Lobe-finned Fish, Amphibians, and Devolutionary Processes

So earlier this week I was thinking about von Baerian development and considering a specifically non-recapitulatory view of The Vertebrate Embryo and was thinking, "Could the majority of fish be considered more specialized in terms of the general vertebrate pattern, with the more amphibian-like fish being the archetype rather than the specialization.

Lo and behold, I came across something in Stuart Kauffman's work that evidence seems to have been mounting that the swim bladder came from the lung, not the other way around!  It would be even more intriguing if the archetype for fish were something even more like an amphibian, say a sarcopterygian like Tiktaalik, rather than the more numerous actinopterygians.  The actinopterygians depict for us how (bony) fish are different from other vertebrates.

Now, is it possible that we could learn things about vertebrate development thinking this way much more efficiently than we could with this quasi-Haeckelian recapitulatory progressivism?

The story as I knew it was that fish somehow developed a swim bladder around the time they replaced cartilage with bone.  Since utility is the mother and father of invention in the Darwinian world, the usefulness of swim bladders for pelagic swimming was sufficient to explain their existence.
 
If it turns out that lungs were adaptation preceding tetrapods by ages, well then that too is just what one expects in Darwinland.

It's important to remember that evolutionary logic means never having to say you're sorry.

. . . in 2009, just three years ago . . . the purported fact that 95 percent of the human genome "might as well not be there" was an embarrassment "for creationists," whom in typical Darwinian fashion Dawkins conveniently conflates with intelligent-design advocates. Junk DNA is just what a Darwinist would expect, in other words. Cut to 2012, and now the evident fact that "junk DNA" isn't junk at all but is instead vital for life has become "exactly what a Darwinist would hope for," namely, "to find usefulness in the living world." That is, heads you lose, tails I win. . . . suspiciously convenient self-contradiction. Ah well, as we knew already, being a Darwinist means never having to say "I was wrong." 

But it's not a self-contradiction. With Evolution-Did-It , whatever direction the evidence points is just what Dr. Pangloss would expect, because evolution as a theoretical framework can accommodate almost anything.

Even if the upward trends of progressive (teleological) evolution makes more sense for the historical framework that the evolutionary research program has pursued, what value might there be in more devolutionary hypotheses being assumed.  (Devolution is not nearly as satisfying and validating to Darwinism as complexity-building evolution.)  The alligator is a reptile that is considered the pinnacle of reptilian evolution nor a transition animal but a sort of devolved representative of warm-blooded archosaurs.  Ceolocanths are believed to be derived from sarcopterygians.  The TTSS1 pump might be better thought of as having devolved from the flagellum.  The mimivirus might best thought of as having devolved from a cellular creature.  Maybe Tiktaalik, like the alligator is a throwback, an amphibian that thinks its a fish.  Cynodonts might be better understood as a devolution from monotremes.

The evolutionary relationships of the fossil [for the eutriconodont mammal Yanoconodon] suggest that either the "modern" middle ear evolved twice, independently or that it evolved and was then lost [i.e devolution] in at least one ancient lineage.

Or maybe some are mosaics because phylic boundaries, as we know them, are a trend, not a rule (otherwise, how would either mosaics or missing links be possible?).  Now, Archaeopteryx is no longer the missing link it pretended to be in our textbooks.  It is, for now, more like the coelocanth or alligator.  It is a mosaic that represents the diversity of archosaurs (though not as radically as ceratopsids which had apparently "recapitulated" and rediscovered their four-footed roots.  And if quadruped representatives of the archosaurs were there all along, like the elusive coelocanths, they simply flew under the paleontological radar and avoided the fossil record as many group likely have.

All this puts me in mind of this structuralist (and narrative-agnostic) formulation by Dr. Richard Sternberg:

The approach I am taking to this problem is a variant of structural realism, by which I mean that biological phenomena are manifestations of logico-mathematical structures. This perspective is orthogonal to the origins debate, if you will, because all historical actualities are understood to be space-time instances of pre-existing non-temporal possibilities. Within this context one can accept all that is empirically valid in evolutionary biology, while not axiomatically dismissing the position that structures as well as their “real” instantiations have an intelligent cause.