there is absolutely nothing funny about it. It is what it is. Patterns never contain symbolism or language that represent something other than themselves, they contain no design whatsoever. That's why patterns aren't design. But design has patterns and design. There is a vast chasm between the simplest possible design and most complex pattern.

 

[australopithecus]

DNA doesn't contain symbolism or language that represents something other than DNA. You can pepper your rhetoric with as much convoluted flowery nonsense all you like, it doesn't make what you say any more true.

DNA is a sequence of nucleotide chains. That's it.

 

[scientia]

No I'm not trying to expand the definition of info to contain patterns.

I have no idea what you are trying to say here. I guess I need to define this: Information specifically with regard to Information Theory is a reduction in uncertainty. This doesn't require any great effort to understand. Suppose I have a box and inside this box is a circle. I drop four colored balls into the box and then ask you if any of the balls ended up inside the circle. You have no idea because you haven't looked into the box. Uncertainty is very high. However, you could probably estimate the probability that one or more balls would end up inside the circle. So, uncertainty is not 100%. Now, if you look inside the box then you know; uncertainty becomes zero. What you gained in eliminating uncertainty is: Information.

You can't use this term the same way when you talk about communication because part of that is maintaining the integrity of the information. Therefore, if you talked about information as a lack of uncertainty you would be using the word "information" twice with two different meanings. In this context, information only refers to the pattern itself, not to the integrity of the pattern.

Information & designs will always contain patterns

Here is where your argument begins falling apart. What you are actually talking about is abstraction. Patterns can be naturally occurring or created by an intelligent agent. Intelligent agent is a super-set that includes not only intelligent organisms but also things initiated, created, or directed by an intelligent organism. For example, a music box is not itself intelligent but they are of course build by people (who are intelligent). Information in the communication sense and design always contain abstraction. Information in the Information Theory sense does not.

but patterns never contain design

Your terminology doesn't make any sense. A pattern would not contain design. On the other hand, a pattern can be associated with with a natural occurrence or an intelligent agent.

design always invokes symbolism

More fuzzy terms. Design always involves some abstraction. The word "symbolism" is not a very good fit.

and language which represents something other than itself

That is the definition of abstraction that I gave earlier. Languages are abstract.

which is what DNA does

I guess I can repeat what I said earlier. A working DNA sequence is a process or functional abstraction. Some of this might be difficult to understand but DNA itself is not a language. The language would actually be composed of whatever functions are contained within the DNA sequence and all of the support structures such as ribosomes. Again, I'm sorry if this seems difficult but the language would be the abstraction of the functional capabilities rather than the DNA itself. Paper and pen are not a language. However, I assume the primary point you were trying to make is that DNA does involve a language which it does.

or Chinese or a pile of sticks re arranged to spell your name. Patterns don't. A snowflake is a snowflake is a snowflake. Or a sand dune. Or a rock etc etc, these will come about when the requisite natural forces occur.

Again, you are repeating what I already said, patterns are not necessarily abstractions.

Once again, do you have an argument dealing with intelligent agents and abstraction?

 

[he_who_is_nobody]

DNA doesn't contain symbolism or language that represents something other than DNA. You can pepper your rhetoric with as much convoluted flowery nonsense all you like, it doesn't make what you say any more true.

DNA is a sequence of nucleotide chains. That's it.

Seconded. The fact that you do not already know this, tigert is why it is obvious that you are new to this.

 

[Deleted member 619]

DNA is a code in precisely the same way that London is a map.

Carry on.

 

[Rumraket]

Here's a convincing argument for God.

(1) The sequence of base pairs in DNA is a code.
Much effort has been made to discredit this statement, unsuccessfully. This statement is fully and explicitly supported in virtually all of the scientific literature since the 1960"²s.

No, the function of the "genetic code" is analogous to a code. DNA is actually just a molecule. This is true by definition since you can have completely uncoded sequences of DNA (they don't transcribe into mRNA that gets translated into protein).

The "genetic code" is and always was intended as a metaphor for how DNA sequence is transcribed and translated into protein sequence.

Genetic Code, Text, and Scripture: Metaphors and Narration in German Molecular Biology

Abstract

This paper examines the role of metaphors in science on the basis of a historical case study. The study explores how metaphors of “genetic information,” “genetic code,” and scripture representations of heredity (i.e. the metaphorical comparison of DNA with text and alphabet) entered molecular biology and reshaped experimentation during the 1950s and 1960s. Following the approach of the philosopher Hans Blumenberg, I will argue that metaphors are not merely a means of popularization or a specific kind of modeling (by building analogies) but rather are representations that can unfold an operational force of their own.
While the influence of cybernetics and information theory on molecular biology is well documented in historical analysis throughout recent years, this paper offers new insights into the metaphysical and religious resonances of textual metaphors in the life sciences. The main focus will be on developments in Germany, in particular on the work of the German biochemist Gerhard Schramm. In this historical case study the interaction between metaphors and experimental practices will be discussed. The paper analyzes different phases in the use of metaphors during the 1950s and 1960s: it will explore how the metaphors of a “genetic alphabet” or of “genetic code” (which were used with an illustrative purpose in the 1950s) developed into a new research program and eventually attained ontological status in the early 1960s. At that time Schramm's use of textual metaphors was reminiscent of nineteenth-century German natural philosophy. In this case, the metaphorical shift shows how the metaphor of a “genetic text” or a “genetic code,” which were central for the emerging molecular biology, drew on older cultural traditions with all of their metaphysical and religious preoccupations.

So it's a metaphor, you're wrong, get over it.

(2) All codes that we know the origin of come from a mind.

Question begging, pure and simple. You first have to actually establish this conclusion, you can't simply assume it as your premise.

First of all, we haven't actually directly witnessed the origin of the genetic code, so you can't assume what you state in this premise. At best, you can construct a statistical syllogism whereby you attempt to claim that "since almost all codes we know of were designed, therefore the genetic code was probably designed", with a certain probability you have to compute.

Now, unfortunately, the exact details of the "genetic code" is that first and foremost, it's not actually a code(see above), but also, there's a pretty significant case to be made for it's evolutionary basis. This significantly alters the strength of any attempted statistical syllogism you might provide, since we now have two competing cases that needs evaluation.

See this:


Since the evolutionary process is an observed empirical fact, and we have a number of evidences for not only observations of the evolutionary process altering the code, but also evidence that hints at how the code actually evolved - the evolution of the genetic code, as opposed to unobserved and untestable supernatural designers working in the ancient geological past, wins out hands down as the more plausible, best evidentially supported and therefore most believable hypothesis for the code's origin.

Much effort has been expended to discredit this statement as well. Assertions have been attempted that gravity, snowflakes, magma flows and the like are codes. But none accurately conforms to Shannon's communication model. Most of the examples cited do not contain an encoding system, and none contain a decoding system.

Here's a natural code: Atomic spectral lines in the electromagnetic spectrum. Unique and specific in profile in all the elements of the periodic table.
Quick example:

So, all the elements of the periodic table have their own "barcode" in the electromagnetic spectrum. This constitutes an empirical counterexample to your claim that all codes are known to originate from "minds", unless you also want to question-beg that emission spectra were designed too.

Do you want to do that? Bring the evidence!

(3) Therefore DNA came from a mind.
The objection to this statement has been that the conclusion is reached inductively. Complaints have been lodged that inductive reasoning is inherently unreliable. But we do observe that the laws of thermodynamics and in fact the majority of known scientific laws are determined inductively and not deductively. If you wish to throw out inductive reasoning, then we can discard almost all scientific knowledge and start all over again and use rocks and sticks to make fire.

Your conclusion doesn't follow from your premises. Premise one is simply outright false, premise two is question begging, and last but not least, we have pretty substantial evidence that the DNA synthesis pathway evolved in an RNA world, since all DNA monomers are synthesised biochemically from RNA precursors in extant cells.