Problems with Naturalism: The Fossil Record

[Josh]

Originally posted 7/1/06:

Next, let's consider the fossil record. What does it really tell us?

Here's a link exploring the issue from the ID perspective:

The Fossil Record
www.ideacenter.org/contentmgr/showdetails.php/id/1232


In summary, though the article states it much more fully, "gradualism" (the idea dominating early Darwinian theory) predicted the life's evolution would occur in tiny step over long periods of time, with natural selection being the key mechanism. Indeed, natural selection, if capable of macro-evolutionary change, would have to have occured slowly with almost imperceptible changes in real time observation.

Because of this, the logical conclusion of early evolutionary theorists was that transitional forms should be evident in the fossil record- that it was only a matter of time until plenty of examples showed up under the paleontologist's pick.

Yet, apart from a very few (and well debated) examples, transitional forms have not been found in the fossil record. This realization lead to the demise of gradualism/natural selection as the primary evolutionary model and spurred new theories attempting to explain the tremendous gaps and sudden appearances of new species in the fossil record followed by stasis, such as Gould and Eldredge's Punctuated Equilibrium. The problem with such theories is that they are not demonstrable mechanisms, but
are tenuous hypotheses (that will be the next subject thread).

What we do see in the fossil record is species appearing, running their course, and going extinct. We see species that make sense for their environment appearing and surviving for periods of time. Yes, we do see a general increase in time of the complexity of species (although we see lots of mass extinction events and simpler life forms replacing more advanced ones many times).

How exactly, Hume, do you think the fossil record is manifest evidence for the occurance of biological evolution? If so, how?

PS: Perhaps you'll appreciate the quote-mining caveat in the article, Hume. Heh heh.

[hume]

Originally posted 7/2/06:

Quote mining is indeed a dangerous activity. As we all know, miners are subject to black lung and accidents.

[hume]

Originally posted 7/8/06

Stuff to keep in mind when discussing fossils:

- most often, fossilized remains are partial. Sometimes what's found is literally a jumbled up mess of fragments from different individuals.

- generally, the part of an animal best-preserved is its bone structure & teeth. Look at this picture of Tiktaalik

evolution.berkeley.edu/evolibrary/news/060501_tiktaalik

(an unusually high-quality fossil as fossils go) and note how the reconstructed model of the beast adds in "fleshy parts" (which are of course conjectures, although they look reasonable; we'll simply never know *exactly* what a Tiktaalik looked like). In rare cases more of an animal is preserved (e.g., Archeopteryx was found in an unusual fine-grained limestone deposit capable of making "casts" of the animal's feathers). This also entails that soft-bodied creatures are rarely preserved even partially.

- no skeleton will be permanently preserved if left exposed to the elements -- rain and wind-driven particles will erode it. Thus most fossils are found embedded in sedimentary rock layers that formed in bodies of water as silt, sand & small rocks sank through the water and settled on the bottom. This explains why most fossils are of marine creatures. For a land animal to be preserved requires quite a lot of luck. It helps to drown or be forcibly buried in disasters like flash floods, landslides, or volcanic ash events -- or, famously, the occasional tar pit accident. Erosion is still a danger even if everything goes well from the outset -- if the rock layer containing that fossil finally reaches the earth's surface and coughs up the fossil, it will be slowly worn away like any other rock.

- Evolutionists aren't bothered that descendants occasionally appear earlier in the fossil record than their likely "parent" species. (If this was true of most cases, that *would* be problematic). When one species branches off from another, the original species doesn't thereby die out -- it normally continues alongside the new species, may even turn out to be more successful in the long run (the parent may "outlive" the child). So you expect an overlap period in which the two species co-exist. If both species are lucky
enough to leave any fossil record at all, the only remnants may happen to be from late in the existence of the parent species, while the child species generates fossils early on; this would cause the older species to appear somewhat later in the fossil record. All else being equal, we do expect to find fossil order matching the order of speciation -- but in many cases all else isn't equal.

- Fossilization is something of a numbers game: a species that's highly successful over a long period will have many individual members and therefore many chances at fossilization.

- One of the likeliest scenarios for the arisal of a new species is when a small group of individuals becomes geographically isolated from the larger population group. This predicted speciation within a small group, in a small area, is coincidentally a scenario that works against the odds of fossil preservation of the individuals in that important "transitional" group, making the preservation of intermediate forms all the more unlikely.

Given all this, what can we reasonably demand from the fossil record? Due to the nature of decay and preservation, we can be confident that even if we find, dig up and reconstruct every remaining fossil and fragment, we'll be left with a signficantly incomplete picture of life's history. The question isn't, "Can you use the fossil record to give absolute proof for evolution?" The question is, "Does the available evidence substantially support this theory?" The answer appears to be, yes.

[Josh]

Originally posted 7/8/06:

Just a small clarification, which I'm sure was assumed in this post: fossils aren't technically 'remains' of animals (as in their bones, etc..), but mineral deposits that have filled in the spaces left by their bones. So, when we're looking at fossils, we're not looking at bones, but bone-shaped mineral deposits.

[hume]

Originally posted 7/8/06:

In another post, you wrote, "I'll grant you an overall shift from 'rudimentary' to 'complex'" in the fossil record. In order for the fossil record to support evolution, I'll need you to grant more than that. If all life shares a common ancestor, we know to expect an overall pattern of fewer and simpler creatures the further we go back in time; but looking at the millions of species around us, it's obvious we'll need to find more than that. If shark and seaweed, bear and bacteria are related, the relation must be very distant; while brown bear and Kodiak or tiger shark and great white must have very close common ancestors.

Thus evolution demands that the fossil record reveal a huge, elaborate branching sequence of species, all connected to each other in logical ways. This includes the expectation that if a distinct morphological feature is found among a wide range of species today (especially across higher taxa), that feature should be found quite far into the past; whereas features that are common only to a few species should be found to be recent. For instance, very few known species have hands with opposable thumbs; whereas, a huge range of species are found to have the ability to convert sunlight to usable energy via photosynthesis. Evolutionists would conclude that you'd have to go back pretty far in earth's history to find the first example of a creature that used photosynthesis, because so many subsequent lines of speciation came to include that feature. Opposable thumbs should not be found anywhere near that long ago; otherwise we'd expect to see such a useful feature spread out among many more creatures by now.

To summarize, evolution requires:

1. a fossil record showing a "tree of life" that logically connects parent to child species & groups

2. a strong correlation between a morphological feature's commonality across species (and especially higher taxa) and that feature's age according to the fossil record

3. some clear examples of intermediate forms linking different species and groups. This does NOT mean evolution requires the discovery of fossil examples of every proposed common ancestor, as the fossil record is significantly and permanently incomplete.

So is this what we find? According to paleontologists, yep.

As for specific examples and discussion, I'll leave 1 and 2 alone -- they're large & complex subjects, but deeply documented in paleontology. For some reason it's number 3 that gets all the attention anyway, so I'll do some kind of a post addressing that ...

[hume]

What's an intermediate form? Evolutionists don't predict or expect an intermediate form to be some sort of mathematical half-way point between two other creatures. No one's looking for an ancient bird ancestor that looks like a reptile with 1/2 of a usable wing, 1/2 of a working beak, and a body 1/2-covered in feathers. What they're looking for is along the lines of what's been discovered so far (to stay with the bird example):

Sinosauropteryx, the earliest yet-discovered feathered dinosaur, was found in China in the 1990s. It's not a bird-like creature except that its body is covered in simple feathers (described as similar to the down of a modern-day kiwi bird).

Microraptor (another Chinese find; 6 almost complete skeletons have been uncovered to date) is classified as a dinosaur, though it has thick feathers, a fan-like tail, even a feathered crest on the head. It has 2 sets of wings (front legs and back); neither set of wings is considered sufficient on its own to support the animal in flight, but by using all 4 legs, Microraptor could have flown, or perhaps simply glided.

Archaeopteryx, one of the famous "missing link" discoveries, is an ancient bird with many dinosaurian features (for instance, it has teeth, it lacks a bill, its brain-case is more reptilian than avian in shape, and unlike birds it lacks a reversed toe for perching; here's an extensive list).

Note that intermediate forms are not the same as common ancestors. Most species are not common ancestors; they either mark one step along a single branch in the tree of life or they represent an evolutionary "dead end." Only at the branching points (nodes) do we expect a common ancestor.

This is not to suggest a line of descent that goes dinosaur to Sinosauropteryx to Microraptor to Archaeopteryx to modern bird. These particular animals may not be in any sort of direct lineal relationship to each other. Any given "intermediate" species could turn out to be *the* common ancestor of the creatures it resembles; more likely it's an evolutionary descendant -- or precursor -- of that actual common ancestor. This is why species living today (a classic being the Platypus) can be called "intermediate forms," even though they are distant descendants of the common ancestor they resemble.

[hume]

Some land mammal-to-whale transitionals (there are others ... I'm just tossing out a few examples to counter this old anti-evolutionist favorite):

Pakicetus - likely candidate for the land mammal ancestor of modern whales. Found in modern-day Pakistan (thus the name). Teeth adapted to hunting fish, skull features like the "anvil" of the middle ear that show unusual features found in whales but not most mammals. Lived in a "low-lying wet terrestrial environment"

Ambulocetus - (name means, "walking whale"): discovered in Pakistan in 1992 (intriguingly, in the same area as Pakicetus, just in more recent rock layers), it's an amphibious mammal, described as a "furry crocodile," better adapted to swimming than walking, but capable of both. It has ear and teeth structures in common with whales.

Basilosaurus - here's an entertaining one. It's the Alabama "state fossil;" apparently the remains were so common at one time that people there used them as furniture. The first truly gigantic whale, it was so long & relatively skinny that it's been called "the closest a whale has ever come to being a snake." Some people suggest the Loch Ness monster might be one of these It's a huge sea mammal, but by modern whale standards rather primitive (no blowhole, incapable of deep diving, no echolocation organ etc.)

[Josh]

7/8/06:

While we might rationally strongly infer a relationship between the wolf-like Pakicetus and whales, it is especially in this commonly touted example of the evolutionary mechanism that we see the lack of any explanatory model.

This transition from Pakicetus to whale took place in only 15 million years. That's really quick- arguably ruling out gradualistic, natural selection alone. What if we turn to a more radical mechanism, perhaps involving beneficial mutation?

There are demonstrable conditions for observable micro-evolution (based on random, beneficial mutations): extremely high populations, extremely short gestation periods, and small body size/complexity (this is why we can only currently observe micro-evolution in microscopic organisms and insects). If these factors are arguably, a must for micro-evolution, then, we might say they're probably pretty darn important for quick macro-evolution. It's interesting, though, that whales are the worst possible candidates along these lines. And after whales start appearing in the fossil record, they go extinct numerous times, only to keep re-appearing, against the odds, it would seem.

The same is true for another evolutionary poster-child: the horse, which is also one of the most fragile animals ever.

[hume]

7/17/06:

"There are demonstrable conditions for observable micro-evolution (based on random, beneficial mutations): extremely high populations, extremely short gestation periods, and small body size/complexity (this is why we can only currently observe micro-evolution in microscopic organisms and insects)."

The key term here being, "observable." This is rather like saying, "there's insufficient evidence that shield volcanoes are formed by lava flows, because the only observable lava flows have been unable to create 10,000-foot high piles of material." If we lived 80,000 years instead of 80, we'd have a different notion of "observable change."


"It's interesting, though, that whales are the worst possible candidates along these lines."

You've got a point there, relatively speaking. The basic reason that a blue whale takes nearly a year to produce one child is that the kid weighs 2.5 freaking tons! Naturally it takes awhile to grow a creature of that size, starting from a couple of single cells. However, there's no reason to assume that Pakicetus or other early-lineage forms had this slow a gestation period, as they weren't enormous creatures. And in any case, as I'll try to show below, the years really stack up, even with a slow gestation rate.


"This transition from Pakicetus to whale took place in only 15 million years. That's really quick- arguably ruling out gradualistic, natural selection alone."

I'll make a weak, back-of-the-envelope attempt to calculate mutations per generation (please note that I DO NOT know what I'm talking about here. This is highly speculative; the intent is only to show that, largely due to the time scales involved -- even in the case of a "swiftly evolving creature" like the whale -- evolution by natural selection is at least a plausible mechanism.)


TOTAL MUTATIONS PER GENERATION:

ASSUME avg. population over this time period of 25000 critters (sometimes many more, often many fewer; but on avg.)
ASSUME that at any given time, 30% of pop. is both of child-bearing age and successfully "hooked up"
= 3750 productive couples

ASSUME gestation period allowing for 1 new generation per year (faster than a blue whale, but lots slower than a fruit fly)
ASSUME litters of 5 children
= each year, 18750 offspring produced

ASSUME 1 mutation per zygote (that's obviously an important assumption; it's in the ballpark of what's been discovered so far in studies of fruit flies etc., tho this is an issue currently under close investigation, partly motivated by the interest in accurately calibrating "molecular clocks" ... in short, this is probably wrong, but probably not way off)
= each year, 18750 new mutations enter the gene pool


WEEDING OUT THE MEANINGLESS / HARMFUL MUTATIONS:

Of course most of those mutations are neutral or harmful. so ... since only perhaps 5% of the typical mammalian genome consists of non-junk DNA (i.e., genes that encode proteins or have other clear functions), assume 95% of mutations are automatically irrelevant. That leaves 937 mutations.

The majority of mutations in protein-encoding genes are silent (they change the gene in a way that results in the identical amino acid production as before the change) -- assume this removes another 65% of remaining mutations, leaving 328. Now (arbitrarily) do sort of a bell curve with these -- figure 5% will be fatally damaging, 20% somewhat harmful, 50% will have such a slight effect that they amount to no change, 20% are somewhat beneficial, and 5% are of undoubted benefit to the creature's chances at survival and/or successful reproduction in the given environment. Toss out everything except that 5%, leaving 16 selectable, beneficial mutations.

Another factor to consider is accidents -- some critters won't make it regardless of their fantastic genes (caught in a landslide or zapped by aliens or something). Also in very rare cases, a highly beneficial mutation will occur in an individual that unfortunately also recieved one of those fatally damaging mutations. To account for these, cut the 16 down to 12.

Thus, in each generation, of the nearly 19000 individuals born, only 12 will avoid random accidents while carrying new mutations that are substantially beneficial in the given environment. Note that this amounts to well under 1/10th of 1 percent of each generation; if you were wandering through the herd introducing yourself, you'd be unlikely to meet even one of these lucky fellows.


MEANINGFUL MUTATIONS STACKED UP OVER THE PERIOD OF EARLY WHALE EVOLUTION:

Let's take it as given that primitive whales developed from Pakicetus-like land mammals over a period of 15 million years. How many significant beneficial mutations, all told, would have occurred over that period of time? Given that every year brings about a new generation, we're dealing with 180 million mutations.

Even if my "estimate" above is considerably too high, we're dealing with millions of mutations. If even a fairly small minority of these mutations were actually selected for and thus successfully preserved and spread throughout the population, generation after generation, you can start to see how extensive adaptations could accumulate over time -- even over a "mere" 15 million years.

[hume]

7/2/0/06:
Hrmm. Re-reading this I see an embarassingly large error in the so-called calculation: I failed to account for the fact that the vast majority of mutations occur in somatic cells, won't be passed along to the next generation, and so are irrelevant to natural selection.

Adjusting: let's assume that 1 percent of all these mutations occur in germ-line cells. That cuts the estimated incidence of selectable beneficial mutations down to 1 for every 2.84 generations.

Still, this multiplies out to nearly 5.3 million events over 15 million years.

Again, this estimate could still be high (or low). But anything like this number of events is certainly lots of material for natural selection to work with.

[Josh]

7/27/06:

It's late. Forgive this thougth if it turns out to not make any sense, but:

another factor that would undercut the numbers on your calculation: these 'positive' mutations don't have an end in mind. So, let's use the wolf to whale scenario. For every positive mutation that would lend itself toward slowly changing this creature into a sea-dwelling mammal, surely there were several other mutations leading it in other directions (a more successful land mammal, a smaller but faster mammal who can avoid predators a little better, etc..) So, if in the end, the sea won out, it would have been based on far fewer mutations than you've put out there.

Actually, there are probably several more factors severely cutting the number we're not thinking of. I've been searching to no avail for a specific 'calculation' of sorts (similar to your example) that I read which attempts to show how radically improbable the wolf to whale change is (btw, I'm saying wolf to whale for economy, knowing full well we're not really talking about modern day wolves and whales). I thought it was in Darwin on Trial (I just read it a couple weeks ago) and I swear it has dissappeared.

Must have been a miraculous intervention, since I can't explain it....

Anyway, we are somewhat in a league not our own here... It's good to think in more realistic terms, though, so thanks for the ballparks.

I think it's enough to know whether (and how many) experts think these kinds of things are problems for evolution or not to gauge whether this is a potential issue or not.

Apparently stuff like this was a shock to at least older evolutionary paradigms (gradualism). I haven't read enough to know if this kind of stuff has some scientists scratching their heads or if new answers have provided more confidence, like yours.

[hume]

Not to keep beating on long-dead whales, but here are a few further considerations (note that these are relevant to the evolution of many creatures, not just whales):


1. "Mutations" = "changes." For a long time, I thought "genetic mutations" referred just to radiation damage (some kind of Cold War obsession I suppose). It turns out this effect, while real, represents only a small proportion of what biologists mean by mutation effects. Chemicals & viruses can also alter genetic material, but by far the most important mutagens are errors in the cell-division process. To change a gene all that's needed is an alteration in the base-pair *sequence* -- it's not even necessary to change individual base pairs, if you have a mechanism for shuffling them around, deleting, inserting, or copying this or that series of base pairs within a gene or between genes. There are several such mechanisms, including Transposons, insertions and deletions, (which can result in frameshift mutations), duplications, chromosomal inversions, and exon shuffling. These various mutagenic effects can interact, e.g. when chemically or radioactively induced point damage causes a cell to attempt DNA repair procedures, some of which are inherently error-prone.


2. Strong selection pressure. Early whale evolution ocurred during the early stages of the "Age of Mammals" (replacing the "Age of Dinosaurs"), following the mass extinction event of 65 million y.a. In short, the Great Dying "emptied the oceans" (greater effect there than on land), especially of large predators (there were still plenty of lil' fishies to eat tho). Mammals were strictly land animals -- but the "biggest opportunity" at that time, especially for a good-sized animal, was in water, so any mammal that could attain an aquatic lifestyle would probably be hugely successful. This created an unusually strong "selection pressure" in favor of aquatic adaptations. Any mutation that provided even a fairly small improvement to (say) swimming ability or breathing efficiency underwater would be especially likely to persist. (By contrast, modern-day humans are probably under virtually no selection pressure at all. We've created societies and technologies, starting w/ agriculture and culminating with medicine, that shield us from the selection effects of all but the most extreme genetic deficiencies, while most beneficial genetic changes are unlikely to make an individual more "reproductively successful" in any measurable degree.)

(It's interesting to consider the logic of major extinction events being followed by periods of unusually strong selection pressure among the surviving creatures -- all those niches to fill! This helps explain, from an evolutionary perspective, the flourishing of new taxa after an event like the Great Dying of the dinosaurs & co.)

[hume]

Bringing us right up to "breaking news" science, the Tiktaalik species was discovered in early 2006 (three substantial fossils were found); it's a 4- to 9-foot fish with notable tetrapod characteristics (kind of a fishy crocodile), and has been offered as a significant transitional form linking fish to land animals.

This diagram is reprinted in a recent "sceptical" treatment of Tiktaalik on the IDEA website (in fact it's currently the website's lead article). Note the rudimentary beginnings of finger bones in Tik's "fin."

Now scroll to bottom of the main article (I'm skipping over the top section discussing a recent hominid discovery, that's too big a topic) to find the writer stating, "if anything, the fin of Panderichthys appears closer to a true tetrapod limb than does the fin of Tiktaalik." Huh? What is he smoking? Panderichthys, in comparison, has a club for a hand. Just above the diagram ("Figure 4") he writes, "I think that Figure 4 ... says it all." It certainly says alot; but what's more important is what it (innocently), and the author (conveniently), leaves out.

The article fails to note most of the significant features of Tiktaalik that identify it as a transitional form -- in particular, according to the researchers who discovered it, the creature could bend its elbow and flex its wrist, so it was probably capable of performing "fish pushups." Since its shoulders couldn't rotate, it wasn't able to walk, tho it could have dragged or (in water) pushed itself along. Tiktaalik is also notable for actually having a neck (basically, fish don't have necks; since they swim by undulating their entire bodies horizontally, it works better to have the head and shoulders fused together), and the neck structure is flexible enough to suggest that it could have lifted its head above water (the ability to do pushups would have been useful in that case). This makes sense in the context of its environment: the shallow stream systems and marsh areas of a river delta. Tiktaalik's rib cage is strong enough to support the animal's weight outside of water, raising the distinct possibility that it was capable of brief trips over land (e.g., to cross a narrow sandbar separating one stream from another). Also notable is the creature's crocodile-like head (evolutionarily speaking, fish gave rise to amphibians, then reptiles).

To summarize, Tiktaalik has ribs, head and neck reminiscent of a land-based tetrapod and the scales, gills and jaw of a fish. Its fins are clearly not tetrapod feet, yet they bear an intriguing resemblance, especially considering the limb's ability to bend and flex. Scrutinize the IDEA article in vain for reference to any of this information, other than a brief comment denying that the creature's fingers resemble a tetrapod's.

I don't want to accuse the article's writer of laziness or dishonesty -- more likely he's so convinced that evolution is a hoax that he only looks at evidence long enough to find some way to quickly dismiss it.

In any case, the article's main point is not to discuss Tiktaalik but rather to "quote mine" for apparent cases of paleontologists caught in "embarrassing admissions" about the paucity of fossil evidence for the fish to tetrapod transition. This seems to be a trademark practice of anti-evolutionists: don't deal clearly, directly and fully with the offered evidence. Instead, rely on your audience's presumed ignorance of the actual details to simply ignore that evidence, then focus on the hunt for misleading bits and pieces of a decades-long discussion in order to present the participants in a bad light.

[Josh]

Is it your impression that the ID movement as a whole opposed common ancestry? (don't have time to look into it).

It just seems strange to me- the case against naturalism would only benefit from accepting it, or at least conceding it. And it seems all the more strange because the ID movement isn't as crimped by specific evangelical paradigms as most 'creationist' organizations.

[hume]

Anti-evolutionists represent quite a range of positions -- for instance, I don't know if Philip Johnson is considered an ID proponent or simply an evolution opponent. Then you've got the Old Earth / Young Earth distinctions. They're all a little mixed together in my mind unfortunately -- partly because at the time I was reading lots of that stuff, I didn't understand evolution enough to be able to recognize a clear distinction between mechanism and outcome. Maybe the ID movement does tend to accept common ancestry, or maybe they're just sort of coy about discussing it? (Surely they're aware that if they made a big deal out of arguing in favor of TCA, this would turn off a good portion of their readership.)

[hume]

I've posted a few samples of what seem to me like legitimate "transitional fossils." Being no expert in any of this stuff, I'm not going to add to the list -- if you're interested in more (and there are many more), you might start here:


en.wikipedia.org/wiki/List_of_transitional_fossils

[Josh]

Hume,

You've done a stellar (and beneficial) job compiling a strong apologetic for common ancestry. Heck, I'm more convinced than ever. Common Ancestry in just the mind of God alone doesn't explain the evidence nearly as well as the idea that all animals are related in some more tangible way.

However, I brought up issues with fossil record that primarily challenge the proposed mechanisms of naturalistic evolution, not common ancestry per se (back to our dilineation of definitions).

For instance, to paraphrase myself, the fossil record challenges gradualism/ natural selection alone as a viable explanation for all the speciation in the fossil record.

For instance, if, as originally expected, species only changed incredibly slowly (100 million years for a perceptible change), then I think we should expect to find much more evidence of transitional forms. That's why I think Darwin himself thought we should be finding lots of them.

But if the changes came way sooner, such as in the Cambrian Explosion in which ALL animal phyla appear within just 6-10 million years, then we might understand why the fossil record doesn't have much in the way of transitional forms. But, since the fossil record shows that species appear quite suddenly, then the old mechanism is shown to be inadequate (thus, punctuated equilibrium and other hypotheses which are really just that- hypotheses). BTW, I'm not saying natural selection isn't occuring, I"m saying it can't explain much of the evolution of life. Almost all the appearances of species in the fossil record are in sudden radiations of new species. I don't see anyone doubting how the fossil record appears in regard to this. I see them trying to find a naturalistic model to explain sudden speciation.

So what I'm saying is that the fossil record may not be a challenge to common ancestry (in fact it appears to bolster the idea), but it does appear to me to be a challenge to any workable naturalistic evolutionary
mechanism. See the quote by Johnson above:
Philip Johnson: a Critique of Mechanism

What do you think about that?