I wrote:
>>Although it may seem trivial, the easiest example of demonstrating
>>that a *single* point mutation can "result in an increase in genomic
>>information" (Spetner's definition, not mine), is a reverse-mutation,
>>which converts a mutated base in a non-functional gene back to the
>>original base. After all, if a point mutation which wipes out the
>>function of a gene is described as losing information, how would we
>>describe a reverse-mutation which perfectly restores the original
>>sequence?
Art, you said:
> That does not qualify as a gain in information, and you know that.
I do not know that.
Spetner CLEARLY indicates that a mutation which reduces or
eliminates a functional gene represents a loss of information.
Unless his measure of information is path dependent, a reverse
mutation must result in an increase of information.
> If the information is already present, deleting then restoring it
> is not what anybody means by an increase in information.
Royal Truman voiced a similar complaint when confronted with this
example. Let's put is this way: If the information was already present
in a gene damaged by a point mutation, why did it require a mutation
to restore function? Clearly something was lacking in organisms
which carried the original mutation which neither you nor Royal seem
willing to call "information".
> Producing the system necessary to convert a reptilian scale into
> an avian feather involves an increase in information of the system.
How so? Given the uncertainties of the Spetnerian measure of information,
couldn't Lee consider that process a loss on information? After all,
it seems to me that birds have lost the ability/information to
produce reptilian scales. They've certainly lost many other reptilian
traits. What is a feather but a damaged and barely functional scale?
See, the problem with Spetnerian metrics is that anyone can play a
different game using the same "rules".
> All the arm-waving in the world will not provide a cogent explanation
> for such an increase in information.
Provide us with a cogent measure of information and the amount of
information "increase" this example involves, and we'll discuss it.
Spetnerian information metrics is nothing but arm-waving, as far as I
can tell from his book. We've been through this before, Art. For
example, in some instances Spetner uses contradictory examples of binding
specificity as a measure of information content. I've documented that
previously. He never demonstrates the linkage to genomic information.
In other parts of the book he mentions that selection can add information
("but no more than one bit", he claims) but never develops it further.
He barely touches on Shannon. He discusses nothing that relates genomic
sequences to selection.
Admittedly, _Not by Chance!_ may be a dumbed-down book for general
consumption, and thus lacking some details, but some of what
he presents is clearly contradictory. Some parts are unsupported.
Other sections employ arguments with faulty premises (the convergent
evolution section is a howler). And his literature search for mutations
that increase information must have been brief because it took me
about 5 minutes to locate examples.
> Even if you are correct in asserting that increases in information
> content in the Spetnerian sense can be demonstrated in, say, bacteria,
> that is not the order of information increase that will work in the
> above example.
I've previously given examples of gene duplication producing
novel genes. Now, Lee might claim that this is simply rearrangement
of previously existing sequences. But these aren't all perfect
duplications; sometimes they involve partial duplications and
fusions between separate sequences.
Another tactic of Spetner is to shift goalposts and ignore contradictory
information. Spetner clearly states that he is aware of no examples
where mutations have "increased the information" of an organism's
genome. Although duplication and recombination are discussed as
possibilities, he dismisses them as either not producing the right
sort of information required or as possibly being non-random.
Aside: I suppose in Lee's mind, "non-random" = "directed", but that's
hardly the case. What it can mean is "biased" mutation. This is never
presented as a likely explanation in his book.
Regards,
Tim Ikeda
tikeda@sprintmail.hormel.com