Slashdot Mirror
Convergent Evolution Upends Honeyeaters' Taxonomy
grrlscientist writes in with a beautiful piece of science, beautifully explicated. The poignant bit is that the birds in question are all extinct. "Every once in awhile, I will read a scientific paper that astonishes and delights me so much that I can hardly wait to tell you all about it. Such is the situation with a newly published paper about the Hawai'ian Honeyeaters. In short, due to the remarkable power of convergent evolution, Hawai'ian Honeyeaters have thoroughly deceived taxonomists and ornithologists as to their true origin and identity for more than 200 years."
You quoted from a non-scientific dictionary. Theory is misunderstood by people like you. In the context of science:
1. You make observations.
2. You make a hypothesis based on your observations.
3. You test you hypothesis to see if it holds its ground.
4. If the hypothesis survives the tests, it becomes a theory, if not it's back to square one.
Another poster, who is probably a biologist, gave two very good answers to your question; as a bioinformatician, I'll give a third. You're right that DNA doesn't lie, but we can have a damn hard time figuring out what it's trying to tell us. There is no universally agreed-upon method for reconstructing phylogenetic trees from sequence data -- Google on "phylogenetic algorithm" to see the enormous number of methods that people have come up with, and what an active area of research this continues to be. Also, the Linnaean taxonomic system, obviously, was not designed with modern genetics in mind, and trying to shoehorn phylogenetic data into this system (which is pretty much what everybody does, even if they're not happy about it) can lead to bizarre results. Until we have what everyone agrees is really a gold standard method for reconstructing ancestral trees, this is the way it's going to be.
The correlation between ignorance of statistics and using "correlation is not causation" as an argument is close to 1.
The Law of Gravitation and the Theory of Gravity are two related but different things.
The Law of Gravitation is (like most laws) is an equation describing the effect of gravity and nothing else.
The Theory of Gravity goes over gravities existence (or not) and how it actually works.
Theories can never become Laws (and Laws were never Theories) because they are two fundamentally different concepts within science.
Thank you for making this post. It's good to see that someone out there understands this basic concept. Although, as a sister post notes, we do have a pretty good idea as to the cause of gravity.
Law of gravity: Objects attract each other according to their mass and the square of the distance between them. (F = GMm/R^2)
Theory of gravity: Gravity is caused by the interaction of gravitons at the quantum level, making gravity one of the basic forces.
Notice that theories describe the how and why of things, laws make concrete facts about the way things are.
Neither scientific laws or scientific theories (not to be confused with traffic laws or layman's theories, which describe different things entirely) are above being revised. In the case of laws, becoming more precise as our instruments become more precise, for instance, I believe that the universal gravitational constant (G, above) has been changed at various times as we are able to get better measurements. Theories, are often refined, partially revised, or scrapped altogether, often, we will use different theories depending on the model we are using (once again, scientific model, not like a car model, or like a car in general, get your analogies away from me! ;-) ).
The most important distinction between laws and theories, is that laws can be proven: measure the gravitational pull between two masses, if it fits the formula, voila! Gravity still works as predicted, the aprocralypse is not upon us yet. Theories, however, can only be disproven, you can say, "Gravitons cannot be the cause of gravity, here is my evidence, which the theory doesn't explain," but you cannot say, "Here is my proof that gravitons cause gravity." (you can try to get proof, but at most you will have a very strong correlation, which, as every slashdotter is fond of pointing out, does not prove causation.
"But," you say, "that means that everything we know could be a lie!" To which I reply, "Yes, yes it does, but that is unlikely, because most theories that actually have a name, have been shown to correlate with the facts very, very consistently, and although this doesn't prove causation, it's a pretty good indication that most of science is true."
The creator of this post (Jacob Smith) hereby releases it, and all of his other posts, into the public domain.
For example - if this or that was proven in science, makes it into a science book, and is proven wrong later... what would you call it?
I'd call it not math. Math is the only place where you can prove things; everywhere else, you can only falsify things.
Look at global warming. Both sides think it's proven.
Ahh, but all the scientists are on the side that thinks it's happening and are arguing a bit over how much is due to us.
"We returned the General to El Salvador, or maybe Guatemala, it's difficult to tell from 10,000 feet"
When your theory is defined such that it includes both sides of interpretation of any given piece of data, it is hard to argue against. (e.g., I heard on the news that it was getting colder and certain ice caps or something were growing, not shrinking, and that that is "exactly what is expected with global warming, because with something like global warming, the unexpected is going to occur."
If that is actually what the "scientist" said, that is indeed moronic. Can you provide a reference? What do you mean by "the news"? Sounds more like some anti-global-warming moron trying to create a straw man.
OTOH, last I looked, there were good (ie. they have made accurate predictions) climate models that predict increasing average global temperatures while simultaneously predicting cooling around the poles. This is good science: show that your model fits some data and then try to understand what's going on. If your model does not make correct predictions (give it (today-n) years of data, and see if it can predict the last n years correctly), you need to change your model. That is exactly what science is.
Evolution is harder, since the physics is infinitely more complex--there are no precise predictive models of evolution in the physics sense, so whatever we see has to be incorporated into the theory without predictive testing. Evolution is thus more useful as a series of observations and a way of explaining them, rather than a proof in any mathematical sense. That doesn't make it wrong.
Elementary evolution theory does make a claim like "All complex life evolved from simpler life." If you now find a fossil of something complex that has no simpler ancestors, then evolution is wrong. Not that you can easily prove the bit about ancestors, but you get the idea. Finding that some complex life has simpler ancestors doesn't prove evolution, but it certainly makes it more plausible. Excuse the very shallow example, but you get the point.
"The biggest problem with communication is the illusion that it has taken place."