LsOL: Entropy is NOT from the 2nd Law, but vice versa

[Yorzhik]

Anytime I use the word information I will never be referring to something that might be generated by noise. Thus Rex and Weaver might have a problem with my common use of the word rather than adhering to what is a rather weird technical usage.

You don't have a problem with Weaver; only Rex.

 

[Yorzhik]

My, what selective reading skills you have.

I can only read what you write.

You took the last sentence of one paragraph, and the first sentence of the next paragraph, and ran them together, discarding the adjacent part which contradicts stripe in saying that noise adds information, and you did not include enough information to know what problem he was discussing. You also didn't provide a useful citation with which I could go find more context. You are obfuscating.

I did quote it wrong, in that I didn't include the new paragraph. That error does not change the meaning of what Weaver said.

I provided a useful citation, which is how you were able to find that I didn't include the next two sentences.

Those two sentences support my position, not yours.

In fact, I quoted it in context, and I provided a link to my source so that you could go find the entire document that I was looking at if you felt that more context was necessary.

No more context was necessary, although the extra context you provided was helpful for my position. Thanks.

The important distinction that I am making is that he didn't say that the added noise is not information. He just said that it is spurious. He also acknowledged that he is using a special meaning of information that can be "good" or "bad", a concept that doesn't transfer to evolution in the same way.

Quite. He said spurious and undesirable for a reason. His special meaning of information is the only meaning of information we can measure. So you can toss out the use of Shannon information if you want to for evolution, but that wouldn't be scientifically sound.

What you have to understand is what problem he is describing when he says that the information introduced by noise is undesirable, and how that problem differs from evolution by natural selection.

Right, and if you want to say that noise is desirable, then you must, by necessity, toss out Shannon information as a measure of the information in the DNA.

Now, tell us, how are you measuring the information content in the DNA?

You seem to be able to read me as selectively as you read Weaver, and perhaps more imaginatively. First, I never said anything one way or the other about Weaver being an authority.

Sorry. I was assuming you were going to be scientific. If you aren't using Shannon information (which, by necessity, makes Weaver an authority), then what information theory are you using?

Rex continues:

I said that your application of his model is incorrect by the model's own terms.

Not at all. If the noise that Weaver talks about is undesirable, I'll take him at his word.

Rex continues:

Second, the point of my post was that Weaver's model does not--, and was not intended by Weaver himself, to model something that is analogous to evolution by natural selection.

If that's the way you see it; fine. Weaver's/Shannon's model doesn't work for evolution. Tell us what information theory you use for DNA?

Rex continues:

He was modelling a situation in which you want the original message to be received with complete fidelity by definition, and it is in these terms which he describes the noise information as undesirable. This model does not address situations where the noise is considered potentially desirable.

I see. That would be a good reason not to use it in evolution. So what information model do you prefer? The "whatever information" model?

Lets try an example. Imagine I'm sending you an email. I write some text, click send, and away it goes. The goal of the underlying technology would be for you to receive the email exactly as I wrote it, and it employs various codings to try to ensure this. Say that I made a typo in my email, a misspelling. And say that, as unlikely as it is, the noise from the communication channel coincidentally fixes my mistake. Within the model that Weaver is describing, this change is undesirable, and spurious, even if in a more objective sense it is more desirable or correct. However, evolution, in contrast, uses the more objective standard of fitness, rather than assuming that the original message is correct by definition. This is the important difference.

THE important difference. Whoo-hoo. We've finally gotten to the important difference.

Your analogy could be even better. I'll explain why it's better afterward: You send an email that says "put tab A in slot B, and then fold both top flaps and turn the box over and put tab C in slot D" but somehow, due to noise, I got the email that said "put tab A in slot B and tape the joint because it will come undone when you go to fold the top flaps, and then fold both top flaps and turn the box over and put tab C in slot D". Let's also assume that the addition was an improvement because tab A kept slipping out of slot B as soon as you let go of it.

It is better because in your example, the standard (the target you say evolution doesn't have) was degraded in your sending - a bad mutation, so the mutation that fixed it wasn't a good mutation, but merely gave you back what you already had. In other words there was no improvement.

If you don't like my analogy we can stick with your's and what the implications mean. Just let me know.

Not so far, and I haven't set out to do so. I've just set out to refute the assertion that such a thing would be precluded by information theory.

You've said quite directly that we are precluding Shannon information theory. So what information theory are you talking about here?

I don't know where the information content came from, or what the process was, but I do know that it is quite easy to find information on Earth, from the Sun or not.

If you don't know, then don't say evolution is scientific. That's what we need to know to lend credence to your theory.

Are you going to tell me what you're talking about, or are you going to make me guess?

ThePhy has been pretty clear that he is fine with using Shannon information for evolution.

 

[Stripe]

You don't have a problem with Weaver; only Rex.

:chuckle: Should I ask him to step outside?