The value of Gitt information

In the Role of Mutations thread Scrutinizer has suggested that Gitt information would be the meaningful measure of information content for genetics rather than Shannon information/entropy. Leaving the comparative merits of Shannon and Gitt information I can't actually see how gitt information can possibly be measured without making a whole host of unwarranted assumption.AIG host a paper by Gitt discussing his measure of information.Gitt puts forward a number of empirical principles concerning information which to me look like nothing but complete assumptions based on a preconcieved endpoint. In what way are these empirical principles? In the absence of a way to actually measure Gitt information, something Gitt unaccountably forgets to provide, all we are left with is subjective pontifications and no way to conceivably support these 'priniciples' empirically.I wonder how Gitt would describe all the information we glean daily from the natural world, perhaps he would ascribe that ultimately to an intelligent source as well or come up with an alternative term for it rather than information.Is there any way to actually measure Gitt information?We will never be able to show creationists an increase in genetic information if they will not agree on a quantititative measure of information they are prepared to accept.Is Gitt information useful as anything other than a vague and shifty goalpost for creationists to use?TTFN,WK

Thread moved here from the Proposed New Topics forum.

I'm going to repeat my usual point on Gitt.According to Gitt: DNA is "mechanically" "understood" and therefore does not contain semantic information - and therefore does not contain Gitt information.

I'll take that as a no then.TTFN,WK

WK, If the above assumptions can't be demonstrated, then ID'ists can't say with any certainty that genetic information exists at all in genetic systems.In this, they find themselves making a circular argument. In order to infer ID based on information existing in DNA etc. they need to show that information exists in genetic systems. They can't do this without assuming their conclusion. Given that new functional genes have been observed to evolve via "statistical" processes, it therefore follows that those genes contain no information.Silly ID'ists.MarkEdited by mark24, : No reason given.

---

There are 10 kinds of people in this world; those that understand binary, & those that don't

So stellar spectra contain no information. The astronomers of the last 150 years will probably be a little surprised at that.

Studying the tracks in the dirt of a barnyard, one might gather some knowledge about the chickens that made them, and in this sense, the tracks might be said to contain information; it just wouldn't be Shannon information. Studying patterns in chicken tracks might be a legitimate approach to discovering some hitherto unseen aspect of chicken behavior. Studying them in the hope of decoding hidden messages from the chickens (or from God or whatever) would be more meaningfully interpreted as an aspect of the observer's behavior, possibly indicating the need for an adjustment in medication.Since Shannon was concerned with communications signals, his starting point was the tacit assumption that the message originated with an intelligent source; it didn't have to be explicitly stated. Gitt ridiculously tries to turn that around, concluding that what is treated as a coded message IS such, and therefore originates with a mental origin.It's laughable.

I don't think Shannon made this assumption, even tacitly. If you read the original Shannon paper you'll see that he defines the communication problem as one of sending messages from a source to a destination. Each transmission sends one message from the message set. He states that meaning is not part of the communications problem (meaning would be associated with an intelligent source), and his definition of a message set does not specify an origin.--PercyEdited by Percy, : Fix typo.

Precisely my point.

Your point was that you were wrong?What Percy wrote gives no indication that Shannon made any implicit assumption that any given message originated with an intelligent source, only meaning.People do apply Shannon entropy analysis to DNA, are you saying that this is a misapplication?TTFN,WK

A likely component of any robust attempt at defining "intelligence" would be the ability to manipulate symbols and to attribute "meaning" to them (and, conversely, defining "meaning" would be tough to do without referring to "intelligence"). Shannon's definition of "message" didn't need to specify an origin, because he was dealing with structures explicitly designed by intelligent beings for the express purpose of exchanging coded messages. See how that works? Absolutely. What's worse is that as it is reasonable to expect anyone clever enough to do so to also be clever enough to know why it's wrong, the very act raises serious questions about such a person's motives. Like the traditional freshman's trick algebra equation which proves that two equals one, or some such thing (impressing only those unable to spot the division by zero), Gitt disguises a trivial bit of sleight-of-hand with a plethora of irrelevance, counting on the eagerness of his intended audience to lead them quickly past it, and directly to his absurd conclusion. Then they all have cookies and punch.

In what way is he dealing with structures 'explicitly' designed by intelligent beings? By generalising his formulation Shannon allows it to cover any situation in which a signal is being transmitted over a noisy channel, that he may have been thinking specifically in terms of telecommunications as for applicability doesn't stop his work from being generally applicable. I'm not talking about Gitt I'm talking about bioinformaticians who measure shannon entropy in genomes (Chen et al, 2005, Schneider 2000) or use Jensen-Shannon divergence as a metric for identifying genomic features such as CpG islands (Luque-Escamilla, et al., 2005).If people are able to successfuly analyse the genome in terms of shannon information why do you feeel it is inapplicable? The transmission of genetic information from one generation to the next seems ideally suited to Shannon's problem of transmitting a signal over a noisy channel.TTFN,WK

That's true, but in applying it more broadly, it's vitally important to confirm that assumptions valid for a telecommunications channel are equally valid for the structures to which it is being applied. That's the "division by zero". Would it be valid for a non-Chinese-speaking person to use the same assumptions when examining a Chinese newspaper as when examining chicken tracks in the barnyard -- even if they were Chinese chickens? Well, an erupting volcano may seem like a violent demonstration of power on the part of an angry god, but a lot depends on your initial assumptions.My problem is in the circular way "information" is defined:What is a code? That which contains information.
What is information? That which is represented by a code.
What is DNA? Coded information.It's an empty tautology -- hence the division by zero.With some effort, we might analyze the topography of a hillside on the assumption that it represented a coded set of instructions specifying the path a given rock should take while rolling down the hill (we could similarly interpret the rock's topograhpy as the instructions for decoding this message -- or, for that matter, we could do it the other way around). A particular bump might be assigned a value such as "turn left five degrees", etc. We might then take a shovel up on the hill and introduce some "noise" by flattening out that bump. But the only possible basis for regarding the bump in its original form as "signal", and the flat spot as "noise" would be that somebody or something cared what path the rock took.Once were done looking at all the charts and graphs and equations, and considering all the nuances of conflating Shannon entropy with Kolmolgorov complexity, and reviewing Kant and Hume, Gitt's argument emerges as a Sunday-school-simple proposition: DNA is a message from God. I can't prove that false, and I won't waste a lot of energy trying. If your initial assumptions make you happy, who am I to screw with that?

Actually, a code is a set of rules for translating information from one form to another. Actually, information is one of a set of messages that you wish to communicate. A code is how you transform that information into a form suitable for transmission. That's correct. This seems a leap of logic unrelated to anything else you said. I think few here would dispute someone's right to believe that DNA is a message from God. It only becomes a dispute when someone asserts that science says it is a message from God.--Percy

Well, if we want to be really pedantic, the strictest possible interpretation of Gitt might be said to include the implication that information (which cannot exist without a code, transmitter, etc) isn't actually information when it isn't being encoded, transmitted, or decoded. What it is the rest of the time, I can't imagine. Apologies for not finding that very... well, informative. "Wish" seems to be the key word there. I guess the whole metaphor must not have worked for you then. I don't have time to unpack it fully right now; maybe this evening. For now, I'll offer this hint: I'm suggesting that DNA transcription is, ultimately, as deterministic a process as is a rock rolling down a hill. What if someone asserts that they believe that science says it's a message from God?