Malaclypse the Younger

I believe there are several simple, naturalistic mechanisms that simply add more base pairs to a DNA molecule. I'm sure our molecular biologists would be happy to tell us more about these mechanisms in great detail.

Adding more base pairs even at random is adding "information" (in the sense that you are adding more bits).

Individuals where the added information is neutral or beneficial will reproduce advantageously vs. individuals with added information that is deleterious. Thus natural selection will ensure that these additions correspond to "good" (or at least non-deleterious) protein-formation and phenotype creation, creating "meaning" (correspondence with some extrinsic process) from the additional information.

By every useful objective definition of "information" the creationist assertion is simply false that natural processes cannot create information.

[ March 18, 2002: Message edited by: Malaclypse the Younger ]</p>

BolshyFaker

Ah thats better, thanx muchly. i just finished reading that Dawkins article as well which has a very helpful paragraph on the evolution of haemoglobin.

Now if only I could take the whole secweb gang into the debate with me

PS - love the <img src="graemlins/banghead.gif" border="0" alt="[Bang Head]" /> Graemlin, doens't it just feel like that all the time?

MrDarwin

Perhaps I'm quibbling, but in discussions like this you need to be careful to distinguish between "gene" and "allele". Most non-scientists and many scientists when they get sloppy (I plead guilty) use the former when they really mean the latter.

Second, natural selection does not remove genes from populations--it removes individuals from populations. In fact, when we compare the genomes of related organisms, we discover that they generally have the exact same genes, but slightly different versions of them (this is the basis for molecular systematics). What natural selection does is increase or decrease the relative frequency of a particular allele in a population. And eliminating an allele completely, even a harmful one (and here again we need to be careful--a trait that is disadvantageous in one environment can be advantageous in another) is extremely difficult. An exception is dominant alleles which confer a definite selective disadvantage on the individual.

Quetzal

Hi Guys, I'm back. Sorry to miss out on the end of the Randman episode. Unfortunately once in a while I have to actually work at this job, so was out of contact since around 10 March.

Anyway, BF, the best refutation I've ever used for the silly creationist arguments about "information" is simply to demand that they show the entire concept is both relevant to biology and provides any real-world, meaningful way of measuring or describing the effect of mutation or changes in genotype (or for that matter, phenotype - which is what NS operates on).

Information is a great analogy for genetics. It provides a convenient way of thinking about genetics. It is, unfortunately, utterly meaningless in the context that most creationists (including Dembski, et al) try and use it. It's really just a metaphor - even Dawkins only uses it in that manner - and like any metaphor is easily stretched to the breaking point. The entire creationist position is spurious. DNA does not "code information" about either the organism or its environment.

Let me try and explain a bit what I'm talking about. (For those whom this is old news, apologies.)

The first step is to take a look at what definition of "information" the cretinist is using.

1. Shannon-Weaver Information. This is really just communications theory - a way of describing the probability that a given set of symbols will be received as sent. Shannon "entropy" increases as the probability of accurate transmission decreases. Entropy rises or falls depending on how you set the probabilities on each symbol. From a biological standpoint, you change this probability depending on how much you describe about the trait, its environment, etc. IOW, the "information" content of a biological system depends on how much weight or influence on a particular trait you assign to environmental factors, and is basically included in the description. Shannon-Weaver is an artefact of the way we look at the system - not an inherent property of the system.

2. Kolmogorov-Chaitan Algorithm Information Theory (AIT). This is usually defined as the shortest possible string of code that can describe a sequence of symbols that defines a system. Again, we're talking about an abstract - not concrete - description of a thing and not the thing itself. AIT does have some relevance to the way we look at DNA, since we can assign a sequence of abstract symbols to DNA and then apply AIT to the symbols. It again refers to a description of the thing, not the thing itself and has NO concrete reality.

3. Semantic information: This is the common usage version referring to the "meaning" of something - form, function, role, task, goal, purpose - and again is purely an abstract way of looking at the thing - it has no reality.

4. Fisher Information. This is a description of the relation between an observer system and an observed state. It basically describes the liklihood that a particular measurement is accurate (accuracy is equivalent to a high amount of information). There is nothing unreal or semantic about the relation, only about how good the values assigned in that relation are. For biological systems, especially DNA, the measure of a process of change means only that the Fisher information of the thing measured itself changes over time. IOW, although a "real" property of the genetic code, as DNA changes through mutation, crossover, whatever, the measure will change right along with it, and hence the "amount" of information.

Information (cretinists' demands to the contrary), is an abstract property of abstract symbols. "Information" neither causes anything to occur nor prevents anything from occurring. It can't - it is purely an analytical tool. Information is simply in our heads or in the way we look at the world. It isn't the world. It can certainly be a useful tool - but like any tool it can be misused. Cretinists love to obfuscate and confuse the ignorant with complicated concepts. Our brains appear to be hard-wired to look at the world in abstract terms. Since we think in abstractions, we seem to demand that abstractions actually exist - "the name of the thing is the thing". I ain't never met one, however.

My advice would be to a) demand a definition of information as your cretinist is using it, and then b) demand your cretinist show a direct, real-world correlation between whatever flavor of information they use and genetics. My guess is you'll have them chasing their tails relatively quickly.

Hope this helps. It's good to be back.