Remember I am not currently affiliated with a university, and my personal library and teaching materials are in storage, about 350 miles away. So I can’t take the time to look up references for you. Off the top of my head, there was an example involving two groups of salmon that was relatively recent, and the apple maggot example (which is on the page I cited of speciation examples in the previous post.)

Re: punctuated equilibrium, remember, it is not necessarily the rate of mutation which changes but the rate of fixation of particular variations into the genome – the rate of selection. Bottlenecks, founder effects, etc. are sufficient to cause rapid selection and elimination of certain variants from the pool.

And EVERY organism is the tip of a phylogenetic branch. If we had a complete phylogenetic tree, it would have no straight lines at all – every branch would be a forking zig-zag covered with finely branching twigs. The straight lines are only there because we do not have details – we are inferring the connection of reptiles to birds and inferring the branch that leads to Archaopteryx based on morphological continuity. Phylogenetic schematics should not be taken to imply that there were “intermediate” species that existed one after the other in a smooth series of unbranching transitions on the way to one final “branch-tip” species.

Thank you for the stimulating discussion. I could not refrain from reacting, though, in spite of your last statement.

First, you said that speciation was actually observed in the wild. Do you have references ? This is an even MORE impressive observation than the artificial speciation, and I would love to read the original papers. I am very surprised, though, by this information, since, again, I never heard about it, and so many biologists argued that speciation through selection was not observed because our time-frame of observation was too short. Again, what makes biologists NOT use these examples in public debates ?! They are so much more convincing than whatever else I heard before.

Now to your other remarks:

“I don’t think I understand your position. If you are talking about a transitional form between two ancestral species that exist in the fossil record, such as a reptile and a bird, we have examples like Archaeopteryx. What you seem to be saying is that Archaepteryx is not good enough because we do not have the species immediately before and after it in the chain, to prove that it is a “transition”. But surely you see that is an extremely demanding expectation, to find three tightly related species like that?”

I should know better than to argue with a biologist, but to the best of my knowledge, Archeopteryx is no more transitional than any other fossil ever found. To the best of my knowledge, again, on any phylogenic tree that I ever saw, Archeopteryx (like all other fossil organisms) lies at the tip of a branch, not somewhere in-between dinosaurs and birds. I am not asking for any step right before or after a given fossil in the chain, this would be ridiculous on two acounts: (i) the probability would be so small that it would be totally impossible; (ii) from my understanding, there is not a unique “before” and “after” form to a given species: the “tree” of evolution seems rather like a very twisted bush !

So, no, the real objection about the gap in fossil records is the plain and simple fact that, again, to the best of my knowledge, NO direct transitional form have ever been found. Only collateral ones.

“There are certainly “gaps” in the fossil record. Stephen Jay Gould, who accepted evolution completely, argued that to explain them, we need to rethink how evolution works. ”

Yes, I am a little bit familiar with the theory of punctuated equilibrium. What I do not really know is the mechanism theorized by Gould to explain the “saltations”. Obviously, it must have something to do with a time-localized selective pressure, since one could hardly imagine why the rate of mutation would be faster than usual at times. Again, any reference you would know of would be welcome.

Alright, that’s all for today !
Be well,

Pierre

“I am very surprised by your statement that forced selection experiments are possibly not the kind of experiments which have a great interest to evolutionary biologists.”

I am not an evolutionary biologist so I can’t speak for them. But given that we have observed speciation in the wild, I just don’t think generating additional instances of speciation in captivity is so wow. Might just be me, though.

This page lists documented examples of speciation with authors & titles, in case you are interested in additional research.

Also, there are a lot more examples out there than what is found on that page. I think a very cool example of non-Drosophila speciation is this one – in which researchers appear to have duplicated a speciation event that happened naturally, by going back to the parent species they thought were responsible. The speciation induced in the lab supports a hypothesis about a speciation event that occurred historically in nature.

I agree that I wish there was more press about evolutionary evidence. These sorts of reports do enter the news, but the media doesn’t seem to want to make a big deal of them. Again, I think it is because they want to see a lizard evolve into a bird, or some other dramatic morphological change, which is obviously unrealistic! Smaller changes which yield very similar species just do not excite the public very much.

I am glad you found the Padian testimony useful. Regarding your points about transitional fossils, what I wanted Padian’s talk to convey is exactly that it depends what you mean by transitional fossils! I would direct you here for a documented list if you found the data in Padian’s talk insufficiently new. I am not completely clear on whether you disagree that such examples are indeed “transitional fossils”. I think they are transitional.

You said:

“there are intermediate forms with preliminary wings, feathers, paddles, eyes or whatever, and this is indeed a known fact. But this, per se, is not a direct proof of a direct filiation between species. And Padian must know that as well as we do (which is why I find his argumentation weak in this regards). Also, he (quite rightly) explains that finding direct ancestors for all species would be totally unrealistic given the scarce and random fossil record which we have access to. This is indeed right, but off the point: no-one asls for a complete genealogy of all known species, people expect to see ONE direct transitional record between two known species.”

I don’t think I understand your position. If you are talking about a transitional form between two ancestral species that exist in the fossil record, such as a reptile and a bird, we have examples like Archaeopteryx. What you seem to be saying is that Archaepteryx is not good enough because we do not have the species immediately before and after it in the chain, to prove that it is a “transition”. But surely you see that is an extremely demanding expectation, to find three tightly related species like that?

There are certainly “gaps” in the fossil record. Stephen Jay Gould, who accepted evolution completely, argued that to explain them, we need to rethink how evolution works. Specifically, evolution must sometimes proceed at a different pace than the gradual slow steady rate that was assumed. Not all biologists agree with Gould on this. But regardless, Gould emphatically did not say that the “gaps” disproved evolution – even though a few people constantly misuse his quotes to imply that! (That was again part of Padian’s testimony.)

Why didn’t Gould throw evolution out entirely? Because he saw a plausible way to reconcile fossil gaps with the overwhelming evidence for evolution as a whole. To be very simple and visual, a theory is like a net. A net is not a continuous sheet of fabric: it has holes. Yet a net can still very successfully hold things, and the measure of its success is how strong it is, and how many separate things (observations) it holds together. Even if a new discovery rips a cord in the net, the whole net does not need to be thrown out – that would be wasteful. it needs to be repaired with an even stronger cord. That’s exactly what the hypothesis-driven model of science does: try to rip holes, and then fix them with stronger cords.

I hope I don’t insult you with this example because I am aware you understand how science works, Pierre. But I think in terms of framing my perspective on evolution, and the relative unconcern I have about “gaps” in the fossil record, the image sums it up quite well!

Personally, I don’t view the bulk of evidence for evolution as being the fossil record, because I am a geneticist. I view the strongest evidence for evolution as the ridiculously powerful predictive usefulness of the evolutionary model in handling genes. Animals that predicted to have diverged long ago have more dissimilar genes than animals that evolutionary theory predicted would be more closely related. When a deleterious gene like that responsible for sickle cell anemia is in a population, and you ask why it hasn’t been selected out by natural selection, lo and behold, it turns out to protect against malaria, which is a much bigger threat in the tropics! Evolution makes so much sense all the time, in every aspect of biology, that claiming it didn’t happen just is incomprehensible to me.

I think I shall rest with that position statement . Please feel free to respond and clarify any of your own points further, but I think that you will get more out of your own informed research than my comments at this point, so I will move on to other posts. Cheers!

-jess

Thank you for taking the time to write this very detailed response.

My reactions, in fully random order:

“So hopefully you see that the majority of drosophila genetics is not aimed at breeding new species.”

I am indeed fully aware of that. There is obviously much more to genetical research than just probing the theories of evolution. But that does not directly relate to the discussion, I think. I am very surprised by your statement that forced selection experiments are possibly not the kind of experiments which have a great interest to evolutionary biologists. Having the possibility of getting direct evidence of the existence of a mechanism which is known only through indirect records … well, if THAT is not of interest to support the present theories of evolution, what is ?

No, what really made me jump on my seat is your next sentence:

“However, drosophila speciation HAS been accomplished.”

Why on Earth is that result not more emphasized to the public !!?? I mean, if there is a single (and easy-to-understand) argument IN FAVOR of neo-darwinian theories, it is THE one. Never ever did I read about these results in the (many) books and journal articles I have read about theories of evolution. All the arguments I have ever read were based on the very convincing yet INDIRECT evidence of molecular biology and paleontology. As I said in my earlier post, there is no serious reason to doubt the modern interpretation of these numerous converging data. But we physicists tend to favor DIRECT evidence over any other kind, when possible. And indeed, it is often much easier to get reproductible direct evidence of a phenomena in physics than it is in evolutionary biology (for obvious reasons). Then, WHY ON EARTH is the artificial laboratory “production” of a speciation process under a forced selection pressure NOT PUT FORWARD by the defendents of neo-Darwinism ? To a physicist like me, again, this is the most powerful argument that has ever been presented to me: speciation CAN happen as a result of random mutations through many generations, since we have been able to reproduce this very mechanism in the lab. QED.

I find this lack of publicity very weird. Maybe it is a matter of social differences between what physicists and biologists expect as evidence in favor of a theory. But to the general public, the artificial selection of new species would be a very strong argument, again.

“As for your other major point:
“In the same way, no “transitional” fossil has been ever found either (I assume my terminology is wrong, but you will get my point). Fossil records in a cladogram are all cousins, but none of them are another one’s mother or son (if I may use that candid image).””

I have read the very interesting testimony you linked to, and actually saved it for my future records (although I already knew most of what was presented, but this presentation is certainly among the most concise and efficient that I know of). Padian does a good job explaining the way the present theories of evolution construct a coherent corpus based on the fossil and molecular evidence known today. But: his argumentation is not totally honnest (or, rather, he seems to be so much involved in the debate that he possibly unvoluntarily misses the real point).

What ID proponents present as a serious concern in neo-darwinian theories is the existence of gaps in fossil records concerning what is called transitionnal fossils (which, as I have read, was indeed the correct phrasing). But as Padian states himself, the answer wether such transitional records exist depends on WHAT you call “transitional”. Basically, his answer (which confirms what I thought I already knew from my previous readings) is that there exists transitional records if you speak of collateral ancestors (what I naively called “cousins”), but that there is indeed a gap in records as concerns direct transitional ancestors of any known species. So the debate becomes one of apples and oranges: there are intermediate forms with preliminary wings, feathers, paddles, eyes or whatever, and this is indeed a known fact. But this, per se, is not a direct proof of a direct filiation between species. And Padian must know that as well as we do (which is why I find his argumentation weak in this regards). Also, he (quite rightly) explains that finding direct ancestors for all species would be totally unrealistic given the scarce and random fossil record which we have access to. This is indeed right, but off the point: no-one asls for a complete genealogy of all known species, people expect to see ONE direct transitional record between two known species. And even that is missing. Padian’s argument is obviously flawed here (and it is just a matter of pure logic, you do not have to have any scientific training to realize this).

To emphasize my point more clearly, let us play the devil’s advocate for a minute (always a healthy exercise). Let us imagine we are proponents of ID, and analyze Padian’s argument. Namely, he says: “Alright, it is technically impossible or highly improbable that we will find the fossil of a direct ancestor representing a transitional form between two related species, but we find many collateral ancestors (namely, forms that are not direct ancestors but “cousins” with obvious intermediate characters”). What may an ID proponent think of that argument ? Here is one possible answer:

“Alright, I see collateral transitionnal forms in cars all the time: all cars have tires, the modern cars have ABS and airbags and GPS, but they still have a steering wheel and pedals, and a gear box, just as the 1905 early cars. In between, you find HUNDREDS of transitionnal collateral “ancestors” of modern cars, each new technical invention being tested in a model, and then possibly re-used and improved in future cars. Does it mean that cars evolve according to natural selection ? Show me a car breeding baby-cars with random mutations, then selected by the pressure of the market, and leading to a brand new BMW after a full century of car “breeding”, and then I will believe that cars are not intelligently designed.”

Alright, this is a very naive and voluntarily stupid (but hopefully funny) way of presenting the argument. But it clearly shows that, on the basis of pure logic, Padian’s argument that there exist collateral transitional forms is by no means a serious PROOF of neo-darwinian theories, and neither is it a serious REBUTAL of ID. A strong piece of evidence, yes. But no more. And an ID proponent would have an easy job rejecting that flawed argument with just a hand-wave.

Whereas a STRONGER argument, to me, would be: “We HAVE FOUND direct transitionnal ancestors between two species: here is Drosophila Melanogaster #***, which has led to these two separate species when submitted for *** years to radom mutations and an imposed selection pressure.”
This, in my views, is a real argument that speciation can occur under the simple combination of random mutations and a selection pressure.

Again, it really feels to me that people are so involved in the political side of the debate that they kind of miss the real point, and are at risk of sounding extremely peremptory. And this does not do any good to science.

“Hope this helps and sorry for the long delay in my reply.”

Do not be sorry, it was nice of yours to take the time to pursue this interesting discussion. I have learnt much. And the existence of an artificial speciation is a BIG news to me. When you have time, please, send me the related references, and I will get them from our university library. Those are things to have at home, definitely.

Best wishes,

Pierre

“Have such long-term experiments been performed ?”

To generate a new species of drosophila by cross-breeding drosophila in the lab? Yes.

It may surprise you that this sort of experiment is not the kind of experiment drosophila geneticists are generally trying to do. In fact, I personally find that sort of experiment relatively uninteresting.

In order to get a new species, you need three things: variation, a selective force to act on the variation, and time. We’ve been breeding drosophila in the lab for a hundred years, and we’ve generated hundreds of mutations (I’m using the royal science “we” here.) But there is usually no selective pressures applied to “evolve” these stocks of flies. In fact, the mutant stocks are generally quite weak and need to be coddled because we are trying to keep alive flies with very bad mutations. Unlike a natural situation, where only the fit individuals would survive predators or disease, we provide abundant food, safety, and ideal growth conditions, and don’t push them to evolve. I never tried to “evolve” a fly stock, nor did I want to!

“But why should the “brute force” radiation-driven mutation be more or less random than the “natural” ones?”

Your question goes with this related question: if mutations are so bad, as I just claimed, how can they lead to new species through evolution? Well, many mutations that occur naturally are lethal, and affected individuals die, often before hatching. But usually mutations are rare – it’s unlikely an individual will get “hit” more than once with a new mutation, so if the mutation it has is not too bad, it survives and introduces its variation into the population. In contrast, when we generate mutations in the lab using brute-force techniques, an individual fly can end up with a large number of mutations. (Flies produced by mutagenesis screens must be outbred to isolate these individual mutations in different lines of progeny, so we can investigate those mutations one at a time, without confusing the effects of several different genes.) Many other fly stocks that have been useful in genetics have very large chunks of DNA missing entirely, called “deletions.” These flies are especially sick; many of these stocks have been lost over the years despite careful husbandry. Such sick animals would be selected out quickly in the wild! In order to keep the sick, “deleted” chromosome in the fly population, so we can study it, we have to use still other mutant chromosomes that have been created in the lab. These “balancer” chromosomes are grossly abnormal – they are actually designed to be so abnormal that we can see at a glance if a fly carries them, and no fly can live with two copies of them. All of this enables standard genetic experiments to be done using visible markerrs, but all of it makes the flies really messed up.

Basically, all your standard stocks of lab flies have a number of mutant genes in them – that is not something that would happen in the wild, because many of them would be selected out. Some of the mutations are very severe – again, we would expect them to be selected out. We actually try to prevent selection, although it does happen despite our efforts, and makes the flies healthier – reducing the severity of our mutations, and ruining our experiments!

So hopefully you see that the majority of drosophila genetics is not aimed at breeding new species. However, drosophila speciation HAS been accomplished. According to the general definition of “species,” the threshold for speciation is when one population cannot breed with another to make viable offspring. Biologists have generated populations of drosophila that cannot breed with each other successfully, and by that definition are new species. I believe Dobzhansky first did this in 1971, in a Nature paper (which unfortunately I no longer have access to since I am not a subscriber): “Experimentally created incipient species of Drosophila.” Other scientists been able to select for traits in the lab and indirectly create populations (descended from common ancestors) that are reproductively isolated from each other, not by mutating them, but by selecting for flies that naturally preferred different habitats or behaved slightly differently.

The results are not flashy and exciting to the average person. We haven’t created a six-winged fly or a giant fly or a Jeff Goldblum-headed fly–anything that screams “I’m a totally different species!” to the layperson. But keep in mind the layperson cannot distinguish many natural species. I can’t tell the difference between species of seagulls to save my life, but they’re clearly different species.

To come back to your concern:

” So although numerous examples of evolution (driven or natural) have been observed directly INSIDE the same species, to the best of my knowledge, not a single example of the appearance of a new species has ever been observed.”

If you accept the general biological definition of species as reproductively isolated populations, actually yes, it has been observed, in drosophila and other experimental models, more than once.

As for your other major point:

“In the same way, no “transitional” fossil has been ever found either (I assume my terminology is wrong, but you will get my point). Fossil records in a cladogram are all cousins, but none of them are another one’s mother or son (if I may use that candid image).”

I am not sure I quite follow what you are saying here. I do not teach paleontology, so I am not good at deciphering a layperson’s questions and answering them in a helpful way. Because I am so unqualified, I would direct you to this testimony by Kevin Padian, a paleobiologist at UC Berkeley who has been teaching the topic for a quasrter century. It’s from a court case involving the teaching of evolution in public schools here in the US. Padian does an excellent job of going over the fossil history in a manner that is neither dumbed down, nor full of jargon. You can also see his responses to cross-examination on some points similar to those you have brought up.

Hope this helps and sorry for the long delay in my reply.

Thank you for your response. I am aware I am asking a lot from you, so be sure I understand that you may be too busy to answer. Besides, it seems that I have lead your post off-track, and this is maybe not what you expected. Sorry about that.

Sure I will be interested to have a specialist enlighten me on all those issues, which I obviously master very poorly. Just be sure to send me a reminder email when/if you write an answer, so I remember to check this entry of your blog. In two weeks- time, I may not be monitoring it spontaneaously.

Be well, and concentrate on your carreer shift: this is what matters now, not my candid questions.

Best wishes,

Pierre

Let me come back to your response:

“It depends on what you mean by “species” (and “complex animal”.) Since even thousands of years of selection have not made breeds of domestic dogs speciate – they can still interbreed with each other, or other canids – that’s a tall order! We have successfully selected for variants of simpler organisms (with much faster generation times) in the lab – bacteria, insects, etc – that are clearly different in terms of behavior, appearance, or other traits (like antibiotic resistance) than the original population. It’s difficult to say if those variants consistute new “species” or not, since the difference between bacterial species is subtle.”

I was expecting the time-length argument. I am well aware of it, and it would be foolish indeed to expect anything like a forced selection of a mammal into a new species in a reasonnable time frame (since indeed, neither horses, sheep or dogs have undergone such species separation despite a good ten thousand year of human selection). Neither was I referring to selection of bacteria, where indeed the notion of “species” is much more vague (or rather, can have several conventional definitions) than in complex animals with sexual reproduction. But in order to stick the latter, short-lived and fast-reproducing insects is an obvious choice for such experiments. Cross-breed drosophiliae for 25 years, and you can easily get several thousands of generations. Add some degree of forced mutations through a mutagenic environment (drugs or radiations), a good deal of arbitrary selection pressure, and one could expect to end up eventually with different species. Maybe you will need to make it 50 years instead of 25, but … if one thinks in terms of number of generation (which, I guess, must be the right approach), then it sounds like a reasonnable endeavor. Have such long-term experiments been performed ?

And about mutagenic conditions: I am by no mean knowledgeable on DNA mutation, so maybe I will make a fool of myself. But why should the “brute force” radiation-driven mutation be more or less random than the “natural” ones ? I mean, it is not like the heavy ion sent into the animals’ cells could choose what gene it may or may not damage. And what about weakning the DNA repair mechanism, in order to favor the stabilization of random replication errors ?

Finally, I have read many accounts of directed selection in animals like dogs, horse, foxes or drosophiliae. But on each occurence, the evolution remains intra-specific: the behavior, appearance and general needs of the animals may be very different, but they remain members of the same species. You an technically still cross-breed a chihuahua and a wolf (although you had better chose the female on the wolf side) . So although numerous examples of evolution (driven or natural) have been observed directly INSIDE the same species, to the best of my knowledge, not a single example of the appearance of a new species has ever been observed. So those selection-driven differences are nothing near a direct experimental observation of a darwinian species selection.

In the same way, no “transitional” fossil has been ever found either (I assume my terminology is wrong, but you will get my point). Fossil records in a cladogram are all cousins, but none of them are another one’s mother or son (if I may use that candid image).

These two elements, in my view, raise serious concerns. As long as nothing better comes up than the present neo-darwinian theory, it is reasonnable to think that the gaps are indeed due to experimental biases (like the too short duration of forced selection experiments, or the small probability of finding a transitionnal fossil). But my guess is … the day somebody manages to create two viable and completely separate species out of a single one by forced selection … Nobel Prize will not be far !

Don’t you ever wish this would happen, and settle the case once and for all ? And is it not the responsibility of the scientist to inform the general public of those gaps too, in spite of the dangerous political use of these which could be made ?

It depends on what you mean by “species” (and “complex animal”.) Since even thousands of years of selection have not made breeds of domestic dogs speciate – they can still interbreed with each other, or other canids – that’s a tall order! We have successfully selected for variants of simpler organisms (with much faster generation times) in the lab – bacteria, insects, etc – that are clearly different in terms of behavior, appearance, or other traits (like antibiotic resistance) than the original population. It’s difficult to say if those variants consistute new “species” or not, since the difference between bacterial species is subtle.

Basically, selection only works on variation that’s already in the population, until new mutations arise. Selection will not drive mutation – it’s pretty constant, random, and only occasionally useful to the organism. So just because we’re trying to select super-fast in the lab for some new species doesn’t mean mutation will speed up to oblige us in completing the process. We can force mutation by bombarding the organisms with X-rays or chemicals – indeed, that’s what we’ve done with fruit flies very successfully. But that’s an extremely crude, blunt-force technique, and you end up with gross deformities and lethality. It would be completely unacceptable to deliberately speed up mutation in that way in dogs, given the low likelihood of hitting on a favorable mutation. Most of the dogs would just die (as already happens naturally when an embryonic dog or human is too abnormal to survive). So to successfully breed a new species in the lab in the “natural” way, which is the point here, you’d have to go through a prohibitive amount of time and organisms.

On a related note, here was a 40-year study that was done in Russia to domesticate foxes, which is one of the longest studies I know involving selective breeding of mammals. In the end, the selected foxes were significantly different in behavior and appearance from the original fox population, in ways that surprisingly resembled domestic dogs. If you haven’t read it, you really should: Early Canid Domestication

I guess this is where I have to partially disagree with you (at least in principle). A gap or lack of evidence remains a problem of methodology as long as one remains in the same base paradigm. But when the issue of methodology has been sorted out without positive results, then such gaps are the root of a paradigm change.

I do not claim that the present gaps in the neo-darwinian theory are sufficient to question its base paradigm (given that, in a way, this paradigm is sufficiently tautological to be undisputable). But since you only find what you are looking for, the fact of discussing the gaps only in terms of methodology is, in my views, a strong scientific flaw, only driven by social issues. We all know there are whims, trends, pre-conceptions and social pressures in science as well as in any other human activity, and we should stay away from them as much as possible.

And trust an agnostic proponent of the scientific method for those things !

My point may be clearer if I base it on the example of the “founding” experiment of Pasteur on micro-organisms. Neither Pasteur nor ayone in his time could see those, so they had to rely on indirect evidence. And as you know, the commonly accepted notion of the time was to believe in spontaneous generation of diseases, and the non-existence of micro-organisms. The basic principle of Pasteur’s experiment was very simple: to compare the evolution of mould in “sterilized” or “unsterilized” containers. But how do you qualify “sterilization” if you are not sure that micro-organisms exist (let alone know their characteristics) ? This is where the paradigm comes into play, with the notion of self-consistency: if you are Pasteur, you are convinced that micro-organisms exist, and you make an educated guess about their resistance to temperature (namely, you assume that boiling water will be sufficient to kill most of these when exposed long enough). And indeed, what you observe is that mold does not appear on “sterilized” containers and does on the others, thus “proving” that there was something invisible in all of them which was killed by boiling water. QED. Now, the problem with this picture is that it holds only to the self-consistency of the global picture. And several methodological biases could have ruined this self-consistency. For instance, what if Pasteur had been confronted to micro-organisms resistant to temperature (as we know thousands of them now) ? What would have been the meaning of his “sterilization” process in this framework, and what conclusions would he have drawn for the appearance of mold in the “sterilized” containers ? Tricky question.

Of course, later experiments brought more and more arguments in favor of the existence of micro-organisms. But this example shows that the image of the “decisive experiment” is almost always a post-hoc romantic reconstruction. There are numerous such examples, for instance those linked to the famed experiments of light deviation by the sun in the solar eclipse in the early XXth century (I think it was around 1917), which is now too often described as “the first experimental proof of relativity theory”.

So what does all this say about theories of evolution and its scientific status ? It says that indeed, experimental gaps are most likely to be due to methodological problems, given the paramount other evidence of a common filiation of life on Earth. But at the same time, it shows that the whole picture holds only on self-consistency, and that this self-consistency can only hold to a point (like you expressed yourself very clearly). Otherwise indeed, we would have a “complete” and “perfect” theory of something, which is intrinsically non-scientific as you stated. So it is the duty of scientists to try to “save” the paradigm by searching for methodological issues on the one hand, and freely think about what adaptation may have to be made to that paradigm to “save” the paradoxical experimental evidence. My feeling, as an “outsider” in biology, is that the present political climate is NOT favorable to that kind of dual-thinking. And it is not good for science, definitely.

Now, what to do ? As an agnostic myself in every domain of life (including science), I guess that the present theories of evolution are about the most convincing ones we have at our disposal, so I see no scientific reason to doubt them (and every reason to try and understand why there is conflicting evidence in some cases). But in my total ignorance of fundamental evolutionnary biology, I will go one step further into conviction the day we can actually experimentally produce a new species of a complex animal by selection. From what I know, this has yet to happen, and it troubles me.

Now, speaking of audiences … I understand your point about the general public being unaware of what a scientific theory is, and making a swift use of the word whenever they want to mock existing theories which do not fit their own beliefs. But I am afraid that there is no other path for us scientists than a constant attempt at educating people about the roots of scientific method. Speak of a Sysiphian effort ! But in some cases, I really think the effort is wasted. As the saying goes: “You don’t teach a pig to sing: you are wasting your time, and it annoys the pig”.