The DinoBase Public Forum

When I first saw the title, I became concerned that you'd fallen into some chronic emotional crisis. Well, hey, you and me both...

Very interesting! Speaking of turtles, I keep reading bits and pieces that there's a theory that turtles are ARCHOSAURS or at least close to them, and not anapsids as usually believed? What's the evidence for this theory, and has it garnered any acceptance in the paleo community?

I think even more intriguing than the question of the origin of life is:

Could/does the same thing happen elsewhere?

Evolution, ultimately, cannot be proved in the same way that other fields of science "prove" their theorems and laws.

The evolution of life on this planet is not a repeatable experiment.

For one, it is very costly to create a planet, then throw in some basic elements like carbon, oxygen, nitrogen, silica, etc., then wait 4.5 billion years to see if accepted theories on phylogeny play out.

For another thing, given that living things are highly opportunistic, the history of life on THIS planet is as much the consequence of plate tectonics, volcanism, the carbon cycle, etc. as it is the variability of DNA.

Until those UFOs stop abducting people and choose instead to talk openly online, that question will never come close to being answered.

David Hone wrote:

I think you are confused there - only in maths can things be proved, and physics, chemistry and the rest cannot prove their tehories any more or less than biology can.

You have also got abiogenesis and evolution mixed up, and we can certainly repeat the experiemnt by trying to replicate the conditions on a younf Earth to see what happens. Sure the smae thing is unlikely to reoccur, but we can look for patterns and clues. We cannot re-run the formation of the solar syatem, but you would not say that the work doen by astrophysicists cannot be analysed or cannot be proven as well as evolution. You learn, make hypotheses, test them, form theories, make predictions and test them. That is what we do, and this can be applied to abiogenesis and evolutionary history as it can any other branch of science.

Well - I don't think you can have one without the other (abiogenesis and evolution, that is), they are inextricably linked. It's not just about whether *can* life arise under a particular set of circumstances, but what path does life follow given that same set of circumstances? Is all life carbon-based? Is bilateral symmetry inevitable? Is the tetrapod design the only one which can evolve once life leaves its fluid origins? How long before "intelligent" life arises? And how does all of it relate to the fact that planet is also changing over time - and that a good deal of that change is due to biotic activity itself?

These are all unprovable, just as one cannot prove whether birds came from dinosaurs, or whether T-rex was a predator or a scavenger.

People can gather fossil evidence and argue and make the best case for their arguments, but they cannot prove their theories.

This is what paleontology is all about: arguing (which is why the field has so many hotheads who shall remain nameless here).

Other sciences, however, *do* have experiments which are repeatable: if I take a measured amount of water, then break it down with electrolysis, I will always end up with two parts hydrogen and one part oxygen, there are no other possible outcomes; if I have dominant Gene A and recessive Gene B, they will be expressed in the offspring accordingly just as Mendel discovered.

You can't do anything like that in paleontology.

You mention that only math is "provable" - well yes - and that's what most of chemistry and physics are based on: mathematical models. The creation of the solar system is not a repeatable experiment, but the model on which it is based consists of complex equations, all of which have been "proven". The same with the theories of Relativity and Quantum Mechanics: impossible to physically illustrate/re-enact these concepts, yet they are accepted theories because the math behind them plays out.

You really can't even "test" anything in paleo - the animals are all dead. Based on presumed phylogenetic relationships deduced from the available fossil record, however, we can make certain predictions about what remains to be discovered - just someone once predicted "Archeopteryx" or "Tiktalik", or even some of the earliest reptiles leading to dinosaurs. (But even then, no one can really "prove" exactly which vertebrate line these animals descended from because we can't resurrect the postulated ancestors, then wait millions of years for them to re-evolve Archy or Tiky.)

No one will ever really know how - or exactly when - the last dinosaur died. People can argue - with very good geological evidence - it was an asteroid or climate change, but the issue remains irresolvable. Unless someone has it on videotape, it will forever remain mystery. And you can't stage an experiment which would prove it (oh you can mathematically model a horrendous asteroid impact - but the exact setting of the horrendous impact remains unknown - were the dinosaurs already gone by that time?)

Last edited by OstromsRazor (2008-06-07 21:35:56)

Well as you said in your post for Answer #4

David Hone wrote:

Testability and repeatability are fundamental to science

And that's exactly my point about paleontology: ultimately, very little is testable, virtually nothing is repeatable. Even among the paleontologists who responded to your question was the general thinking that alot is speculative and must remain that way.

David Hone wrote:

So models are proof are they? I don't think so. We have models of evolution that work on maths too, yet you don't seem to think these count.

Well - in a previous post in this thread you cited that nothing is provable except in math. My point is that concepts in chemistry and physics play out because they are built on mathematical proofs. There is nothing like that in paleontology. There is no universally fundamental equation like E=mc2 or Avogradro's law.

The closest thing we have is "Cope's Rule" which is a concept so steeped in 19th century thinking so as to be virtually archaic - yet there are still those who support it.

David Hone wrote:

I really think you need to learn more about science works.

I know how science works. In fact, I've worked as a scientist. On a daily basis we would gather our data, make predictions, then change them and test our ideas/predictions again. Paleontology doesn't work like that - it's all about explaining what has passed.

The problem too is that so much of it is subjective. As you also mentioned in your post to question #4:

David Hone wrote:

...that goes too for basic descriptions and the associated interpretations of skeletons – is that a partial femur or a distorted humerus, is the lacrimal really missing or just crushed and / or dissolved, is that a fenestra in the skull or a hole left by damage?

Only in paleontology, only paleontology. A chemist would never misidentify a carbon atom for an oxygen atom - even if isotopes were involved. A physicist would never mistake up for down.

But much in paleontology is interpretation - not calculation. And thus - that bone - a fundamental unit of a vertebrate skeleton - will always carry with it the semblance of a femur to some, a humerus to others.

Last edited by OstromsRazor (2008-06-10 06:23:23)

The scientific method is the same in any science, from organic chemistry to linguistics: you propose a series of hypotheses and then you test them.  The results that you acquire from your experiments allow you to reject, or not reject, the null hypothesis based on a degree of probability.  Really, 'proving' or 'disproving' doesn't come into it - you can only say that, given your results, there is a very great likelihood of something.  Palaeontology (when carried out rigorously) works like this too. We propose a series of hypotheses and we set out to test them. We gather data and we analyse it. Our results let us conclude that our originial hypotheses were correct, or incorrect, within a certain degree of probability. Sure - we have to battle with small sample sizes and data that is sometimes ambiguous. But that does not change our methods or, necessarily, our results.  And the great thing is that new data is continually coming to light, allowing us to test our original hypotheses against new discoveries.

...Actually, I had a bit of a revelation here but I think people (even some scientists) are confusing science vs non-science with quantitative science vs qualitative science.

Although I have distinguished the natural history side of palaeo from the science side of palaeo in my previous post, I must make it clear here that it is actually more appropriate to call them qualitative and quantitative palaeontology. The natural history aspect, i.e. qualitative science, of palaeontology is still a science, perhaps albeit being a "soft" science. Let's say that we find a new fossil and we initially think it is a new species of genus A. So we compare this new specimen with known specimens of genus A and also with specimens of other genera, let's say B and C. After extensive observations, measurements, etc. we find that this new specimen is unique from genus A and probably is not a member of this genus. The same goes for genera B and C so we can conclude with some confidence that this new specimen probably represents a new genus that we shall call D from now on. This conclusion is based on evidences gathered from extensive comparative anatomy. So the hypothesis was that this new specimen was a new member of genus A but the evidence didn't support that, instead we take an alternative hypothesis that this is a new genus, which it seems likely from its uniqueness as compared to other genera B and C. This leads us to conclude that this specimen likely represents a new genus. This conclusion can be rejected like any other scientific conclusions, perhaps with a subsequent discovery of a new specimen or perhaps by another worker who investigates this in much more detail with information that we may have overlooked.

Another example would be simple functional morphology vs biomechanics. With increasing advancement and decreasing costs in computer technology, we can use more sophisticated biomechanical approaches to help us understand quantitatively the relationship between morphology and function. However, much more simpler qualitative approaches can also provide us with similar understandings. For instance we compare general snout morphology of modern crocs and compare that with spinosaurs - coupled with some acid-etched scales, we can infer with some certainty that spinosaur snout morphology is indicative of piscivory. This is a likely conclusion given the evidences but can be rejected if new information or more rigorous testing show otherwise. Another example of functional morphology is looking at muscle attachments and determining if one morphology was more efficient with its force input vs force output. For instance if one species had muscle attachments further down the arm as compared to another species, then you'd know that the first species could have a much more efficient arm flexion compared to the second. You can go on and discuss arm flexion efficiency with respect to prey handling or digging or whatever - and formulate a working hypothesis that can be further tested by biomechanics for instance.

Thus, qualitative approaches are what some would call a "soft" science in that it doesn't test hypotheses with enough rigor. However, qualitative approaches are usually the necessary steps leading to more rigorous quantitative analyses and thus cannot be criticised for its "softness".

Why not research and read the literature on the subject - I would especially point you to Kingsolver & Pfennig in 2004. You are challenging something that is already well documented in the literature, and with almost the exact requirements you are demanding. The only thing missing I think is the designation of a new species trough size increase, but species are rarely if ever defined by size alone and certainly there are cases of enormous jumps in size on occasion that could warrant definition of speciation as it would prevent cross-breeding (e.g. the giant feral cats of Australia).

Manabu Sakamoto wrote:

You don't "prove" anything in any field of science.

Ultimately, usually, you propose a theory which is then confirmed - or at least agreed upon - when the results get repeated as they are tested independently by others.

This is what sunk "cold-fusion" a couple of decades ago: no one could repeat the experiment.

Manabu Sakamoto wrote:

It is actually rare to find a purely descriptive paper these days, most are combined with some sort of analytical study, such as, cladistics, biogeography, biomechanics, testing Cope's rule, looking at diversity, etc, all of which employ the scientific method.

Well - exactly - a paper concerned with something like biomechanics is steeped in physics, and the physics portion of the discussion certainly requires quantitative analysis. But the strictly paleo-portion of the analysis still lingers in the realm of the speculative: no one say with any certainty what the animal did during its life because they are still too many unknowns (i.e., metabolism, environmental geochemistry, etc.).

Cladistics? Hah! Consider how many solutions exist when constructed by different paleontologists. They interpret - yes that's a key word here - interpret relationships among lines of organisms (ichthyosaurs are sauropterygians - no wait they're ichthyopterygians - no wait they're an order - no wait they're a sub-order - no wait they evolved from Labyrinthodont amphibians...). All of this remains arguable - not testable - (I'm not using the word "provable" mind you - but "testable"). Indeed what test could one concoct to determine where those "fish-reptiles" came from? What test whose results would give the same result when performed by other paleontologists?

Last edited by OstromsRazor (2008-06-11 00:06:15)

I am not going to get drawn itno this, but:
1. A rule does not have to be universal to be a rule. Read up on scientific rules.
2. Yes of course the conclusiosn are in a conditional tense, have you not read the above? Science does not work in absolutes, merely evidence.
3. They repeatedly observe an increse in size in multiple species over multiple generations.
4. You want more time? Science has only really systematiclly been going for 150 years, wait another 500 and see what is happening. Or check the fossil record. You do not seem to think that counts, but everyone else does.
5. Many formats in journals do not allow for a 'conclusions' section, and even fewer offer "watertight" results. This is biology, not maths, remember?

Another even later entry from a British museum curator working in the UK:

I reckon it must be still abiogenesis... evolutionary theory can't provide an answer, nor can any other branch of biology or chemistry yet. I'm not sure it's even a viable question for palaeontology!