Monday, August 29, 2005

A post from ID the Future

Here is a post from ID the Future, about Daniel Dennett's op-ed piece. I would like to hear critical responses from opponents of ID. Can Dennett be defended?


http://www.idthefuture.com/index.php?p=700&more=1&c=1&tb=1&pb=1#more700

18 comments:

Jason Pratt said...

Sorry for lack of comments--busy-ness, plus head cold...

The original op-ed text can be found here:

http://www.nytimes.com/2005/08/28/opinion/28dennett.html?pagewanted=print&oref=login

Probably better to discuss that, than to work at secondhand.

(Good post below, though, Victor... {s})

Victor Reppert said...

The idthefuture article links to the Dennett piece, right?

Steven Carr said...

'Yet natural selection has never been demonstrated to produce even one new species, much less new organs and body plans--the "ingenious designs" to which Dennett refers. Natural selection, like artificial selection, produces only minor changes in existing species.'

I guess that means that the Ebola virus was intelligently designed.

Where do these new diseases like HIV come from, if natural selection cannot produce new lifeforms?

Blue Devil Knight said...

I agree with Ahab, but as for one of the criticisms of Dennett, I certainly agree that it is almost always fishy to claim that a structure is an example of poor "design". There is a whole list of such examples at talk.origins (e.g., though I find such lists uncompelling because evolution is usually more clever than we are (Orgel's second rule). While I am sure there do exist inefficient and foolish structures from an engineering perspective (perhaps the appendix), we don't understand enough physiology and ontogeny to say something is doing nothing.

Isn't it kind of ironic to have Dennett essentially arguing for the existence of spandrels? Hee hee.

Dennett's piece, overall, was an excellent polemic.

Blue Devil Knight said...

An excellent case for useless or poorly "designed" phenotypic traits can be found here. It has almost convinced me my original skepticism was wrong.

Jason Pratt said...

Yes, the idthefuture article links back to the Dennett op-ed. Just figured its link should be more up-fronted, if that's what's going to be primarily discussed. {shrug}

BDK: I'm inherently suspicious of polemic even from my own side of the aisle. Dennett isn't helping with his routinely polemical contributions.

Nor am I saying this to discount the good points he does occasionally make, as in the section Ahab quoted at length. (Though I'm a bit suspicious about Dennett's implication that Dembski didn't go on to explain _why_ he thought Schneider's response was only hair-splitting. Anyone have access to that exchange, for checking? Dembski's reported phraseology looks on the face of it like he's about to _correct_ the common outsider's perception that it's hair-splitting, i.e. it _isn't_ really hair-splitting. Dembski would hardly consider _himself_ to be an outsider on the subject, after all.)

Speaking of mere polemic... {wry g}

Ahab asked: {{What is this magical process, that ID seems to assume, which prevents all these minor changes from producing a new species?}}

Probably the same magical process you yourself emphasized, in trying to defend the apparent lack of transitional species. (Remember, back when you insisted we put on our scientist caps and use our imaginations to come up with all sorts of reasons why near-relative new species would have difficulty interbreeding with established populations? {g})

I don't consider handfuls of mutants to constitute a viable new "species". Neither did Sir Ronald Fisher or George Gaylord Simpson (who were hardly basing their conclusions on "the mention of kinds in the Genesis account of creation [as] a leftover from the earlier scientific creationism that got tossed out of the schools for being overtly religious.")

Granted, these fellows may be somewhat out of date. {shrug} But if leading evolutionists (including ones who helped revise the basis theory to NDT) have insisted that natural factors would make it exceedingly difficult for mutations to survive to become viable species (thus explaining, in passing, why we _don't_ find the numbers of transitional species the theory would otherwise seem to predict as a result of the process--an explanation that would appear to still have popular credence, else you wouldn't have used it yourself so forcefully), then I don't think it's fair to simply assume IDers are merely pulling Genesis out of their butts as their hidden justifications. They might actually be paying attention to implications (even if rather overstating their case).

In passing, I'll also mention (again) that the sort of 'demonstration' Mr. Witt may have had in mind, can't really be expected to _ever_ happen in support of the theory. (Heck, even Richard Dawkins agrees with _that_--or at least he did back in 1996. {shrug})

I'm saying this in order to call for some fairness toward NDT proponents, too: it isn't right to ask for a kind of demonstration that either cannot be expected (probablistically) to happen under observation, or else would happen under conditions anathema to the proposal (rendering the result invalid with respect to the basis theory.)

That does _NOT_ mean the basis theory is altogether undemonstrable in principle. It just means the types of demonstration have to be carefully made, with an eye toward the restrictions involved (including in the conclusions and their implications.)

Jason Pratt said...

Opps! Memory error. _Very_ sincerely sorry. (Sorry to BDK, too, come to think of it...)

I'm a layman in it myself, which is why I try my best to agree with the specialists wherever I can find leeway _to_ agree with them. That even includes Dennett (I'm certainly willing to consider him a specialist.) My most significant disagreements with him stem from hearty agreements with him, actually. (Same is true about my various disagreements with some theologians.) It depresses me to _have_ to disagree; I _like_ scientists. (And love one, too. {s}) And I get more depressed when I have a head-cold. (and when it's her birthday tomorrow, and I can't be with her... {sigh})

{{If common descent is true, then it seems to me that the concept of 'species' is perhaps more fluid or hard to pin down than might commonly be supposed.}}

I agree, to some significant extent that would have to be true. (Dawkins has made the same point rather well, as I recall.)

{{As far as transitional fossils go: can't we in one sense look at all fossils as being transitional?}}

Sure. Assuming (universal) common descent is true, we'd have to look at them that way.

But that wasn't what I was talking about (um, with BDK originally I guess. {g}) Gradualistic NDT necessarily implies that we should have scads of interbreedible transitional species around us right _now_. But we don't.

BDK rightly pointed out, in reply, that technically they may in fact be here, but that we'd necessarily have a hard time figuring out if there were, if they had significant breeding problems even between near-related species; which problems can be easily imagined.

I agreed with that, and pointed out that we don't even have to go outside the species for such examples (wolves technically could breed with dachsunds, but as a practical matter it isn't going to happen.)

The problem (somewhat tautologically {s}), is that solving one problem in a theory, by ramping up a related and even more basic problem in the theory, doesn't really make the theory more feasible. Those transitional species have to be able to interbreed _very_ successfully with the larger population pool, for gradualistic NDT to have any chance to work as a significant contributor to the development process (much moreso as the _only_ contributor).

So even if we did experiments confirming that technically squids and octopi (for instance) or wolves and foxes or chimps and humans could successfully breed (at the genetic level)--and I honestly have no idea that this _hasn't_ been confirmed; although I suspect I'd've heard about it by now if it had--the fact is that we still don't find close relations being practically able to breed. (Our own domestic breeding programs tend to obscure this; they can't be used as prevalent _natural_ examples, though otherwise they do have some real importance in helping demonstrate some of the functionality of the biological breeding system.)

The normal pattern should be: Species A and Species X are non-interbreedible with each other, but interbreedible with all species in between (and all in between with all others in the set, including A and X, minus any species which have since gone extinct, of course). Humans, for instance, should be a rare exception (maybe because we killed off all other capable interbreeders, or some combination of similar factors.) But we aren't. Our situation is the common rule. The close interbreeder species are the exception.

And in the cases we _do_ know, the results (though admittedly usually quite healthy otherwise) are (almost?) always infertile. (As BDK also pointed out. {s})

Problems like this shouldn't be blinked aside. But it's easy to mistake this problem with some proposal that isn't really (or as much of) a problem; and while those other less problematic problems (or pseudo-problems) _are_ more easily dealt with, their resemblance to _this_ problem makes it more difficult to clearly address.


{{[B]y what principle does ID make the claim that micro-evolutionary variations can't result in new species coming into existence?}}

I don't think they claim any principle strictly prevents this (not usually, anyway). Their problems with micro-ev are much the same as their problems with macro-ev: it's the technical, and in some cases the practical, feasibility that seems very lacking. (Philosophical criticisms are in another boat, and not strictly ID's balliwick.)

It's difficult to overstate how much the ID crew _agrees_ with the NDT proponents, about which parts of the biological mechanism work and to what extent. This is important to keep in mind, but advocates on either side tend to obscure this. Moreso on the NDT side, I've found; but there's some blame to be spread on the other side, too. Still, the ID crew has a definite stake _in agreeing_ as much as possible with the NDT about the technical aspects, and even about the biological principles involved.

Mainly the disagreements seem to be focused on how effectively different elements of the mechanism contribute to the overall picture; and what the total result of accounting for the piece-fitting (micro and/or macro-level) entails.

Some IDers have a bad habit of trying to jump from a tally of genuine problems with the theory, to God as a conclusion. (I have similar criticisms about most, if not all, uses of cosmological arguments that I've seen.) And this might truly stem from religious overeagerness (plus a sense of responsibility on saving people from what amounts to gnostic damnation scenarios); but I think some of it is simply a reaction to all-or-nothing hyper-positive bullyragging by some NDT proponents. It's simply irresponsible to hammer home an acceptance of a theory as being settled beyond all reasonable doubt, when there _are_ genuine controversies about the theory, even among its own proponents. It smacks of fundamentalism in the worst sense: ALL GOD'S WORD OR NOT AT ALL GOD'S WORD! THAT SETTLES IT! (THOOM!)

(And, of course, one can trace reactions to just such theological bullyragging among a significant number of NDT proponents who are doing much the same thing in much the same language. "For _this_ state of affairs," in the spirit of Dorothy Sayers, "I'm inclined to blame the orthodox." {s!})


Jason

Blue Devil Knight said...

Jason has the following argument:

1. Gradualism implies that we should presently see (lots and lots) of "transitional forms" interbreeding.
2. Transitional forms interbreeding is not widespread.
3. Therefore, there is a serious problem with Darwinian gradualism.


In favor of brevity I didn't do the above justice in my first response, so I'll try to be a little more explicit here. This is a fun argument, but not right because the second premise is wrong. In fact, the prediction from premise 1 is probably being confirmed in your backyard right now by a bunch of insects mating.

Below, I have given the main shape a proper response would take. I have ignored lots of details, some which I expect an intelligent person can supply, and others that are hopefully irrelevant such as the existence of different sexes. A more technical and thorough discussion could be found in any introductory population genetics text, required in most evolutionary biology curricula (unless you live in Kansas).

To repeat, premise 2 above is false. I would restate its falsification as, "Potential transitional forms are interbreeding every time two individuals with different genetic fitnesses interbreed." This happens constantly in nature, so premise 2 is false.

The above merely restates the falsity of premise 2 without much justification. A slightly more detailed justification follows.

Let's consider first an easy case for me, that of allopatric speciation, caused by geographical separation and subsequent genetic divergence (e.g., the different but closely related species of fruit flies on the different Hawaiian islands). We wouldn't expect much interbreeding of intermediate species in such cases, because of physical separation. Let's call the two separated populations X1(t) and X2(t) (in the literature they are referred to as 'allopats'). As t increases, X1 and X2 will be less and less likely to interbreed (if they happen to bump into one another). Once the probability of having viable offspring is zero, then X1 and X2 are different species. This example is easy for me, as I have constructed a geographic barrier that disallows interbreeding. I mention this case because it is so common in nature (as anyone who has moved from the east to west coast and looked at birds and rodents can see), and because it is a useful way to introduce notation I will use in what follows.

Now for the more relevant scenario. Imagine (over aeons) that one species becomes another species under a gradualist type of scenario (which is what Jason is attacking). That is, we have species X1 and species X2, and X2 is a descendent population of X1.

By assumption, members of X1 and X2 would not produce viable offspring. However, members of X1(t1) and X1(tn) can, so they are technically members of the same species [note: X1(tn) is the last population in which at least one member could breed with at least one member of X1(t1)].

Let's focus on X1(t1) and X1(t2), where the time difference between t1 and t2 isn't all that big (say, 5 years). Clearly X1(t1) and X1(t2) will share a whole lot more genotype than X1(t1) and X1(tn). That is, the t1 and t2 populations are likely very similar genetically, with minor variations in alleles and their frequencies (very much like allopats after very small separation in time). We can even assume that X1(t1) and X1(t2) are populations that contain no unique genes or alleles (alleles are gene variants such as blue/brown variants of 'eye color genes'). However, even if only the *frequency* of alleles is different in X1(t1) and X1(t2), then evolution has happened. Evolution, by definition, is simply genetic change over time. Natural selection, at least in part, works by selecting genetic variations that are already there in the pre-existing population.

If it turns out that X1(t2) leads to X2 via selection processes, then we have a case of a transitional form (X1(t2)) breeding with an ancestral form (X1(t1)). [Because we have assumed the t1 and t2 populations are pretty much the same genetically (minor allelic variations), then of course members of X1(t1) and X2(t2) can interbreed ].

One problem is that we can't really say interbreeding between genetically different indiduals is a case of breeding between "transitional forms" unless we know that they are leading to new species, so evolutionary biologists don't talk this way because they are rightly being cautious. Hence, the phrase 'potential transitional forms' in my bumper sticker above. Also, since we don't assume all evolution is speciation (most evolution is neutral genetically), they tend to not fetishize speciation at all, but just focus on breeding patterns among genetically diverse populations that are continuously evolving.

So, Jason's prediction is basically right and is verified all the time, though it has a strange focus on 'transitional forms' which puts the attention on speciation instead of evolution more generally. To accommodate this idiosyncracy, I stated things in terms of 'potential' transitional forms, though nobody in evo bio would talk this way.

Jason Pratt said...

BDK (and hereafter): {{In favor of brevity I didn't do the above justice in my first response}}

No kidding. {g} You basically told me, _of course_ we'd have a hard time detecting them, _because_ of all the problems they'd have in interbreeding (so put on our scientist caps and imagine the difficulties, etc.) Emphasizing the interbreeding difficulty as a defense, is _rather_ a different line of response than claiming that they are certainly widespread (i.e. that premise 2 is wrong).

But I'm good to go either way. If this new defense works better, great--I assure you, it won't bother me, and I'll be glad to agree with it.


In passing, I'm a little fuzzy on how the existence of different sexes is _irrelevant_ to any question of transitional species from _interbreeding_. On the contrary, it seems clear there wouldn't be any problem (or even pseudo-problem, if pseudo it is) without sexual reproduction. An asexual bacterium undergoes a type of mitosis during reproduction; the daughter cell could be any level of different, without there being the slightest problem interfacing with another bacterium for future generations. No meiosis, no interbreeding; no interbreeding, no even potential problem with interbreeding.

But maybe what you meant, is that it would be irrelevant to point out how intensely obviously relevant sexual reproduction is to this question. {shrug?}



{{"Potential transitional forms are interbreeding every time two individuals with different genetic fitnesses interbreed." This happens constantly in nature, so premise 2 is false.}}

Premise 2 wasn't about _potential_ transitional forms. It was about _actual_ transitional forms. Let's not confute the two. We ought to be in some position to say whether a form is transitional or not. (I do accept your point, though, about begging the question in calling it "transitional", per se; and so biologists would be properly cautious about calling a species transitional. I'll get to that later.)

For better clarity (because I'm bad about slipping in the description of this myself): a transitional _species_ is something of a misnomer anyway. In our typical taxonomies, we define speciation by whether there are definite populations which _do not_ mate between populations, even if given the opportunity and even if genetically they could. (If I recall correctly, genus would be the level at which they couldn't even genetically mate.)

To coin a term, the transitional population problem is with _sub_-species. More on this below.


{{We wouldn't expect much interbreeding of intermediate species in such cases, because of physical separation.}}

True; consequently, it's a non-problem for the claimed problem. {g} The populations are entirely and absolutely distinct.

Put another way, we wouldn't expect much interbreeding of intermediate species in such cases, because technically speaking there wouldn't _be_ any _intermediate_ species (between Island A and Island B population). The lack of intermediate species between the populations would, I agree, eliminate any problems regarding intermediate species relationships. {g}

What's happening in the _common_ population (including the common populations of the fruit flies), is what concerns us here--and where we should be focusing our attention, in addressing this problem. (Though I do readily admit that it's helpful to begin with this example first for purposes of _distinction_ from my actual contention.)

In the case of common populations (say, of African lions), I agree that the non-interbreedibility effect could still happen; and if it did, then it would still most likely be facilitated through geological separation of some sort (or so I'm willing to suppose, probably due to range issues.)

Form A in West Africa is never exposed to form C in East Africa, and so eventually the lion populations could become distinct enough that they _wouldn't_ breed (even though technically they could) if they _did_ happen to be brought together; while meanwhile (unlike the absolute separation of the Hawaiian fruit fly example) both A and C populations would be interbreeding with B form in Central Africa, with which they both maintain contact.

My problem, is with the evident regular lack of B.

It's always imaginable, and even plausible, that between one thing and another the various population forms could become non-interbreeders (and so become different species, then different genuses--sorry, I've forgotten what the plural is there {g}--and so on). It can even be clearly inferred from examples like the fruit flies.

But that isn't the question.

The question is whether we're actually finding this happening, on a sufficiently routine basis, within a _common_ population, with _mediant_ sub-species: how often are we finding a situation where Eastern Diamondback rattlesnakes and Western Diamondbacks reject mating with each other when given an opportunity (even if maybe they technically could), but they _are_ _both_ routinely mating with an established population of Central Diamondbacks (plus X number of other established forms, like say Timber Rattlers)?

That's the transitional situation you _aren't_ describing.

But it's what the theory _necessarily_ implies we should be finding.

So for example:

{{This example [of allotropic fruit flies] is easy for me, as I have constructed a geographic barrier that disallows interbreeding. I mention this case because it is so common in nature (as anyone who has moved from the east to west coast and looked at birds and rodents can see)}}

Sure. Which is why my problem _isn't_ with that. I want to know how often we're finding cases of Pennsylvania bluebirds and California bluebirds not breeding with one another (even when given an opportunity), but both of them breeding with established Arizona bluebird populations, and Texas bluebirds, and East Tennessee bluebirds, etc (when given opportunities).


I'm not just pulling this problem out of the blue--or out of ID apologist books, either. {g} One of the reasons Gould and Eldridge suggested punctuated equilibrium (which they would say was happening in the cases you're describing), was to try to account for the _evident lack of transitional species_ (or sub-species rather).

Their proposition looked very good at first, because anyone can see that it does happen that way--when (tautologically {g}) it _does_ happen. (It also helps ease some of the problems Fisher et al have brought up regarding the improbability of mutation establishment in significantly large established populations.)

But punct-eq doesn't wash as being the main form of gradualistic NDT--and that's what they were trying to propose it as. They _had_ to try to propose it as the _main_ process (they thought) because the lack of existant transitional populations is _so_ evident. (Plus they were trying to account for some fossil distribution evidence, as I recall; which punct-eq _is_ probably good for, up to a point. Punct-eq is still grad-NDT, btw.)


{{However, even if only the *frequency* of alleles is different in X1(t1) and X1(t2), then evolution has happened. Evolution, by definition, is simply genetic change over time.}}

I know this is going to sound picky, but that's a definition of "evolution" which absolutely no one anywhere is denying happens (so far as I know anyway). But gradualistic NDT proponents are definitely _NOT_ simply saying that genes change in populations over time; otherwise there would be nothing to distinguish grad-NDT from Lamarckism or saltationism (much less from ID and various types of theistic evolutionism.)

Mere genetic change is _mutation_, not _evolution_. It would have to account for _development_ of _species_ (not mere change within species) even to be considered mutationistic evolution. (Which I expect you actually know, and so is somewhat beside the point; I just want to ensure we don't slip into thinking that this alone establishes any kind of viable evolutionary process per se.)


{{Natural selection, at least in part, works by selecting genetic variations that are already there in the pre-existing population.}}

I'm not sure what else natural selection is supposed to be doing (in everything I've ever read, the selection of genetic variations already in the pre-existing population is basically _the whole point_ to natural selection); but that's closer to an actually evolutionary process. (Which is why Darwin started with that.)

But even mutation + natural selection, as a combined process, isn't _successfully_ evolutionary until it results in development of species. (It might not even qualify as evolutionary if the process only accounted for development of complexity--though it would have to do that, too, at least under NDT.)


{{If it turns out that X1(t2) leads to X2 via selection processes, then we have a case of a transitional form (X1(t2)) breeding with an ancestral form (X1(t1)).}}

And breeding with X2, and all other transitional form populations in between.

What you haven't told me, is that we _commonly_ find these distinct interbreeding populations (barring X1 and X2 of course). I can imagine them being out in my backyard anyway, but that isn't the same thing. {g}

Are we finding plenty of cases where beetle Type-A is interbreeding with beetle Type-B which is interbreeding with beetle Type-C, though C and A _aren't_ interbreeding (even when they're given an opportunity)?

Beetle types A and B, or B and C, can be found easily enough in my backyard (I suppose? You do understand I'm talking about diverse established phenotypically common population pools here, not just individual beetles with even the slightest differences in genetic structure from each other, right?)

I've gotten the impression, even from proponents of NDT, that we _aren't_ finding this _other_ situation, though: that if I go out to my backyard, I _won't_ be finding distinct phenotypic grasshopper populations A and C not interbreeding while both are interbreeding with distinct phenotypic population type B. I might find A and B established phenotypic populations interbreeding (maybe??); and I'll (definitely!) find B and C populations _not_ interbreeding.

But that's it. From what I've understood, I _won't_ find A, AB, B (much less AB1...ABn, where each distinct population is interbreeding with each other and with A and B, given the opportunity, but A and B aren't.)

Yet gradualistic NDT (and most other theories of evolution) necessarily implies I ought to be finding situation A, AB1...ABn, B, with some significant frequency.


{{One problem is that we can't really say interbreeding between genetically different indiduals is a case of breeding between "transitional forms" unless we know that they are leading to new species}}

Point taken: we'd be begging the question if we said that existant shark population B was _transitional_ between existant populations A and C, if all that was happening was that A and B, and B and C populations were interbreeding.

Similarly, we really shouldn't be talking about _transitional_ forms in the fossil record (and drawing conclusions therefrom), unless:

a.) we can ascertain that Allosaur population B was breeding with non-interbreeding populations A and C;

and

b.) we can ascertain that (for instance) B descended from A, and C from B.

That's why it should be a lot easier to demonstrate transitional populations in _existant_ species, than in extinct ones.

Except, apparently it isn't. {s}


{{Also, since we don't assume all evolution is speciation (most evolution is neutral genetically), they tend to not fetishize speciation at all, but just focus on breeding patterns among genetically diverse populations that are continuously evolving.}}

That's true, too--except, we really shouldn't say populations are evolving unless they're actually speciating. That's begging the question again.

(And "neutral evolution", isn't evolution at all. Least, that's what Dawkins would say; and I think he actually makes sense there. {g} Evolution doesn't occur, one way or another, until a positive increase in selectivity value occurs. And even _that_ kind of evolution, though real, is no good, if it doesn't establish in a population.)


{{So, Jason's prediction is basically right and is verified all the time, though it has a strange focus on 'transitional forms' which puts the attention on speciation instead of evolution more generally.}}

It's more than a little strange to hear a proponent of a theory of species development, say that _I'm_ being strange to put the attention on speciation. Sure, there _are_ other things worth considering in the mechanism, too; but it's the speciation that's the point. The development of species was Darwin's point (and Lamarck's and others' before him); it was Huxley's point; it was Weismann's point; it was De Vries' point; it was the point of Stebbens, Dobzhansky, Mayr, (Julian) Huxley, Simpson, Jepsen, Fisher and Wright (when they revised Darwinism into the NDT synthesis); it was the point of Watson and Crick; it was the point of Gould and Eldridge; and it is absolutely _still_ the point of proponents of the theory (including popular apologists like Dennett and Dawkins).

Some of them focused their work on disparate pieces of the theory--which is the only feasible way to scientifically study it, and I'm not knocking that--but insofar as the sum total goes, they agreed, and still agree, with the title of Darwin's main work: _The Origin of Species_ (by Means of Natural Selection, etc.)


So, I think I'm in bounds pointing out an evident problem with the speciation that the theory (as it currently stands) would entail. Since I'm focusing on a particular issue, and that issue has something to do with the speciation process per se, then by tautology it shouldn't be strange if in doing so I focus on the speciation. And you _do_ agree with me, that the theory _does_ entail these transitional forms.

But you aren't showing me the _results_. The problem _isn't_ about x-number of individual forms slightly different from one another. The problem is that the transitional _populations_ should be there, and really they should be evident to us. We couldn't expect to be able to identify all existant transitional sub-species populations, between any given species A and X, but we ought to be able to identify some significant number of them, on a pretty regular basis.


Jason

Blue Devil Knight said...

I'm not going to generate another long post on this topic. My post contains enough information for someone to make up their mind. As I mentioned in my post, my previous response wasn't adequate, and I still take it that this one is. I'll just point out one problem with Jason's response:

I said:
{{Also, since we don't assume all evolution is speciation (most evolution is neutral genetically), they tend to not fetishize speciation at all, but just focus on breeding patterns among genetically diverse populations that are continuously evolving.}}

Jason replied:
That's true, too--except, we really shouldn't say populations are evolving unless they're actually speciating. That's begging the question again.

(And "neutral evolution", isn't evolution at all. Least, that's what Dawkins would say; and I think he actually makes sense there. {g} Evolution doesn't occur, one way or another, until a positive increase in selectivity value occurs.


In modern evolutionary biology evolution is synonymous with genetic change over time. Natural selection is but one mechanism of genetic change over time: it is one of many ways to get a change in allele frequencies. Speciation is a special case of evolution. Neutral changes are as well. They may not be examples of natural selection, but the species is not the genus.

If by 'transitional forms' you mean populations with gross phenotypic differences, then see my first post (the inadequate one, precicely because it focused on such cases). The present post is a better response, because it sets up the problem correctly in terms of genetic changes rather than "gross phenotypic changes." By the time "gross phenotypic changes" have occurred, you have an instance (in the terminology of my previous post) of X(t1) versus X(tn). For obvious reasons, that can be thought through based on my previous post, such populations would be much less apt to interbreed than X(t1) and X(t2).

Again, if you have a genuine scientific interest in this topic, I would suggest a population genetics text, such as this one.

Blue Devil Knight said...

I said:

For obvious reasons, that can be thought through based on my previous post, such populations would be much less apt to interbreed than X(t1) and X(t2).

So as to avoid Jason posting a book about this line (:P) let me spill the beans aboutt he obvious reason. One obvious reason is that tn-t1 >>> t2-t1, so there will be no individuals in common in the two populations. Also, by that time, after all the evolution that happened between x1(t1) and x1(tn) to lead to the brink of species-dom, the genetic differences would be so great that the probability of producing viable offspring would be very low, even if the two populations could come in contact (if, say they were two allopats).

Blue Devil Knight said...

Another great introductory population genetics text, at a slightly less technical level than the on I previously mentioned is here.

It is the standard undergraduate text. Any serious scientific criticisms of the type Jason makes would have to be familiar with this work. I have described a *TINY TINY* bit of the kinsd of material they discuss (in inadequately nontechnical terms). It is a fascinating and beautiful field.

Blue Devil Knight said...

Another book that is due out in January here. It promises to be excellent: Templeton, the author, wrote some amazing early criticisms of the mitochondrial Eve hypothesis (which my undergrad advisor was involved in advancing). Eve is basically corroborated nowadays (~15 years later), but Templeton helped kept everyone methodologically honest.

Jason Pratt said...

So, in (non-book-length) summary:

BDK has explained (and I agree, and always did agree, since that was part of my whole point) that the basis theory predicts we should have plenty of mediant interbreeding sub-populations within a common population (when there _is_ a common population--punctuated equilibrium situations like Hawaiian fruit flies being a special case where no such mediant interbreeding groups could be expected to occur anyway, consequently there wouldn't be any problem with an evident lack of them. Which I never claimed there would be, in such special cases, though I still appreciate the example.)

In fact, the number of mediant interbreeding sub-populations ought to be significant greater than the number of sub-populations on the fringes of the population continuum; which, though they interbreed very well with the nearer variants, don't interbreed well with the ones on the other side of the continuum (consequently are beginning to speciate in relation to those further along toward the opposite side.)

(Furthermore, for carefulness sake, we shouldn't call the mediant interbreeding sub-populations ABx 'transitional', even though we know they pretty much have to be, until we can ascertain which direction the transition is going.)


However (BDK has explained), by the time these mediant interbreeding sub-populations would otherwise become observable to us (though having long since become observable to natural selection), they're no longer interbreeding well at all.

Consequently, we can't in fact point reliably to any significant number of these situations, as such (A, ABx, B; or X1, X1(t1...n), X2); which is why, despite having textbooks handy where such examples might have been expected to be recorded, BDK didn't simply answer my contention (transitional forms should be common _but_ are rare instead) by providing a good ten examples of this sort of situation.

(e.g. Eastern Diamondbacks don't breed with Western Diamondbacks when given an opportunity, though both of them breed pretty well with identifiable mediant sub-populations, which in turn breed _very_ well with each other, and pretty well with the Eastern and Western variants; and here are the mediant interbreeding sub-populations. Instead, the Eastern and Western Diamondback rattlesnakes are clearly there, but we can't point to the transitional forms between them still in existence, and can't really be expected to be _able_ to point to them.)


However, even though these mediant interbreeding sub-populations are effectively invisible to scientific observation (thus no particular examples can be readily given from the population-study literature), nothing is wrong with the basis theory, and the basis theory says they should be there (in common speciation situations). So we can imagine that potential forms exist (i.e. potentially identifiable, except not really either in principle or in practice, but actually existent not just potentially so) and are doing their interbreeding business, even in my backyard.

(Which, as it happens, means we'll also never be in any position to officially call them transitional: we can't observe them, and so have no way to tell which way they're transiting.)

Molecular taxonomy would help with this, giving us data for inferences, to validate via experimental breeding situations; and then we would have a steadily growing set of examples, which BDK could have supplied from the literature--but he didn't.

I can't tell from his explanation whether something inherently prevents this, too; or if the field is so new that no one has been able to get reliably verifiable results yet from experimental testing; or if he thought actual examples would be irrelevant to the topic; or if providing actual examples when I was implicitly and explicitly asking for them over a period of several days and posts never occurred to him.

So I won't count this against his explanation.



Consequently, my second premise (transitional forms are in fact rare when they should be common) is clearly false. Ergo, my conclusion (something is wrong with the basis theory) is also false.


In short: nothing is wrong with the basis theory, thus the mediant interbreeding sub-populations will in fact be there even though we can't point to them, thus they do commonly exist and aren't rare, thus I'm wrong about this indicating that something is wrong with the basis theory.


As any intelligent person would be able to see.


Well... perhaps it would be more charitable to rephrase:

The basis theory not only predicts such mediant interbreeding sub-populations will in fact be there, but also predicts that they will not be scientifically observable. Consequently, even though we can't point to them as specific examples, that doesn't mean anything is wrong with the basis theory.

So, to whatever extent the basis theory can otherwise be shown to reliably work, we can be assured that those mediant interbreeding sub-populations will in fact be there, even though we can't see them. Thus I shouldn't bring up their apparent lack against the basis theory.

Eh... ... {pondering that for a while}

Okay, I might be willing to buy that.

(In theology, I would call this faith. {g})

Jason

Blue Devil Knight said...

Jason, if you think I need to cite scientific literature to justify the claim that individuals with different fitnesses are breeding all the time, then your cockiness so far exceeds your knowledge that I am wasting my time.

Blue Devil Knight said...

Jason,

If you are serious about the science, which it is now clear you are not, you should start with any of the book-length treatments I gave, which with this recent study:
http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=pubmed&dopt=Abstract&list_uids=16136812&query_hl=1

The hallmark of the charlatan creationist is extreme ignorance coupled with similar heights of arrogant confidence. It is one of the least respectable aspect of this fringe movement, which I obviously have little patience for.

Jason Pratt said...

BDK,

The article you referenced does not address my topic. It's about hybridization causing increased reproductive problems in individual hybrids, compared to individuals of the host races, when all three are exposed to food they aren't used to.

The hybrids had (on average) 30% more difficulty breeding than members the host races, when their normal food of MacNab cypress was replaced by incense cedar. I thought it was interesting that "gene flow between host races may be reduced because the correlation between preference and performance that was previously described in parental populations, is essentially broken by hybridization." The immediate implication, in other words, is that interbreeding between races may be expected to make for increased vulnerability to damage in hybrid population numbers due to change in environmental resources. (Forister's study of the holarctic skipper looked interesting, too, and maybe more pertinent to my topic, even if only incidentally. I wonder if that paper's co-author, AM Shapiro, is related to Robert Shapiro?)

Anyway: providing an example of artificially created hybrids having their own worse problems breeding with everyone, in the case of a shared resource substitution, is _not_ the same as providing an example of two species which _don't_ create hybrids when they interbreed (because they never do interbreed) interbreeding pretty well with a mediate subspecies which in turn breeds quite well with everyone (including the speciating fringe populations).


For what it's worth, I was (and still am) serious about maybe being willing to agree with the formulation I reached from putting together your explanations (after charitably removing the accidental circularity of one of your earlier arguments). I was (and still am) also seriously prepared to accept examples from you (or anyone) of the mediant sub-population situation I was talking (and asking) about.

If Gillespie's text (which you recommended) happens to mention them, great; I've added it to my bookpile, so I'll find out whenever I work my way down to it.

(Which may be sooner than later, since I can tell from the book's description that he might in fact address the topic more-or-less incidentally, and even incidental confirmation would be fine with me. Though nothing I've yet read anywhere has bothered to confirm the situation, even in passing, so I'm dubious Gillespie will be any different. But maybe he will. {shrug})

I would, however, prefer if you would at least read the abstracts of your own examples (if not my comments) more closely, if you're going to provide them in the future. Otherwise it only looks as though you merely did a wordsearch for 'interbreed' and 'mediate', and copy-pasted the first example that contained any cognates of those words, without actually reading it yourself. (Maybe you have little patience for reading abstracts, too.)

Jason

Blue Devil Knight said...

Jason,

I suggest you read the paper: I have actually read the paper, and did before I posted the link to the abstract. You obviously have misunderstood the paper by just reading the abstract. These are naturally occuring populations of butterflyes, some individuals which they picked with which to perform an elegant series of experiments. The entire paper can be found here:

http://www.bioone.org/bioone/?request=get-document&issn=0014-3820&volume=059&issue=05&page=1149

It is just a window into a larger literature on this topic.