Kotatsu
Phthirapterist
You veer off when saying "they are the expressions of the same preserved genes." He does not say that. Could be. But that's adding a little bit of your paradigm and making an assumption.
My apologies. They would be developments by a mechanism Davison calls "a complete mystery" of the same information which has been carried in these (and all other lineages) since time immemorial in unspecified information carriers.
Btw, are they expressions of the same preserved genes?
Why do you ask someone whose understanding of these matters is so demonstrably inferior to your own?
Meanwhile, could you explain what purpose the following extract from Davison's paper serves, from a purely scientific point of view (emphasis as in original):
"The very word evolution derives from the Latin evolvo meaning to unfold as the pages of a book. Needless to say, a book has, by definition, already been written."
I find it highly curious that a professor in biology has never seen a notebook.
Why would they have to?
Because they have different functions.
Based only on "Ontogeny...", we know that Davison is interested exclusively in function. Structure, development, homology -- these are all concepts that Davison has either not heard about, don't understand, or don't care about. Sure, there are references to both developmental characters (both "higher" insects and nemerteans are said to have imaginal discs) and structure (e.g., the eggs of vertebrates), but no attempt is made to show that these comparisons are relevant.
If I were to write a paper criticizing evolutionary theory, I would expand on what evolutionary theory says about these perceived similarities, how evolutionary theory explains them, cite the evidence that this interpretation of the natural world is wrong, and propose a more solid hypothesis to explain the same phenomenon. This would be the more rudimentary form of scientific critique that I would expect to be taken seriously by other scientists.
By contrast, Davison mentions that two taxa -- or "kinds" really -- have a feature in common that he believes is the same thing, because they share the same function. He then proceeds to the next example, and the next, and the next, which would be acceptable, if at the end there would be some exploration of the evolutionary hypotheses for these patterns, some data that falsifies these hypotheses, and a proposal for a more robust hypothesis that would explain them.
We don't get that from Davison. Once he is finished listing his examples, he dives directly into arguments from personal incredulity -- "Rather than assuming independent invention of these remarkable parallels, it seems to me more reasonable to postulate these events resulted from the activation (derepression) of an enormous yet clealry limited stockpile of potentialities which were available when those events took place." and so on -- and then a page-long appeal to perhaps consider God, because a lot of people we connect with intelligence mentioned God in their work at some point.
Well, there is also the charming statement that "How such transformations were effected remains, of course, a complete mystery", meaning that even having gone through all these examples and -- hopefully -- having read, but not cited, all the relevant literature, Davison still has no clue about how his own mechanism is supposed to work, but as "macroevolution seems no longer to be in progress, we may never be able to resolve that issue."
So we know that to Davison, function is the only thing that matters, regardless of structure. Same function = same information that is preformed. Does the animal you're studying have a "kind of placenta"? Then it "exhibits characters of [...] placental animals". Structure, development, homology, and other factors never come into the picture.
Since similar functions are evidence for the information being preformed, and all organisms contain all the information required to develop any organism, then I believe it follows that there is an implied need for separation between the preformed information needed for the function of a wing and that needed for the function of a fin, as there are plenty of organisms that have one but not the other. Otherwise, developing the information for a finwing into either a fin or a wing would lead to these being randomly developed in any given population of birds or fish.
Of course, the same preformed information could result in a wing or a fin depending on reliance of the completely mysterious mechanism that is supposed to derepress these potentialities in the first place. Like many smart people throughout pre-genetic history, we could for instance consider God.
You are oversimplifying Davison's claims
This is impossible. His claims are already so childishly simple that nothing that contains difficult words like "gene" or "homology" could possibly be a simplification.
Prove it. That's not what standard evo theory says, at least not right after the isolation. Dinwar showed no such thing.
I will certainly try, and hope that someone else will be able to correct me if I am wrong.
Let us assume that we have a population of size X, which is genetically identical except for Y individuals that each have a genome that is different from that of the non-Y individuals, and also from each other, so that no member in Y has the same genome. Let us assume that Y is randomly distributed in the population. We can simplify the diversity (D) as:
D = (Y1 + Y2 + Y3 + Y4 + ... + Yn + 1)/X = (#Y+1)/X
A change occurs, and a random subsample is isolated from X. We call this subsample X', and put a ' on all members of Y that are randomly included in the isolated population. The diversity for this isolated population can be simplified as:
D' = (Y'1 + Y'2 + Y'3 + Y'4 + ... Y'n + 1)/X' = (#Y'+1)/X'
X' is by definition always smaller than X, as it is an isolated subset of it. Y', however, being a randomly selected subsample of Y, may be equal to (if all of Y is in the isolated X') or smaller than (if only some Y are isolated) Y.
It follows that D' = D (the genetic variation is identical) if for instance:
(#Y'+1)/X' = (#Y+1)/X
That D' < D (the genetic variation is lower) if:
(#Y'+1)/X' < (#Y+1)/X
And that D' > D (the genetic variation is higher) if:
(#Y'+1)/X' > (#Y+1)/X
That is, if the proportion of Y' to X' is higher than the proportion of Y to X was, genetic variation has increased in the isolated subpopulation. If X = 100 and Y=25, this means that every fourth individual in X has a unique genetic sequence. A random subsample of X' = 10 where, by chance, we get Y' = 5 means that every second individual has a unique genetic sequence, which means that the variation of X' -- measured in this clumsy way -- has doubled. In the extreme case of belonging to Y meaning that you are more likely to end up in Y', we could have a population X' = 10 which contains only Y' individuals, in which case the genetic variation that is 100%, as opposed to the initial 25%.
This primitive model ignores, of course, phylogenetic relatedness within Y and between Y and X, geographical substructuring of X prior to isolation, and the effect on the lineage's survivability of Y and X, respectively.