Richard Dawkins has a new book, in which he sets out to prove, once and for all, that evolution is a fact `in the same sense as it is a fact that Paris is in the northern hemisphere’, or, as he calls it, a Theorum.
Here he is talking about a prediction made by Darwin and Wallace after observing the striking length of the nectaries of Angraecum Sesquipedale (it’s an orchid ) that an insect capable of extending to such a length must exist. The insect was indeed discovered forty years later, and aptly baptized praedicta.
This little example gives the lie, yet again, to the allegation that evolutionary science cannot be predictive because it concerns past history
(The Greatest Show on Earth, Kindle location 718)
Two comments come to mind. First, this prediction, insightful as it is, has little to do with the core assertions of evolutionary science — that all species evolved from one another, and that this process is driven by random mutations and natural selection. Indeed, it seems to me that Darwin was not appealing to any of these ideas when he made it, but based it solely on the fact that the plant fertilizes via insect pollination and that the nectar serves to lure these insects. I don’t care or know much about creationists, but my guess is that this fact will be gladly adopted by the most diehard of this lot, since it perfectly fits their framework of design and purpose in Nature.
Second, when critics complain about the poor predictive power of evolutionary science, they mean I think falsifiable prediction. An assertion that an insect will be discovered at some unspecified time in the future is not falsifiable, because you can never go wrong. Either the insect is found, in which case you made a prediction that turned out to be true or the insect has not been found yet, but this does not serve to weaken the theory of evolution. Case in point: Angraecum longicalcar, a cousin of the A. Sesquipedale whose nectaries are too long even for the praedicta. The insect whose tongue is large enough to suck the nectar from the longicalcar was not found yet, but Dawkins is not a bit embarassed by this fact. On the contrary, he only mentions it as another manifestation of the predictive power of evolution.
6 comments
September 28, 2009 at 2:43 pm
Nate
I hope he did a good job. I felt his last book (The God Delusion) suffered from inconsistent arguments. Do you know any authoritative and firm but fair reviews out there?
October 1, 2009 at 6:19 pm
Chris Edmond
Dawkins does not use the *language* of falsification, but in the book he does give numerous examples of falsifiable predictions. Easiest to grasp, I guess, is the example due to Haldane: no rabbits in the precambrian era. Anyone who finds a fossilized rabbit dating to the precambrian will falsify evolution.
And of course evolution makes a vast multitude of such predictions at the levels of species, organisms and genes, all of which are falsifiable.
October 2, 2009 at 5:19 pm
Eran
This takes me away from my original post, but I would argue that Haldane’s famous retort to the falsifiablity challenge is unsatisfactory.
First, why should a precambrian rabbit falsify the theory of evolution ? Our view about the timing of arrival of different life forms to this planet is not a deduction from Darwinian Evolution but a reflection of what we found in the fossil record. Thus, if it turns out that mammals came a couple of hundreds millions year before we thought, this fact per se is no problem to evolutionary theory.
However, the imaginary precambrian rabbit would indeed present a problem for evolution theory since, according to this theory, species once lost should not pop up on earth again 700 million years later.
But note first that to view this as a blow to evolution theory we need to agree that inexistence of rabbits in the fossil record during these 700 million years proves their inexistence on the planet. But such an argument is the creationist’s favorite weapon, and is (rightly in my view) dismissed by evolutionists.
Moreover, the precambrian rabbit is not a falsification of the core assertions of Darwinians evolution, but of the theory that describes the timing of appearance of different life forms on this planet. The fact that species do not re-appear after extinction is not a prediction that was first made as some deduction from Darwin’s theory about gradual changes in species by mutations and selection, but an observation that was already known to Darwin and he mentioned it as a natural phenomena that is well explained by his new theory.
Haldane essentially says that the type of fossils we found until now would be the same type of fossils we will continue to find. That consistent evolutionary stories can be told that fits with what we have found until now is a plus, but the prediction itself has little to with evolutionary theory. Yes, it falsifies something, but that something is not the theory of random mutations and natural selection.
October 4, 2009 at 3:24 pm
Daniele
I guess that the theory of random mutations and natural selection (that you mention in your last sentence) could be easily tested in the laboratory (and it has been, I guess). You pick a sample of individuals from one specie and put them in an habitat where they can thrive. Then you introduce some variation in their habitat that create pressure to their existing abilities (e.g. a new species or a different climate). If you observe the characteristics of the population changing in time, as a result of random mutations that get transmitted to next generations, you conclude that the theory is valid.
Instead, if what you have in mind is a test about a theory on the origin of species in the planet earth, then one must be content with imperfect tests (see Haldane), or “thereof one must be silent”. In fact, there is no way we can satisfactorily falsify a statement about the past in the scientific sense, that is by experiment. For an extreme example, suppose that I have a theory that tells you that Earth’s gravity has been constant (approximately) in the last few days. What would be a falsifiable prediction of my theory?
October 4, 2009 at 11:27 pm
Eran
I don’t think assertions about the past are inherently non-falsifiable. Precambrian rabbit would falsify the theory that lost species do not pop up again on earth. Theories about the age of the planet and its climatic history have implications (assuming the laws of Nature) about the distribution of radioactive isotopes on earth today. The theory about common descent of living species has falsifiable implications in terms of similarity between the DNA sequences. However, the theory that random mutations and natural selections drives the evolution of species don’t have such implications. If that theory would `take more risks’ and commit to some explicit mathematical description of the random process that generates the mutations then this theory would probably also have some falsifiable implications.
No theory — not even physics theories with their astonishing capacity to predict — reaches the falsifiability ideal in its purest form. This is true for theories about the past as well as for theories about the present. at the very least we always take some metaphysical assumptions for granted like some constancy in the laws of Nature, and some trust in outcomes of other people’s experiments and in our own memory. But the assertion that the evolution of living species was driven solely by random mutation and selection is very far from this ideal.
October 5, 2009 at 7:19 pm
Daniele
I think we agree that the definition of “fact” must be somewhat vague and that “truth” is something that is not well defined outside formal systems.
Why then should not we call a fact the statement “the evolution of living species was driven solely by random mutation and selection”, given the observation (that seems to me widely accepted in the scientific community) that all living species that have been closely observed appeared to evolve solely by random mutation and selection (while no other species that has been observed that evolved in a different manner)?
Said that, I agree that it would be better to call the thing “hypothesis” or “theory”, rather than “fact”. But of course this caution should also apply to the laws of physics.