Principia

Hi everybody,

I am transferring a topic from the Evo-Cre forum to here because the subject matter turned out to be more philosophical than I had thought.

At the ARN site, mturner is arguing that the use of 'information' in scientific discussions is nebulous. Specifically, he aims to categorize information as one that is 'meaningful' versus one that is 'physical.' Thus, meaningful information is one that possesses meaning outside of the medium on which the information is encoded. He argues that DNA falls in this category because it has meaning in the sense it directs protein synthesis and its meaning is thus not the same as the its nucleotide constituents (the information). 'Physical' information is thus all other information that has no meaning, in the sense that nothing is conveyed beyond the actual encoding.

From this argument, he then argues that meaningful information can only be encoded and decoded by 'intelligences.' Thus, the meaning in DNA is 'understood' by the 'intelligence' in proteins. Of course, the ultimate conclusion is that life must have been intelligently designed.

I have skipped several steps because they are unclear to me. However, if any here would like to peruse these links:
<a href="http://www.arn.org/ubb/Forum1/HTML/001953.html" target="_blank">http://www.arn.org/ubb/Forum1/HTML/001953.html</a>
<a href="http://www.arn.org/ubb/Forum1/HTML/001956.html" target="_blank">http://www.arn.org/ubb/Forum1/HTML/001956.html</a>

and perhaps tell me how to better argue what I feel is intuitively nonsensical (or perhaps why it should be sensible), I would really appreciate it.

Thanks,
Scientiae

John Page

I had a quick look at the links. I think mturner is correct if one is dealing with entities that do not have information processing capabilities. i.e. what they process is reality.

IMO, the adjective "information" is applied subjectively by us humans as 'information processors'. Information is only meaningful in the context of the intelligence that performs the 'information processing'. So, a lump of wood contains no information but a scientist may collect information about it. A scientist may also acquire information about gibbons, and gibbons about scientists, but these sets of information only have meaning within the context of the respective 'information processors'.

I think people generally, not just scientists, confuse the situation somewhat by talking about e.g. "the information goes from a to b" or "the crystal knows what temperature it is". Obviously there is a borderline as to what an information processor is - I don't understand DNA but I believe its a complex organic molecule that just "reacts", on the other hand maybe it does have a mind of its own.

excreationist

To make this less confusing, I suggest that the states of unobserved physical systems not be referred to information at all.

So there is only one kind of information - this involves symbols.

So some basepairs in DNA can symbolise or refer to an amino acid that has to be manufactured. The physical state of an isolated photon is not a symbol for anything. It doesn't refer to anything else. It is just a photon.

So the "meaning" is just what the symbol refers to, in the context of the information processing system. In the case of English words, the information processing system also includes our brains and the patterns we have learnt during our lifetimes. We might hear or read an English word and dozens or hundreds of direct and indirect associations can be triggered. So the word can have a very rich meaning. On the other hand, in programming languages and in DNA, the meaning has a very straightforward "mapping" where certain symbols always refer to particular things.

And about "meaningless" words. I believe that it isn't possible for information to be meaningless. I think that if it is meaningless, then it isn't information. For example, the word "foogaw" could be said to be meaningless by people. Well the word doesn't refer to any particular concept, but it *does* refer to a certain pronounciation and as a written word, we can associate with many things, such as words that look similar. Even a handful of sand can become meaningful when we use our brains to try and find patterns in it. When we find patterns in it, information is created in our brains - we create new memories which are encoded in neurons. These neurons are symbols that refer to patterns.
So in our brains, we form new memories that we use to interpret the sand. We might find patterns between this handful of sand and our past experiences. And so this sand becomes a rich symbol for us - referring to elements of our past experiences. If it doesn't refer to anything else, a handful of sand is just a handful of sand - not information.

[ March 10, 2002: Message edited by: excreationist ]</p>

owleye

I'm not entirely sure what the problem is, even from the distinction that mturner makes, but I think that information theory as it is understood in the work of Shannon, helps enormously. Fred Dretske, a philosopher of some note at Stanford, has developed an information-based epistemology that helps, I think, to clarify the role that meaning and information play in knowledge acquisition. As you may know, Shannon differentiates meaning from information. Information can be coded because it represents a measure of the possibilities inherent in observations. Observations, in addition to providing knowledge of what is given in experience, provide us with clues (i.e., information) about other things. If I hear a bid of 2-clubs from East, this is supposed to provide me with information on what is in East's hand (as well as what might be in others' hands). There are a finite number of arrangements of cards, and this can be coded so that information is revealed through whatever source there is.

Meaning, on the other hand is different. When asked for the meaning of 2-clubs, I don't seek what the bid tells me (since that is based on what is in my hand and other clues), but rather what the bid signifies under the condition of the auction sequence. As such, meanings may be informative, but not such that we can decipher anything more than the extent of all the possibilities latent in it. It remains for the knowledge seeker to acquire enough information from the cues given or sought to reduce the entire set of possibilities to one.

Meaning, then, provides us with what counts as a possible thing. Information is what is extracted from the world, conveyed through some medium and is coded within us in some fashion that serves to assist us in acquiring knowledge (of the individual) that what is given in experience qualifies as being one of those possibilities (or of its class) and provides clues to something else The moving cloud of dust on the horizon is my uncle Fred coming for dinner because there is enough information conveyed by the observation of a moving cloud of dust coupled with other information that I have already stored that allows me to conclude that uncle Fred is coming for dinner.

I'm not quite certain why information coded in genes cannot be understood in the same way. The only difference that I can tell is that genes are relatively blind by comparison with human cognition. That is, they depend more on having a narrow range of possibilities to their expression.

owleye

excreationist

owleye:
Have a read of my post. I'm saying that information is just symbols or references to things or patterns/concepts other than the physical medium itself.

The common definition of "Information is what is extracted from the world" is a bad one I think.

"Information can be coded because it represents a measure of the possibilities inherent in observations."

I don't see how that applies to DNA. DNA is accumulated by random recombinations, mutations and natural selection. This mechanism is a form of information processing IMO, and DNA is information (it is symbolic).

How does DNA "represent a measure of the possibilities inherent in observations?"

In DNA, there is a one to one mapping of a symbol and its meaning. So there are no possibilities or ambiguities. Programming languages also are unambiguous to the compiling program.

Principia

I think I've tried to argue that, but mturner's argument is tantamount to defining an IDer into existence.

His argument:

1) DNA is information with meaning.
2) Meaning can only be encoded/decoded by an intelligence
C: DNA information demonstrates ID

Fishy?
SC

[ March 11, 2002: Message edited by: Scientiae ]</p>

John Page

Yes, the premise #1 is unproven (although not necessarily untrue).
#2 implies DNA has intelligence, I'd be interested if anyone has evidence to show it is more than an organic molecule.

excreationist

I agree. I use the term "symbol" and "information" pretty interchangeably and symbols have to have a meaning (the thing they refer to. DNA is a string of words (symbols) that refer to amino acids. A decoding/manufacturing mechanism translates the symbols into the amino acids.

Yeah, I think that information requires an information processing system and this is what intelligence is. (Though the sophistication can vary) I think mutation, natural selection, DNA recombinations and DNA copying are the processes involved that encoded new versions of the DNA. The information was processed - and it evolved. I'd call this about the least sophisticated form of intelligence that is possible - its like a simple computer - that can decode information though it evolves.

I don't think it is totally impossible for an information processing system (that is self-replicating) like DNA to happen by chance. It *might* be extremely unlikely though. (But maybe there was some self-organisation involved when life began)

owleye

excreationist...

"Have a read of my post."

I hope I have the right one. So here goes.

Having read your post it's possible that what your understanding of meaning and information is approximately the same as how I understand it, but you use a different vocabulary. But let's take a closer look.

"To make this less confusing, I suggest that the states of unobserved physical systems not be referred to information at all."

Since my post did not speak of unobserved physical systems I take it you were not addressing me. But let's see where this goes.

"So there is only one kind of information - this involves symbols."

Ok. (I'm ignoring the 'so' part of this here and in subsequent statements.)

"So some basepairs in DNA can symbolise or refer to an amino acid that has to be manufactured."

This doesn't seem right to me. The chosen language directs us to a causal sequence having but one outcome (by your use of "has to be"). I do not thereby see any information content. If, on the other hand, I am unaware of the DNA sequence, but only know how it works (what theoretical meaning there is in DNA sequencing), observing only the phenotypic amino acid, I could take this as a clue (a piece of information) that the DNA sequence that produced it is from one of the those that the theory predicts. To decide which one, I would need more information. Note that this does not specifically depend on "unobserved physical states" but rather to the theory by which we understand how to interpret observations.

"The physical state of an isolated photon is not a symbol for anything. It doesn't refer to anything else. It is just a photon."

As in the previous case, however, we can take a colored mark on a medium as a piece of information that tells us that (in conjunction with a lot of theory) that it is a photon having a particular frequency that caused the mark.

"So the "meaning" is just what the symbol refers to, in the context of the information processing system."

Ok, but let's see.

[] (The following is now taken from your specific response to me.)

"I'm saying that information is just symbols or references to things or patterns/concepts other than the physical medium itself."

In conjunction with your position on meaning, I can probably agree with this, but I think you have your references backwards. At least this is how I interpreted your two examples in the prior post.

"The common definition of "Information is what is extracted from the world" is a bad one I think."

I'm not sure what your objection is, so let me hold off on responding until I find out.

"I don't see how that applies to DNA. DNA is accumulated by random recombinations, mutations and natural selection. This mechanism is a form of information processing IMO, and DNA is information (it is symbolic)."

If I use your vocabulary (symbols whose meaning is what the symbols refer to), then I would say that phenotypes are symbols and what they refer to are genotypes. Thus, the 'information processing system' that I would be referring to is not the one you select. Rather it is the process by which phenotypes are produced by genotypes. To be an information processor, of course, it has to select information from its environment, much of which, I believe, is already determined by other genetic features, in consideration of its high degree of adaptation to the environment, and through delicate controls, produces the phenotype it is designed to do.

"In DNA, there is a one to one mapping of a symbol and its meaning. So there are no possibilities or ambiguities. Programming languages also are unambiguous to the compiling program."

To the extent to which there is a similarity between biological information processing systems and programming systems, I could agree with this. However, I may not be agreeing with you about what constitutes the symbol and what constitutes its meaning. In particular, I'm not sure that your use of 'programming languages' quite captures what is meant by an information processing system. This is not to say that language in its broadest sense does not form part of information processing systems. However, in so far as you characterize it above, I'd say the compiler is the information processor, processing information it receives in the from of a program given to it allegedly in some language to which the compiler can be said to convert what it finds to some other form that other processors can deal with. Each piece of information picked up by the compiler is processed according to a rule whose meaning is deciphered and pinned down according to higher rules, thereby removing ambiguity, if it can. This is not entirely unlike what our cognitive system works, though it must be confessed that we live with a fair amount of ambiguity in our lives, unlike computer programs who are not much able to tolerate it.

owleye

Vorkosigan

mturner used to post here at Infidels. He turned out to be surly, confused, insulting and obnoxious, and eventually left.

1) DNA is information with meaning.
2) Meaning can only be encoded/decoded by an intelligence
C: DNA information demonstrates ID

He's just trying to define ID into existence. All three premises are false.

Let's start by looking at what ID is. It is a problem of "intentions." For the IDer is not merely claiming that the universe was Designed, he's claiming that the current outcome is intended. This is a much more serious claim than the mere claim that it was Designed. After all, the Design could be the result of some random process instituted by a Designer, or a complex automatic process, like bees communicating with each other. So IDers really have some serious uphill sledding to do, and if they can do it with simple verbal tricks, then all the better.

You just have to keep a simple fact in mind: the universe was not Designed, and any argument that says so is wrong. Any definitional argument that does is simply a word game, and you should be able to spot the problem with it by focusing on the issue of intentions.

He uses "meaning" because it is an ambiguous word. It can refer to information that is either intentional or accidental. What he is really doing is not redefining "information," but "intentions." He's attempting an end run around the problem of intentions.

Consider the two questions:

1) What is the meaning of this letter you left on my desk?

2) What is the meaning of this strata superimposed here?

Note that "meaning" can refer to the significance of information that was intentionally encoded, and information that is there because we humans organize reality around us. By using the word "meaning" mturner has hornswoggled you into confusing one with the other.

To stop a silly ID argument, you just have to focus on "intentions." How does anyone know why the universe was Designed? There's no way to look at any given artifact and know what it is for, unless you have some context to make judgements. When we find a neandertal scraper, we infer it was a scraper because we know something about the needs and behavior of H. sapiens. Many other artifacts baffle, and the archaeological literature is rife with examples of things whose function is pretty much only guessed at. And those are from humans, which we understand.

If you can turn up some of the literature on the Calico Site, you can view the controversy over the stone tools there. Louis Leakey pulled 11,000 tools out of the cobble there, but hardly anyone believes that they are tools. Statistical studies seem to prove that they are not. Thus, you must judge. Did Leakey fuck up 11,000 times? Or are the statistical methodologies wrong somehow? Note that this is with Homo, a genus we have intimate knowledge of. But mturner wants you to buy his arguments about an unknown Designer.

Only now mturner is making the following claims.

1) The universe is an artifact
2) The current shape of the universe is intentional.

Ask him to prove, using empirical evidence, that (1) and (2) are both true. It is impossible to get to (1), as every IDer knows. You can't demonstrate the universe is an artifact, because there is no context that allows you to.

Michael

[ March 11, 2002: Message edited by: turtonm ]</p>